Food flavour technology

0 26 0
  • Loading ...
1/0 trang

Thông tin tài liệu

Ngày đăng: 21/09/2019, 21:37

favour technology in food and beveragetraing in food flavourthe requirement to develop new flavour sensory technique for development of beverage and othersFood flavour technology is of key importance for the food industry. Increasingly, food products must comply with legal requirements and conform to consumer demands for “natural” products, but the simple fact is that, if foods do not taste good, they will not be consumed and any nutritional benefit will be lost. Food Flavour Technology Second Edition Edited by Andrew J Taylor and Robert S.T Linforth Division of Food Sciences, University of Nottingham, UK A John Wiley & Sons, Ltd., Publication This edition first published 2010 C 2010 Blackwell Publishing Ltd Blackwell Publishing was acquired by John Wiley & Sons in February 2007 Blackwell’s publishing programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom Editorial offices 9600 Garsington Road, Oxford, OX4 2DQ, United Kingdom 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic books Designations used by companies to distinguish their products are often claimed as trademarks All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners The publisher is not associated with any product or vendor mentioned in this book This publication is designed to provide accurate and authoritative information in regard to the subject matter covered It is sold on the understanding that the publisher is not engaged in rendering professional services If professional advice or other expert assistance is required, the services of a competent professional should be sought Library of Congress Cataloging-in-Publication Data Food flavour technology / edited by Andrew J Taylor and Robert S.T Linforth – 2nd ed p cm Includes bibliographical references and index ISBN 978-1-4051-8543-1 (hardback : alk paper) Flavour Flavouring essences Flavour–Analysis I Taylor, A J (Andrew John), 1951- II Linforth, Robert S T TP418.F65 2010 664 07–dc22 2009028000 A catalogue record for this book is available from the British Library Set in 10/12 pt Times by Aptara R Inc., New Delhi, India Printed in Singapore 2010 Contents List of contributors Preface Creating and formulating flavours John Wright 1.1 Introduction 1.1.1 A little history 1.2 Interpreting analyses 1.3 Flavour characteristics 1.3.1 Primary characters 1.3.2 Secondary characteristics 1.3.3 Taste effects 1.3.4 Complexity 1.3.5 Flavour balance 1.3.6 Unfinished work 1.4 Applications 1.4.1 Ingredient factors 1.4.2 Processing factors 1.4.3 Storage factors 1.4.4 Consumption factors 1.5 Flavour forms 1.5.1 Water-soluble liquid flavours 1.5.2 Clear water-soluble liquid flavours 1.5.3 Oil-soluble liquid flavours 1.5.4 Emulsion-based flavours 1.5.5 Dispersed flavours 1.5.6 Spray-dried flavours 1.6 Production issues 1.7 Regulatory affairs 1.8 A typical flavour 1.9 Commercial considerations 1.9.1 International tastes 1.9.2 Abstract flavours 1.9.3 Matching 1.9.4 Customers 1.10 Summary References xi xiii 1 3 6 8 10 10 11 11 11 12 13 13 13 14 15 16 16 19 19 20 21 22 22 23 iv Contents Flavour legislation Jack Knights 2.1 2.2 2.3 2.4 2.5 2.6 Introduction Methods of legislation Legislation in the United States International situation: JECFA Council of Europe European community 2.6.1 Background – national to EU legislation 2.6.2 The 1988 Council Directive 2.6.3 Smoke flavourings 2003 Directive 2.6.4 Developments 2008 onwards 2.7 Current EU Situation and the future References Basic chemistry and process conditions for reaction flavours with particular focus on Maillard-type reactions Josef Kerler, Chris Winkel, Tomas Davidek and Imre Blank 3.1 Introduction 3.2 General aspects of the Maillard reaction cascade 3.2.1 Intermediates as flavour precursors 3.2.2 Carbohydrate fragmentation 3.2.3 Strecker degradation 3.2.4 Interactions with lipids 3.3 Important aroma compounds derived from Maillard reaction in food and process flavours 3.3.1 Character-impact compounds of thermally treated foods 3.3.2 Character-impact compounds of process flavours 3.4 Preparation of process flavours 3.4.1 General aspects 3.4.2 Factors influencing flavour formation 3.4.3 Savoury process flavours 3.4.4 Sweet process flavours 3.5 Outlook References Biotechnological flavour generation Ralf G Berger, Ulrich Krings and Holger Zorn 4.1 4.2 4.3 4.4 Introduction Natural flavours: market situation and driving forces Advantages of biocatalysis Micro-organisms 4.4.1 Biotransformation and bioconversion of monoterpenes 4.4.2 Bioconversion of C13 -norisoprenoids and sesquiterpenes 4.4.3 Generation of oxygen heterocycles 24 24 24 26 27 28 30 30 31 40 41 47 48 51 51 51 54 58 61 62 65 65 70 74 74 74 78 80 80 81 89 89 89 90 91 91 95 96 Contents 4.4.4 4.5 4.6 4.7 4.8 4.9 Generation of vanillin, benzaldehyde and benzoic compounds 4.4.5 Generation of miscellaneous compounds Enzyme technology 4.5.1 Liberation of volatiles from bound precursors 4.5.2 Biotransformations 4.5.3 Kinetic resolution of racemates Plant catalysts 4.6.1 Plant cell, tissue and organ cultures 4.6.2 Callus and suspension cultures 4.6.3 Organ cultures 4.6.4 Plant cell biotransformations Flavours through genetic engineering 4.7.1 Genetically modified micro-organisms 4.7.2 Isolated enzymes from genetically modified micro-organisms 4.7.3 Plant rDNA techniques Advances in bioprocessing 4.8.1 Process developments in microbial and enzyme systems 4.8.2 Process developments of plant catalysts Conclusion References v Natural sources of flavours Peter S.J Cheetham 5.1 Introduction 5.2 Properties of flavour molecules 5.2.1 Flavour perception 5.2.2 Differences in sensory character and intensity between isomers 5.2.3 Extraction of flavours from plant materials 5.2.4 Commercial aspects 5.2.5 Economic aspects 5.2.6 Safety aspects 5.3 Dairy flavours 5.3.1 Background 5.3.2 Cream and butter 5.3.3 Cheese 5.4 Fermented products 5.4.1 Hydrolysed vegetable proteins 5.4.2 Chocolate 5.4.3 Tea 5.4.4 Coffee 5.4.5 Beer 5.4.6 Wine 5.4.7 Sweeteners 5.5 Cereal products 97 99 101 101 101 103 104 104 105 105 107 107 108 109 110 112 112 114 114 115 127 127 129 129 141 142 146 147 147 147 147 148 149 151 151 152 153 154 154 156 158 158 vi Contents 5.6 Vegetable sources of flavour 5.6.1 Spice flavours 5.6.2 Mushroom 5.6.3 Garlic, onion and related flavours 5.6.4 Brassica flavours, including mustard and horseradish 5.6.5 ‘Fresh/green/grassy’ 5.6.6 Nuts 5.6.7 Other vegetables 5.6.8 Fermented vegetables 5.7 Fruit 5.7.1 Apples 5.7.2 Pears 5.7.3 Grapefruit 5.7.4 Blackcurrant 5.7.5 Raspberry 5.7.6 Strawberry 5.7.7 Apricot and peach 5.7.8 Tomato 5.7.9 Cherry 5.7.10 Tropical fruit flavours 5.7.11 Vanilla 5.7.12 Other fruits 5.7.13 Citrus 5.7.14 Citrus processing 5.8 Other flavour characteristics 5.9 Fragrance uses 5.10 Conclusion References Useful principles to predict the performance of polymeric flavour delivery systems Daniel Bencz´edi 6.1 6.2 6.3 6.4 6.5 Overview Introduction Compatibility and cohesion Sorption and swelling Diffusion and release References Delivery of flavours from food matrices Saskia M van Ruth and Jacques P Roozen 7.1 Introduction 7.2 Flavour properties 7.3 Thermodynamic aspects of flavour delivery 7.3.1 Definition of gas/product partition coefficients and activity coefficients 7.3.2 Types of binding 159 159 161 161 163 164 164 165 165 165 166 167 167 167 168 168 169 169 169 170 170 171 171 172 174 174 175 175 178 178 178 179 182 184 187 190 190 191 191 191 193 Contents 7.3.3 Lipid–flavour interactions 7.3.4 Carbohydrate–flavour interactions 7.3.5 Protein–flavour interactions 7.4 Kinetic aspects of flavour delivery 7.4.1 Principles of interfacial mass transfer 7.4.2 Liquid food products 7.4.3 Semi-solid food products 7.4.4 Solid food products 7.5 Delivery systems: food technology applications 7.6 Conclusions References Modelling flavour release Robert S T Linforth 8.1 Introduction 8.2 Equilibrium partition models 8.2.1 The air/water partition coefficient 8.2.2 Estimation of Kaw using QSPR 8.2.3 Effect of lipid on volatile partitioning 8.2.4 QSPR estimation of the air/emulsion partition coefficient 8.2.5 Internet models and databases 8.3 Dynamic systems 8.3.1 Modelling flavour release from a retronasal aroma simulator 8.3.2 Non-equilibrium partition modelling of volatile loss from matrices 8.3.3 Modelling the gas-phase dilution of equilibrium headspace 8.3.4 Modelling the gas-phase dilution of equilibrium headspace above emulsions 8.3.5 Modelling the rate of volatile equilibration in the headspace above emulsions 8.4 In vivo consumption 8.4.1 Modelling release from emulsions during consumption 8.4.2 Effect of gas flow on volatile equilibration above emulsions 8.4.3 Modelling volatile transfer through the upper airway 8.4.4 Non-equilibrium partition model for in vivo release 8.4.5 Modelling flavour release using time–intensity data 8.4.6 QSPR of in vivo volatile release from gels 8.5 Conclusion References Instrumental methods of analysis Gary Reineccius 9.1 Analytical challenges 9.2 Aroma isolation 9.2.1 Aroma isolation methods based on volatility vii 194 195 196 197 198 200 200 201 202 203 203 207 207 208 208 209 211 212 213 214 214 215 216 218 219 220 222 222 223 223 224 224 226 227 229 229 231 231 viii Contents 9.3 9.4 9.5 9.6 9.7 9.8 9.9 10 9.2.2 Aroma isolation methods using solvent extraction 9.2.3 Solid-phase micro-extraction 9.2.4 General considerations in preparing aroma isolates 9.2.5 Aroma isolation summary Selection of aroma isolation method 9.3.1 ‘Complete’ aroma profile 9.3.2 Key components contributing to sensory properties 9.3.3 Off-notes in a food product 9.3.4 Monitoring aroma changes in foods 9.3.5 Using aroma compound profiles to predict sensory response 9.3.6 Summary comments on isolation methods Aroma isolate fractionation prior to analysis 9.4.1 Fractionation of concentrates prior to analysis Flavour analysis by gas chromatography 9.5.1 High-resolution gas chromatography 9.5.2 Gas chromatography–olfactometry 9.5.3 Specific gas chromatographic detectors Flavour analysis by HPLC Identification of volatile flavours 9.7.1 Gas chromatography 9.7.2 Infrared spectroscopy 9.7.3 Mass spectrometry Electronic ‘noses’ Summary References 237 238 241 241 242 242 243 243 244 244 245 245 245 249 249 250 254 254 255 255 256 257 261 262 262 On-line monitoring of flavour processes Andrew J Taylor and Robert S.T Linforth 266 10.1 10.2 266 268 268 268 269 270 270 270 271 272 272 275 275 276 277 10.3 10.4 Introduction Issues associated with in vivo monitoring of flavour release 10.2.1 Speed of analysis 10.2.2 Analysis of different chemical classes 10.2.3 Sensitivity 10.2.4 Identification of analysed compounds 10.2.5 Interfering factors 10.2.6 Non-volatile tastants Pioneers and development of on-line flavour analysis On-line aroma analysis using chemical ionisation techniques 10.4.1 Analysis via atmospheric pressure chemical ionisation 10.4.2 Analysis via PTR 10.4.3 Analysis via selected ion flow tube 10.4.4 Calibration 10.4.5 Suppression 10.4.6 Assigning ions to compounds for unequivocal identification 10.4.7 Summary 277 279 Contents 10.5 10.6 10.7 11 279 280 281 283 285 285 286 289 290 290 Sensory methods of flavour analysis Ann C Noble and Isabelle Lesschaeve 296 11.1 11.2 296 296 296 298 299 301 301 303 304 304 304 304 304 305 306 308 308 308 309 309 311 312 314 314 11.3 11.4 11.5 11.6 11.7 11.8 12 Analysis of tastants using direct mass spectrometry Applications 10.6.1 Breath-by-breath analysis 10.6.2 Flavour reformulation in reduced fat foods 10.6.3 Flavour release in viscous foods 10.6.4 Measuring aroma release in ethanolic beverages 10.6.5 Monitoring flavour generation on-line 10.6.6 Rapid headspace profiling of fruits and vegetables Future References ix Introduction Analytical tests 11.2.1 Discrimination tests 11.2.2 Intensity rating tests 11.2.3 Time–intensity rating 11.2.4 Taste–smell interactions 11.2.5 Descriptive analysis 11.2.6 Quality control tests Consumer tests 11.3.1 Purpose of consumer tests 11.3.2 Methods Sensory testing administration 11.4.1 Facilities 11.4.2 Test administration 11.4.3 Experimental design Selection and training of judges 11.5.1 Human subject consent forms and regulations 11.5.2 Judges Statistical analysis of data 11.6.1 Analytical tests 11.6.2 Consumer tests Relating sensory and instrumental flavour data Summary References Brain imaging Luca Marciani, Sally Eldeghaidy, Robin C Spiller, Penny A Gowland and Susan T Francis 319 12.1 12.2 319 320 320 321 323 Introduction Cortical pathways of taste, aroma and oral somatosensation 12.2.1 Basic brain anatomy and function 12.2.2 Central gustatory pathways 12.2.3 Central olfactory pathways x Contents 12.3 12.4 12.5 Index 12.2.4 Central oral somatosensory pathways 12.2.5 Interaction and association of stimuli Imaging of brain function 12.3.1 Methodologies to image brain function 12.3.2 Functional magnetic resonance imaging 12.3.3 fMRI design for flavour processing 12.3.4 Behavioural data and subject choice 12.3.5 Measurement limitations Brain imaging of flavour 12.4.1 Brain imaging of taste 12.4.2 Brain imaging of aroma 12.4.3 Imaging cortical associations 12.4.4 Texture and the ‘taste of fat’ 12.4.5 The issue of the ‘super-tasters’ Future trends References 325 325 327 327 328 338 341 341 343 343 343 344 345 345 345 346 351 List of Contributors Daniel Bencz´edi Firmenich SA, Corporate Research and Development, Switzerland Jack Knights Duston, Northampton, UK Ralf G Berger Institut făur Lebensmittelchemie Gottfried Wilhelm Leibniz Universităat Hannover, Germany Ulrich Krings Institut făur Lebensmittelchemie Gottfried Wilhelm Leibniz Universităat Hannover, Germany Imre Blank Nestl´e Product Technology Centre, Orbe, Switzerland Isabelle Lesschaeve Wine Aroma Wheels, Davis, CA, USA Peter S.J Cheetham Hatton Park, Warwick, Warwickshire, UK Tomas Davidek Nestl´e Product Technology Centre, Orbe, Switzerland Sally Eldeghaidy Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK Susan T Francis Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK Penny A Gowland Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK Josef Kerler Nestl´e Product Technology Centre, Orbe, Switzerland Robert S.T Linforth Samworth Flavour Laboratory, Division of Food Sciences, University of Nottingham, Loughborough, Leics, UK Luca Marciani Nottingham Digestive Diseases Centre NIHR Biomedical Research Unit, Nottingham University Hospitals, University of Nottingham, Nottingham, UK Ann C Noble Wine Aroma Wheels, Davis, CA, USA Gary Reineccius University of Minnesota, Food Science and Nutrition, St Paul, MN, USA Jacques P Roozen Institute of Food Safety, Wageningen University and Research Center Wageningen, The Netherlands Robin C Spiller Nottingham Digestive Diseases Centre Biomedical Research Unit, Nottingham University Hospitals, University of Nottingham, Nottingham, UK xii List of Contributors Andrew J Taylor Division of Food Sciences, University of Nottingham, Sutton Bonington, Loughborough, UK Saskia M van Ruth Institute of Food Safety, Wageningen University and Research Center Wageningen, The Netherlands Chris Winkel Givaudan UK Ltd, Ashford, Kent, UK John Wright Princeton, NJ USA Holger Zorn Institut făur Lebensmittelchemie und Lebensmittelbiotechnologie, Justus-Liebig-Universităat, Gieòen, Germany Preface Food Flavour Technology was originally designed as a textbook to give a broad introduction to the formulation, origins, analysis and performance of flavours Since 2002, when the book was first published, there have been developments in several areas, which necessitated a review of the book’s content Specifically, there have been developments in the science and technology available for the study of flavour, changes in European regulatory processes and changes in consumer attitudes to food flavour The original chapter headings have been retained, as all of them are still relevant, but the chapters have been revised by the authors to include new material that has appeared since 2002 Some new chapters have also been added The aim of the book is to provide coverage of flavour technology topics that are relevant to scientists who are beginning to specialise in the area Information on flavour research can be found in research papers published in scientific journals, but flavour researchers also like to present results at conferences and there is a wealth of information available in conference proceedings such as those from the Weurman, Wartburg and American Chemical Society symposium series The chapter authors have tried to incorporate all this information into the chapters, so as to give a good overview of the science available on a particular topic The creation of flavourings is the starting point for the book as this outlines the methodology and constraints faced by flavourists This is followed by a second set of constraints that are the result of the new European flavour legislation This is a very new area where there is still much discussion as to how the regulations will be, and should be, interpreted, and there are the usual inconsistencies and omissions that will be discovered and debated over the next few years The origins of flavours are described in three chapters covering thermal generation, biogeneration and natural sources The current consumer trend is to demand ‘natural’ ingredients in foods, and flavour manufacturers have adjusted their raw materials and processes to comply with this need as well as complying with the cost issues Delivery of flavours using encapsulation or through an understanding of the properties of the food matrix is described in the next two chapters, and this section is followed by chapters describing the different ways to analyse flavours using instrumental, modelling and sensory techniques Two new chapters have been added to introduce experimental techniques that are useful to the study of flavour On-line flavour monitoring has been established for over 10 years and has been used to study a range of flavour processes Measuring aroma release during eating and probing the link between the flavour profiles produced in vivo and the resulting sensory perception of the flavour has been one aspect that has received considerable attention The effect of reactants and process conditions has been studied in thermally generated flavours, and on-line analysis also provides a high-throughput technique In situations where there is a need to analyse hundreds or thousands of samples (e.g individual fruits to study the link between fruit flavour and plant genetics), on-line analysis can gather large quantities of data to understand the complexities of plant breeding The other new chapter describes xiv Preface the techniques available to image the signals in the brain during food consumption and how the data can be used to study the perceptual process Brain imaging is still relatively new in flavour studies, and the challenge is to carry out the experiments so as to obtain high-quality data and then to interpret the data to understand how the measured brain activity relates to perception While the book describes the availability of science and technology to help the flavour industry, consumer attitudes in some parts of the world are limiting the uptake of these ideas It is difficult to generalise these attitudes, but there seems to be a fear that food is no longer wholesome and that some of our current disease states (especially obesity) are the fault of the food and flavour industries In this atmosphere, innovation needs to show some direct benefit for the consumer as well as the manufacturer, but the mood may change if the predicted changes in energy availability and climate take place and food becomes limiting Against this rather negative mood, there are some interesting new aspects that may help us develop flavours in a more positive way The discovery of both taste and odour receptors in the gut, followed by evidence that the sweet taste receptor is actively involved in glucose uptake, offers new potential to link flavour, not just with food intake but also with food uptake Already there are patents covering the use of antisweet compounds such as lactisole to decrease glucose uptake in the gut, and the notion that flavours could be designed to influence nutrient intake through intake and uptake is interesting and one that merits intensive study As ever, the only certainty in flavour research is that there will be changes and that work will be needed to apply these changes so as to produce acceptable flavours The book editors and chapter authors hope that this book will assist future generations in this goal 1 Creating and formulating flavours John Wright 1.1 INTRODUCTION There are many different approaches to flavour creation and no one approach has a monopoly on the truth Any successful technique must simply recognise the fundamental structures of flavours and then proceed logically to the goal Some flavourists rely totally on blotters (strips of filter paper that are used to assess the odour of a mixture by sniffing) Some never touch them and make everything up to taste Some flavourists throw most of the ingredients in at the start and some prefer to build up the composition step by step Arguments about the logic, or lack of logic, inherent in some of these creative approaches miss the point I have known good flavourists who use techniques that seem to me to be impossibly complicated and impractical What all successful flavourists have in common is the ability to imagine the interactions between a very complex blend of raw materials and to use intuition and creative originality to fashion a work of art Many successful flavourists are trained as scientists, but some had no scientific training whatever Scientific method alone, without the spark of creativity, would mean that a single flavour would be a lifetime’s work 1.1.1 A little history The flavour industry originated in the latter half of the nineteenth century with essential oil distillation and botanical extraction as the main sources of raw materials, often with a strong link to the pharmaceutical industry Simple chemicals were available by the turn of the century, and during the first half of the twentieth century the fledgling flavour industry was increasingly driven by chemical research For the flavourist of those times (who was often a pharmacist or chemist-turned flavourist), the task of making flavours was purely creative Very little was known of nature, other than the major components of essential oils and a very limited number of chemicals that had been isolated from food and successfully identified Most new chemicals that were synthesised had no possible value in flavours The few that proved useful became the starting point for the synthesis of every possible related compound Thus, the available raw materials were concentrated in a few obvious areas Flavours created in this era were often not very close to the character of the real food, but some of them displayed real creativity and became accepted standards in their own right The advent of gas chromatography and mass spectroscopy marked a real turning point for the industry For the first time it was possible to see, in some detail, the chemicals used by nature to flavour food The advance was, understandably, treated with some caution What had been a purely creative and artistic profession could possibly be reduced to analytical Food Flavour Technology routine The early analyses quickly dispelled all concerns On reconstitution it was never possible to recognise anything more than a passing resemblance to the original target Relieved flavourists quickly settled back to the old routine, but the more astute among them recognised a few diamonds in the mud Among the first useful results from the new analytical techniques were pyrazines and unsaturated aliphatic alcohols Chemicals such as trimethylpyrazine gave a true-to-nature roasted note to nut and chocolate flavours Earlier flavours had been forced to rely on oldfashioned phenolic compounds such as dimethyl resorcinol Dimethyl resorcinol provided a hint of roasted character but, at the same time, drowned the flavour in an uncharacteristic rubbery phenolic soup cis-3-Hexenol gave an authentic green note to a multitude of fruit flavours, which previously had to depend on methylheptine carbonate to achieve a modicum of freshness (although tinged with melons and violets) Many of the failings of the early analytical techniques have now been overcome Analyses are still not easy to interpret and different techniques can give very contradictory results, but they should form the starting point for the work of a good flavourist 1.2 INTERPRETING ANALYSES For virtually all flavours the nucleus is nature We may or may not aim to reproduce nature accurately, but fully understanding nature is essential even for a caricature Analysis is therefore the first step Usually, several different types of analyses will be available (see Chapters and 10 for details of the different flavour analyses available) Headspace analyses emphasise the more volatile components and are relatively true to the character of the food being analysed The quantification of headspace analyses can usually be improved by applying vapour pressure correction factors Early headspace analyses lacked detail and failed to capture less volatile components, but these shortcomings have now been largely overcome Extract analyses are less accurate and contain more artefacts They are often representative of a rather cooked character, but they emphasise the less volatile components Stir bar sorptive extraction is a good, nonintrusive, analytical technique and offers a wide-ranging analysis of liquids Specialised analyses are often carried out to investigate the high-boiling components and also the sulfur and nitrogen compounds The flavour of food will often vary depending on the plant variety as well as the growing or cooking conditions, and many analyses will quantify these differences In consequence, the flavourist will often first have to correlate a wealth of information about the target food The correlated list can be daunting, often running into many hundreds of different chemicals The quantification used by the flavourist should be derived from the best of the headspace and stir bar results, corrected for vapour pressure, with extract results pressed into service for the less volatile chemicals – an impossibly complex problem on the face of it The ‘trick’ of being a successful flavourist hinges on the ability to imagine the smell of complex mixtures, but a mixture of several hundred ingredients is far too complex to imagine The first priority is to simplify the problem Simplification can be carried out in three stages The first stage is relatively easy Many of the chemicals that have been found will be present well below their threshold levels and it might seem safe to ignore them all Some caution is needed because synergistic and additive effects are common The best approach at this stage is to build in a comfortable margin of error and retain any questionable chemicals Creating and formulating flavours The second stage of simplification is to eliminate those chemicals that are likely to be artefacts Artefacts can be present in the original food, produced during the separation process prior to analysis or produced during the actual analysis Again, in cases of doubt, retain rather than discard The final stage of simplification is to reject those notes in the target food that are genuinely present but are not desirable Examples would be the trace by-products of fermentation and enzymatic browning in fresh fruits Even the simplified analysis will usually be of daunting complexity At this stage it is beneficial to try to reconstitute the analysis by mixing the flavour components in the proportions identified by the various analyses and then smelling and/or tasting the mixture The result is certain to be disappointing, but it will serve to clarify the key aroma characteristics of the target food Sometimes it is feasible to recreate the conditions of the original analysis using the reconstitution Reanalysis will highlight the odd errors of identification, but it will invariably give a much improved quantitative base to start from A second reconstitution may now give a recognisable product, but not one that anybody would be remotely happy to buy It is time to abandon the strictly scientific approach and move on to the more abstract creative approach 1.3 FLAVOUR CHARACTERISTICS Smelling and tasting the target food will give the flavourist a good idea of which aroma characteristics are important Reconstituting the analysis will clarify this assessment even further and may well add a few unexpected notes The aroma characteristics can be divided into two broad categories, primary and secondary characters 1.3.1 Primary characters Primary characters are essential to the recognition of the target food They constitute the basic skeleton of the flavour Good examples are ‘violet’ (␣-ionone) in raspberries and ‘clove’ (eugenol) in bananas It is impossible to create a realistic flavour without some contribution from these notes Secondary characters are not essential for recognition but contribute an optional descriptive characteristic Good examples are ‘leaf green’ (cis-3-hexenol) in strawberries and ‘dried’ (2-methylbutyric acid) in apricots In both cases it is perfectly possible to make good, authentic flavours without these notes Their effect is simply to vary the type of flavour to green strawberries and dried apricots, respectively Strictly speaking, the primary characteristics can also be regarded, in some circumstances, as having secondary characteristics as well A raspberry flavour with unnaturally emphasised ␣-ionone will smell distinctly violet This is not a problem because the object of this exercise is, once again, simplification It allows the flavourist to balance the primary characteristics in isolation and leave the secondary characteristics for later Flavours vary greatly in the complexity of their primary recognition characteristics The simplest example, at first sight, is probably vanilla On its own, the chemical vanillin smells recognisably of vanilla For many vanilla flavours in common use worldwide, this is all the primary character needed Where consumers are accustomed to a more complex flavour, such Food Flavour Technology as the character of real vanilla beans, vanillin alone will not suffice to build a recognisable skeleton Strawberry is a more complex flavour, and a more complex mixture of notes is required to achieve a recognisable flavour In this example, ‘peach’ (␥ -decalactone), ‘fruity’ (ethyl R butyrate), ‘guava’ (methyl cinnamate) and ‘candy’ (Furaneol ), blended in the correct proportions, would be the primary characters for the strawberry flavour skeleton Some of the primary characteristics will be simple and will be represented by just one chemical in the analysis Others may be more complex and may be represented by several chemicals An example is the ‘peach’ note in fruit flavours Major contributors to this note in many fruit products are ␥ -decalactone and ␥ -dodecalactone Both chemicals have a similar ‘peach’ odour, but the taste characteristics intensify (and the odour strength decreases) with increasing molecular weight When several chemicals contribute, the balance between the different components may need to be adjusted from that indicated by the analysis Fortunately, that task can often be deferred until the basic skeleton of the flavour has been devised It is easy to introduce unnecessary complication at this stage In our peach example, we will find numerous additional lactones of similar structure in an analysis and it is tempting to think of them as part of a very complex primary characteristic In reality, the additional lactones are not essential for peach recognition and are secondary notes The flavourist is now ready to begin the real creative work The objective is to achieve the best possible combination of what is now a reasonably limited number of chemicals to obtain a recognisable flavour skeleton The analysis can be taken as a starting point, but it is no more than that Even if the analysis is entirely quantitatively accurate, which is unlikely, it is still probably a long way from the optimum blend It represents, at best, a specific example of the target food rather than one with every characteristic optimised – something that never quite occurs in nature Ultimately, individual notes should be emphasised or reduced to make the flavour more attractive than the specific example of nature that has been analysed It is possible, at this point, to try to take a relatively scientific approach and blend the two most important components first The next step would be to determine the best level for the third component, and so on The problem with this approach is that the presence of the third component alters the ideal balance between the first two components The scientific approach rapidly becomes unimaginably complex and impossibly time-consuming The best approach is to plunge in, taking the analysis as a starting point, and experiment with blends to understand the role of each of the primary characteristics Speed is normally vital for commercial reasons, but it is also vital if the flavourist is to remain fresh and able to smell accurately For that reason it is best just to use blotters at this stage and to experiment with large rather than cautious changes If an addition is overdone, it can be blended back very quickly If it is underdone, it is a slow process to carry on adding small quantities and there is a very real risk that the nose will fatigue to the chemical being added 1.3.2 Secondary characteristics Once the basic skeleton has been built, the flavourist has to concentrate on the more complex secondary characteristics These can generally be worked on in groups Green notes, for example, usually contain several subcategories and many different chemicals Our strawberry flavour would almost certainly contain the common ‘leaf green’ character cis-3-hexenol, but it could also contain lesser quantities of ‘fruity green’ (cis-3-hexenyl acetate), ‘apple green’ (trans-2-hexenal), ‘melon green’ (melonal; 2,6-dimethyl-5-heptenal), ‘unripe green’ Creating and formulating flavours (hexanal) and ‘tropical green’ (cis-3-hexenyl butyrate) Once again the empirical approach is used to optimise this blend Working through all the secondary characteristics will probably take some time It is still best to use a blotter at this stage and to experiment with large rather than small changes By now, the first stirrings of pride should be evident It is time to taste the flavour Tasting solutions should be simple and appropriate If, for example, the target is a fruit and contains sugar and acid, then the taster should contain sugar and acid for the flavour to be appreciated accurately Forget the end application at this stage Two problems are apparent The first, and most obvious, is that the balance between the components will seem a little different in aqueous solution from the way it appeared on the blotter This is something flavourists learn to allow for when using blotters and is usually only a problem for trainees Blotters offer three great advantages to the flavourist They allow a very quick evaluation of each flavour They also allow the simple comparison of many variants Blotters uniquely offer a panorama of different aspects of your flavour as they air off and the more volatile components evaporate This is a big advantage because it allows you to smell ‘through’ the flavour as it evaporates The odour approximates that experienced in a simple taster for a relatively short time, usually about 5–10 minutes after dipping The second problem is that some of the real taste (as opposed to odour) characteristics may be partially or even totally missing For some flavours, such as roast beef, the taste element is obviously vital Even when it is not so obviously important, for example in bananas, it is still surprisingly vital to the realism of the flavour Correcting the taste imbalance is the next step in the flavour creation process 1.3.3 Taste effects Taste effects are normally confined to individual flavouring ingredients that are highly water soluble or have a high molecular weight Research on taste has lagged far behind that on odour, so natural extracts are still widely used to confer subtle taste effects Maltol is a good example of a water-soluble taste effect ingredient Maltol has a pleasant candyfloss odour, and a lingering sweet aftertaste, and is claimed to have flavour-enhancing properties (Labbe et al., 2007) It forms an important part of the aroma of a number of flavours, but the use of maltol as a taste ingredient dwarfs its use as an odour ingredient Ethyl maltol is stronger than maltol, has similar taste and odour characters, but is not found in nature Furaneol is even stronger than ethyl maltol and is found widely in nature The only drawback to the use of Furaneol is that it can be easily oxidised Vanillin is another water-soluble ingredient frequently used for its sweet taste effect and vanilla odour Vanillin is widely found in nature and can be integrated into many flavour types The taste effect of high-molecular-weight ingredients can be illustrated by the lactones in dairy flavours The two most important lactones in all dairy flavours are ␦-decalactone and ␦-dodecalactone ␦-Decalactone provides an excellent creamy odour in dairy flavours ␦-Dodecalactone has a similar odour but has only about 10% of the odour strength of ␦-decalactone The two ingredients have similar costs, and if odour were the only consideration, it would not make any sense to use ␦-dodecalactone The higher molecular weight of ␦-dodecalactone gives it a noticeable creamy, oily taste If cost were no object, the best combined taste and odour results would be achieved by a mixture of ten parts of ␦-dodecalactone and one part of ␦-decalactone 6 Food Flavour Technology Many high-boiling, nature-identical chemicals have been little used in flavours because of the historical emphasis on odour rather than taste They can often play a very useful role in enhancing taste characteristics even though they have little or no effect on the odour of the flavour A wide range of natural botanical extracts have useful taste characteristics Kola nut extract has a good astringent character, ginger extract has a hot character, Saint John’s bread extract has an attractive fruity sweetness and gentian extract has a lingering bitterness All of these extracts also possess noticeable odours, and care must be taken to blend in their odour when they are added to a flavour for their taste effect 1.3.4 Complexity Flavour formulations vary radically in complexity The simplest flavour can be based on just one component Many flavours, just like nature, contain hundreds of ingredients Which is best? Very simple flavours have been popular since the earliest days of the flavour industry Vanillin, isoamyl acetate and benzaldehyde have been the most popular single-component examples Very simple flavours may represent an attractive caricature, but they never taste like a real food At the other extreme, very complex flavours often lack impact and can taste flat and characterless Complex flavours can be deliberate (the result of slavishly following every detail of an analysis) or accidental (the result of lazy blending of flavours and intermediates) If a natural character is desired, then the optimum level of complexity is often the minimum number of components required to prevent the taster from perceiving the individual characters This level of complexity can vary from perhaps as few as 15 components in simple fruit flavours to up to 100 in the most complex flavour of cooked food There are, however, some important exceptions to this rule The key problem with complex flavours is that a mixture of two chemicals usually smells weaker than the sum of its parts The perceived intensity of flavour chemicals has a logarithmic rather than a linear relationship with concentration At low concentrations, near the threshold, the logarithmic relationship does not hold because the chemical is not perceived at all until it reaches the threshold level At high concentrations the relationship also does not hold because the nose fatigues to the stimulus The lower extremes of the concentration scale explain synergistic effects, which otherwise appear to contradict the rule that a mixture smells weaker than the sum of its parts (see Keller and Vosshall, 2004, for more information on measuring odour psychophysics) Traces of components that, tasted individually, would be well below their threshold level can thus have significant positive effects in mixtures At the other extreme, it is unwise to use so much of any single ingredient that the taster will quickly become fatigued A mixture of two or more chemicals with complementary odours can often give better results 1.3.5 Flavour balance Evaluating the flavour in tasters may also involve quite a number of modifications to improve the overall balance of the flavour Once you have something you are basically happy with, it is a good idea to try out variations of concentration of the flavour in the taster This is a little known, but extremely critical, way of evaluating a new flavour Creating and formulating flavours Most flavours in nature are not particularly sensitive to changes in concentration If you add twice as many apricots to a yoghurt, apart from the added acidity and sweetness, the yoghurt just tastes twice as strongly of apricots The flavour does not become unbalanced Most flavours created by humans not fare nearly so well It is possible to draw an analogy with jigsaw puzzles An ingredient that does not have a counterpart in nature, in the flavour being created, can be seen as a large misshapen piece in the jigsaw puzzle Not only is that specific piece out of place, but it also forces many of the other components out of balance It may be possible, with enough effort, to get this flavour to taste right in a specific application and at a specific dose rate The flaws will immediately become obvious if the application or the dose rate is varied because the apparent strengths of the different components will not change in unison A prime example of an ‘alien’ unbalancing ingredient is ethyl methyl phenyl glycidate (strawberry aldehyde) This chemical is seductively attractive to flavourists because it smells more like strawberries than any other ingredient they have It is very hard to turn your back on something that seems likely to give your flavour such a great start It is not found anywhere in nature and it is certainly not found in strawberries As we saw earlier, the natural character recognition skeleton of strawberry is a combination of ‘peach’, ‘fruit’, ‘guava’ and ‘candy’ primary characteristics Ethyl methyl phenyl glycidate has a very complex odour, with a little of each of these notes ‘Peach’ and ‘guava’ dominate and the chemical also has a strong ‘jammy’ character It follows that if ethyl methyl phenyl glycidate is used in a strawberry flavour, it is impossible to build up the rest of the character recognition skeleton in the correct balance and it is also impossible to avoid some degree of ‘jammy’ character This phenomenon is a powerful argument for using only those ingredients that are found in nature in the target flavour This is undoubtedly the ideal, but as long as the odour character of a potential raw material is close to that of a naturally occurring ingredient, it is often possible to use it effectively This sort of substitution would be very desirable if the naturally occurring ingredient were prohibitively expensive, impossible to make or very unstable Tasting the flavour at double the optimum dose rate will make unbalanced components horribly obvious Once those problems are corrected, the flavour should, at last, be something that is ready to show to other people As with everything else involved in flavour creation, opinions vary radically about when and how to solicit opinions from other flavourists, nonflavourists and sensory panels One thing is certain – I not know of a single instance of a really successful flavourist who works in complete isolation 1.3.6 Unfinished work An old saying cautions that you should ‘never show fools or children unfinished work’ Like many old sayings, it has an uncomfortable kernel of truth It certainly highlights a real dilemma for the aspiring flavourist Successful flavourists must be able to memorise and recognise a formidable range of raw materials They must also have the ability to imagine the effect of complex mixtures and the creative spark to use these talents to make original flavours A further essential requirement for this formidable being is an abundant helping of self-confidence By self-confidence I certainly not mean arrogance Input from others is vital and it should never be treated with contempt Self-confidence is essential to keep the flavourist sane in the face of well-meaning, but often contradictory, suggestions and criticism 8 Food Flavour Technology Help for a trainee during the early stages of the creation of a new flavour is really the preserve of a mentor who is deeply involved in the project There are always many different possible approaches to any problem, and it may not be obvious to other flavourists in which direction the trainee is trying to go Their advice in the early stages of a project is likely to be wildly contradictory Advice once the flavour has taken shape can be sought from a wide variety of sources Other flavourists can be very helpful in a number of ways They can give quick, and often accurate, assessments based solely on blotters They will often have original ideas of raw materials to try out Some will work and some will not, but the extra source of ideas is invaluable Other flavourists can often pick out mistakes that the originator has missed or, more frequently, has become too saturated with the flavour to notice It is important to recognise that the advising flavourist is often making an impromptu suggestion based on a quick evaluation However good the flavourists, their suggestions are not necessarily gold dust Sensory panels, especially expert panels, can be a valuable source of guidance on matches, hedonic ratings of new flavours and profiling Panels are especially helpful in matching work No flavourist is ever completely satisfied with a match of another flavour and a panel provides a reality check Preference mapping, linked to profiling, can provide real insight into the best way to optimise a flavour for a specific consumer group Simple sensory panels should be avoided as they all tend to lead in the direction of a bland, uninteresting flavour that offends nobody but, equally, excites nobody (see Chapter 11 for more detail on sensory testing methodologies) Other, noncreative, staff can also be a useful source of criticism It is often helpful to involve applications and sales staff It is, after all, very difficult for sales staff to sell something that has not first been sold to them Sensory panels and noncreative staff can rarely comment on blotters or simple tasters The flavour must first be applied to a realistic end product 1.4 APPLICATIONS All flavours are used in end products that impose some requirements on the finished flavour because of interactions with the finished food These interactions can be broadly grouped into four categories – ingredients, processing, storage and consumption 1.4.1 Ingredient factors The most important factor is the fat content of the finished product Flavour chemicals vary in polarity and consequently in fat solubility Taste thresholds in fat are much higher than in water The partition of different components in a flavour may vary, and this can alter the balance of the perceived aroma It is often possible to adjust the formulation, and the methods described in Chapter 10 have been used to measure aroma release and then rebalance flavours in foods with different fat contents (Shojaei et al., 2006) However, an alternative approach is to avoid drastic differences in the polarities of the flavour components In all foods containing fat, added flavour will slowly partition between the fat and the aqueous phases on storage This effect can be partly avoided by adding separate flavours to the fat and the aqueous Creating and formulating flavours phases, but this is a laborious approach and will rarely be sufficiently accurate to avoid subsequent partition effects Care must be taken in application trials to store the finished food sufficiently long before tasting to allow the partition of the flavour to be substantially completed The lipophilic gum base in chewing gum has an effect similar to that of fat, but the problems are aggravated because the flavour is gradually extracted by chewing If there are differences in the polarities of the flavour components, the chewing gum will appear to taste mainly of the most polar components at the start of chewing Eventually only the nonpolar chemicals will be extracted A completely fat-soluble flavour may be necessary for some applications At the other extreme, entirely water-soluble flavours are essential for clear soft drinks In both cases it is difficult to produce a balanced profile within a restricted range of polarity In these examples it is sometimes helpful to depart from the essentially naturebased approach we have used so far All flavourists should keep a reference record of the characteristics of all the raw materials they have encountered This database can usually be used to find a chemical with a similar odour character to a problem raw material but with different physical or chemical properties Natural extracts and oils often contain chemicals with widely differing polarities They can be processed by distillation, solvent extraction and chromatography to reduce these differences The most common example of this type of process is the deterpenisation of lemon oil Lemon oil contains about 90% of terpene hydrocarbons, which are nonpolar, low boiling and susceptible to oxidation, and contribute little to the overall flavour character The oil also contains about 6% of oxygenated chemicals, which are polar, relatively high boiling and less susceptible to oxidation, and provide most of the flavour character The level of lemon oil that would be required to impart an acceptable flavour level to lemonade would result in a level of terpene hydrocarbons in the drink well in excess of their limit of solubility The oxygenated chemicals would be readily soluble at this level, so a clear drink could be obtained by removing the hydrocarbons from the oil Chromatography and solvent extraction are obvious possibilities Distillation also works because of the difference between the boiling points of the terpene hydrocarbons and most of the oxygenated chemicals Some loss of the true lemon character is inevitable owing to processing and the small, but significant, flavour contribution made by the hydrocarbons This is more than justified by the gain in stability to oxidation Solvent extraction generally gives better results than distillation because this method retains the most volatile aliphatic chemicals, which are responsible for the fresh, juicy character of many citrus oils, especially orange oil Solvent extraction is discussed in more detail later in this chapter Distillation, if it is used to produce a terpeneless oil, unfortunately also removes the high-boiling antioxidants that are present in cold-pressed citrus oils Major components of a flavour may themselves cause problems in a finished food These problems are often changes in texture or in the stability of emulsions The solvents are the most likely culprits, and in many instances a change of solvent will provide a cure Where flavour dose rates are very high, particularly in chewing gum, individual flavour chemicals may also be responsible When this happens, the flavour can often be modified, but sometimes the only possible solution is to modify the formulation of the application This may also be an issue if the flavour necessarily contains large quantities of a food additive, for example, an acid This could happen, for instance, in a natural flavour containing significant quantities of concentrated fruit juices 10 Food Flavour Technology Carbohydrates in a finished food may have a binding effect under certain conditions, but this is not a frequent problem The most obvious example is the loss of flavour in bread on storage because of flavour binding to the helices in starch molecules Flavour binding by proteins is a more serious problem Most protein molecules are folded in such a way that the nonpolar amino acid side chains are on the inside and the polar groups are on the outside Flavour chemicals can interact with the nonpolar interior regions of the protein and cause it to unfold They can be bound into the protein by absorption at the protein surface or inclusion in the nonpolar interior Protein flavour binding is most evident in processes involving heat and is most pronounced with carbonyl flavour chemicals Chemical interactions and partitioning effects make it essential to wait at least 24 hours before tasting some applications 1.4.2 Processing factors Minor effects from processing include those from filtration, aeration and freezing, but by far the most important factor is heat This may cause chemical changes in the flavour, but the main problem is the loss of volatiles This may have the effect of reducing the fresh top note of a flavour If the key recognition chemicals have widely different boiling points, heat could render the flavour unrecognisable The choice of solvent can reduce this problem In some instances, volatile chemicals may be replaced by higher boiling analogues It is usually possible to change the balance of the flavour to allow for differential losses, but this solution gives a flavour that is suitable for only a limited range of applications In processes involving considerable heat, such as bakery and extrusion, the best solution is multiple encapsulation In this process a spray-dried flavour is coated with a high-melting-point fat This process protects the flavour until the fat melts Chapter introduces the key concepts required to encapsulate flavours effectively 1.4.3 Storage factors Some wines and cheeses improve with age, but they are the exceptions rather than the rule Flavour stability in an application should ideally at least match the shelf-life of the food itself When a flavour is added to food, some chemical changes, such as the hydrolysis of acetals, occur quite quickly There are frequently subtle differences in flavour character after only one day In the longer term, oxidation is responsible for most of the changes in flavour during storage When most flavour chemicals oxidise, the effect is simply perceived as loss of flavour because the flavour chemical has a much stronger odour than its oxidation products When some incidental components, such as the hydrocarbons in a lemon flavour, oxidise, the effect is often perceived as an off-note Substitution of flavour components with more stable alternatives and the use of antioxidants usually reduce the problem to manageable proportions Migration of flavour chemicals into or through food packaging materials can sometimes occur It may lead to a detectable loss of flavour or cross-contamination problems A change in packaging material is the best cure, but, if this is not possible, it may be practical to reformulate the flavour without the problem of chemicals This may change the profile of the flavour Creating and formulating flavours 11 Tea bags present a very specific packaging problem because of the size of the holes in the tea bags Liquid flavour can be spread directly onto tea leaves, but this leaves the flavour very prone to oxidation and evaporative losses Spray-dried flavours need to be agglomerated to prevent them from falling through the holes in the tea bags Tea dust can be included in the agglomerated flavour to give it a similar appearance to the tea in the bags Tea used in tea bags is generally of small particle size, and care needs to be taken to ensure that the size of the agglomerated flavour particles matches that of the tea 1.4.4 Consumption factors Many of the processing problems can resurface when the food is consumed This is particularly common when a powder flavour is used in a dry convenience food The final factor that may influence the formulation of the flavour is the temperature at which the finished food is consumed At temperatures below room temperature, as in the case of ice cream, the intensity of the whole flavour is reduced The intensity of the most volatile chemicals is reduced relatively more than the rest of the flavour, and they may need to be increased Caution should be exercised because the food can warm up in the mouth Foods that are consumed hot are more difficult to flavour The high temperature increases the intensity of the flavour, particularly the more volatile chemicals At the same time it may cause a relatively greater loss of the same components 1.5 FLAVOUR FORMS Liquid and powder flavours can be split into a number of major types, each of which poses some specific problems for the flavourist 1.5.1 Water-soluble liquid flavours These are by far the most common types of flavours The flavour chemicals and natural components are dissolved in a simple solvent, most commonly propylene glycol, triacetin or ethanol, with the possible addition of water If the flavour contains significant amounts of solids, such as vanillin or maltol, then the quantities added must remain well within their limit of solubility Storage conditions can be much harsher in real life than in a laboratory, and a large safety margin should be built in The same consideration should be applied to the nonsolid components of the flavour, but problems are not as common in this area Propylene glycol is usually the solvent of choice It is stable, virtually characterless in use, and confers some stability in applications involving heat processes The drawbacks of propylene glycol are that it is not a strong solvent, that it is not natural, that the levels of use are restricted in some countries, and that it forms acetals and ketals quite readily with carbonyl flavour chemicals Acetal and ketal formation can be inhibited by the addition of water to the flavour, but this makes an already weak solvent even weaker Acetals and ketals can actually be useful in some applications because they may protect the parent carbonyl from oxidation during storage of the flavour They will later break down to release the parent carbonyl in many applications in the presence of water The most serious problem deriving from acetal and ketal formation is that many acetals and ketals are only poorly soluble in 12 Food Flavour Technology propylene glycol This can result in the puzzling phenomenon of an initially clear flavour, gradually phase separating and forming two layers Ethanol is also widely used, especially where there are no duty handicaps (many countries impose high taxes on ethanol) It is relatively stable, has a mild but pleasant character in use, is readily available in a natural form, and can be diluted significantly with water The drawbacks are relative instability in applications involving heat processes, religious restrictions in some countries, flammability and the formation of acetals and ketals The last factor is less important because ethanol is a strong solvent and can still remain effective if sufficient water has been added to inhibit acetal and ketal formation Triacetin is not a solvent of choice for most applications It is not very water soluble, and when it does dissolve, it decomposes to glycerol and acetic acid Triacetin has a slight bitter taste and acetic acid has a noticeable ‘vinegar’ odour Triacetin can be the solvent of choice when water solubility is not critical (as in many confectionery applications), when propylene glycol is restricted, when the components of the flavour will not dissolve readily in ethanol or propylene glycol and, most importantly, when propylene glycol has an undesirable effect on the texture of the finished food Chewing gum is the most important example Propylene glycol hardens chewing gum, but triacetin acts as a plasticiser Water is not added to triacetin-based flavours because it would hydrolyse the triacetin Other solvents may be useful in specific cases Triethyl citrate is similar in many respects to triacetin It is poorly water soluble, but is odourless and confers heat stability The major difficulty with triethyl citrate is the bitter aftertaste, which severely restricts the level of use Diacetin is also similar to triacetin but is generally less effective Glycerol is a very weak solvent but can be used effectively in conjunction with ethanol in natural extracts to confer some heat stability Lactic acid is not generally a very effective solvent but can be useful, in mixtures, for some problematic raw materials, especially maltol Benzyl alcohol has a faint floral character and is a good solvent but is prone to oxidation to benzaldehyde Benzyl benzoate is stable and relatively odourless It can be used in solvent mixtures, especially for oil-soluble flavours, but has an unpleasant flavour at high levels Many of these lesser solvents are not universally recognised as solvents They may be permitted as flavouring ingredients, but care must be taken of the level of use 1.5.2 Clear water-soluble liquid flavours This category is very similar except for the requirement that the end product, usually a beverage, should be crystal clear Most flavour raw materials are entirely water soluble at their normal level of use The exceptions are limited to a few chemicals that need to be used at relatively high levels (usually esters), chemicals that can form insoluble polymers on storage and terpene hydrocarbons Terpene hydrocarbons are found in many natural essential oils They have limited use as flavouring ingredients (there are exceptions, such as myrcene), and they are prone to oxidation The hydrocarbons can be removed from essential oils by distillation, chromatography or solvent extraction The most effective method is solvent extraction (often called ‘washing’) because it causes least change in the character of the original oil The most effective solvent is a mixture of ethanol and water, but propylene glycol can also be used The extraction is carried out by dissolving the oil (for example, orange oil) in ethanol, adding water to throw out the hydrocarbons (commonly called ‘terpenes’), chilling the mixture and allowing it to stand for days The terpenes float to the top of the mixture, which can then be drawn off Creating and formulating flavours 13 and filtered A little extra alcohol is added as the final stage to prevent the flavour becoming cloudy if it is stored in the cold The process can be speeded up by the use of a coalescer, a metal mesh that coalesces the oil droplets Propylene glycol ‘washings’ are difficult to make because of the viscosity of the solvent and the small amount of water that can be added The use of a coalescer is virtually essential to make propylene glycol-based ‘washings’ A surprising, but effective, alternative way to produce clear beverages is through the use of low-payload, small particle size emulsions The flavour must contain only very limited amounts of terpene hydrocarbons for the process to work This method is widely used for cola flavours 1.5.3 Oil-soluble liquid flavours Oil-soluble flavours are needed where the end product is an oil or a fat They are also used where the end product cannot tolerate water Both ethanol and propylene glycol contain small amounts of water, so these solvents cannot be used in water-sensitive products such as chocolate Natural or synthetic (medium-chain triglyceride) vegetable oils can be used as solvents The problems are susceptibility to oxidation (for the natural oils) and poor solvent power Many of the chemicals that are important for taste effects are highly polar and poorly soluble in oils Some of the minor solvents discussed earlier, such as benzyl benzoate and triethyl citrate, can be particularly effective in oil-soluble flavours They all have some drawbacks and may be more effective when used as mixtures If this does not work, one possible solution is to dispense with traditional solvents altogether and use the major components of the flavour to dissolve the solids This is not always possible without adding excess quantities of weaktasting esters such as ethyl acetate When it can be done, the resulting flavour may be highly concentrated and very difficult to dose accurately in an industrial environment Essential oils can be effective ‘solvents’ for some oil-based flavours This is especially true of citrus flavours The natural oil gives a realistic, complex background and added flavour ingredients give powerful specific character 1.5.4 Emulsion-based flavours Emulsions, based for example on orange oil, are often used to give cloud to a beverage, but they can also be a cheap and effective way of delivering a flavour where cloud is not an issue The water-soluble components, such as vanillin, can be dissolved in the gum solution (typically gum arabic or modified starch is used as emulsifiers), and the remaining components can be mixed together to form an oil phase, which is then emulsified Potential problems include the clumping or separation of the oil phase, the hydrolysis of susceptible flavour ingredients and the microbiological stability of the emulsion over an extended period, especially once the container has been opened Ideally, for these applications, the oil phase should constitute around 5% and certainly not more than 10% of the emulsion 1.5.5 Dispersed flavours Dispersions are a similar, cheap and cheerful, way of delivering flavours in powder form If all the ingredients are solids, they may be mixed together and diluted with a carrier such 14 Food Flavour Technology as lactose If some of the ingredients are liquids, they are mixed together and spread on the carrier before the solids are mixed in This method works if all the ingredients are relatively high boiling and not susceptible to oxidation Even so, it produces flavours with a relatively short shelf-life and it is difficult to mix the flavours so that they are entirely homogeneous 1.5.6 Spray-dried flavours Spray-drying is the method of choice for powder flavours (see Chapter for the mechanisms of flavour encapsulation) The flavour is typically emulsified in an aqueous gum solution, and then dried by spraying into a hot chamber This method is preferable to dry mixing because the resulting flavour is stronger and much more stable to evaporation and oxidation Spray-drying works so well because the sprayed droplets form a semipermeable shell very quickly, long before most of the water has evaporated The semipermeable shell allows water to pass, continuing the drying process, but is relatively impermeable to most flavour components This is true of even the smallest flavour molecules, such as ethyl acetate and acetaldehyde Only a small proportion of the ethyl acetate or acetaldehyde added to a flavour survives spray-drying, but, without the effect of the semipermeable shell, logic would dictate that the loss on drying would be virtually 100% Flavourists are not expected to be experts in the area of spray-drying and the many variants of this technique, but they should know enough to get the best out of the process The first issue is the way the flavour is added to the emulsion The criteria are much the same as those for liquid emulsions, except that there is no need for the emulsion to be stable in the long term The ingredients of the flavour should be split into water-soluble and oil-soluble keys The water-soluble components should be dissolved in the gum solution The oil-soluble components should be emulsified in the resulting mixture This emulsion does not need to be stable for longer than it takes to dry the batch, but it should be emulsified to a reasonably small and uniform particle size Poor emulsification will result in more surface oil, flavour loss and susceptibility to oxidation Some solvents should not be used in the formulation of the keys Ethanol will increase the flavour loss, and propylene glycol (in more than trace amounts) will make the powdered flavour hygroscopic Triacetin works well in most instances The maximum loading of the oil phase is around 30% of the dry weight, but drying losses increase steeply after 20%, as does the amount of surface oil For cost-effectiveness, 20% is a good maximum to aim for The second issue is the composition of the gum solution Gum acacia is the most widely used material, although some modified starches can give equally good results Gum acacia varies widely in quality and care should be taken to buy 100% pure gum from a reputable source The cheap gum that has been cut back with filler is always of poor value The flavourist should be free to control the proportion of pure gum used It is a waste of money to use 100% gum acacia as the carrier In spray-dried flavours, 30% is the absolute maximum quantity of gum needed for even the most challenging applications In many cases as little as 10% is all that is needed to form the semipermeable film during drying The filler, usually maltodextrin, is important because a high dextrose equivalence is needed to make the shell of the spray-dried particle less permeable to oxygen One unintentional advantage of using reduced levels of gum acacia is that the viscosity of the emulsion is lower This allows the solid content of the emulsion to be increased, while still keeping to a viscosity level that can be handled readily Higher solids mean more throughput and less energy costs Creating and formulating flavours 15 The third issue is the processing conditions They should ideally be set for each flavour The emulsion should not be warm because this will damage the flavour The inlet temperature must be adjusted so that particles hitting the sides of the drier not stick In general, the best results are obtained with the highest inlet temperatures and the highest throughput The outlet temperature should be reduced as much as possible, but not so much that the spray-dried particles contain significant moisture when they leave the drier 1.6 PRODUCTION ISSUES One of the most difficult challenges facing flavour companies is the link (or lack of it) between the creative flavourists and the production staff It is possible to sidestep the issue by introducing a complete department to sort out problems, but it is obviously much better not to have them in the first place More thought should be given during the training of flavourists to the possibility of a spell in production QC training is often included, but nothing beats getting your hands dirty and learning about the practical issues first hand The first issue is the total number of raw materials available for use A sensible number can be reached by adding to the number of GRAS (generally recognised as safe; see Chapter for information on the legal status of flavours) and European raw materials (around 3500), the number of sensible variations of natural products (around 500) and the legal variations (natural, organic, kosher, etc.) (around 1500) A sensible maximum is 6000 Not many companies can boast such a small list, but the cost, quality and service problems associated with large raw material lists are formidable The total number of flavour formulations is also often quoted by production as a key problem metric, but it is only a problem if the operations function is so inefficient that it is necessary to keep stock of finished flavours Very few flavours are now sold off-the-shelf and the tailor-made flavour is becoming the standard It makes much more practical sense, and it is much more cost-effective, to concentrate on controlling the number of raw materials The second real issue is the number of raw materials in any individual formulation It is simply not possible to justify more than 100 ingredients in a flavour Depending on the type of flavour, the optimum number of ingredients can vary between 15 and 100 but in most cases the best effect is obtained using between 15 and 50 ingredients The cost of compounding and the service problems associated with very complex flavours are both serious issues Accurate compounding in a production environment is very different from the situation in a laboratory The use of solutions should be tailored to production needs and kept to a minimum necessary for accurate weighing Old solutions should be discarded The use of a single key ingredient may be helpful in some instances to separate out all the very lowvolume items Outside this restricted context, the use of keys and the blending of flavours in general are real headaches for production They are also, frankly, indicative of lazy work on the part of the flavourist concerned The correct compounding order may be obvious to the flavourist, but it must be specified in a formulation to assist production Other important notes are the need to filter (which can often be avoided by better selection of raw materials or more careful formulation) and full details of any processes The originating flavourist should always be involved in the quality control testing of the first batch in case there are problems scaling up the flavour 16 Food Flavour Technology 1.7 REGULATORY AFFAIRS Flavourists should receive extensive training on regulatory issues, not simply the widely varied global flavour regulations but also the implications for finished foods and labelling (see Chapter for further information on the safety and legislation of flavours) With the current time scales for projects, it is not practical to expect that a final regulatory check should be anything more than a safety net A generally conservative approach should be taken, and wherever possible GRAS ingredients should be used The regulations in Europe and the US are increasingly well harmonised, so this restriction is usually practical For any country, the IOFI (International Organisation of the Flavour Industry) guidelines represent the minimum standard, irrespective of the lack of local regulations Natural certification of raw materials should not be accepted without critical evaluation Natural standards vary by country and common sense should be applied 1.8 A TYPICAL FLAVOUR Raspberry flavour is a good learning tool It is relatively simple, but not so simple that it does not contain a multitude of useful lessons To illustrate the process of flavour creation, we will work on an imaginary, but typical, customer project The task at hand is to create a nature-identical flavour, with a profile, that the customer has described as true to nature, fresh and red The end use is hard candy Let us imagine that the flavourist has two analyses to work from One derived from the analysis of an extract from the fruit and the other derived from the analysis of the headspace over the fruit In these two, hypothetical analyses, 362 different chemicals have been identified, 271 in the extract analysis and 203 in the headspace analysis In both cases the quantification is expressed as a percentage of the total volatiles The headspace analysis is also quantified with an added vapour pressure correction Table 1.1 gives the quantification of those chemicals from the analyses that we will consider using to create a simple flavour The first step is to decide which chemicals in either analysis represent primary characteristics In the case of raspberries, the violet note is clearly essential The analyses contain both ␣-ionone and ␤-ionone ␣-Ionone has a clean ‘violet’ note and ␤-ionone has a ‘violet’ note in addition to a strong ‘cedar’ note To keep things simple there is an obvious temptation Table 1.1 Flavour components identified in analyses of raspberry flavour Cost (in order of appearance) Extract (%) Headspace (%) Vapour pressure adjusted (%) ␣-Ionone ␤-Ionone 4-Hydroxyphenylbutan-2-one Damascenone Dimethyl sulfide Acetyl methyl carbinol Ethyl acetate cis-3-Hexenol cis-3-Hexenyl acetate ␦-Decalactone 4.00 1.80 0.50 0.05 0.02 0.50 5.00 8.00 0.02 0.60 0.70 0.50 — 0.02 1.50 0.20 9.80 0.60 0.04 — 8.000 9.500 — 0.150 0.001 0.002 0.040 0.030 0.010 — Creating and formulating flavours 17 to ignore the complications of ␤-ionone and work with ␣-ionone alone This simplification might work, but it is probably unwise The ‘cedar’ note of ␤-ionone generates a ‘pippy’ or ‘seedy’ effect in raspberry flavours This note is hardly a primary characteristic, but it is normally attractive If ␤-ionone is ignored at this stage, then a later correction to add a ‘seedy’ note will necessitate a rebalancing of the ‘violet’ character The next step is to establish an estimate of the correct concentration of the 70/30 mixture (these proportions are derived from the extract analysis) of ␣-ionone and ␤-ionone using a simple taster A good starting level would be 0.25 ppm (part per million or milligram/ kilogram) The most common dilution of flavours in beverages is 0.05% rtd (ready to drink) This is equivalent to 0.035% ␣-ionone and 0.015% of ␤-ionone in the flavour The other primary characteristic is not quite so easy to identify The flavour of ␣-ionone alone is simply ‘floral, violet’ The missing character should confer a specifically ‘berry’ note The only feasible candidate in the analyses is 4-hydroxyphenylbutan-2-one This chemical has a distinct ‘berry’ aroma, even a specific hint of raspberries Again trial and error can be used to establish a good balance between these two chemicals A good starting level would be around 2% in the flavour, but later in the process this will prove to be too high (once other ingredients with somewhat similar characteristics have been added) and the final level is 1% At this stage we already have a recognisable raspberry skeleton We can move on to the optional, secondary components The customer wants a true-to-nature character, but also describes the target as ‘red’ and ‘fresh’ Neither of these descriptors is very specific, so the flavourist has to try to guess the customer’s wishes This dilemma is very common and illustrates the need to work with customers to establish specific descriptors ‘Red’ can reasonably be taken to mean red raspberries rather than black (so no musk character), ripe rather than unripe (so ripe, ‘fruit’ notes and restricted ‘green’ notes) ‘Fresh’ can be taken to mean an absence of ‘jammy’ or ‘cooked’ notes It might also indicate high ‘green’ notes, which certainly confer freshness A more moderate level of ‘green’ notes is probably a good idea because ‘red’ was also specified High levels of ‘green’ notes, especially ‘raw, green’ notes, give an unripe effect A good choice for the red ‘fruit’ note is the ‘damson’ character of damascenone This chemical is found widely in nature and is, justifiably, a favourite with flavourists The ‘damson’ character of damascenone adds richness and a deep ‘fruity’ character to our fledgling raspberry flavour Taking the extract analysis as a guide (damascenone is fairly high boiling), the levels of the ionones used in the flavour indicate a level of 0.0004% of damascenone in the flavour This seems very low indeed To obtain the correct character we must increase the level in the flavour to 0.04% Dimethyl sulfide is also an excellent ripe ‘fruit’ note in dilution, although at high levels it has a ‘cabbage’ character and can make the flavour seem cooked and ‘jammy’ The addition of dimethyl sulfide also improves our flavour dramatically, but 0.01% is the most we can add before the character becomes slightly ‘jammy’ This is, however, far more than the level indicated by the vapour pressure-corrected headspace analysis The ‘buttery’ note of acetyl methyl carbinol will also, surprisingly, add to the ‘ripe’, ‘red’ character The concentration that would be required in the flavour, on the basis of the amount found in the extract analysis, relative to ␣-ionone, is around 0.005% This level works well and provides the required note The final ‘fruity’, ‘red’ note is ethyl acetate The headspace analysis indicates a very low level, but the extract analysis (which would be expected to give a low result) indicates a level broadly similar to that of ␣-ionone Increasing that level a little to 0.10% in the flavour gives an attractive result 18 Food Flavour Technology The most obvious green note is cis-3-hexenol, but this chemical has a ‘leaf, green’ character, similar to fresh-cut grass Like damascenone, cis-3-hexenol is found very widely in nature and is often the first choice when a ‘fresh’ character is desired We could add a low level of cis-3-hexenol, but we would run the risk of introducing an ‘unripe’ note A much better choice for this flavour would be cis-3-hexenyl acetate, which has a softer ‘fruity, green’ character and very little unripe note The analyses would indicate a low level of cis-3hexenyl acetate, but a higher level is necessary because we are not adding any cis-3-hexenol In practice, the ideal level is 0.02% in the flavour This flavour will smell reasonable but taste very thin Only two components of the flavour so far have a significant taste effect – damascenone and 4-hydroxyphenylbutan-2-one The flavour has a degree of ‘berry’ depth of taste, but needs added ‘sweet’ character The addition of 2% of maltol (not found in the analysis) will help to solve this problem, but is obviously not the ideal solution Maltol has a ‘candyfloss’ aroma and imparts a lingering sweet aftertaste It adds depth, but is too a simple character One other addition that will help to add depth is ␦-decalactone This chemical has a ‘creamy’ character and trial, and error establishes an ideal concentration in the flavour of 0.05% This level is higher than that indicated by the extract analysis The flavour we have developed thus far is much too simple and will taste like an obvious mixture of separate notes The addition of a small amount of jasmine absolute (0.02%) will add complexity and traces of desirable ‘berry’ (from benzyl acetate), ‘lavender’ (from linalool), ‘animalic’ (from indole) and ‘jasmine’ (from methyl jasmonate) notes The final stage in the creation of our very simple raspberry flavour is to make allowances for the processes involved in making hard candy The only factor is heat, so the most volatile components must be increased to allow for the losses in processing Ethyl acetate should be increased to 0.20%, dimethyl sulfide to 0.02%, acetyl methyl carbinol to 0.009% and cis-3-hexenyl acetate to 0.03% In real life, the process of developing this flavour would be much more complicated, but this simple example serves to illustrate the principles involved The composition of the final flavour is compared to the two analyses in Table 1.2 The analyses help, but they are a long way from the quantification of the finished flavour This example illustrates the high level of creative input, even when analyses are taken as the starting point Table 1.2 Comparison of raspberry flavour analysis (from Table 1.1) with formulation of a raspberry flavour suitable for hard candies Cost (in order of appearance) Extract (%) VP adjusted (%) Flavour (%) ␣-Ionone ␤-Ionone 4-Hydroxyphenylbutan-2-one Damascenone Dimethyl sulfide Acetyl methyl carbinol Ethyl acetate cis-3-Hexenol cis-3-Hexenyl acetate Maltol ␦-Decalactone Jasmine absolute 4.00 1.80 0.50 0.05 0.02 0.50 5.00 8.00 0.02 — 0.60 — 8.000 9.500 — 0.150 0.001 0.002 0.040 0.030 0.010 — — — 0.035 0.015 1.000 0.040 0.020 0.009 0.200 — 0.030 2.000 0.050 0.020 Creating and formulating flavours 19 1.9 COMMERCIAL CONSIDERATIONS 1.9.1 International tastes We are all accustomed to rapidly increasing globalisation, and with it the assumption that one product can be sold in all markets There are cases where, with sufficient advertising, this is manifestly true In most cases, however, the assumption does not hold Regional tastes for most of the key flavour types still override global stereotypes The regional tastes are often derived from historical familiarity and may fade in time, but, for now, they are very important The main regional preferences for the most important flavour categories are summarised as follows: Beef : Roast beef is the preferred profile in the UK Grilled beef reigns supreme in the US, and in much of Asia boiled beef is the main profile Cheese: Cheddar is far and away the most important type of cheese in terms of flavour sales Blue cheese and Parmesan are very small categories compared with Cheddar Cheddar can be broken down into two main types by region: sharp and mild Sharp Cheddar is best defined (ironically) by aged US or Canadian Cheddar cheese and represents the target profile in Europe The taste preferences of US consumers are very different In this region a mild, creamy, buttery character is preferred Cherry: In Europe the hawthorn note of Morello cherries is preferred, but in the US benzaldehyde is the prominent character Chocolate: Milk chocolate predominates in most markets outside Europe and the milk component often has a cooked character Some popular milk chocolates also have an added signature note such as cinnamon or almond Dark chocolate is popular in Europe and can have pronounced burnt and bitter characteristics in this market Lemon: In the UK especially, but also in much of Continental Europe, a high citral level is liked The European taste also likes an exaggerated level of jasmine character In the US, lemon is milder and more floral, and in much of the rest of the world citral is the defining note, sadly, sometimes accompanied by the waxy character of oxidised oil Mango: As with most other tropical fruits the situation is the exact reverse of berry flavours The genuine character, with its strong terpene, skin note is preferred in Asia and Latin America In Europe and the US a pale imitation flavour is preferred, with much reduced skin and sulfur notes and an emphasis on melon and peach Milk: Fresh milk and dairy flavours are optimum in the US and Europe In Asia a boiled, condensed milk note is preferred, and in Latin America the even more caramelised ‘dulche de leche’ is ideal Orange: In Europe there is the strange contradiction that the flavour of processed juice is liked, presumably for nostalgic reasons, together with the pungent, fresh note of acetaldehyde An exaggerated hint of violet is also liked in many orange flavours for confectionery Fresh juice character is popular in the US and, for the rest of the world, cold-pressed orange peel oil is the most popular character Raspberry: In the US a strong violet character is preferred, but in Europe this note is muted and balanced by fresh and green characters In Asia real raspberries are a rarity and an old-fashioned candy character is preferred Strawberry: At first sight, strawberry would seem to be easily standardised Not so It is one of the most difficult flavours to fit into its many regional variations In the US, strawberry is generally sweet and slightly jammy Green notes are not liked In most of Europe the preferred character is fresh and distinctly green Within Europe, the French taste is 20 Food Flavour Technology for a pronounced jasmine note and the Spanish taste is for strawberry jam In Asia the preferences are more abstract and old-fashioned because of relative unfamiliarity with the real fruit Vanilla: Alcoholic genuine vanilla extract, with rum and fruity overtones, defines the US taste In France the taste veers towards a creamy hawthorn note, similar to the rare Tahitian natural vanilla extract In Germany a hint of balsam is appreciated, and in much of northern Europe, a simple vanillin taste is preferred The UK preference is for a distinctly buttery note 1.9.2 Abstract flavours Most flavours have an obvious natural target The flavour may be realistic or have a degree of abstraction, especially for children’s products Many flavours, such as lemon-lime, are blends of recognisable natural targets Very few flavours are basically abstract The most important examples are cola and tutti-frutti Cola flavours are all quite complex blends The main characters are distilled lime oil, cassia oil, nutmeg oil and vanilla extract In some instances, the caramel colour also contributes a characteristic flavour The flavour ages very quickly in the bottling syrup because of the high level of phosphoric acid Matching cola flavours is especially difficult because of the noticeable change in flavour in the syrup and the need to age the syrup before evaluation Tutti-frutti flavours, in contrast, are quite widely varied The main characters are banana, orange, pineapple, vanilla and berries They can be classified into two broad types – those based on banana and those based on berries The banana family is usually built around isoamyl acetate and the berry family is usually built around ␣-ionone There are other interesting, but less commercially important, abstract flavours Root beer was originally based on sassafras oil, vanilla and methyl salicylate, as well as a host of minor ingredients Sassafras oil has not been used for many years (because it contains safrole over which there are safety concerns) and the substitutes vary in effectiveness Sarsaparilla and dandelion and burdock are similar products Cream soda flavours were similarly modified in line with regulatory requirements and now consist of vanillin with lactone-based hay and cream notes Cachou flavours are intensely perfumed and are often based on combinations of violet, rose and musk Another, accidental, category of abstract flavours is the group of the best of the flavours from the early years of the industry They were often not exactly recognisable, but were triumphs of artistry over paucity of raw materials and became standards in their own right Blackcurrant and cherry flavours are good examples Early blackcurrant flavours were based on one simple raw material, buchu leaf oil This raw material reproduced nothing more than the ‘catty’ character of blackcurrants and it was unpleasantly harsh and minty Creativity improved this simple base by adding a mixture of ingredients, most importantly vanillin and ␣-ionone The harshness was covered and the pleasant abstract confection is still the basis of many blackcurrant-flavoured foods today The character is not close to blackcurrants, but it is an instantly recognisable standard Early cherry flavours relied on benzaldehyde in a similar way and were not particularly attractive Creativity gradually improved on this ingredient and evolved a complex flavour with anisaldehyde and para-methoxy acetophenone as the main additions This flavour type is exceptionally attractive It is not close to any known variety of cherry, but it is still immensely popular Creating and formulating flavours 1.9.3 21 Matching No project is less welcome to the typical flavourist than one that requires matching The very idea of matching someone else’s work is profoundly unattractive Improving on someone else’s work is quite another matter and represents a real challenge, but simple matching is boring Behind the flavourist’s manifest hostility is not just the simple lack of challenge and novelty; there is also the commercial fact that matching work represents by far the least rewarding use of creative time Most matching projects are obtained by novice sales staff as a way of gaining entry to the account Most successful matches are simply used to pressure the existing supplier to reduce prices Very few product managers will risk changing the flavour of a successful consumer product to save a few cents, especially when it must involve expensive consumer trials Some very successful flavour companies will generally not accept matching projects, and it is not evident that their customer standing has suffered as a result Despite all the objections, there are occasionally good reasons to carry out matching work The customers may have become genuinely hostile to their current supplier and wish to change at any price The need to carry out matching may also derive from a reduction in the number of suppliers in a core supplier programme The customer may wish to duplicate a competitor’s existing consumer product, although this type of project is more often directed towards beating the existing product Matching generally starts with an analysis Normally, it will be a direct analysis of the existing flavour, if the customer is serious In some cases it will be a consumer product Once the analysis has been completed, it should be reconstituted and reanalysed (in consumer product if necessary) to pick up errors in identification and quantification This corrected analysis represents the starting point for the flavourist The target flavour will often contain natural extracts and essential oils, so the first job is to allocate all the components derived from natural sources correctly This is usually a matter of experience and the analyst should also be able to provide some guidance The trickiest part is usually trying to determine which processes (solvent extraction, concentration, etc.) have been applied to the natural raw materials Trial and error, as well as quite a few reanalyses, overcome this hurdle The major chemical components should be identified and quantified from the analysis It is often helpful to carry out liquid chromatography of the main chemicals to improve the accuracy of the quantification One issue that is often forgotten is the need to get the solvents in liquid flavours identified and quantified correctly Flavourists sometimes assume that they will not be important in the end product That is not always true, but incorrect solvent balance always hinders rapid evaluation of matches on blotters All that remains are the trace components Unfortunately, this is often where most of the problems lie Traces of sulfur chemicals, for example, can be very difficult to pick up on analysis, but can be a vital part of the flavour Matches are often carried out under severe time pressure and it is easy to become stale and run out of ideas Involving other flavourists is a must and can be especially helpful in generating ideas about the identity of missing trace components It is also useful to have available a bank of analyses of flavours from the same competitor In many companies the same ideas are trotted out with surprising regularity Sensory panel work is essential to validate the accuracy of the final match It may also help persuade the customer to accept the change Caution should be exercised about blindly 22 Food Flavour Technology accepting routine panel results If a very close match is required, an expert panel may be needed 1.9.4 Customers A flavour is, sadly, nothing, if it is not sold Part of the tremendous ‘buzz’ of being a successful flavourist is the feeling of having created something really good The other part of the ‘buzz’ is seeing your product on the supermarket shelves, enjoyed by thousands, or perhaps millions, of people The best flavourists not divorce themselves from customer involvement and the art of selling Most successful flavourists reach the inescapable conclusion that they are the best judge of the market and the best flavour for a specific project This conclusion has some basis in fact, but the sad truth is that customer involvement is usually the key to success Customers usually know their own market best and especially their own brands They should be encouraged to guide the creative process and take a genuinely active part in the overall profile of the finished flavour One additional advantage of this approach is that the customers buy into the process and regard the resultant flavour as ‘theirs’, not without some justification The only barrier to this approach is the communication problem If the project has passed through intermediaries (sales, marketing, etc.) then communication is very difficult, however thoroughly the project information has been gathered The use of intermediaries also wastes a considerable amount of time that could otherwise be put to good creative use The only practical solution is direct contact between the flavourist and the application specialists in the customer’s laboratories Descriptive terms need to be defined and understood and simple examples (such as cis-3-hexenol for ‘leaf green’) can help a great deal Knowledge of the customer’s application processes is also needed, and the involvement of an application specialist from the flavour company is often vital Problems are often caused by the interactions between the flavour and the application ingredients and processes In many cases a small change to the customer’s formulation or process can save the day Sensory evaluation can play a useful part if it is used in the right context This is particularly true when expert panels are used to divine the precise flavour profiles preferred by a target market segment This is usually a specialised area where the flavour company probably has more depth of knowledge than the customer Most projects go through a number of iterations before they are concluded successfully, and it is vital in this process to remember that ‘the customer is always right’ 1.10 SUMMARY Flavour creation is still more of an art than a science Science provides a vital understanding of nature and a broad palette of raw materials Science may also provide insights into the preferences of a target group of consumers and some understanding of the mechanisms of taste and smell, but science cannot yet replace the intuitive, creative skills of a good flavourist Readers requiring more information on the art of flavour creation can find more information in Flavor Creation (Wright, 2004) Creating and formulating flavours 23 REFERENCES Labbe, D., Rytz, A., Morgenegg, C., Ali, S and Martin, N (2007) Subthresold olfactory stimulation can enhance sweetness Chem Senses 32, 205–214 Keller, A and Vosshall, L.B (2004) Human olfactory psychophysics Curr Biol 14, R875–R878 Shojaei, Z.A., Linforth, R.S.T., Hort, J., Hollowood, T.A and Taylor, A.J (2006) Measurement and manipulation of aroma delivery allows control of perceived fruit flavour in low and regular fat milks Int J Food Sci Technol 41, 1192–1196 Wright, J (2004) Flavor Creation Allured, Carol Stream, Illinois 2 Flavour legislation Jack Knights 2.1 INTRODUCTION Flavourings represent a class of food additives that has had comparatively little legislation This is probably due to the relatively large number of ingredients used in flavourings, the very small quantities involved and the difficulty of regulating added substances (which are also naturally present in foodstuffs) by analytical methods It was estimated about 50 years ago that only 50 defined flavouring substances were used to the extent of over 50 kg/year in flavourings in the UK (a corresponding figure for the USA appears to be about 200 substances) These figures represent an average consumption of about mg/person per year, at which level even many of the natural toxic materials, such as arsenicals or cyanides, are harmless One of the characteristics of food additives in general, and flavouring substances in particular, is the low degree of risk they pose to the consumer in spite of the public perception of their undesirability Around 1965, national regulators in several countries began to become interested in controlling food flavourings, although they had no evidence of untoward risks from their use However, they felt that, without detailed knowledge regarding the composition of flavourings, they were laying themselves open to criticism by food activists Initially, it was believed that all compound flavourings would have to be registered and approved by the appropriate authority, but the flavour industry managed to convince the legislative authorities that this was unnecessary and impractical Currently, flavour legislation is carried out at national and international levels The background history to flavour legislation is important as it explains why there are different systems in place and why they differ in their approach The following sections will explain the situation in the USA, in the international arena and in Europe Since the US system has been fairly stable for some years, in terms of its approach and procedures, the main emphasis is on the European Union (EU) situation where the formation of a large trading body over the last 40 years has led to attempts to harmonise legislation in many areas EU harmonisation is a slow process and despite a directive in 1988, the legislation proposed in 2008 will not be fully in force until 2011, so it is necessary to consider existing legislation (Section 2.6.2) and future legislation (Section 2.6.4) 2.2 METHODS OF LEGISLATION There are fundamentally two basic methods of legislation: Flavour legislation 25 (1) Positive: Only those materials that are listed are permitted to be used, to the exclusion of all others This type of legislation has the disadvantage that research into possible new materials is frustrated by the requirement for publication before use, unless the time required for commercialisation is built into the regulation However, regulatory authorities generally prefer positive listing because they believe that it offers the consumer maximum protection and is relatively easy to police (2) Negative: All materials are allowed except those that are listed In the case of food legislation, this needs to be supplemented by a statement that none of the materials used are injurious to health This firmly puts the requirement for safety of the materials onto the manufacturer, but is strongly objected to by most national legislative bodies on the basis that they not have control over what is used The system encourages the research and development of novel materials to provide more authentic flavourings, to the overall benefit of the consumer The above-mentioned systems may be combined so that some groups of materials are controlled by positive lists while other groups are covered by negative lists This is still the system in use in the EU, although it is gradually being superseded by total positive lists The relative effects on research of positive and negative legislation systems are illustrated in Fig 2.1 The figure shows the number of new flavour compounds discovered year by year on the basis of published data (patents and journal papers) and an estimate of the postulated numbers of new compounds, i.e those additional compounds discovered by flavour houses but not published The number of compounds added to the USA GRAS list is also shown The difference in the number of compounds discovered and those accepted on the GRAS (positive legislation) list suggests that the GRAS process does not encourage the development of new flavouring compounds In contrast, many of the new compounds found use as flavours in Europe, where the legislative system was more flexible Fig 2.1 Total volatile compounds in food, 1965–1990, derived from Volatile Compounds in Food (TNO Biotechnology and Chemistry Institute, Utrechtseweg 48, 3700 AJ Zeist, The Netherlands), and from FEMA/GRAS listing, 1965–1990 ‘Postulated’ is estimated as 10% non-published flavour industry research , postulated; , published; , GRAS listed 26 Food Flavour Technology Since 2000, most of the research into new flavouring substances by the flavour industry has been abandoned due to the impending provision of published positive lists, the loss of confidentiality and the exorbitant costs of toxicological clearance of very minor food-related substances 2.3 LEGISLATION IN THE UNITED STATES The system that has been adopted in the USA since 1965 is the positive list The initial work was undertaken by the USA Food and Drug Administration (FDA) as part of Title 21 of the Code of Federal Regulations Some 27 flavouring substances were evaluated and classified as generally recognised as safe (GRAS), presumably on the basis of long usage without untoward effect At that time the Flavor and Extract Manufacturing Association (FEMA) proposed to assist the FDA by using independent experts to evaluate the other flavouring ingredients that were known to be in use at the time, classifying most of them as GRAS and allocating the numbers 2001-3124 to them (Hall and Oser, 1965, 1970) FEMA later set up a panel of independent experts to evaluate flavouring materials on behalf of the FDA (Hallagan and Hall, 1995) This group was later named FEXPAN and consists of independent international toxicologists who undertake the safety evaluation of novel flavouring ingredients This procedure has continued to the present and there have been further 16 reports allocating GRAS status to a further 1542 flavouring ingredients (Oser and Hall, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979; Oser et al., 1984, 1985; Burdock et al., 1990; Smith and Ford, 1993; Smith et al., 1996, 1997, 2001, 2003, 2005, 2009; Newberne et al., 1998, 1999, 2000; Waddell et al., 2007) It should be noted that GRAS listing applies to all ingredients for use in flavourings, including non-flavouring materials such as solvents, carriers, emulsifiers and antioxidants The GRAS number makes no distinction between natural and artificial ingredients in the list; a given substance has the same entry irrespective of its status However, foods (e.g butter) may be used in the formulation of flavourings without being GRAS listed An entry in the GRAS list is accompanied by average usual and average maximum levels in ppm (parts per million equivalent to milligram/kilogram) that may be used in 34 categories of foodstuffs These levels are provided for novel flavour materials by the submitter and agreed by FEXPAN In the USA, the only alternative designations are ‘natural flavor’ and ‘artificial flavor’, and the latter has to be displayed directly with the name of the food as well as in the ingredients list It is therefore a highly undesirable designation for a food and is avoided wherever possible The term ‘natural flavor’ or ‘natural flavoring’ is defined in Title 21 of the Code of Federal Regulations, Chapter §101.22(a)(3) as the essential oil, oleoresin, essence or extractive, protein hydrolysate, distillate, or any product of roasting, heating or enzymolysis, which contains the flavouring constituents derived from spices, fruits or fruit juice, vegetable or vegetable juice, edible yeast, herbs, barks, buds, roots, leaves or similar plant materials, meat, seafood, poultry, eggs, dairy products or fermentation products thereof, whose significant function in food is flavouring rather than nutritional Natural flavours include the natural essence or extractives obtained from plants listed in §182.10, 182.20, 182.40, 182.50 and Part 184 of this chapter and the substances listed in §172.510 of this chapter (Code of Federal Regulations, 1990, Title 21) Flavour legislation 27 It has become normal in the USA to designate various categories of natural flavourings as follows: r r r FTNF (from the named fruit): As the name suggests, these consist solely of extracts or distillates derived from the named fruit For instance, strawberry FTNF could consist of concentrated strawberry juice with added strawberry distillate It may not contain material from any other natural source Flavourings of this category tend to be very expensive in use (they are usually very weak) and not very stable WONF (with other natural flavourings): These must contain more than 51% derived from the named source but may contain other natural flavouring ingredients For instance, strawberry WONF could consist of 51% concentrated strawberry juice fortified with other fruit juices or natural chemicals These flavourings are still expensive in use because of the price of the named ingredient Natural flavour: These must contain only natural ingredients, but the type or source is not defined For instance, natural strawberry flavour may contain ingredients from any source so long as they are classified as natural The solvent or carrier has to be on the GRAS list (or it can be a food), but this does not affect the natural or artificial status of the flavouring In recent years, a great deal of research effort has been directed to the preparation of natural versions of many of the significant, defined chemicals that are important in flavours This has included such techniques as heat-induced reactions often at elevated temperatures and high pressures (see Chapter 3) and direct esterification using enzymes as catalysts (see Chapter 4), on the assumption that if the reactants are natural, then the finished product may be so designated It should be noted that process flavours and smoke extracts are also regarded as natural under US law, whereas in Europe both these categories are separately designated and specifically non-natural 2.4 INTERNATIONAL SITUATION: JECFA JECFA is the Joint Food and Agriculture Organisation of the United Nations (FAO)/World Health Organisation (WHO) Expert Committee on Food Additives It was set up in 1956 to evaluate the safety of food additives, residues of veterinary drugs in food, and naturally occurring toxicants and contaminants in food JECFA serves as a scientific advisory body to FAO, WHO, their member states and the Codex Alimentarius Commission, primarily through the Codex Committee on Food Additives and Contaminants (CCFAC), regarding the safety of food additives including flavouring substances The committee establishes acceptable intakes on the basis of toxicological data and related information on the substance being evaluated It also develops general principles for assessing the safety of chemicals in food The requirement to keep up-to-date with scientific developments in toxicology and related disciplines necessitates a constant review of evaluation procedures (Munro et al., 1999) JECFA has to date evaluated some 600 flavouring substances and, more importantly, indicated that many flavouring agents are used in such small quantity that a full evaluation may not be appropriate 28 Food Flavour Technology JECFA evaluations and procedures are of international significance since they are not influenced by any particular group JECFA evaluations are taken into account by both FEXPAN and the European Food Safety Authority (EFSA) in their evaluation of the safety of flavourings for the USA and Europe, respectively 2.5 COUNCIL OF EUROPE The Council of Europe ad hoc Working Party on Natural and Artificial Flavouring Substances was set up in 1965 as a subsidiary body to the Sub-Committee on the Health Control of Foodstuffs The work carried out by this group had no legal standing but has been important in shaping the EU legislation described in Section 2.6 The initial aims of the Council of Europe were r r to draw up a list of natural and artificial flavourings that could be used in foodstuffs without hazard to public health; to draw attention to those flavourings that presented hazard to public health It is important in the context of the Council of Europe lists to understand the definitions they have used in classifying flavourings in their initial publication (Council of Europe, 1974) (i) A flavouring is a substance that has predominantly odour-producing properties and that possibly affects the taste (ii) A natural flavouring is a substance obtained from vegetable and sometimes animal sources, exclusively through the appropriate physical processes Those biological processes that occur spontaneously, and roasting, are assimilated to physical processes (iii) An artificial flavouring is a substance that has flavouring properties and that has been obtained by a chemical process This term includes (a) substances that exist in natural products; (b) substances not present, or as yet undiscovered in natural products These definitions are unsatisfactory in a number of ways and were modified in the 1981 publication (Council of Europe, 1981) as follows These definitions refer to materials and flavouring substances considered acceptable as flavourings and not necessarily apply to flavourings found in commerce: (i) Flavouring properties are those that are predominantly odour-producing and that possibly affect the taste (ii) A flavouring substance is a chemically defined compound that has flavouring properties It is obtained either by isolation from a natural source or by synthesis (iii) Natural sources of flavourings are products of plant or animal origin from which flavourings may be obtained exclusively through appropriate physical processes or by biological processes that occur spontaneously (e.g fermentation) (iv) Natural flavourings may be defined as complex mixtures derived from natural sources that have flavouring properties Flavour legislation 29 These definitions were carried on more or less unchanged in the 1992 edition (Council of Europe, 1992) In the first two editions of the list, the natural source materials were classified into the following groups: N1: Fruits and vegetables or parts thereof consumed as food No restriction on the parts used under the usual conditions of consumption is proposed N2: Plants and parts thereof, including herbs, spices and seasonings commonly added to foodstuffs in small quantities, the use of which is considered acceptable with a possible limitation of an active principle in the final product N3: Plants and parts thereof that, in view of their long history of use without evidence of acute untoward effects, are temporarily acceptable for continued use, in the traditionally accepted manner, in certain beverages and other foodstuffs The Committee of Experts, however, stresses that insufficient information is available for an adequate assessment of their potential long-term toxicity N4: Plants and parts thereof that are used for flavouring purposes at present but cannot be classified owing to insufficient information A number of N2 and N3 were listed with active principles such as coumarin or safrole, and these data have been the basis of Annex II in the EU Flavourings Directive The 1992 edition does not have a section on natural source materials In the 1974 and 1981 editions, the artificial flavouring substances were classified into lists as follows: Flavouring substances that may be added to foodstuffs without hazard to public health Flavouring substances that may be temporarily added to foodstuffs without hazard to public health Flavouring substances not fully evaluated This category was dropped from the 1981 edition despite containing a number of commonly used flavouring substances, several of which were on the FEMA/GRAS list The 1994 edition took a totally different approach to the listing of artificial flavouring substances Here, they were classified into chemical groups and for each substance a summary consisting of name, category, structure, Council of Europe, FEMA and CAS numbers, upper levels of use, natural occurrence and toxicity data was provided A further series of reports on natural source materials has been published (Council of Europe, 2000, 2007, 2008) Guidelines on the production of flavouring preparations by enzymatic and microbiological processes (Council of Europe, 1998b) and tissue culture (Council of Europe, 1998a) also have been published The Council of Europe Expert Committee has also produced guidelines on the production of thermal process flavourings, their ingredients and processing conditions (Council of Europe, 1995) as follows: (1) Ingredients added prior to processing (a) A protein nitrogen source (b) A carbohydrate source 30 (2) (3) (4) (5) Food Flavour Technology (c) A fat or fatty acid source (d) Other ingredients; herbs and spices and their extracts, water, thiamin, ascorbic, citric, lactic, fumaric, succinic and tartaric acids, guanylic and inosinic acids, inositol, lecithin, pH regulators and siloxanes as antifoaming agents Ingredients added after processing (a) Flavourings (b) Authorised food additives Processing conditions (a) The temperature of the product should not exceed 180◦ C (b) The duration of thermal processing should not exceed 15 minutes at 180◦ C with correspondingly longer times at lower temperatures; heating up and cooling down time should be as short as practical (c) The pH during processing should not exceed the value of 8.0 (d) Flavourings and food additives should only be added after processing is complete; flavour enhancers may be added before processing but only in minimum amounts necessary for flavour generation Purity criteria (a) Heavy metals as in EU directive (b) Benzo[a]pyrene not more than µg/kg (c) Benzo[a]anthracene not more than µg/kg (d) Amino-imidazo-azaarenes not detectable by the most sensitive routine method available Safety evaluation As a minimum requirement, the following toxicological studies should be performed: (a) A gene mutagenicity test (b) A test for chromosome damage in vivo or in vitro (c) A 90-day feeding study in animals As stated at the beginning of this section, the Council of Europe lists have no legal standing but the work of the Council of Europe Committee of Experts has provided much of the basis for the current EU legislation on flavourings Many members of the committee are also members of the Flavourings Sub-Committee of the EU Scientific Committee for Food, responsible for the evaluation of flavouring substances for the current EU positive list 2.6 EUROPEAN COMMUNITY 2.6.1 Background – national to EU legislation Several countries in Europe have had flavouring regulations in some form going back many years In the former West Germany, the initial regulation governing the use of artificial flavouring substances was enacted in 1959 (Essenzen VO, 1959) In this sense, the term ‘artificial’ refers to substances that are not chemically identical to materials present in natural products This regulation listed only five artificial flavouring substances that were permitted and a short list of processed foods in which they were allowed This regulation was amended in 1970 (Essenzen VO, 1970) to allow several more artificial substances in the same restricted list of foods Flavour legislation 31 In the UK in 1965, the Ministry of Agriculture, Fisheries and Food (MAFF) published a report on flavouring agents (MAFF, 1965) and a further report in 1976 (MAFF, 1976) Both these reports recommended that flavouring ingredients should be controlled by positive listing, but no action was taken to implement this However, the latter report acknowledged that substances chemically identical to natural substances (‘nature identical’) were different from artificial flavouring substances It also drew attention to the work of the Council of Europe in the flavouring field In 1988 the first attempt was made to harmonise the flavour regulations of the member states of the European Community This took the form of a Council Directive on the approximation of laws of the Member States relating to flavourings and to source materials for their production (EC, 1988a) The primary reason for the Directive was the protection of human health, but within these limits, to also take account of economic and technical needs 2.6.2 The 1988 Council Directive The 1988 Directive was intended to lay down a framework for flavour legislation It specified the procedures and principles that future legislation should consider and these included the following: (a) (b) (c) (d) (e) (f) (g) General purity criteria Definitions Labelling Appropriate provisions for the inventories created by Decision 88/389/EEC (EC, 1988b) Specific purity and microbiological criteria Limitation of certain components of vegetable or animal raw materials Drawing up of lists of additives, solvents and diluents for flavourings The following sections consider only those articles (EC, 1988a) that deal with nonprocedural matters and elaborate the principles and the potential effects of legislation on the food and flavour industry 2.6.2.1 Article (definitions) Article of the 1988 Council Directive attempted to define some terms that would help set the scope (Section 2.6.2.2) of the future legislation The text below lists the key definitions that were proposed with some commentary to explain the reasons for writing specific definitions in this manner and how theses definitions (and omissions) led to a working framework: This Directive shall apply to ‘flavourings’ used or intended for use in or on foodstuffs to impart odour and/or taste, and to source materials used for the production of flavourings For the purposes of this directive: (a) ‘Flavouring’ means flavouring substances, flavouring preparations, smoke flavourings, process flavourings or mixtures thereof Unlike most earlier definitions and normal commercial practice, this definition does not include the solvent or carrier but only the active part of the finished flavouring 32 Food Flavour Technology (b) ‘Flavouring substance’ means a defined chemical substance with flavouring properties which is obtained: (i) By appropriate physical processes (including distillation and solvent extraction) or enzymatic or microbiological processes from material of vegetable origin either in the raw state or after processing for human consumption by traditional foodpreparation processes (including drying, torrefaction and fermentation), (ii) By chemical synthesis or isolated by chemical processes and which is chemically identical to a substance naturally present in material of vegetable or animal origin as described in (i), (iii) By chemical synthesis but which is not chemically identical to a substance naturally present in material of vegetable or animal origin as described in (i) These three definitions refer to flavouring substances that, in practice, are called natural, nature identical and artificial, respectively The initial part of the definition refers to a defined chemical substance but is understood to include a defined mixture of defined substances, e.g citral, which is a mixture of cis- and trans-isomers The natural substance definition includes non-foods, e.g cedarwood, and thus the nature-identical definition includes non-food-identical substances There is no definition of ‘a traditional food preparation process’ or ‘appropriate physical process’, which leaves it open to individual interpretation: (c) ‘Flavouring preparation’ means a product, other than the substances defined in (i) above, whether concentrated or not, with flavouring properties, which is obtained by appropriate physical processes (including distillation and solvent extraction) or by enzymatic or microbiological processes from material of vegetable or animal origin, either in the raw state or after processing for human consumption by traditional food-preparation processes (including drying, torrefaction and fermentation) This definition covers such products as essential oils, concentrated essential oils, essential oil terpenes and isolates, oleoresins, resinoids, absolutes, extracts and tinctures of natural source materials (including the extraction solvent if this is a flavour carrier), distillates of natural source materials and fruit juices either concentrated or not used for their flavouring properties It suffers from the same imprecision as in the foregoing paragraph: (d) ‘Process flavouring’ means a product which is obtained according to good manufacturing practices by heating to a temperature not exceeding 180◦ C for a period not exceeding 15 minutes a mixture of ingredients, not necessarily themselves having flavouring properties, of which at least one contains nitrogen (amino) and another is a reducing sugar This definition is defective in that it does not apply to the majority of commercial process flavours, which are heated for 1–4 hours, even though at lower temperatures By negotiation with the commission, it has been unofficially agreed that longer times at lower temperatures are appropriate and a doubling of the time for each 10◦ C decrease in temperature is acceptable Flavour legislation 33 The definition does not specify what other flavouring ingredients may be added or whether sources that provide reducing sugars are covered (e) ‘Smoke flavouring’ means a smoke extract used in traditional foodstuffs smoking processes This is an impossible definition since traditional smoking processes not use smoke extract It is assumed that it refers to smoke generated by a method similar to that used in traditional smoking processes that is then condensed or absorbed in a carrier to form a smoke extract This is the meaning that was used by the European flavour industry but has now been amended by the 2003 Smoke Extract legislation (see Section 2.6.3) Flavourings may also contain foodstuffs as well as additives necessary for the storage, use, dissolution or dilution, or processing aids where these are covered by other Community provisions These materials are not within the ambit of paragraph 2(a) above and thus not form part of the flavouring but are additives to it 2.6.2.2 Article (scope) Having set the definitions in the previous section, the scope of the directive was then described as follows: The Directive shall not apply to: r r r Edible substances and products intended to be consumed as such, with or without reconstitution Substances which have exclusively a sweet, sour or salt taste Material of vegetable or animal origin, having inherently flavouring properties, where they are not used as flavouring sources The first bullet point excludes such products as fruit juices and the third bullet herbs and spices The position of flavour enhancers such as monosodium glutamate and other amino acids is not clear They are not really in accordance with the second bullet point since they certainly have some taste apart from being salty The latest inventory of flavouring substances includes some amino acids and they are also included in the proposed raw materials for the preparation of process flavourings 2.6.2.3 Article (restricted substances) As described in Section 2.2, the directive included a general statement that flavourings should be safe for human consumption and that certain categories of compounds should be avoided because of their toxicity Article describes these limitations as follows: 34 Food Flavour Technology Table 2.1 Maximum limits for certain undesirable substances present in foodstuffs as consumed as a result of the use of flavourings (EC, 1998; Annex I) Substance Foodstuffs (µg/kg) Beverages (µg/kg) Benzo[a]pyrene Benzo[a]anthracene 0.03 0.06 0.03 0.06 Member States shall take all measures necessary to ensure that: (a) Flavourings not contain any element or substance in a toxicologically dangerous quantity r Subject to any exceptions provided for in the specific criteria of purity referred to in Article 6(2) third indent, they not contain more than mg/kg of arsenic, 10 mg/kg of lead, mg/kg of cadmium and mg/kg of mercury (b) The use of flavourings does not result in the presence in foodstuffs as consumed of undesirable substances listed in Annex I [see Table 2.1] in quantities greater than those specified therein These limits are extremely low when compared with the levels of the same compounds in the surface layers of smoked or roasted meats (around ppm) (c) The use of flavourings and of other food ingredients with flavouring properties does not result in the presence of substances listed in Annex II [see Table 2.2] in quantities greater than those specified therein Annex II represented the position as at March 2001 Subsequently, capsaicin, estragole and methyleugenol have been added to the list 2.6.2.4 Article (inventories) This article is concerned with making appropriate decisions concerning the proposed inventories of flavour compounds created by Council Decision 88/389/EEC (EC, 1988b) as follows: The Commission shall, within 24 months of the adoption of this Decision and after consultation of the Member States, establish an inventory of: (a) Flavouring sources composed of foodstuffs and of herbs and spices normally considered as foods (b) Flavouring sources composed of vegetable or animal raw materials not normally considered as food (c) Natural flavouring substances (d) Synthetic flavouring substances chemically identical to substances in foodstuffs, herbs and spices (e) Synthetic flavouring substances chemically identical to substances in vegetable or animal raw materials not normally considered as foodstuffs, herbs and spices, artificial flavouring substances (f) Source materials used in the production of smoke and process flavourings and the reaction conditions under which they are prepared Flavour legislation 35 Table 2.2 Maximum limits for certain substances obtained from flavourings and other food ingredients with flavouring properties present in foodstuffs as consumed in which flavourings have been used (EC, 1998; Annex II) Substance Foodstuffs (mg/kg) Agaric acid 20 Beverages (mg/kg) 20 Exceptions and/or special restrictions 100 mg/kg in alcoholic beverages and foodstuffs containing mushrooms Aloin ␤-Asarone 0.1 0.1 0.1 0.1 50 mg/kg in alcoholic beverages mg/kg in alcoholic beverages and seasonings used in snack foods Berberine Coumarin 0.1 0.1 10 mg/kg in alcoholic beverages 10 mg/kg in certain types of caramel confectionery 50 mg/kg in chewing gum 10 mg/kg in alcoholic beverages Hydrocyanic acid 1 50 mg/kg in nougat, marzipan or its substitutes or similar products mg/% alcohol by volume/kilogram in alcoholic beverages Hypericine 0.1 0.1 mg/kg in canned stone fruit 10 mg/kg in alcoholic beverages mg/kg in confectionery Pulegone 25 100 250 mg/kg in mint- or peppermint-flavoured beverages Quassine 5 10 mg/kg in confectionery in pastille form 50 mg/kg in alcoholic beverages Safrole and isosafrole 1 mg/kg in alcoholic beverages with less than 25% alcohol by volume 350 mg/kg in mint confectionery mg/kg in alcoholic beverages with more than 25% alcohol by volume 15 mg/kg in foodstuffs containing mace and nutmeg Santonin 0.1 0.1 mg/kg in alcoholic beverages with more than 25% alcohol by volume Thujone (␣ and ␤) 0.5 0.5 mg/kg in alcoholic beverages with less than 25% alcohol by volume 10 mg/kg in alcoholic beverages with more than 25% alcohol by volume 25 mg/kg in foodstuffs containing preparations based on sage 35 mg/kg in bitters None of the above-listed substances may be added as such to flavourings or to foodstuffs They may be present in a foodstuff either naturally or following the addition of flavourings prepared from natural raw materials Groups (c)–(e) have been combined into a single inventory under Commission Regulation 2232/96 (EC, 1996) This regulation required member states to notify the commission of those flavouring substances allowed in their territory In practice the local flavour industry, under the guidance of the European Flavour and Fragrance Association (EFFA), supplied a list of the substances that were being used in their territory to the local regulatory organisation for that 36 Food Flavour Technology body to examine and, if appropriate, pass on to the commission by 23 November 1996 The Commission was required to consolidate the submissions into a single inventory representing the flavouring substances that were currently being used in the EU (EC, 1999) It was permitted to disclose substances under code to maintain confidentiality of their identity (EC, 1998), but this route has been taken only for about 12 substances The Standing Committee for Foodstuffs was required to examine and, if necessary, amend the inventory by 23 September 1999 The final inventory contained approximately 2500 substances and contained almost 500 substances that were not on the US GRAS lists The final permitted list would be dependent on safety evaluation of each individual substance by the Scientific Committee for Food This was delegated to the flavouring subcommittee, which examined the substances by chemical grouping under the Scientific Co-operation Procedure (SCOOP) (EC, 1994) The data required for evaluation were provided by EFFA on the form illustrated in Fig 2.2 The substances contained in the inventory were meant to be permitted in each member state of the EU until a final permitted list was established in about 2005 However, both Germany and Italy maintained their restricted lists of artificial flavouring substances that were in force before 1988 Some preliminary progress has been made on group (f) The European Commission has not yet published any guidelines on process flavourings, but it is likely that they will follow the Council of Europe recommendations The flavour industry stance, as agreed by International Organisation of the Flavour Industry (IOFI, 1989), was somewhat similar to those guidelines except for item ‘1 (d) Other ingredients’ where ‘herbs and spices and their extracts’ was expanded to include ‘flavouring substances identified therein’ There can be no argument that this significantly increased the risk, since the presence of these materials is already permitted owing to their occurrence in herbs and spices An initial draft document concerning smoke flavourings (EC, 2001) set out the types of wood that could be used to produce the smoke used in the preparation of the smoke extract, the conditions under which the smoke could be generated, the general specification for the product and the toxicological data necessary for its approval The way the document was written suggested that specific smoke flavourings from specified suppliers were the only ones that would be approved This is currently the position in Sweden, where there is a positive list of named smoke flavourings to the exclusion of all others This is not in accord with the Community anticompetition legislation 2.6.2.5 Article (additives for flavourings) As mentioned previously, legislation should consider not only the active flavouring compounds but also the other materials used to preserve, stabilise and deliver the flavourings in foods This article requires that the following lists of authorised substances be agreed: r r r Additives necessary for the storage and use of flavourings Products used for dissolving and diluting flavourings Processing aids where these are not covered by other community provisions The first two bullet points above have proved extremely difficult in obtaining agreement with the member states Discussions have been taking place with the commission since 1990 to try to thrash out a system that will satisfy all parties involved These have consisted of permitting Flavour legislation Butyl but-2-enoate Flavouring Substance EFFA No 0301 FEMA No - CoE No EINECS No 2-966X CAS No 591-63-9 JECFA No - FL No Chemical name on register Butyl but-2-enoate IUPAC name 2-Butenoic acid, (E)-, butyl ester Synonyms Crotonic acid, butyl ester, (E)- Physical form Liquid 09.324 Boiling point ( °C, 76 Torr) 80(42 T) Melting point ( °C) Ref index lower value (20 °C) 1.425 Ref index upper value Density lower value (25 °C) Minimum assay value Food category Sensory 95 Chemical group 1.431 Density upper value Insoluble in water Solubility 37 Beverages, excluding dairy products B1’721’863 Identity test Solubility in ethanol mL in mL 95%EtOH Origin (Codex, CAC) nat./nat.ident Normal dosage in ppm Maximum dosage in ppm 25 Volume of use by EFFA kg/a 14 Colourless liquid with fruity banana odour description Food source Pawpaw (Asimina Triloba Dunal.) Interpretive study in ppm 0.024 2000-3 Recent studies Butyl but-2-enoate Volume of use by EFFA: kg/year 14 Application Fig 2.2 Example of initial data submission required by SCOOP for flavouring substances, and volume of use by EFFA 38 Food Flavour Technology Food category Normal dosage Maximum dosage (ppm) (ppm) 35 25 10 50 35 10 50 25 10 50 Meat and meat products 10 Fish and fish products 10 Soups, sauces and seasonings 25 10 50 25 20 100 25 Dairy products (excluding those listed elsewhere) Fats and oils and fat emulsions Edible ices including sherbets and water ices Processed fruits and vegetables Sugar and chocolate confectionery Cereals and cereal products Bakery wares Foodstuffs intended for particular nutritional purposes Beverages excluding dairy products Ready-to-eat savouries Composite foods (e.g casseroles) Fig 2.2 (Continued ) automatically all the additives that are permitted for use in food or those that are quantum satis and the others by positive list Neither of these alternatives is regarded as satisfactory, mainly on the grounds that they would allow additives that are not permitted in a particular foodstuff to be added without declaration through the flavour composition The discussion is ongoing and agreement is of considerable importance since individual member state rules are being used to frustrate free trade in flavourings Article also deals with methods of sampling and analysis and microbiological criteria, none of which has been implemented to date 2.6.2.6 Article (labelling) One of the biggest consumer issues with food is understanding its composition and safety Consumer pressure groups are demanding more information on packs of food so that consumers can ‘make a choice’ although the history of food labelling in the EU shows how Flavour legislation 39 these good intentions can backfire The introduction of the E number system in the EU was designed to give consumers confidence that additives in their food have been fully tested and are safe for consumption Instead, the popular press convinced consumers that E numbers meant ‘chemicals’ in their food and that E numbers should be avoided This has led to food manufacturers in several countries either using trivial names in the ingredient list, e.g ascorbic acid (which is perceived as good by the consumer) rather than E 300, or simply by removing additives with E numbers It is unlikely that these moves by the food industry have had any significant benefit for the consumer Applying this background to flavourings that contain many different chemicals creates serious problems both for regulation and for consumer understanding Article therefore attempted a solution, which provided consumer safety without entailing a huge long list of compounds on packs of finished food (1) Flavourings not intended for sale to the final consumer may not be marketed unless their packaging or containers bear the following information, which should be easily visible, clearly legible and indelible: (a) The name or business name and address of the manufacturer or packer, or of a seller established in the community (b) Either the word ‘flavouring’ or a more specific name or description of the flavouring This enables descriptions such as ‘lemon flavouring’ or ‘natural flavouring’ if appropriate (c) Either the statement ‘for foodstuffs’ or a more specific reference to the foodstuff for which the flavouring is intended (d) A list in descending order of weight of the categories of flavouring ingredients present as follows: r Natural flavouring substances r Nature-identical flavouring substances r Artificial flavouring substances r Flavouring preparations r Process flavourings r Smoke flavourings (e) In the case of mixtures of flavourings with other substances referred to in Article 6, a list in descending order of weight in the mixture of: r The categories of flavourings classified as in (d) above r The names of each of the other substances or materials or their E numbers It is not clear whether this means a single list or two lists In general, the flavour industry uses the latter (f) An indication of the maximum quantity of each component contained in Annexes I and II or sufficient information to enable the food producer to comply with the limits for the finished food (g) An indication identifying the consignment (h) The nominal quantity in units of mass or volume (2) The word ‘natural’, or any other word having substantially the same meaning, may only be used for flavourings in which the flavouring component consists of exclusively flavouring preparations and/or natural flavouring substances 40 Food Flavour Technology Paragraph means that flavourings containing process flavourings or smoke flavourings cannot be labelled as natural, in contrast to the US position The above requirement is complied with in all EU member states except Italy, where flavourings containing nature identical flavouring substances are also designated as natural If the sales description of the flavouring contains a reference to a foodstuff or flavouring source, the word ‘natural’, or any word having substantially the same meaning may not be used unless the flavouring components have been isolated solely or almost solely from the flavouring source concerned This means that in ‘natural lemon flavouring’ the flavouring components have to be natural and derived solely or almost solely from lemon The term ‘almost solely’ is not defined but, by agreement with the 1988 Commissioner, a level of greater than 90% from the named source would be acceptable to both the commission and industry This leaves the question of how to designate natural flavourings that are not derived solely or almost solely from the named source This is of course a matter for the individual Member States In the UK, wordings such as ‘natural lemon flavour flavouring’ or ‘natural flavouring lemon type’ have been used, but the matter has never been tested in the courts to decide whether these are acceptable (3) By derogation from paragraph 1, the information required by paragraphs 1(d), (e) and (f) may appear merely on the trade documents relating to the consignment supplied prior to the delivery, provided the phrase ‘intended for the manufacture of foodstuffs and not for retail’ appears in a conspicuous part of the packaging or container of the products in question It is not clear whether this phrase should replace that in paragraph 1(c) or be in addition to it It is normal for the former to be used (4) Member States shall refrain from laying down requirements more detailed than those contained in this article, concerning the manner in which the particulars provided for are to be shown These particulars shall be given in terms easily understood by the purchaser This shall not prevent them being given in various languages This provision has been ignored by at least one member state by invoking the Reserved Dairy Descriptions Directive to prevent the use of milk, cream, butter or cheese in the description of flavourings After complaints by other member states this was abandoned, as was the insistence that solely their own language be used on label of flavourings for their country 2.6.3 Smoke flavourings 2003 Directive In 2003, the European Parliament and the Council produced a new document (EC, 2003), covering smoke flavourings as defined in Directive 88/388/EEC It was felt that the protection of human health with respect to smoke flavouring was inadequate and these should be regulated separately The intention of the document was to authorise and regulate certain smoke preparations to the exclusion of all others The regulation defined ‘primary smoke condensates’ and ‘primary tar fractions’ and referred to derived products therefrom The conditions of generation of the primary products were defined as controlled burning, dry distillation, or treatment with superheated steam in a controlled oxygen environment with a maximum temperature of 600◦ C The smoke is condensed and may be physically treated Flavour legislation 41 to achieve phase separation, isolation and/or purification The primary products must not contain more than 10 ␮g/kg benzo[a]pyrene and 20 ␮g/kg benz[a]anthracene In order to obtain authorisation for a primary product, an application must be made to the competent authority in a member state giving information on the type of wood used, the detailed production process, the quantitative and qualitative composition including its variability and analytical methodology Also required is information on the intended use levels in, or on, specific food or food categories and detailed toxicological data The data provided will be evaluated by EFSA that will provide an opinion as to the safety of the primary product and any conditions or restrictions to its use The European Commission will take note of the EFSA opinion and, if acceptable, will authorise the use of the primary product, giving it a unique identifying code This code has to be used in all transactions involving the primary product including product derived therefrom The only stage at which this code does not have to be displayed is on food products intended for direct sale to the final consumer If smoke flavourings are used as ingredients of other flavourings and impart a smoky flavour, they must be declared separately in the ingredients list This must include their identifying code except in the ingredients list of food intended for direct sale to the final consumer 2.6.4 Developments 2008 onwards In mid-2008, the Council of the European Union produced drafts of four regulations, commonly termed the ‘Food Improvement Agents Package’ These draft regulations mark a shift in the approach to flavour legislation which is now contained in an overall scheme that includes food additives and enzymes The first of these documents, the Common Authorisation Procedure (EC, 2008a), describes the establishment of a common authorisation procedure for food additives, food enzymes and food flavourings after adoption of a Regulation of the European Parliament and of the Council The second document, ‘Chapter I General Principles’ outlines what the regulation is about and introduces a common assessment and authorisation procedure (hereinafter referred to as the ‘common procedure’) The common procedures will apply to food enzymes (EC, 2008b), food additives (EC, 2008c) and food flavourings including food ingredients with flavouring properties (EC, 2008d) and specifically exclude smoke flavourings falling within the scope of Regulation (EC) 2065/2003 (as discussed in Section 2.6.4) The third document ‘Chapter II Common Procedure’ lays down the procedural arrangements for updating the lists of authorised substances and the criteria, according to which, substances can be included on the lists The common procedure may be started on the initiative of the commission or following an application, either by a member state or by an interested party The commission shall seek the opinion of the EFSA, but this may not be required if the update does not have an effect on human health EFSA is required to give its opinion within months of a valid application and forward it to the commission, the member states and where applicable, the applicant This period may be extended where EFSA requests additional information by agreement with the applicant Within months of EFSA giving its opinion, the commission shall submit a draft regulation updating the community list There is a provision for confidentiality with the submission but since this may not include the name and clear description of the substance, it is of little value particularly in the cases of flavouring substances The fourth document ‘Flavourings Regulation’ is specific for flavourings (EC, 2008d) and was published on 31 December 2008 but is intended to apply from 20 January 2011 42 Food Flavour Technology Certain aspects apply from 20 January 2009 and compliance with the community list shall apply from 18 months after its date of application In some ways this regulation is similar to Directive 88/388 (EC, 1988a) but with significant differences, and a comparison of the previous sections with those below will highlight the changes 2.6.4.1 Article (scope) The scope has been modified compared to that described in Section 2.6.2.2 by including food ingredients with flavouring properties The regulation also applies to foodstuffs containing flavourings and/or food ingredients with flavouring properties as well as to their source materials It does not apply to smoke flavourings or substances having exclusively a sweet, sour or salty taste, raw foods, non-compounded foods or mixtures of spices and/or herbs, mixtures of tea and mixtures for infusion as such as long as they have not been used as food ingredients In the new regulation, ‘flavourings’ may be made or consist of flavouring substances, flavouring preparations, thermal process flavourings, smoke flavourings (even though the regulation does not apply to them) and two new categories – ‘flavour precursors’ and ‘other flavourings’ – and mixtures of the above-mentioned categories 2.6.4.2 Article (definitions) The following are the key differences between the situation explained in Section 2.6.2.1 and the 2008 proposals: ‘Flavouring substance’ is defined as before but no distinction is made between nature identical and artificial flavouring substances ‘Natural flavouring substance’ is defined as before but with the proviso that if it is prepared for human consumption, it must be by using one of the processes defined in Annex II (see Table 2.3) ‘Flavouring preparation’ is defined as before but with the above Annex II proviso These not require evaluation and approval if they are derived from food Table 2.3 List of traditional food preparation processes (ANNEX II; EC 2008a) Chopping Heating, cooking, baking, frying (up to 240◦ C at atmospheric pressure) and pressure cooking (up to 120◦ C) Coating Cooling Cutting Drying Evaporation Distillation/rectification Emulsification Extraction, incl solvent extraction in accordance with Directive 88/344/EEC Fermentation Grinding Infusion Microbiological processes Peeling Pressing Roasting/Grilling Steeping Filtration Maceration Mixing Percolation Refrigeration/Freezing Squeezing Flavour legislation 43 Table 2.4 Conditions for the production of thermal process flavourings and maximum levels for certain substances in thermal process flavourings (ANNEX V; EC, 2008a) Part A: Conditions for the production (a) The temperature of the products during processing shall not exceed 180◦ C (b) The duration of the thermal processing shall not exceed 15 minutes at 180◦ C with correspondingly longer times at lower temperatures, i.e a doubling of the heating time for each decrease of temperature by 10◦ C, up to a maximum of 12 hours (c) The pH during processing should not exceed the value of 8.0 Part B: Maximum levels for certain substances Substance Maximum levels (µg/kg) 2-Amino-3,4,8-trimethylimidazo[4,5-f ] quinoxaline (4,8-DiMeIQx) 2-Amino-1-methyl-6-phenylimidazol [4,5-b] pyridine (PhIP) 50 50 ‘Thermal process flavouring’ is defined as before but is subdivided by its ingredients being food and/or source materials other than food If the ingredients are food and the finished flavouring complies with Annex V (Table 2.4), these not require evaluation and approval All other thermal process flavourings require evaluation and approval ‘Smoke flavouring’ is defined as in Regulation (EC) 2065/2003 (Section 2.6.3) ‘Flavour precursor’ shall mean a product, not necessarily having flavouring properties itself, intentionally added to food for the sole purpose of producing flavour by breaking down or reacting with other components during food processing It may be obtained from food or source materials other than food If obtained from food, these not require evaluation and approval ‘Other flavouring’ shall mean a flavouring added or intended to be added to food in order to impart odour and/or taste not covered by the other definitions ‘Food ingredient with flavouring properties’ shall mean a food ingredient other than flavourings, which may be added to food for the main purpose of adding flavour to it or modifying its flavour and which contribute significantly to the presence in food of certain naturally occurring undesirable substances These not require prior evaluation and approval ‘Appropriate physical process’ shall mean a physical process that does not intentionally modify the chemical nature of the components of the flavouring, without prejudice to the listing of traditional food preparation processes in Annex II, and does not involve, inter alia, the use of singlet oxygen, inorganic catalysts, metal catalysts, organometallic reagents and/or UV radiation The processes are listed in Table 2.3 Flavourings may contain food additives as permitted by Regulation (EC) 1333/2008 Annex III Part (EC, 2008c), and/or other food ingredients incorporated for technological purposes 2.6.4.3 Articles and (evaluation and approval) All categories of flavourings listed in the foregoing section require prior evaluation and approval unless specifically defined out 2.6.4.4 Article (presence of certain substances) In a similar way to previously (Section 2.6.2.3, Table 2.2), there is a list of substances that may not be added to food as such (Table 2.5; Annex III EC, 2008d) There is also a list of certain 44 Food Flavour Technology Table 2.5 Substances and limits in flavourings (Annex III; EC, 2008d) Part A: Substances that shall not be added as such to food Agaric acid Aloin Capsaicin 1,2-Benzopyrone, coumarin Hypericine ␤-Asarone 1-Allyl-4-methoxybenzene, estragole Hydrocyanic acid Menthofuran 4-Allyl-1,2-dimethoxybenzene, methyl eugenol Pulegone Quassin 1-Allyl-3,4-methylene dioxy benzene, safrole Teucrin A Thujone (alpha and beta) Part B: Maximum levels of certain substances, naturally present in flavourings and food ingredients with flavouring properties, in certain compound food as consumed to which flavourings and/or food ingredients with flavouring properties have been added Maximum Name of the Compound food in which the presence of level substance the substance is restricted (mg/kg) ␤-Asarone 1-Allyl-4-methoxybenzene Estragole Hydrocyanic acid Menthofuran 4-Allyl-1,2dimethoxybenzene Methyleugenol Pulegone Alcoholic beverages Dairy products Processed fruits, vegetables (including mushrooms, fungi, roots, tubers, pulses and legumes), nuts and seeds Fish products Non-alcoholic beverages Nougat, marzipan or its substitutes or similar products Canned stone fruits Alcoholic beverages Mint/peppermint-containing confectionery, except micro breath-freshening confectionery Micro breath-freshening confectionery Chewing gum Mint/peppermint-containing alcoholic beverages Dairy products 1.0 50 50 50 10 50 35 500 3000 1000 200 20 Meat preparations and meat products including poultry and game 15 Fish preparations and fish products Soups and sauces Ready-to-eat savouries Non-alcoholic beverages 10 60 20 Mint/peppermint-containing confectionery, except micro breath-freshening confectionery Micro breath-freshening confectionery Chewing gum Mint/peppermint-containing non-alcoholic beverages Mint/peppermint-containing alcoholic beverages 250 2000 350 20 100 Flavour legislation Table 2.5 (Continued ) Quassin 1-Allyl-3,4methylenedioxy benzene, safrole Teucrin A Thujone (alpha and beta) Coumarin 45 Non-alcoholic beverages Bakery wares Alcoholic beverages Meat preparations and meat products including poultry and game 0.5 1.5 15 Fish preparations and fish products Soups and sauces Non-alcoholic beverages Bitter-tasting spirit drinks or bitter Liqueurs with a bitter taste Other alcoholic beverages 15 25 5 Alcoholic beverages except those produced from Artemesia species Alcoholic beverages produced from Artemesia species Non-alcoholic beverages produced from Artemesia species Traditional and/or seasonal bakery ware containing a reference to cinnamon in the labelling Breakfast cereals including muesli Fine bakery ware with the exception of traditional and/or seasonal bakery ware containing a reference to cinnamon in the labelling Desserts 10 35 0.5 50 20 15 substances, naturally present in flavourings and food ingredients with flavouring properties, which are limited in certain compound foods to which they have been added (Annex III Part B) The maximum levels not apply to compound foods that are not listed or where no flavourings have been added, and the only food ingredients with flavouring properties that have been added are fresh, dried or frozen herbs and spices 2.6.4.5 Article (the use of certain source materials) There is also a list of source materials that may not be used for the production of flavourings (Table 2.6, Annex IV Part A) and flavourings derived from certain source materials for which there are specified conditions of use (Annex IV Part B) 2.6.4.6 Article 10 (community list of flavouring and source materials) As stated earlier, all flavouring substances, whether natural or not, have to be evaluated by EFSA and approved by the European Commission as laid down in the common authorisation procedure (Regulation (EC) 1331/2008) The approved substances, including conditions of use if appropriate, will form the community list Only flavouring substances that are contained in the community list may be used in flavourings The community list of flavouring substances is likely to apply from the middle of 2012 The regulation refers to all the classes of flavouring ingredients requiring evaluation and approval, but to date no attempt has been made to undertake the evaluation of other than flavouring substances 46 Food Flavour Technology Table 2.6 List of source materials to which restrictions apply for their use in the production of flavourings and food ingredients with flavouring properties (Annex IV; EC, 2008d) Part A: Source materials that shall not be used for the production of flavourings and food ingredients with flavouring properties Source material Latin name Common name Tetraploid form of Acorus calamus L Tetraploid form of calamus Part B: Conditions of use for flavourings and food ingredients with flavouring properties produced from certain source materials Source material Latin name Common name Conditions of use Quassia amara L and Picrasma excelsa (Sw) Quassia Flavourings and food products with flavouring properties produced from the source material may only be used for the production of beverages and bakery wares Laricifomes officinales (Vill.: Fr) Kotl Et Pouz or Fomes officinalis White agaric mushroom Flavourings and food products with flavouring properties produced from the source material may only be used for the production of alcoholic beverages Hypericum perforatum L Teucrium chamaedrys L St John’s wort Wall germander 2.6.4.7 Articles 14 and 15 (labelling of flavourings not intended for sale to the final consumer) The general labelling requirements are similar to the present legislation (Section 2.6.2.6) with the addition of a date of minimum durability or use by date 2.6.4.8 Article 16 (specific requirements for the use of the term ‘natural’) This differs significantly from the present flavouring legislation and the new definition of natural is given below The term ‘natural’ may only be used if the flavouring component comprises only flavouring preparations and/or natural flavouring substances This is the same as previously The term ‘natural flavouring substance(s)’ may only be used for flavourings in which the flavouring component contains exclusively natural flavouring substances The term ‘natural’ may only be used in combination with a reference to a food, food category or a vegetable or animal flavouring source if the flavouring component has been obtained exclusively or by at least 95% w/w from the source material referred to The present legislation requires 90% from the specified source material The label description shall read ‘natural “food(s) or food category or source(s)” flavouring’ Flavour legislation 47 The term ‘natural “food(s) or food category or source(s)” flavouring with other natural flavourings’ may only be used if the flavouring component is partially derived from the source material referred to, the flavour of which can easily be recognised The term ‘natural flavouring’ may only be used if the flavouring component is derived from different source materials and where a reference to the source materials would not reflect their flavour or taste 2.6.4.9 Article 29 (designation of flavourings in the list of ingredients on finished foods) Directive 2000/13/EC Annex III (Section 2.6.2.6) shall be replaced by the following: (1) Flavourings shall be designated by the terms – ‘flavourings’ or a more specific name or description of the flavouring; – ‘smoke flavouring(s)’ if the flavouring component contains smoke flavourings and imparts a smoky flavour to the food This means that, if the flavouring contains a smoke flavouring as defined in Regulation (EC) 2065/2003 and imparts a smoky flavour to the food, both ‘flavourings’ and ‘smoke flavouring(s)’ must appear in the ingredients list on the food (2) The term ‘natural’ for the description of flavourings shall be in accordance with Article 16 of Regulation (EC) 1334/2008 2.7 CURRENT EU SITUATION AND THE FUTURE The introduction of the 2008 draft Regulations means that EU flavour legislation is currently between the old and the new systems with new regulations being phased in up to 2012 This makes a complex situation for food and flavour manufacturers The new regulations have cleared up some of the omissions in the earlier regulations and provided clearer definitions in some areas, but it is very difficult to write legislation that covers the huge range of flavour compounds found in foods and all the potential usages that may arise Inevitably, there will be further amendments when such issues are recognised There appears to be a further problem, namely, whether flavourings that comply with the new Regulation and not the old Directive are legal before the date of implementation of the new Regulation This has yet to be resolved by the European Commission It is widely recognised both among regulators and the international flavour industry that the lack of harmonisation of flavouring legislation across the world represents a significant barrier to trade The acceptance in one country of a flavouring material banned in another cannot possibly be on the basis of safety; it is usually on the basis of commercial interest The International Organisation of the Flavour Industry (IOFI) has been pressing legislators to accept flavouring materials on the basis of safety evaluations performed for other groups, e.g the acceptance by European Scientific Committee for Food of evaluations done by FEXPAN The international nature of the flavour industry is providing data to FEXPAN to enable the evaluation of the 500 or so additional flavouring substances that are on the European inventory so that they may be granted GRAS status for use in the USA 48 Food Flavour Technology REFERENCES Burdock, G.A., Wagner, B.M., Smith, R.L., Munro, I.C and Newbeme, P.M (1990) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances 15 Food Technol 44(2), 82 Code of Federal Regulations Title 21 (1990) Food and drugs Chapter 1, §101.22 (a) (3) Council of Europe (1974) Natural Flavouring Substances, Their Sources, and Added Artificial Flavouring Substances, Council of Europe, Maisonneuve, Paris Council of Europe (1981) Flavouring Substances and Natural Sources of Flavourings, 3rd edn, Council of Europe, Maisonneuve, Paris Council of Europe (1992) Flavouring Substances and Natural Sources of Flavourings, 4th edn, Council of Europe, Maisonneuve, Paris Council of Europe (1995) Guidelines on the Production of Thermal Process Flavourings, Council of Europe Publishing, Paris, ISBN 92-871-2811-1 Council of Europe (1998a) Guidelines for Flavouring Preparations Produced by Plant Tissue Culture, Council of Europe Publishing, Paris, ISBN 978-92-871-3738-8 Council of Europe (1998b) Guidelines for Flavouring Preparations Produced by Enzymatic or Microbiological Processes, Council of Europe Publishing, Paris, ISBN 978-92-71-2586-6 Council of Europe (2000) Natural Sources of Flavourings – Report No 1, Council of Europe Publishing, Paris, ISBN 978-92-871-4324–2 Council of Europe (2007) Natural Sources of Flavourings – Report No 2, Council of Europe Publishing, Paris, ISBN 978-92-871-6156-7 Council of Europe (2008) Natural Sources of Flavourings – Report No 3, Council of Europe Publishing, Paris, ISBN 978-92-871-6422-3 EC (1988a) Council directive of 22 June 1988 on the approximation of the laws of the member states relating of flavourings and to source materials for their production (88/388/EEC) Official Journal of the European Communities, No L 184/61(15/7/88) See also UK implementation: The Flavouring in Foods Regulations S.I 1992 No 1971, The Flavouring of Food (Amendment) Regulations S.I 1994 No 1486 EC (1988b) Council decision of 22 June 1988 on the establishment, by the commission, of an inventory of the source materials and substances used in the preparation of flavourings (88/389/EEC) Official Journal of the European Communities, No L 184/67 (15/7/88) EC (1994) European Commission decision of 20 September 1994 establishing the inventory and distribution of tasks to be undertaken within the framework of co-operation by member states in the scientific examination of questions relating to food (1994/652/EC) Official Journal of the European Communities, No L 253/29 (29/9/94) EC (1996) European Parliament and Council Regulation (EC) of 28 October 1996 laying down a community procedure for flavouring substances used or intended for use in Foodstuffs (2232/96) EC (1998) European Commission Recommendation 98/282/EC EC (1999) European Commission decision of 23 February 1999 adopting a register of flavouring substances used in or on foodstuffs (1999/217/EC) Official Journal of the European Communities, No L 084/1 (27/3/99) amended by Decision 2000/489/EC, Official Journal of the European Communities, No L 197/53 (3/8/00) EC (2001) European Commission Working Document WGF/007/01 Draft European Parliament and Council Regulation for smoke flavourings used or intended to be used in or on foodstuffs (22 February 2001) EC (2003) Regulation No 2065/2003 of the European Parliament and of the Council of 10 November 2003 on smoke flavourings used or intended for use in or on foods EC (2008a) Regulation No 1331/2008 of the European Parliament and of the Council of 16 December 2008 establishing a common authorisation procedure for food additives, food enzymes and food flavourings EC (2008b) Regulation No 1332/2008 of the European Parliament and of the Council of 16 December 2008 on food enzymes and amending Council Directive 83/417/EEC, Council Regulation (EC) No 1493/1999, Directive 2000/13/EC, Council Directive 2001/112/EC and Regulation(EC) No 258/97 EC (2008c) Regulation No 1333/2008 of the European Parliament and of the Council of 16 December 2008 on food additives Flavour legislation 49 EC (2008d) Regulation No 1334/2008 of the European Parliament and of the Council of 16 December 2008 on flavourings and certain food ingredients with flavouring properties for use in and on foods and amending Council Regulation (EEC) No 1601/91 Regulations (EC) No 2232/96 and (EC) No 110/2008 and Directive 2000/13/EC Essenzen VO (1959) Verordnung uber Essenzen und Grundstoffe (Essenzen VO) vom 19 Dezember 1959 (BGB1.IS.747) Essenzen VO (1970) Verordnung uber Essenzen und Grundstoffe (Essenzen VO) Der Neufassung vom Oktober 1970 (BGBL I S 1389) Hall, R.L and Oser, B.L (1965) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances Food Technol 19(2), Part 2, 151; FEMA GRAS Substances Food Technol 24(5) 25 Hall, R.L and Oser, B.L (1970) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances Food Technol 19(2), Part 2, 151; FEMA GRAS Substances Food Technol 24(5) 25 Hallagan, J.B and Hall, R.L (1995) FEMA GRAS – a GRAS assessment program for flavor ingredients Regul Toxicol Pharmacol 21, 422–430 IOFI, International Organisation of the Flavour Industry (1989) Code of Practice Section III IOFI guidelines for the production and labelling of process flavourings (October 1989) MAFF (1965) Food Standards Committee Report on Flavouring Agents, HMSO, London MAFF (1976) Food Additives and Contaminants Committee Report on the Review of Flavourings in Food, FAC/REP/22, HMSO, London Munro, I.C., Kennepohl, E and Kroes, R (1999) A procedure for the safety evaluation of flavouring substances Food Chem Toxicol 37, 207–232 Newberne, P., Smith, R.L., Doull, J., Goodman, J.I., Munro, I.C., Portoghese, P.S., Wagner, B.M., Weil, C.S., Woods, L.A., Adams, T.B and Hallagan, J.B (1998) GRAS flavoring substances 18 Food Technol 52(9), 65 Newberne, P., Smith, R.L., Doull, J., Goodman, J.I., Munro, I.C., Portoghese, P.S., Wagner, B.M., Weil, C.S., Woods, L.A., Adams, T.B., Hallagan, J.B and Ford, R.A (1999) Correction to GRAS flavoring substances 18 Food Technol 53(3), 104 Newberne, P., Smith, R.L., Doull, J., Feron, V.J., Goodman, J.I., Munro, I.C., Portoghese, P.S., Waddell, W.J., Wagner, B.M., Weil, C.S., Adams, T.B and Hallagan, J.B (2000) GRAS flavoring substances 19 Food Technol 54(6), 66, 68–70, 72–74, 76–84 Oser, B.L., Ford, R.A and Bernard, B.K (1984) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances 13 Food Technol 38(10), 65 Oser, B.L., Ford, R.A and Bernard, B.K (1985) FEMA GRAS Substances 14 Food Technol 39(11), 108 Oser, B.L and Hall, R.L (1972) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances Food Technol 26(5), 35 Oser, B.L and Hall, R.L (1973) FEMA GRAS Substances Food Technol 27(1), 64 Oser, B.L and Hall, R.L (1974) FEMA GRAS Substances Food Technol 27(11), 56 Oser, B.L and Hall, R.L (1975) FEMA GRAS Substances Food Technol 28(9), 76 Oser, B.L and Hall, R.L (1976) FEMA GRAS Substances Food Technol 29(9), 70 Oser, B.L and Hall, R.L (1977) FEMA GRAS Substances 10 Food Technol 31(1), 65 Oser, B.L and Hall, R.L (1978) FEMA GRAS Substances 11 Food Technol 32(2), 60 Oser, B.L and Hall, R.L (1979) FEMA GRAS Substances 12 Food Technol 33(7), 65 Smith, R.L and Ford, R.A (1993) Recent progress in the consideration of flavoring ingredients under the Food Additive Amendment FEMA GRAS Substances 16 Food Technol 47(6), 104 Smith, R.L., Newberne, P., Adams, T.B., Ford, R.A., Hallagan, J.B and the FEMA Expert Panel (1996) GRAS flavoring substances 17 Food Technol 50(10), 72 Smith, R.L., Newberne, P., Adams, T.B., Ford, R.A., Hallagan, J.B and the FEMA Expert Panel (1997) Correction to GRAS flavoring substances 17 Food Technol 51(2), 32 Smith, R.L., Doull, J., Feron, V.J., Goodman, J.I., Munro, I.C., Newberne, P.M., Portoghese, P.S., Waddell, W.J., Wagner, B.M., Adams, T.B and McGowen, M.M (2001) GRAS flavoring substances 20 Food Technol 55(12), 34–36, 38, 40, 42, 44–55 Smith, R.L, Cohen, S.M., Doull, J., Feron, V.J., Goodman, J.I., Marnett, I.J., Portoghese, P.S., Waddell, W.J., Wagner, B.M and Adams, T.B (2003) GRAS flavoring substances 21 Food Technol 57(5), 46–48, 50, 52–54, 56–59 50 Food Flavour Technology Smith, R.L., Cohen, S.M., Doull, J., Feron, V.J., Goodman, J.I., Marnett, I.J., Portoghese, P.S., Waddell, W.J., Wagner, B.M and Adams, T.B (2005) GRAS flavoring substances 22 Food Technol 59(8), 24–28, 31–32, 34, 36–62 Smith, R.L., Waddell, W.J., Cohen, S.M., Feron, V.J., Marnett, L.J., Portoghese, P.S., Rietjens, I.M.C.M., Adams, T.B., and Taylor, S (2009) GRAS flavoring substances 24 Food Technol 63(6), 46–105 Waddell, W.J., Cohen, S.M., Feron, V.J., Goodman, J.I., Marnett, L.J., Portoghese, P.S., Rietjens, I.M.C.M., Smith, R.L., Adams, T.B., Lucas Gavin, C., McGowen, M.M and Williams, M.C (2007) GRAS flavoring substances 23 Food Technol 61(8), 22–24, 26–28, 30–49 3 Basic chemistry and process conditions for reaction flavours with particular focus on Maillard-type reactions Josef Kerler, Chris Winkel, Tomas Davidek and Imre Blank 3.1 INTRODUCTION Maillard reaction technology is used by the flavour and food industry for the production of process/reaction flavours or generating flavour upon food processing (in-process flavour generation) Process flavours are complex building blocks that provide similar aroma and taste properties to those found in thermally treated foodstuffs such as meat, chocolate, coffee, caramel, popcorn and bread The Maillard reaction between a reducing sugar and a food-grade nitrogen source is the principal underlying reaction, which is responsible for flavour and colour development However, Maillard-type reactions may also give rise to undesirable molecules that need to be limited using mitigation concepts This review provides a summary of general aspects of the Maillard reaction in flavour formation in view of reinforcing distinct desirable flavour notes An overview of important aroma compounds of thermally treated foodstuffs and process flavours is given In addition, the patent literature and other publications relating to reaction flavour production and their process conditions are discussed This chapter focuses on Maillard-type reactions and only partially deals with other reactions occurring during process flavour preparation (e.g lipid oxidation) 3.2 GENERAL ASPECTS OF THE MAILLARD REACTION CASCADE The ‘Maillard reaction’ is of great importance for flavour and colour formation of thermally treated foodstuffs It is a complex cascade of many different types of reactions rather than one single reaction type, even though it is initiated by an amino-carbonyl reaction step The thermal generation of flavours in foods, process flavours and model systems has accordingly been the subject of many symposia and reviews (e.g Parliment et al., 1989, 1994; Weenen et al., 1997; Reineccius, 1998; Tressl and Rewicki, 1999; Cerny, 2007; Yeretzian et al., 2007) A first milestone in the history of Maillard chemistry was the publication of the wellknown Hodge scheme (Hodge, 1953) Although Hodge’s studies focused only on Maillard browning, this scheme provided a framework that also covered important reaction routes for the formation of aroma compounds The work of Hodge triggered a large number of studies on the elucidation of important intermediates and pathways of the Maillard reaction (see reviews by Ledl and Schleicher, 1990; Tressl and Rewicki, 1999) Isotopic labelling of sugars and/or amino acids in conjunction with GC-MS (gas chromatography–mass spectrometry) analysis (Tressl et al., 1993; Gi and Baltes, 1995; Keyhani and Yaylayan, 1996) and trapping of 52 Food Flavour Technology reactive intermediates (Nedvidek et al., 1992; Hofmann, 1999) are key techniques for the improved understanding of Maillard reaction pathways Improvements and refinements of the Hodge scheme were presented by Tressl et al (1995) Their scheme provides an excellent overview of Maillard reaction pathways leading to the formation of volatile compounds For the optimisation of reaction flavours, however, a strong emphasis is required on those routes that are involved in the generation of key aroma compounds This can be achieved by first evaluating the character-impact compounds of a process flavour or model system using a combination of sensorial and instrumental analysis on the basis of the odour activity value concept (Grosch, 1994) The mechanistic studies can then be focused on the key substances only The work of Hofmann (1995) is an excellent example of such an approach Figure 3.1 gives an overview of the pathways that are involved in the formation of important aroma compounds during Maillard reaction There are three main routes involved in flavour generation All three routes start with imine formation between a reducing sugar and an amino acid The Amadori (derived from aldoses) or Heyns (derived from ketoses) rearrangement products are important intermediates of the early phase of the Maillard reaction Route A (see also Section 3.2.1) leads to the formation of 1- and 3-deoxyosones, which on cyclisation, reduction, dehydration and/or reaction with hydrogen sulfide result in heterocyclic aroma compounds Route B (see also Section 3.2.2) is characterised by fragmentation of the sugar chain through retro-aldolisation or ␣-/␤-cleavage By aldol-condensation of two sugar fragments or a sugar fragment and an amino acid fragment, heterocyclic aroma compounds are generated on cyclisation, dehydration and/or oxidation reactions Alternatively, the fragments can react with hydrogen sulfide and form very potent alicyclic flavour substances Route C (see also Section 3.3) involves the so-called Strecker degradation of amino acids, which is catalysed by dicarbonyl or hydroxycarbonyl compounds The reaction Amino acid + Sugar Amadori/Heyns rearrangement compounds - amino acid Fragmentation products Retro-aldol or α (β)-cleavage 1- and 3-Deoxyosones A - H2S reduction H2S, NH3, amino acid fragments Condensation products Amino acid Cyclisation products Strecker degradation - H2S reduction dehydration B H2S, NH Cyclisation products C Flavour substances Fig 3.1 Major pathways for the formation of flavour substances during Maillard reaction A, B and C denote the three key pathways Basic chemistry and process conditions for reaction flavours 53 Fig 3.2 Reaction of glucose (I) and glycine leading to the Amadori compound N -(1-deoxy-D-fructos-1yl)glycine (V, open-chain form) and related degradation reactions (adapted from Davidek et al., 2002) is a ‘decarboxylating transamination’ and the resulting Strecker aldehydes are potent flavour compounds Strecker aldehydes can also be formed directly from Amadori rearrangement products (ARPs) or Heyns rearrangement products (HRPs) A more detailed, but still simplified scheme is depicted in Fig 3.2 (Davidek et al., 2002) showing the formation of Amadori compounds, N-substituted 1-amino-1-deoxy-ketoses (V) representing an important class of Maillard intermediates (Ledl and Schleicher, 1990) They are formed in the initial phase of the Maillard reaction by Amadori rearrangement of the corresponding N-glycosylamines (II), the latter obtained by condensation of amino acids and aldoses such as glucose (I) as shown in pathway A The importance of Amadori compounds stems from the fact that their formation as well as decomposition can be initiated under mild conditions Thus, the formation of Amadori compounds represents a low-energy pathway of sugar degradation The chemistry of Amadori compounds has recently been reviewed (Yaylayan and Huyghues-Despointes, 1994) Degradation of the Amadori compound V by 1,2enolisation (pathway B) and 2,3-enolisation (pathway D) leads to the formation of 3-deoxy2-hexosulose (VII) and 1-deoxy-2,3-hexodiulose (XII), respectively, as already suggested by 54 Food Flavour Technology Hodge (1953) In parallel to these pathways, other ␣-dicarbonyls can be formed by enolisation For example, transition metal-catalysed oxidation of 1,2-enaminol IV can lead via pathway C to osones such as glucosone (IX) Pathway E gives rise to the 1-amino-1,4-dideoxy-2,3diulose (XIII) by elimination of the C4-OH group of the 2,3-endiol (VI) If the iminoketone (VIII) formed by oxidation of 1,2-enaminol (IV) is not hydrolysed, then Strecker aldehydes can be formed in the course of pathway C by direct oxidative degradation of the Amadori compound and decarboxylation of X, as proposed by Hofmann and Schieberle (2000a) The so-called carbon module labelling (CAMOLA) technique (Schieberle et al., 2003; Schieberle, 2005) has been introduced as an advanced tool to quantify the relative contribution of carbohydrate fragments (e.g C1–C4 fragments derived from Route B in Fig 3.1) in comparison with transient intermediates with intact carbon chain configuration (e.g deoxyosones; formed via Route A in Fig 3.1) in the generation of Maillard-derived aroma compounds The authors used a model system containing 1:1 mixtures of unlabelled and 13 C6 -labelled glucose as well as unlabelled proline and then measured the labelling pattern (12 C6 , 13 C3 and 13 C6 ) of the caramel-like odourant 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol R ) by GC-MS The study revealed that, under dry heating conditions, Furaneol was generated entirely via the intact sugar skeleton, whereas in aqueous solution, 63% of the furanone was derived from the recombination of two C3 -fragments This work can be considered as another milestone in the history of research on Maillard chemistry The CAMOLA technique has recently also been applied for studying the formation of furan in both model systems and foodstuffs (Limacher et al., 2008) 3.2.1 Intermediates as flavour precursors ARPs and HRPs are relatively stable intermediates and have been detected in various heatprocessed foods (Eichner et al., 1994) Since ARPs and HRPs can easily be synthesised (Van den Ouweland and Peer, 1970; Yaylayan and Sporns, 1987), their potential as flavour precursors has been evaluated in several studies Doornbos et al (1981), for example, found that the ARP derived from rhamnose and proline is a useful precursor for the generation of the potent caramel-like odourant 4-hydroxy-2,5-dimethyl-3(2H)-furanone ARPs and HRPs have also been reported to be good precursors for Strecker aldehydes and, in the absence of oxygen, also for 1- and 3-deoxyosones (Hofmann and Schieberle, 2000a) At higher pH values, ARPs and HRPs easily undergo cleavage of the carbohydrate chain, yielding fission products such as 2,3-butanedione and pyruvaldehyde (Weenen and Apeldoorn, 1996) When cysteine is heated with reducing sugars, thiazolidine carboxylic acids (TCAs) are formed instead of ARPs or HRPs (de Roos, 1992) TCAs are relatively stable in anionic form, which is probably the main reason for the inhibitory effect of cysteine in Maillard reactions, especially at higher pH This can also explain the more efficient formation of important meat sulfur compounds at low pH (Hofmann and Schieberle, 1998a) A research disclosure (Anonymous, 1979), however, describes the use of TCAs as precursors for meat flavours TCAs of glyceraldehyde, fructose or xylose were reacted as such or in conjunction with organic acids (e.g succinic acid, malic acid and citric acid) or fatty acids (e.g oleic acid and linoleic acid) Using a pH of 6–7 and temperatures between 50 and 100◦ C, the TCA of glyceraldehyde and cysteine was reported to result in a beef-like aroma, whereas TCAs of fructose or xylose and cysteine yielded ‘meaty/savoury’ flavours Other important Maillard reaction intermediates are the deoxyosones In general, 1-deoxyosones are more important flavour precursors than 3-deoxyosones, whose formation is favoured under neutral/slightly basic and acidic pH, respectively Although 1-deoxyglucosone has been synthesised by Ishizu et al (1967), 1-deoxyosones are too Basic chemistry and process conditions for reaction flavours 55 unstable to be used as precursors 3-Deoxyosones, however, are more stable and are easily obtainable from compounds such as difructoseglycine (Anet, 1960) The structure and reactivity of various 3-deoxyosones have been extensively studied by Weenen and Tjan (1992, 1994) and Weenen et al (1998) By using various H NMR and 13 C NMR techniques, the authors showed that 3-deoxypentosone and 3-deoxyglucosone consist almost exclusively of monocyclic and bicyclic (hemi)acetal/(hemi)ketal structures 3-Deoxypentosones and 3deoxyhexosones are good precursors for furfural and (5-hydroxymethyl)furfural under acidic conditions Under basic conditions, they undergo cleavage of the carbohydrate chain and can form pyrazines in the presence of an N-source Hofmann and Schieberle (2000b) showed that acetylformoin, which is formed from 1deoxyhexosone, is an effective precursor for 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) The amounts of Furaneol obtained from acetylformoin were significantly enhanced in the presence of reductones such as ascorbic acid or methylene reductinic acid as well as the Strecker-active amino acid proline The reaction between acetylformoin and proline also resulted in high amounts of the cracker-like odourant 6-acetyltetrahydropyridine (ACTP) A number of articles and patents report the production of meat-like aromas by reacting 4-hydroxy-5-methyl-3(2H)-furanone (norfuraneol) with hydrogen sulfide or cysteine Van den Ouweland and Peer (1968) were first to file a patent on the use of this precursor system to prepare 3-mercaptomethylfurans, which exhibit meat-like character Later, several studies identified sulfur-containing aroma compounds derived from the reaction of norfuraneol and hydrogen sulfide (e.g Van den Ouweland and Peer, 1975; Whitfield and Mottram, 1999) The latter authors showed that this precursor system is capable of producing compounds such as 2-methyl-3-furanthiol (MFT) and 2/3-mercapto-3/2-pentanone in relatively high amounts These thiols are also key aroma compounds of heated meat (Kerscher and Grosch, 1998) Shu and Ho (1989) and Zheng et al (1997) studied the reaction of the methyl homologue 4hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) with hydrogen sulfide or cysteine, which also gave rise to meat-like aromas Among the identified sulfur-containing volatiles, however, the methyl homologue of MFT (2,5-dimethyl-3-furanthiol) was not found in the reaction mixtures Unilever patented processes for the preparation of savoury flavours using the precursor systems and reaction conditions shown in Fig 3.3 (Turksma, 1993; Rosing and Turksma, 1997) When 2,5-dimethyl-2-(2-hydroxy-3-oxo-2-butyl)-3(2H)-furanone (diacetyloligomer; R and R = CH3 ; process A in Fig 3.3), which can be obtained by heating 2,3-butanedione under acidic conditions (Doornbos et al., 1991), is reacted with cysteine and hydrogen sulfide, high amounts of 2,5-dimethyl-3-furanthiol (DMFT) are generated (Turksma, 1993) For example, almost 40% yield of DMFT was obtained after hour at 120◦ C using a polar organic solvent, acidic conditions and super-atmospheric pressure (100–2500 kPa) DMFT, which was found to have a ‘meaty taste and roasted meat aroma’ was also formed from 2,5-dimethyl-3(2H)-furanone Using similar conditions as for the diacetyloligomer, Rosing and Turksma (1997) reacted 4-hydroxy-2,5-dimethyl-(2-hydroxy-3-oxo-2-butyl)-3(2H)furanone (Fig 3.3; R1 − R3 = CH3 , R4 = acetyl; process B) with cysteine and hydrogen sulfide This process resulted in a flavouring with sweet, onion-like, meaty aroma, odours attributed to high amounts of 2,5-dimethyl-4-mercapto-3(2H)-furanone and 2,5-dimethyl4-mercapto-3(2H)-thiophenone However, these two sulfur compounds were not found to significantly contribute to the flavour of heated meat Mottram et al (1998) filed a patent on flavouring agents that serve as precursors for generating cooked (e.g cooked meat) flavours in foodstuffs in situ The authors claim 56 Food Flavour Technology Precursor O A R Main odorant SH R' Cysteine + H2S O O OH R, R’ = CH3 or H HO B R1 R O HO O R2 Glycerol 90–120°C 1–3 hours, 100–2500 kPa SH Cysteine + H2S OH O R3 R4 Glycerol 90–120°C 1-3 hours, 100–2500 kPa R1 X R2 R1 = CH3 , C2H5 or H R2 = alkyl, C1-4 or H R3 = alkyl, C1-5 R4 = organic radical, C1-6, H or O-atoms X = O or S Fig 3.3 Precursors, reaction conditions and main aroma compounds of two savoury process flavours (according to Turksma, 1993 (A), and Rosing and Turksma, 1997 (B)) a long list of precursor substances that are capable of developing flavour during microwave cooking or conventional oven cooking with reduced cooking times This list comprises several sulfur compounds such as hydrogen/ammonium/sodium sulfides, cysteine, thiamine, onion and garlic as well as ‘non-sulfur-containing post-rearrangement Maillard products such as furanones (e.g 4-hydroxy-5-methyl-3(2H)-furanone, 4-hydroxy-2,5dimethyl-3(2H)-furanone, 2-methyl-4,5-dihydro-3(2H)-furanone), pyranones (e.g maltol, 5-hydroxy-5,6-dihydromaltol), 3-deoxyglucosone, ketones (e.g cyclotene) and aldehydes The precursor mixtures were encapsulated or spray-dried and applied to the foodstuffs through dusting or inclusion prior to the heat treatment In two other studies, a similar precursor system consisting of unsaturated aldehydes and hydrogen sulfide resulted in aroma blocks with deep-fried notes (Van den Ouweland, 1989; Zhang and Ho, 1989) As shown in this chapter, meat-like flavours can be obtained from the reaction of intermediate precursor systems such as 4-hydroxy-5-methyl-3(2H)-furanone (norfuraneol) and cysteine Cerny and Davidek (2003) investigated the efficiency of this precursor system in the generation of the meat-like odourants 2-methyl-3-furanthiol (MFT) and 3-mercapto2-pentanone (MP) relative to their formation from ribose/cysteine Reacting 13 C5 -labelled ribose together with unlabelled norfuraneol and cysteine under aqueous conditions (pH 5, 95◦ C for hours), the authors showed that mainly 13 C5 -labelled MFT was formed, suggesting that norfuraneol is less important as intermediate of MFT In contrast, MP was found unlabelled and hence originated from norfuraneol On the basis of these results, as well as additional mechanistic studies using the CAMOLA technique (heating of ribose and [13 C5 ]ribose (1 + 1) with cysteine under above-mentioned conditions), a new reaction pathway for the formation of MFT and MP from ribose via 1,4-dideoxyosone was proposed (Fig 3.4) Basic chemistry and process conditions for reaction flavours 57 Ribose O OH O OH OH O OH OH O O O 1,4-Dideoxyosone + H2S + H2S O SH OH O OH OH OH SH O O - H2O - H2O + H2S O SH OH SH OH OH SH O O - H2O SH - H 2O + Cys - H2O O O 2-Methyl-3-furanthiol SH HO N O OH SH - CO2 - H2O SH 3-Mercapto-2-pentanone Red SH NH2 N SH SH O NH + H 2O + H2O SH - NH3 SH SH O Fig 3.4 Proposed formation pathway for 3-mercapto-2-pentanone and 2-methyl-3-furanthiol from ribose and cysteine via the 1,4-dideoxyosone route (adapted from Cerny and Davidek, 2003) 58 Food Flavour Technology 3.2.2 Carbohydrate fragmentation Carbohydrate fragments have been found to originate from deoxyosones, ARPs or HRPs, as well as from the sugar directly (Ledl and Schleicher, 1990; Weenen, 1998) (Fig 3.1) The extent of the sugar cleavage reactions depends on the pH and on the reaction medium, with fragmentation favoured at higher pH values (pH ≥7) and in aqueous systems Using isotopic labelling of the sugar molecule in conjunction with GC-MS analysis, C5 /C1 , C4 /C2 and C3 /C3 fission reactions were established (review by Tressl et al., 1995) The proposed cleavage routes involve retro-aldolisation, vinylogous retro-aldolisation, ␣- and ␤-dicarbonyl cleavage (reviewed by Weenen, 1998; Tressl and Rewicki, 1999) Retro-aldolisation is by far the most accepted fragmentation route Studying the generation of acetic acid (which is a major carbohydrate fragmentation product and also a good marker for the 2,3-enolisation pathway, as exclusively formed from 1-deoxy-2,3-diuloses) during Maillard reaction of glucose and glycine under aqueous conditions (90–120◦ C, pH 6–8), Davidek et al (2006a) revealed that the organic acid is mainly formed through a hydrolytic ␤-dicarbonyl cleavage pathway (Fig 3.5) The authors also evidenced that ␤-dicarbonyl cleavage, which can be seen as an acyloin cleavage or a reverse Claisen-type reaction, represents a general cleavage route for diacylcarbinol intermediates of the Maillard reaction under aqueous conditions and that the frequently reported hydrolytic ␣-dicarbonyl cleavage can be ruled out as a sugar fragmentation mechanism Furthermore, a new sugar fragmentation pathway has been suggested to occur under oxidative conditions (Davidek et al., 2006b) by oxidative ␣-dicarbonyl cleavage with a Baeyer–Villiger-type rearrangement as key steps The sugar degradation pathway will mainly depend on the reaction conditions (Fig 3.6) Carbohydrate cleavage products have been analysed by several authors (Nedvidek et al., 1992; Weenen and Apeldoorn, 1996; Hofmann, 1999) Since these intermediates are reactive dicarbonyl and hydroxycarbonyl compounds, trapping agents such as 1,2-diaminobenzene and ethoxamine hydrochloride were used to transform them into stable quinoxaline and O-ethyloxime derivatives, respectively Weenen and Apeldoorn (1996) studied the formation of glyoxal, methylglyoxal, 2,3-butanedione and 2,3-pentanedione in both Maillard and caramelisation reactions The results are shown in Table 3.1 The study revealed that sugar fragmentation is highest in the presence of a Strecker-inactive amine functionality (cyclohexylamine), followed by a Strecker-active amino acid (alanine) Without amine (caramelisation reaction), the extent of fragmentation was even lower and no detectable amounts of 2,3-butanedione and 2,3-pentanedione were observed The yields of the pentanedione were relatively high in alanine model systems, indicating that the Strecker aldehyde, acetaldehyde, is involved in its formation In addition, the ARP of glucose and alanine was found to be an efficient ␣-dicarbonyl precursor, whereas the 3-deoxyglucosone/alanine reaction mixture yielded only low concentrations of fission products (Table 3.1) The latter result is in good agreement with the finding that 3-deoxyglucosone is also a poor pyrazine precursor (Weenen and Tjan, 1994) Hofmann (1999) studied the time course of the formation of carbohydrate degradation products in thermally treated solutions of either xylose or glucose with alanine The author showed that during the first 10 minutes of the Maillard reaction, glyoxal is the most abundant fragmentation product from both xylose and glucose Its formation can be explained by retroaldol cleavage of 2-xylosulose or 2-glucosulose After 20 minutes of the Maillard reaction, the main fission products were found to be methyl glyoxal and hydroxy-2-propanone, with xylose yielding higher amounts than glucose One year later, Yaylayan and Keyhani (2000) H C OH CHO + H CH3 CH3COOH + H C OH H C OH CH2OH CHO 1-Deoxy-2,3-tetrodiulose CH2OH C O CH3 C O C O H C OH CHO CH3 C O H C OH + H + H C OH CH2OH CH3COOH + H C OH CHO 2-Hydroxy-3-oxobutanal OH B H Acetylformoin OH C O CH C OH C O CH3 CH2OH CHO CH3COOH + H C OH CH3 C O H C OH OH CH3 C O CHO H C OH 1 H C OH H C OH + CH3COOH 2-Hydroxy-3-oxobutanal C Acetylformoin OH + C O H C O CH Fig 3.5 Formation of acetic acid from glucose via 1-deoxy-2,3-hexodiulose as the key intermediate by the hydrolytic ␤-dicarbonyl cleavage mechanism, indicating the various possible degradation pathways (adapted from Davidek et al., 2006a) 2-Hydroxy-3-oxobutanal OH C O Erythrose Tetrulose CH2OH C O CH2OH CH2OH CH3 C O H C OH H C OH H C OH C OH CH3 C OH H C OH C=O + H C OH CHO H C OH H2O CH3COOH CH OH + CH3COOH A H + 1-Deoxy-2,4-hexodiulose CH2OH CH2OH 1-Deoxy-2,3-hexodiulose C O H C OH H C OH H C OH H C OH OH C O CH C O C O CH Basic chemistry and process conditions for reaction flavours 59 60 Food Flavour Technology CH3 COOH C O C O A1 H C OH + H C OH H C OH CH2OH H C OH CH3 CH2OH Erythronic acid COOH 1-Deoxy-2,3-hexodiulose Acetic acid CH2OH + CH3 B1 C O H C OH CH2OH C O Tetrulose H C OH C O H C OH CH2OH CH3 + B2 C O CH2OH 1-Deoxy-2,4-hexodiulose Acetol COOH H C OH CH2OH CH3 Glyceric acid H C OH C O C O H C OH CH2OH A2 CH3 + H C OH COOH Lactic acid 1-Deoxy-3,4-hexodiulose Fig 3.6 Scheme summarising possible pathways of sugar fragmentation via 1-deoxyhexodiuloses A1 and A2, oxidative ␣-dicarbonyl cleavage; B1 and B2, hydrolytic ␤-dicarbonyl cleavage by nucleophilic attack of OH− at the C-2 and C-4 carbonyl group, respectively (adapted from Davidek et al., 2006b) performed a mechanistic study to determine the origin of Maillard intermediates such as glycolaldehyde, methylglyoxal, hydroxy-2-propanone and 3-hydroxy-2-butanone 2,3-Butanedione (diacetyl) and 2,3-pentanedione are aroma-active carbohydrate cleavage products, contributing to a sweet-caramel odour of coffee (Grosch, 2001), and, for example, in the presence of hydrogen sulfide or cysteine, they are precursors of important sulfur aroma compounds such as 2-mercapto-3-butanone and 2/3-mercapto-3/2-pentanone (Hofmann, 1995) Yaylayan and Keyhani (1999) investigated the origin of these two dicarbonyl compounds in glucose/alanine Maillard model systems under pyrolytic conditions Using labelled glucose or alanine, the authors showed that 90% of the formed pentanedione requires the participation of the C2/C3 atoms of alanine, whereas diacetyl was derived from the sugar chain only This result is supported by the finding of Hofmann (1995) that Basic chemistry and process conditions for reaction flavours Table 3.1 61 Formation of ␣-dicarbonyl products (according to Weenen and Apeldoorn, 1996) α-Dicarbonyl products (µg) Amine Carbohydrate Glyoxal Methyl glyoxal 2,3-Butanedione 2,3-Pentanedione No No No No No Alanine Alanine Alanine Alanine Alanine Cyclohexylamine Cyclohexylamine Cyclohexylamine Cyclohexylamine Cyclohexylamine Glucose Fructose Xylose 3-Deoxyglucosone Fru-Alaa Glucose Fructose Xylose 3-Deoxyglucosone Fru-Alaa Glucose Fructose Xylose 3-Deoxyglucosone Fru-Alaa 26 28 62 23 103 58 45 27 16 81 618 691 591 317 509 11 15 17 57 101 43 28 81 56 67 865 1104 925 583 454 – – – – – – – – 98 41 22 28 11 81 227 265 614 146 232 18 38 25 42 21 22 39 89 101 25 39 N -1-(deoxy-d-fructosyl)-l-alanine (Amadori rearrangement product of glucose and alanine) 2,3-pentanedione is formed by reacting acetaldehyde and hydroxy-2-propanone Schieberle et al (2003) applied the CAMOLA technique to a Maillard model system of glucose (13 C6 + 12 C6 = + 1) and proline and showed that 2,3-butanedione was formed by 87 and 13% through the recombination of C3 + C1 and C2 + C2 fragments (Route B in Fig 3.1), respectively As a result, no diacetyl is generated from the intact carbon chain (Route A in Fig 3.1) 3.2.3 Strecker degradation In the presence of ␣- or vinylogous dicarbonyl compounds, ␣-amino acids can undergo a ‘decarboxylating transamination’, which results in the formation of aldehydes with one carbon atom less than the amino acid (Strecker aldehydes) (Schăonberg and Moubacher, 1952) The Strecker degradation of amino acids is a key reaction in the generation of potent aroma compounds during Maillard-type processes (Ledl and Schleicher, 1990) (see also Fig 3.1) Certain amino acids (leucine, valine, methionine or phenylalanine) are known to produce Strecker aldehydes with significant odour strength such as 3-methylbutanal, methylpropanal, methional or phenylacetaldehyde These aldehydes have been confirmed as key contributors to many thermally processed foods (Hofmann et al., 2000) Besides aldehyde formation, Strecker degradation also contributes to flavour formation during Maillard reaction by reducing dicarbonyls to hydroxycarbonyls (e.g formation of 1,4-dideoxyosone from 1deoxyosone) (Nedvidek et al., 1992) or by generating ␣-aminocarbonyl compounds, which are pyrazine precursors (Weenen and Tjan, 1994) Weenen and van der Ven (1999) studied the formation of phenylacetaldehyde in Maillard model systems, including reactions of phenylalanine with various sugars, ␣-dicarbonyl and hydroxycarbonyl compounds as well as ARPs The authors found that methyl glyoxal was the most efficient dicarbonyl compound for the formation of phenylacetaldehyde, followed by 3-deoxyerythrosone, glyoxal, 3-deoxyxylosone and 3-deoxyglucosone Hydroxycarbonyl compounds such as dihydroxyacetone and glyceraldehyde also yielded high amounts of the 62 Food Flavour Technology HO HO OH OH O2,Me2+ COOH COOH OH N H HO H O O + HO H+ OH H Me + COOH OH N HO OH N HO Me2+,H2O2 O O O OH OH HO O OH HO OH H 2O N N OH COOH CO2,H2O O O OH HO H OH O NH2 Fig 3.7 Formation of phenylacetaldehyde via an oxidative degradation of N -(1-deoxy-D-fructosyl)-Lphenylalanine (Fru-Phe) (adapted from Hofmann and Schieberle, 2000a) Strecker aldehyde Sugars were less reactive, with the reactivity decreasing in the order erythrose, xylose, fructose and glucose In addition, the authors showed that the Amadori compound Fru-Phe is a superior phenylacetaldehyde precursor to the corresponding sugar/amino acid mixture This finding was also confirmed by Hofmann and Schieberle (2000a) In addition, Hofmann and Schieberle (2000a) revealed that the yields of phenylacetaldehyde formed from the Amadori compound Fru-Phe were significantly increased in the presence of oxygen and copper (II) ions On the basis of the observation that 1,2-hexodiulose is also generated in high amounts under these conditions, they proposed a mechanism for the formation of phenylacetaldehyde from Fru-Phe (Fig 3.7) Hofmann et al (2000) additionally found that in the reaction of phenylalanine and glucose, considerable amounts of phenylacetic acid were generated While the formation of phenylacetaldehyde showed an optimum pH of and was not influenced by oxygen, the acid was most abundant at pH in the presence of oxygen and copper (II) ions The authors proposed a mechanism for the generation of phenylacetic acid that involves a similar oxidation step to that shown in Fig 3.7 Cremer and Eichner (2000) studied the influence of the pH on the formation of 3-methylbutanal during Maillard reaction of glucose and leucine They showed that the formation rate of the aldehyde was higher at pH than at pH or This was consistent with the high degradation rate of the Amadori compound Fru-Leu at pH 7, suggesting that the ARP was a good precursor for 3-methylbutanal However, Chan and Reineccius (1994) showed that the optimal reaction conditions of different Strecker aldehydes vary, owing to differences in stability of the aldehydes 3.2.4 Interactions with lipids In addition to the Maillard reaction, lipid oxidation is another major reaction occurring in process flavour production and food systems Both reaction cascades include a whole network of different reactions in which extraordinary complex mixtures of compounds are obtained, triggering important changes in food flavour, colour, texture and nutritional value, with desirable and undesirable consequences In addition, both reactions are intimately interrelated as shown in Fig 3.8 (Zamora and Hidalgo, 2005) and the products of each reaction influence the other Furthermore, there are common intermediates and products in both pathways The existing data suggest that both Maillard reaction and lipid peroxidation are so closely interrelated that they should be considered simultaneously to understand the Basic chemistry and process conditions for reaction flavours 63 Lipid Reducing sugar Amino compound N-Substituted glycosylamine Oxidation Amadori rearrangement Oxidation Lipid hydroperoxides 1-Amino-1deoxy-2-ketose Hydroperoxide decomposition Sugar dehydration Sugar fragmentation Sugar enolisation Strecker degradation Lipid peroxyl and hydroxyl radicals Cleavage Formation of stable compounds Polymerisation Ketones Alcohols Epoxides Dimers Polymers Glyoxal Methylglyoxal Others Aldehydes Ketones Alcohols Epoxides Hydrocarbons Acids Cyclic peroxides Hydroperoxy compounds Cleavage Taste Pyrrole polymerisation Antioxidants • Schiff base of HMF or furfural • Reductones • Fission products • Aldehydes • Others Hydroxyalkylpyrroles Aldol condensation Carbonyl-amine polymerisation Aroma Volatile and non-volatile monomers Carbonyl-amine polymerisation Coloured compounds Aldol condensation Carbonyl-amine polymerisation Neo-formed toxicants Fig 3.8 Known interactions between Maillard reaction and lipid oxidation pathways in nonenzymatic browning development (adapted from Zamora and Hidalgo, 2005) reaction mechanisms, kinetics, and products in the complex mixtures of carbohydrates, lipids and proteins occurring in food systems and process flavours In these systems, lipids and carbohydrates are competing in the chemical modification of amino compounds (e.g proteins and phospholipids) Therefore, although there are significant differences between Maillard reaction and lipid peroxidation, many aspects of both reactions can be better understood if they are included in only one general carbonyl pathway that can be initiated by both lipids and carbohydrates (Hidalgo and Zamora, 2005) As an example, 2-pentylpyridine has been reported (Henderson and Nawar, 1981) as an interaction product of linoleic acid and valine, with 2,4-decadienal and ammonia being the key intermediates (Fig 3.9) Its formation was studied by Kim et al (1996) in model systems Lipid COOH O Oxidation Linoleic acid NH2 R COOH (E,E)-2,4-Decadienal Maillard NH3 reaction Amino acid (asparagine, glutamine) Ammonia [O] N NH N H Fig 3.9 Formation of 2-pentylpyridine from 2,4-decadienal and an amino acid (adapted from Henderson and Nawar, 1981) 64 Food Flavour Technology by reacting 2,4-decadienal with amino acids (glycine, aspartic acid, asparagine, glutamic acid and glutamine) at 180◦ C for hour (pH 7.5) The relative yields of alkylpyridine formation from the reactions were asparagine Ͼ glutamine Ͼ aspartic acid Ͼ glutamic acid Ͼ glycine When amide-15 N-labelled glutamine and asparagine were heated with 2,4decadienal, the relative contribution of amide nitrogens to the formation of alkylpyridine was determined Approximately half the nitrogen atoms in 2-pentylpyridine formed from asparagine, originated from the amide nitrogens of asparagine, whereas when glutamine was the reactant, almost all the nitrogen atoms came from the amide nitrogens in glutamine The above-mentioned results may indicate that both free ammonia and ␣-amino groups bound in amino acids can contribute to the formation of alkylpyridines, but free ammonia does so more effectively The formation of Strecker aldehydes in the presence of lipid oxidation products is another example, illustrating the interactions between Maillard and lipid intermediates Strecker degradation of amino acids is one of the most important reactions leading to final aroma compounds in the Maillard reaction Hidalgo and Zamora (2004) have studied the reaction of 4,5-epoxy-2-alkenals with phenylalanine In addition to N-substituted 2(1-hydroxyalkyl)pyrroles and N-substituted pyrroles, which are major products of the reaction, the formation of both the Strecker aldehyde, phenylacetaldehyde, and 2-alkylpyridines was also observed The aldehyde was only produced from the free amino acid (not esterified), suggested to be produced through imine formation, which is then decarboxylated and hydrolysed (Fig 3.10) This reaction also produces a hydroxyl amino derivative, which is the origin of the 2-alkylpyridines These data indicate that Strecker-type degradation of amino acids occurs at low temperature by some lipid oxidation products This is a proof of the interrelations between lipid oxidation and Maillard reaction, which are able to produce common products by analogous mechanisms However, recent results suggest that, analogously to carbohydrates, certain lipid oxidation products may also degrade certain amino acids to R1 H O H2N O − H2O + O R1 OH N O O O H − CO2 O R1 NH2 H + + H2O R1 N OH OH 13 14 R1 N 15 Fig 3.10 Strecker-type degradation of phenylalanine produced by 4,5-epoxy-2-alkenals (adapted from Hidalgo and Zamora, 2004) Basic chemistry and process conditions for reaction flavours 65 undesirable compounds, e.g vinylogous derivatives such as styrene (Hidalgo and Zamora, 2007) Simulating food flavours by the process flavour approach requires precursors and recipes that are close to food or which well represent the composition of food Therefore, lipids are key ingredients in the reaction flavour system to obtain boiled chicken notes Similarly, polyphenols play an important role in generating cocoa flavour The role of these specific components (lipids, polyphenols, vitamins) may be (i) generating new specific flavour compounds (Whitfield, 1992) as compared to the pure Maillard system (e.g 2-pentylpyridine) or (ii) intervening in the Maillard reaction cascade and thus alerting the overall flavour composition The latter is probably of higher relevance 3.3 IMPORTANT AROMA COMPOUNDS DERIVED FROM MAILLARD REACTION IN FOOD AND PROCESS FLAVOURS In the flavour industry, process flavours are often developed by empirical means, i.e the reaction conditions are optimised using organoleptic evaluation However, this approach should be accompanied by analytical evaluation of the products Knowledge of the important aroma compounds derived from the Maillard reaction in food and process flavours is essential in order to focus investigations into reaction mechanisms enhancing the key aroma compounds In addition, the knowledge of their formation pathways and key intermediates allows the development of multi-step approaches (e.g two-step reactions), which aim at optimising reaction conditions of each of the flavour generation stages (e.g from sugar/amino acid mixture to deoxyosones, from deoxyosones to target flavour compounds) One of these steps (often the first step) can also be a biotransformation that yields an intermediate that, on thermal treatment, releases the key aroma compound An example for such an approach is the biogeneration of 2-(1-hydroxyethyl)-4,5-dihydrothiazole through yeast fermentation, which releases 2-acetylthiazoline on microwave heating (Bel Rhild et al., 2002) It is recommended that the evaluation of character-impact aroma compounds should be based on a combination of instrumental and sensorial analysis A well-established approach involves the determination of the odour activity values (ratio of concentration and odour threshold values) of aroma compounds (Grosch, 1994) This requires quantitative analysis of a selected number of aroma compounds, the importance of which has been screened by GColfactometry (GC-O) Suitable screening techniques such as aroma extract dilution analysis or headspace dilution analysis are available (Ullrich and Grosch, 1987; Guth and Grosch, 1993a, b) As these methods not consider interactions of different aroma compounds, they should be combined with organoleptic evaluations of reconstituted model mixtures Our selection of important Maillard-derived aroma compounds (see Sections 3.3.1 and 3.3.2) is primarily based on results of GC-O techniques or other sensory evaluations or on quantitative data 3.3.1 Character-impact compounds of thermally treated foods Character-impact compounds of thermally treated foods that are formed during Maillard reaction are summarised in Table 3.2 Their identification in foodstuffs such as meat, bread, 3-Mercapto-2-pentanone 2,5-Dimethyl-3-furanthiol 2-Methyl-3-(methylthio)furan 2-Methyl-3-(methyldithio)furan 2-Methyl-3-(methyltrithio)furan 2-Furfurylmethyl disulfide 2-Methyl-3-furylthioacetate 1-(2-Methyl-3-furylthio)-ethanethiol 4-Hydroxy-2,5-dimethyl-3(2H )thiophenone Methional 2-Acetyl-2-thiazoline 10 11 12 14 18 17 16 3-Hydroxy-4,5-dimethyl-2(5H )-furanone (sotolon) 2-Ethyl-4-hydroxy-5-methyl-3(2H )furanone 3-Hydroxy-4-methyl-5-ethyl-2(5H )furanone (abhexon) B: Compounds containing oxygen 15 4-Hydroxy-2,5-dimethyl-3(2H )-furanone 13 Meaty, roast beef Caramel-like, fruity 3-Mercapto-2-butanone Seasoning-like Caramel-like, sweet Seasoning-like Caramel-like, strawberry-like Roasty, popcorn-like, burnt Cooked potato-like Meaty, onion-like, coffee-like Roasty, brothy Cooked meat-like Cooked meat-like Meaty, thiamine-like Meaty, sweet, sulfury Sulfury, catty Sulfury, catty Meaty, sweet, sulfury Roasty, sulfury 2-Furfurylthiol Odour description Compound 7.5 (26) 1.15 (26) 0.3 (5) 10 (26) 1.0 (25) 0.2 (18) 0.05 (20) — — — — 0.004 (16) 0.05 (15) 0.018 (1) 0.7 (1) 3.0 (1) 0.01 (1) 0.007 (1) Odour threshold in water (␮g/kg)b Coffee (26), chocolate (17), cocoa (17) Beef (3, 9, 19), chicken (5), coffee (7), French fries (21), tea (29), chocolate (17), cocoa (17), yeast (40) Coffee (7) Beef (3, 9, 19), chicken (6), coffee (7), beer (27), popcorn (27), bread (24), chocolate (17), sesame (28), French fries (21), tea (29), yeast (40) Beef (3, 9), chicken (4, 5), sesame (10), fish (22) Beef (3, 9, 19), chicken (4, 5), pork (6), coffee (7), French fries (21), potato chip (18), fish (22, 23), bread (24), yeast (40) Yeast (20) Yeast (20) Yeast (20) Beef (12) Beef (12) Cocoa (17), chocolate (17), meat (41) Yeast (15) Chicken (4) Beef (3, 4, 13), chicken (3, 14), yeast (40) Beef (12) Meat (3–6, 9), yeast (40), coffee (7, 8), sesame (10), popcorn (11) Meat (2–6), yeast (20, 40), coffee (7, 8) Detected inb Important aroma compounds derived from the Maillard reaction in various thermally treated foods.a A: Compounds containing sulfur 2-Methyl-3-furanthiol No Table 3.2 66 Food Flavour Technology 2,3-Diethyl-5-methylpyrazine 2-Ethenyl-3,5-dimethylpyrazine 2-Ethenyl-3-ethyl-5-methyl-pyrazine 2-Acetylpyrazine 29 30 31 32 Roasty, sweet, nutty Earthy, roasty Earthy, roasty Earthy, roasty Earthy, roasty 62 (38) — — 0.09 (7) 0.4 (18) 0.16 (7) 1.6 (35) 0.1 (33) 30 (32) 15 (32) 25 (9) (18) 0.7 (23) 0.35 (5) Sesame (28), popcorn (39), bread (36) Coffee (7), French fries (21) Coffee (7) Beef (19, 25), chicken (14), coffee (7), French fries (21), cocoa (17), chocolate (17), sesame (10), popcorn (36), bread (37), yeast (40) Bread (24), French fries (21), cocoa (17), chocolate (17), popcorn (36) Beef (19, 25), chicken (14), coffee (7), sesame (10), bread (24), French fries (21), cocoa (17), chocolate (17), popcorn (36) Bread (24), popcorn (11) Bread (24), rice (34), popcorn (11), sesame (10), beef (2, 3), French fries (21), yeast (40) Fish (22, 23), coffee (7), bread (24) Beef (3, 9), bread (24), coffee (7), fish (22, 23), Chocolate (17), cocoa (17), yeast (40) Beef (9), chicken (5), coffee (8), French fries (21), fish (31) Beef (2, 19), cocoa (17), chocolate (17), bread (24), French fries (21), coffee (7), tea (29), yeast (40) Beef (19), chicken (5), bread (24), chocolate (17), coffee (7), French fries (20), tea (29) Meat (3, 5), yeast (40), chocolate (17), cocoa (17), coffee (7), French fries (21), bread (24), tea (29), beer (30) b The sensory significance of the aroma compounds was assessed by quantitative data or by GC-O techniques 1, Hofmann (1995); 2, Gasser and Grosch (1988); 3, Kerscher and Grosch (1997); 4, Gasser and Grosch (1990); 5, Kerler and Grosch (1997); 6, Gasser and Grosch (1991); 7, Grosch (2001); 8, Semmelroch and Grosch (1995); 9, Guth and Grosch (1994); 10, Schieberle (1993a); 11, Schieberle (1991a); 12, Mottram and Madruga (1994); 13, Guth and Grosch (1993a); 14, Kerler (1996); 15, MacLeod and Ames (1986); 16, Schieberle et al (2000); 17, Schnermann and Schieberle (1997); 18, Guadagni et al (1972); 19, Kerler and Grosch (1996); 20, Werkhoff et al (1991); 21, Wagner and Grosch (1997); 22, Milo and Grosch (1993); 23, Milo and Grosch (1996); 24, Rychlik and Grosch (1996); 25, Cerny and Grosch (1993); 26, Semmelroch et al (1995); 27, Schieberle (1993b); 28, Schieberle (1996); 29, Guth and Grosch (1993b); 30, Schieberle (1991b); 31, Milo and Grosch (1995); 32, Blank et al (1992); 33, Buttery et al (1983); 34, Buttery et al (1982); 35, Buttery and Ling (1995); 36, Schieberle and Grosch (1987); 37, Schieberle and Grosch (1994); 38, Teranishi et al (1975); 39, Schieberle (1995); 40, Măunch and Schieberle (1998); 41, Madruga and Mottram (1995) a 2-Ethyl-3,6-dimethylpyrazine 28 Earthy, roasty 2-Ethyl-3,5-dimethylpyrazine Buttery, green 27 2,3-Pentanedione 24 Buttery Roasty, cracker-like 2,3-Butanedione 23 Solvent-like 6-Acetyltetrahydropyridine Acetaldehyde 22 Honey-like, sweet, flowery 26 Phenylacetaldehyde 21 Malty, fruity, pungent Roasty, popcorn, bread-like Methylpropanal 20 Malty, cocoa-like C: Compounds containing nitrogen 25 2-Acetyl-1-pyrroline 2/3-Methylbutanal 19 Basic chemistry and process conditions for reaction flavours 67 68 Food Flavour Technology coffee, cocoa, chocolate, sesame, popcorn, French fries, tea, fish and yeast, as well as odour description and odour thresholds of the aroma substances, is covered In the following discussion of Table 3.2, emphasis is given to important aroma compounds of meat, yeast, coffee and bread The meaty character of boiled beef, pork or chicken is mainly due to sulfur compounds such as MFT, 2-furfurylthiol, 3-mercapto-2-butanone, 2/3-mercapto-3/2-pentanone, DMFT, methanethiol, hydrogen sulfide and methional (Gasser and Grosch, 1988, 1990; Mottram and Madruga, 1994; Kerscher, 2000) The precursors of these aroma substances in meat are known to be free or bound C5-sugars such as ribose, ribose phosphate and inosine monophosphate as well as sulfur-containing compounds such as thiamine, cysteine, glutathione and methionine After the publication of the patent of Morton et al (1960), in which meat aroma formation established from the reaction of cysteine and ribose was described, a great number of patents and publication on meat-like process flavours followed (see Section 3.4.3) It is well known that species-specific differences in the aroma of cooked meats such as beef and chicken are mainly due to concentration and composition differences in lipidderived flavour substances Kerscher and Grosch (1998) and Kerscher (2000) confirmed these findings and showed that significant differences also exist for Maillard-derived aroma compounds Cooked beef, for example, was found to contain higher amounts of the sulfur compounds MFT and 2-furfurylthiol as well as the caramel-like 4-hydroxy-2,5-dimethyl3(2H)-furanone, whereas MP and methional were more important in cooked chicken The character-impact compounds of yeast extracts were found to be very similar to those of cooked meat, which is due to similar pools of Maillard precursors Măunch and Schieberle (1998) reported high odour activity values for the sulfur compounds MFT, 2-furfurylthiol, MP and methional, the Strecker aldehydes 3-methylbutanal and phenylacetaldehyde, as well as the furanones 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) and 3-hydroxy-4,5dimethyl-2(5H)-furanone (sotolon) Werkhoff et al (1991) additionally identified 2-methyl3-(methylthio)furan, 2-methyl-3-furylthioacetate, 1-(2-methyl-3-furylthio)ethanethiol and 4hydroxy-2,5-dimethyl-3(2H)-thiophenone (thiofuraneol) in yeast The authors claimed that these sulfur compounds contribute significantly to the meaty character of yeast Besides yeast, onion extracts are also used in process flavourings Widder et al (2000) have identified 3-mercapto-2-methylpentan-1-ol as new powerful aroma compounds in both process flavourings (containing onion extract) and raw onions This odourant that exhibits a pleasant meat broth, sweetish, onion and leek-like character (at a low concentration of 0.5 ppb in water), is suggested to be formed by aldol condensation of two molecules of propanal, followed by hydrogen sulfide addition at the double bond to yield 3-mercapto-2-methylpentanal The aldehydes are finally reduced enzymatically to the corresponding alcohol The Maillard reaction is also a key reaction in flavour formation during roasting of coffee The precursor pool in green coffee comprises a complex mixture of various soluble sugars such as glucose, fructose, galactose and sucrose In addition, the amount of polymeric arabinose and rhamnose was found to decrease during roasting, which indicates that these sugars are also involved in caramelisation and Maillard processes (Tressl, 1989) The total amino acid content drops by about 30% during roasting Especially, the amino acids – lysine, serine, threonine, arginine, histidine, methionine and cystine – are degraded to a high extent during the roasting process (Belitz et al., 2008) Semmelroch and Grosch (1995) reported the simulation of the aroma of Arabica and Robusta coffee brews using reconstituted mixtures of 23 aroma compounds The authors showed that the Maillard products – 2furfurylthiol, Furaneol, sotolon, methanethiol, 2,3-butanedione, 2,3-pentanedione, 2-ethyl3,5-dimethylpyrazine (EDMP), 2,3-diethyl-5-methylpyrazine (DEMP), methylpropanal and Basic chemistry and process conditions for reaction flavours OH 69 O O HO + N H Alapyridaine COO O HO O N O N + 5-MPC OH COOH OH O N H COOH O N O HO OH OH Dru-Glu 1-Oxo-2,3-dihydro-1Hindolizinium-6-olate 5-MPF Fig 3.11 Chemical structures of some recently identified taste molecules formed by Maillard-type reactions Fru-Glu, N -(1-deoxy-D-fructos-1-yl)-L-glutamic acid; 5-MPC, 5-methyl-2-(1-pyrrolidinyl)-2-cyclopenten-1one; MPF, and 5-methyl-4-(1-pyrrolidinyl)-3(2H )-furanone 3-methylbutanal – contribute to the flavour of coffee brews They also investigated the aroma differences between Arabica and Robusta coffee The more earthy/roasty and less caramel character of Robusta was found to be due to higher concentrations of the pyrazines EDPM and DEMP as well as to lower amounts of Furaneol and sotolon, respectively The flavour of cereal products, especially of bread, has been studied extensively, and the results have been reviewed (Grosch and Schieberle, 1997) 2-Acetyl-1-pyrroline (ACP) and 6-acetyltetrahydropyridine (ACTP) are responsible for the pleasant roasty character of wheat bread crust and popcorn However, these compounds are not important in wheat bread crumb and rye bread Both ACP and ACTP are generated by a reaction of proline with reducing sugars or sugar breakdown products (Schieberle, 1990) When ornithine instead of proline was reacted, only ACP was formed The fact that yeast contains relatively high amounts of ornithine explains why ACP concentrations in bread are strongly dependent on the amount of yeast used in the baking process (Grosch and Schieberle, 1997) Other important Maillardderived aroma compounds of wheat and rye bread are Furaneol and the Strecker aldehydes methional, 3-methylbutanal and methylpropanal They contribute to the caramel-like and malty aroma of bread Another breakthrough has been the identification of new taste-active molecules and taste modifiers as result of Maillard-type reactions (Fig 3.11), which contribute to the overall flavour perception (Hofmann, 2005) As an example, Alapyridaine has been identified in Maillard model systems containing hexose sugars and alanine as well as in beef broth as an essential compound enhancing the sweet taste and umami character in beef broth (Ottinger and Hofmann, 2003; Soldo et al., 2003) Glycoconjugates of glutamic acid, namely the N-glycoside dipotassium N-(d-glucos-1-yl)-l-glutamate and the corresponding Amadori compound N-(1-deoxy-d-fructos-1-yl)-l-glutamic acid (Fru-Glu), were found by systematic sensory studies to exhibit a pronounced umami-like taste, with recognition taste thresholds of 1–2 mmol/L, close to that of monosodium glutamate (MSG) (Beksan et al., 2003) Contrary to MSG, they not show the sweetish and slightly soapy by-note, but evoke an intense umami, seasoning, and bouillon-like taste Added to a bouillon base, which did not contain any taste enhancers, both glycoconjugates imparted a distinct umami character similar to the control sample containing the same amount of MSG on a molar basis (Schlichtherle-Cerny et al., 2002) The Amadori compound Fru-Glu has been reported in dried tomatoes as an example (Eichner et al., 1994) Furthermore, thermal treatment of aqueous solutions of xylose, 70 Food Flavour Technology rhamnose and l-alanine led to a rapid development of a bitter taste of the reaction mixture (Frank et al., 2003) Liquid chromatography/mass spectrometry (LC/MS) and nuclear magnetic resonance (NMR) spectroscopy revealed 1-oxo-2,3-dihydro-1H-indolizinium-6olates as the key compounds and most significant contributors to the intense bitter taste of this process flavour mixture Thermally treated glucose/L-proline mixtures that contain ‘cooling’ compounds were recently reported These Maillard systems generate 5methyl-2-(1-pyrrolidinyl)-2-cyclopenten-1-one (5-MPC) and 5-methyl-4-(1-pyrrolidinyl)3(2H)-furanone (MPF) as key compounds contributing to the cooling sensation without imparting aroma notes (Hofmann et al., 2001; Ottinger et al., 2001a) They have also been found in dark malt (10–100 ␮g/kg) (Ottinger et al., 2001b) 3.3.2 Character-impact compounds of process flavours Meat-like process flavours are often prepared by reacting cysteine and/or thiamine with sugars, although pentoses such as xylose and ribose are preferably used There is a range of additional precursor sources such as pectin hydrolysates (C-source), hydrolysed vegetable proteins and wheat protein hydrolysates (N-sources) as well as hydrogen sulfide, inorganic sulfides, onion, garlic and cabbage (S-sources) The products serve as building blocks in the creation of meat flavours (see also Section 3.4.3) They can also be described as middle notes that are combined with base notes (mainly taste compounds and taste enhancers) and top notes (mainly compounded flavourings) (savoury flavour pyramid according to Yeretzian et al., 2007) The important sulfur-containing compounds in process flavourings derived from either cysteine/ribose or thiamine reaction systems are quite similar (Table 3.3) The aroma of both cysteine- and thiamine-based process flavours is determined by MFT, 3-mercapto-2-butanone, 2/3-mercapto-3/2-pentanone, 2-methyl-3-thiophenethiol and 2-thenylthiol (Găuntert et al., 1992, 1996; Hofmann and Schieberle, 1995, 1997) The same compounds are also responsible for the meaty character of process flavours that are based on -inosine monophosphate (5 -IMP) and cysteine (Zhang and Ho, 1991; Madruga and Mottram, 1998) Although -IMP is more abundant than ribose in raw meat, it is a much poorer precursor for these sulfur compounds than ribose, when reacted with cysteine (Mottram and Nobrega, 1998) Other character-impact compounds that are formed primarily in thiamine or thiamine/cysteine reaction systems are 2-methyl-4,5-dihydro-3-furanthiol, 1(methylthio)ethanethiol, mercaptoacetaldehyde and 2-methyl-1,3-dithiolane (Găuntert et al., 1996) Many of the thiols mentioned above are also important aroma substances in glucose/cysteine or rhamnose/cysteine process flavourings (Hofmann and Schieberle, 1997) However, both reaction systems contain other characteristic aroma compounds For example, 2-(1-mercaptoethyl)furan and its thiophene derivative are only formed from glucose and cysteine, whereas 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) and 3-hydroxy-6-methyl2(2H)-pyranone belong to key odourants of the rhamnose system The latter two substances are responsible for the strong caramel and seasoning-like character of rhamnose/cysteine process blocks, which render them very suitable for application in beef flavours Another compound having seasoning-like character is 3-hydroxy-4,5-dimethyl-2(5H)-furanone (sotolon) Sotolon was found to contribute to the aroma of various cysteine-derived reaction flavours (Hofmann and Schieberle, 1995, 1997) In contrast to the other compounds mentioned above, the formation of sotolon is less influenced by the type of sugar 2-Thenylthiol 11 2-Methyl-3-thiophenethiol 2-Methyl-4,5-dihydro-3-furanthiol Methional Bis(2-methyl-3-furyl) disulfide 3/2-Mercapto-2/3-pentanone 10 3-Mercapto-2-butanone Mercaptoacetaldehyde Mercapto-2-propanone Sulfury, roasty Meaty, sulfury Meaty, sulfury Meaty, sulfury Cooked potato-like Sulfury, catty Sulfury, catty Sulfury, putrid Cabbage-like Sulfury, roasty, coffee-like 2-Furfurylthiol Meaty, sulfury, sweet A: Compounds containing sulfur 2-Methyl-3-furanthiol Odour description Character-impact compounds of process flavourings.a Compound No Table 3.3 0.042 (1) 0.00002 (10) — 0.02 (1) 0.2 (8) 0.7 (1) 3.0 (1) — — 0.01 (1) 0.007 (1) Odour threshold (␮g/kg) in waterb (Continued ) Cysteine/ribose (2, 6), cysteine/glucose (4), thiamine/cysteine (5), cysteine/IMP (6, 7), cysteine/ribose-5-P (6) Cysteine/ribose (2, 3, 6), glutathione/ribose (3), thiamine (5), cysteine/IMP (6), cysteine/ribose-5-P (6) Thiamine (3, 5), thiamine/cysteine (5) Cysteine/ribose (2, 6), cysteine/IMP (6), cysteine/ribose-5-P (6), thiamine/methionine (5) Methionine/ribose (9), thiamine/methionine (5) Cysteine/ribose (2, 3, 6), cysteine/glucose (4), cysteine/rhamnose (4), glutathione/ribose (3), thiamine (3, 5), thiamine/cysteine (5), cysteine/IMP (6), cysteine/ribose-5-P (6) Cysteine/ribose (2, 6), cysteine/glucose (4), cysteine/rhamnose (4), thiamine/cysteine (5), cysteine/IMP (6, 7), cysteine/ribose-5-P (6) Cysteine/IMP (7) Thiamine/cysteine (5) Cysteine/ribose (2, 3, 6), cysteine/glucose (4), cysteine/rhamnose (4), glutathione/ribose (3), thiamine (3), cysteine/IMP (6, 7), cysteine/ribose-5-P (6) Cysteine/ribose (2, 3, 6), cysteine/glucose (4), cysteine/rhamnose (4), glutathione/ribose (3), thiamine (3, 5), thiamine/cysteine (5), cysteine/IMP (6, 7), cysteine/ribose-5-P (6) Detected inb Basic chemistry and process conditions for reaction flavours 71 Roasty, popcorn-like Caramel-like, sweet, meaty 1-(Methylthio)ethanethiol 2-Methyl-1,3-dithiolane Hydrogen sulfide Methanethiol Ethanethiol 2-Acetyl-2-thiazoline 5-Acetyl-2,3-dihydro-1,4-thiazine 4-Hydroxy-2,5-methyl-3(2H )thiophenone 17 18 19 20 21 22 23 24 Sulfury, burnt Roasty, popcorn-like Sulfury, putrid Sulfury, putrid Sulfury, egg-like Sulfury Thiamine-like, meaty Sulfury, burnt 2-(1-Mercaptoethyl)thiophene 2-Methyltetrahydrothiophen-3-one 15 Sulfury, burnt Sulfury, roasty Sulfury, roasty Odour description 16 5-Methyl-2-thenylthiol 2-(1-Mercaptoethyl)furane 13 5-Methyl-2-furfurylthiol 12 14 Compound (Continued ) No Table 3.3 24.0 (1) 1.25 (1) 1.0 (13) — 0.2 (12) 10 (11) — — — 0.038 (1) 0.022 (1) 0.049 (1) 0.048 (1) Odour threshold (␮g/kg) in waterb Cysteine/glucose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Thiamine/cysteine (5) Thiamine/cysteine (5), thiamine/methionine (5) Cysteine/ribose (2, 3), thiamine (5), thiamine/cysteine (5), cysteine/IMP (6, 7), cysteine/ribose-5-P (6) Cysteine/glucose (4) Cysteine/glucose (4) Cysteine/rhamnose (4) Cysteine/rhamnose (4) Detected inb 72 Food Flavour Technology 3-Hydroxy-6-methyl-2(2H )-pyranone 4-Hydroxy-5-methyl-3(2H )-furanone (norfuraneol) 2-Ethyl-4-hydroxy-5-methyl-3(2H )furanone (homofuraneol) 3-Hydroxy-2-methyl-4(4H )-pyranone (maltol) 3-Hydroxy-4,5-dimethyl-2(5H )-furanone (sotolon) 6-Acetyltetrahydropyridine 2-Acetylpyridine 33 34 Roasty, caramel-like Roasty, burnt, caramel-like Roasty, popcorn-like Earthy, roasty Seasoning-like Seasoning-like Caramel-like Caramel-like, sweet Caramel-like, burnt chicory Caramel, strawberry-like 19 (23) 1.6 (22) 0.1 (21) 0.09 (20) 15.0 (1) 0.3 (19) 35000 (17) 1.15 (14) 8500 (1) 10 (14) Proline/glucose (15) Proline/glucose (15) Proline/glucose (15) Cysteine/glucose (4), cysteine/rhamnose (4) Cysteine/rhamnose (4) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4) Serine/maltose (18), proline/lactose (18) Cysteine/rhamnose (4) Cysteine/ribose (2), proline/xylose (16) Cysteine/ribose (2), cysteine/glucose (4), cysteine/rhamnose (4), proline/glucose (15), proline/rhamnose (16), lysine/rhamnose (16) b The sensory significance of the aroma compounds was assessed by quantitative data or by GC-O techniques 1, Hofmann (1995); 2, Hofmann and Schieberle (1995); 3, Gasser (1990); 4, Hofmann and Schieberle (1997); 5, Găuntert et al (1996); 6, Mottram and Nobrega (1998); 7, Zhang and Ho (1991); 8, Guadagni et al (1972); 9, Meynier and Mottram (1995); 10, Buttery et al (1984); 11, Pippen and Mecchi (1969); 12, Guth and Grosch (1994); 13, Cerny and Grosch (1993); 14, Semmelroch et al (1995); 15, Roberts and Acree (1994); 16, Decnop et al (1990); 17, Pittet et al (1970); 18, Fickert (1999); 19, Kerler and Grosch (1997); 20, Grosch (2001); 21, Buttery et al (1983); 22, Buttery and Ling (1995); 23, Teranishi et al (1975) a 2,3-Diethyl-5-methylpyrazine 2-Acetyl-1-pyrroline 31 32 C: Compounds containing nitrogen 30 29 28 27 26 B: Compounds containing oxygen 25 4-Hydroxy-2,5-dimethyl-3(2H )-furanone (Furaneol) Basic chemistry and process conditions for reaction flavours 73 74 Food Flavour Technology Furaneol, which is also a key ingredient of caramel-like process flavours, can be efficiently prepared by reacting rhamnose or other 6-deoxyhexoses with lysine, proline or hydroxyproline (Decnop et al., 1990) Another important caramel-like aroma compound, 3-hydroxy-2-methyl-4(4H)-pyranone (maltol), is a key substance of process flavours that are prepared from disaccharides (maltose or lactose) and proline or serine (Fickert, 1999) In thermally treated solutions of proline and glucose or fructose, ACP and ACTP have been evaluated as important aroma compounds, contributing a roasty, popcorn-like odour to these process flavours (Roberts and Acree, 1994; Schieberle, 1995) ACP and ACTP have also been reported to be character-impact compounds of various thermally treated cereal products (see also Section 3.3.1) In addition, Roberts and Acree (1994) showed that Furaneol and 2-acetylpyridine contribute to the flavour of the proline/glucose model system 3.4 PREPARATION OF PROCESS FLAVOURS 3.4.1 General aspects In Europe, flavourings that are obtained by thermal treatment of a reducing sugar and a foodgrade nitrogen source such as amino acids, peptides, food proteins, hydrolysed vegetable proteins (HVPs) and yeasts are referred to as thermal process flavourings (official terminology revised by EU in 2008; see Chapter for more details) These products have been designated as a separate class of flavours and are identified as complex mixtures that have been converted to flavours by heat processing In the US, the term ‘process flavours’ does not exist in regulatory terms Maillard reaction flavours are considered natural or artificial flavours, depending on whether the starting materials and process are considered natural The International Organisation of the Flavour Industry (IOFI) has established a guideline for manufacturers of process flavours This guideline is part of the IOFI’s Code of Practice and defines the types of raw materials and general reaction conditions (for instance, a maximum temperature/time treatment of 180◦ C/15 minutes, pH ≤8) (reviewed by Manley, 1995) The US Department of Agriculture, however, did not set a guideline for manufacture but established labelling criteria for materials used to produce a process flavour (Lin, 1995) Although processed flavours that are prepared according to the IOFI guidelines were considered GRAS (generally recognised as safe) in the US in 1995, the regulatory status of Maillard reaction flavours still lacks clarity with respect to the GRAS specification This is due to lack of information on whether processed flavours contain heterocyclic amines in amounts sufficient to affect their safe use in foods Maillard reaction technology is commonly used by the flavour industry to produce complex building blocks that provide similar aroma and taste properties to thermally treated foodstuffs such as meat, chocolate, coffee, caramel, popcorn or bread Although flavour formation during the Maillard reaction is quantitatively a minor pathway, process flavours are very important to the flavour industry This is because these complex blocks exhibit unique flavour qualities, are difficult to copy and are relatively cheap, being based on low production costs and high flavour potency of the aroma compounds formed 3.4.2 Factors influencing flavour formation The factors that influence flavour formation and, thus, the sensory properties of process flavours, are the type of sugar and amino acid, pH, reaction media, water activity as well as Basic chemistry and process conditions for reaction flavours Table 3.4 75 Flavour types of processed sugar–amino acid model mixtures Amino acid Temperature (◦ C) Flavour description Referencesa Glucose Ribose Ascorbic acid Ascorbic acid Glucose Cysteine Cysteine Threonine Cysteine Serine or glutamine or tyrosine 100–140 100 140 140 100–220 Meaty, beefy Meaty, roast beef Beef extract, meaty Chicken Chocolate 1, 1, 2, 1 Glucose Glucose Glucose Ribose or xylose Leucine Threonine Phenylalanine Threonine 100 100 100–140 140 Chocolate Chocolate Floral, chocolate Almond, marzipan 3 1 Glucose Glucose Proline Proline or hydroxyproline 100–140 180 Nutty Bread, baked 3, Glucose Glucose Xylose Ribose Glucose Glucose Glucose Glucose Glucose Alanine Lysine Lysine Lysine Valine Arginine Methionine Isoleucine Glutamine or asparagine 100–220 110–120 100 140 100 100 100–140 100 —b Caramel Caramel Caramel, buttery Toast Rye bread Popcorn Cooked potatoes Celery Nutty 1, 3 1, Sugar a 1, Lane and Nursten (1983); 2, Morton et al (1960); 3, Herz and Schallenberger (1960); 4, McKenna (1988); 5, Apriyantono and Ames (1990); 6, Yaylayan et al (1994) Microwave heating (640 W for 2–4 minutes) b temperature and time (see reviews by Shibamoto, 1983; Reineccius, 1990) In general, the sensory quality of a process flavour is less influenced by the type of sugar than by the amino acid Several authors (Herz and Schallenberger, 1960; Lane and Nursten, 1983; Yaylayan et al., 1994) studied the variety of odours produced in Maillard model systems, comprising two or more components in reaction systems containing various sugars (or ascorbic acid) with each of the protein-derived amino acids The flavour characters of some of these processed sugar–amino acid model mixtures are summarised in Table 3.4 Cysteine is the favoured amino acid to produce meat-like flavours, both on heating with reducing sugars or alone (Lane and Nursten, 1983) These authors also obtained chicken and beef aromas by reacting, respectively, cysteine and threonine with ascorbic acid Chocolate flavours can be prepared by heating glucose with amino acids such as serine, glutamine, tyrosine, leucine, threonine or phenylalanine (Herz and Schallenberger, 1960; Lane and Nursten, 1983) Phenylalanine also gives rise to a floral aroma (in reaction with glucose or alone), whereas threonine yields nutty aromas when reacted with ribose or xylose Proline is the favoured amino acid for the production of bread-like and baked flavours (Lane and Nursten, 1983; Yaylayan et al., 1994) However, Schieberle (1992a) showed that the yeastderived amino acids ornithine and citrulline are even more effective precursors for ACP, which is a key aroma compound of bread crust Herz and Schallenberger (1960) reported the generation of rye bread and popcorn aromas by heating valine or arginine with glucose In 76 Food Flavour Technology addition, alanine and lysine were found to give caramel flavours, and glutamine and arginine, nutty flavours Besides the type of sugar and amino acid, pH is another important factor determining aroma of process flavours It is well known to the flavour industry that meat flavours are preferably prepared at low pH (4–5.5), whereas roast and caramel flavours are obtained under neutral or slightly basic conditions Madruga and Mottram (1995) as well as Hofmann and Schieberle (1998a) showed that important sulfur-containing compounds in meat, such as MFT, 2-furfurylthiol and 2-methyl-3-(methyldithio)furan, are preferably formed at a pH of 3–4 Sensorial evaluations of thermally treated model mixtures of ribose or -IMP and cysteine revealed that the highest scores for boiled meat character were obtained when the reactions were carried out at pH 4.5 (ribose) and (5 -IMP) (Madruga and Mottram, 1998) Reaction media and water activity of the Maillard reaction systems are additional factors that influence aroma generation Besides buffered aqueous solutions, solvents such as propylene glycol, glycerol, triacetin or fats and oils, as well as their emulsions or mixtures with water, are used Vauthey et al (1998), for example, filed a patent on the generation of roast chicken aroma using a cubic phase system This system was prepared by introducing a melted monoglyceride (saturated in C16 and C18 ) into an aqueous phosphate buffer solution Compared to the same reaction in phosphate buffer, the flavour formed in the cubic system was more intense, corresponding to higher amounts of sulfur compounds such as MFT Shu and Ho (1989) investigated the reaction of cysteine and 4-hydroxy-2,5-dimethyl3(2H)-furanone in varying proportions of water and glycerol They found that a superior roasted/meaty character was obtained in the aqueous system The influence of the water activity on pyrazine formation during Maillard reaction was studied by Leathy and Reineccius (1989a) The authors observed that pyrazine formation was optimal at an Aw of about 0.75 Schieberle and Hofmann (1998) compared the character-impact compounds of cysteinebased process flavours formed in aqueous solution and under dry heating conditions In a cysteine/ribose model system, dry heating yielded higher amounts of key odourants with roasty notes such as 2-furfurylthiol, 2-acetyl-2-thiazoline and 2-propionyl-2-thiazoline as well as 2-ethyl- and 2-ethenyl-3,5-dimethylpyrazine, whereas the meat-like sulfur compounds MFT and MP were found in comparable or lower concentrations, respectively Their study also revealed that the amounts of the 3-deoxyosone-derived compounds 2-furfural and 5-methylfuran-2-aldehyde were significantly higher in the dry-heated model systems, whereas the formation of the 1-deoxyosone-derived 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) was enhanced in aqueous solution An explanation for this finding could be that, under dry heating conditions, caramelisation processes are favoured relative to Maillard reaction In caramelisation processes, 1,2-enolisation of the sugar molecule is preferred over 2,3-enolisation, leading to the formation of high amounts of 3-deoxyosones (Kroh, 1994) Apart from the precursor composition (e.g availability of amino acids as reactants), pH and the reaction medium, other reaction parameters such as the presence of catalysts (e.g phosphates) as well as temperature and time have a major effect on the sensory properties of process flavourings Many of these parameters have extensively been studied for the generation of the caramel-like smelling Furaneol from 6-deoxyhexoses (Schieberle, 1992b; Havela-Toledo et al., 1997, 1999; Hofmann and Schieberle, 1997, 1998b; Schieberle and Hofmann, 2002) For example, the yield of Furaneol from rhamnose (pH 7, 150◦ C, 45 minutes) was increased about 40 times when the malonate buffer was replaced with phosphate buffer (Schieberle, 1992b) Even a higher increase of its yield (70-fold) was observed when the pH of rhamnose/cysteine system (145◦ C, 20 minutes) was increased from to Basic chemistry and process conditions for reaction flavours 77 (Schieberle and Hofmann, 2002) Recently, Illmann et al (2009) used a fractional factorial design to identify critical reaction parameters affecting kinetics of Furaneol generation from rhamnose under cooking conditions (120◦ C) The importance of the reaction parameters was found to decrease in the following order: phosphate concentration > concentration of precursors Ͼ pH Ͼ rhamnose to lysine ratio The experimental design approach achieved very high yields of Furaneol (about 40 mol%) The type of amino acid was also shown to affect the yield of Furaneol from rhamnose Lysine was most efficient in generating the caramel-like odourant, followed by alanine, serine, glycine and threonine On the other hand, much lower yields were obtained in the presence of proline and especially cysteine (Davidek et al., 2009) The generation of 3-hydroxy-2-methyl-4(4H)-pyranone (maltol), another important caramel-like aroma compound, was shown to strongly depend on the reaction media Rather low yields of maltol were obtained when lactose was heated with proline in aqueous systems (13.5 mmol/mol lactose) or when the precursors were dry heated (7.2 mmol/mol lactose) However, the yield of maltol increased significantly when water was replaced by propylene glycol (70 mmol/mol lactose; Cerny, 2003) In addition, the knowledge of the reaction kinetics of aroma compounds helps to explain the influence of temperature and time on their formation (Reineccius, 1990) Although flavour formation is a multi-step reaction sequence, Arrhenius kinetics have been found to describe flavour formation well in model and real food systems Stahl and Parliment (1994) used an ingenious device to obtain clear time/temperature conditions for model systems and determined the activation energies of flavour compounds Leathy and Reineccius (1989b) showed that the sensory quality of a product is less influenced by temperature/time in a model system that is designed to result in similar key aroma compounds (e.g pyrazines) This can be explained by the fact that these compounds have similar activation energies However, their study focused on dialkylpyrazines and did not consider the more potent trialkylpyrazines, which were suggested to have different formation pathways (Amrani-Hemaimi et al., 1995; Schieberle and Hofmann, 1998) As a result, flavour generation during Maillard reaction is in most cases strongly influenced by temperature and time, also because Maillard reaction flavours are complex mixtures of different classes of key aroma compounds Lee (1995) proposed a different approach, on the basis of differential equations, to describe Maillard kinetics The model produced was able to simulate the Maillard reaction not just as a function of time and temperature but also as a function of reactant concentrations and pH From this simulation, the relative amounts of reactants could be plotted against time Examples given include the concentrations of aldose and Amadori compounds over a 10-hour reaction and the time profiles of enolisation compounds from both the 1,2- and 2,3-pathways under defined pH conditions The use of stable intermediates as flavour precursors and/or the application of multistep reactions are another way of optimising processing conditions on the basis of the required application Blank et al (2003a,b) compared the generation of odourants in Maillard reaction systems containing glucose and proline (Glc/Pro) or the corresponding Amadori compound fructosyl-proline (Fru-Pro) The major odourants found in both systems were similar and included ACP, 4-hydroxy-2,5-dimethyl-3(2H)furanone (Furaneol), acetic acid, 3hydroxy-4,5-dimethyl-2(5H)furanone (sotolon), 2,3-butanedione and ACTP However, their concentrations as well as their contributions to the final flavour differed For example, the formation of Furaneol was favoured from Fru-Pro, namely at pH and On the other hand, the reaction system Glc/Pro gave rise to relative high yields of ACP and ACTP Although both roasted-smelling popcorn odourants showed high sensory relevance in both reaction systems, 78 Food Flavour Technology they dominated especially the aroma of the Glc/Pro process flavouring These results also indicate that there is no benefit in using Amadori compound for the formation ACTP and ACP The following sections give a survey of the patent literature in the area of savoury and sweet process flavours Emphasis is given to meat-like process flavours 3.4.3 Savoury process flavours The great majority of patents based on Maillard reaction technology have been directed to the production of meat-like process flavours Most of these reaction flavours indicate cysteine and thiamine as the essential sulfur-containing precursor compounds In 1960, the basic concept of Maillard flavour technology was beginning to emerge with the Unilever patent of Morton et al (1960) The authors disclosed Maillard processes for the production of cooked beef and pork flavours by reacting ribose or mixtures of ribose and glucose with cysteine and either additional amino acids or deflavoured protein hydrolysates from cod fish flesh, casein, groundnut or soya The group of additional amino acids was consisted of ␤-alanine, glutamic acid, glycine, ␣-alanine, threonine, histidine, lysine, leucine, serine and valine All flavours were prepared in water at a pH between and and a temperature around 130◦ C Using similar reaction conditions, May and Morton (1960) and May (1961) also prepared meat flavours through reacting cysteine with glyceraldehyde or furfural in combination with deflavoured cod fish hydrolysates or amino acid mixtures Jaeggi (1973) patented processes for boiled and roasted beef flavours, which were also based on the reaction of ribose and cysteine However, the inventor used methionine and proline as additional amino acids and carried out the reaction in glycerol or groundnut oil instead of water Methionine as the sole sulfur source was also reported to result in beef flavourings when reacted with xylose and cysteine-free hydrolysed plant protein (Van Pottelsberghe de la Potterie, 1973) Tandy (1985) prepared process flavourings with white chicken meat character using leucine and cysteine in combination with the reducing sugars arabinose and glucose In addition, he found that the use of rhamnose instead of glucose (as well as the addition of serine) provided a more aromatic, characteristic white meat chicken flavour International Flavors and Fragrances (IFF) filed a number of patents on the preparation of chicken, beef and pork flavours using cysteine in combination with thiamine, often in carbohydrate-free systems This demonstrates that thiamine is capable of replacing carbohydrates by providing similar intermediates and aroma compounds as found in Maillard systems Intense beef flavours, for example, were obtained by refluxing cysteine, thiamine and carbohydrate-free vegetable protein hydrolysate (HVP) in water or water–ethanol mixtures (IFF, 1965) The addition of beef tallow was found to result in a beef flavour with a ‘pan-dripping’ character In addition, chicken flavourings were prepared by heating cysteine, thiamine and HVP in combination with other ingredients such as ␤-alanine, glycine and ascorbic acid, whereas pork flavours required the addition of methionine and lard Similar processes for chicken, beef and pork flavours were disclosed in other IFF patents (IFF, 1967; Giacino, 1968a, 1969; Katz and Evers, 1973) Chicken aroma was found to be improved, for example, by adding diacetyl and hexanal (Giacino, 1968a) or mercaptoalkanones (Katz and Evers, 1973) to the processed flavours Dihydroxyacetone, pyruvic acid or pyruvic aldehyde in combination with thiamine and HVPs were claimed to result in beef flavourings with improved cooked note (Giacino, 1968b) Kerscher (2000) investigated the analytical assessment of the species-specific character of beef, chicken and pork His results (see Section 3.3.1) cannot explain the choice of the Basic chemistry and process conditions for reaction flavours 79 ingredients for the preparation of chicken, beef and pork flavours that are disclosed in the IFF patents mentioned above, but are in agreement with the findings of Chen and Tandy (1988) The authors developed species-specific beef and chicken flavourings through oxidation of oleic and linoleic acids, respectively Their processes involved heat treatment of oleic or linoleic acid in the presence of air at high temperatures of about 300◦ C as well as trapping of the resulting aroma fraction in cold traps The authors also claimed flavour blocks that resembled roast, grilled, bloody and braised beef in different fractions of the distillate of oxidised oleic acid In terms of alternative sulfur sources to cysteine and thiamine, Giacino (1970) found that process flavours with similar characters were obtained when cysteine was replaced by taurine Patents of the Corn Products Company also describe the production of Maillard reaction flavours using taurine in combination with HVPs and xylose (Corn Products Company, 1969; Hack and Konigsdorf, 1969) From a scientific point of view, however, the finding that cysteine can be replaced by taurine has to be questioned, because there are no studies that report the generation of important meat aroma compounds from taurine In addition, Tai and Ho (1997) could detect only trace amounts of volatile sulfur compounds in a Maillard model system containing cysteinesulfinic acid and glucose Broderick and Linteris (1960) used derivatives of mercaptoacetaldehyde such as 2,5-dihydroxy-1,4-dithiane, diethyl- or dithioacetals and hemimercaptals as precursors to impart meat-like flavour to canned simulated meat and vegetable products on sterilisation The patent of Heyland (1977) involved the use of hydrolysed onion, garlic and cabbage in combination with HVP, ribose and beef fat for the preparation of beef flavours Other flavour companies developed meat flavourings with sulfur sources such as hydrogen, sodium or ammonium sulfides (Godman and Osborne, 1972; Gunther, 1972), methionine (Van Pottelsberghe de la Potterie, 1972) or egg white (Theron et al., 1975) Yeast extracts or yeast hydrolysates have traditionally been used either as precursors for the thermal generation of meat flavourings or as taste-enhancing ingredients, in blends with process flavours The advantage of yeast extracts is that they are a relatively cheap, natural source of amino acids and thiamine In addition, their high content of glutamate and -ribonucleotides, particularly inosine -monophosphate and guanosine -monophosphate, provides complexity, body and flavour enhancement Nestl´e, for example, filed several patents in which the preparation of beef and chicken process flavours using yeast extracts was disclosed (Nestl´e, 1966; Rolli et al., 1988; Cerny, 1995) Cerny (1995) also developed bouillon flavours using yeast cream, which is enriched in hydrogen sulfide Such a yeast cream was obtained by incubating baker’s yeast with elemental sulfur In order to obtain complete meat flavouring products, process flavours are blended with several other ingredients, which provide aroma (e.g compounded aroma blocks referred to as top notes or topnote flavours), taste, taste enhancement, mouth feel and body (referred to as base notes) Besides topnote flavours, yeast extracts, hydrolysed vegetable proteins and the monosodium salts of glutamate, inosinate and guanylate, ingredients such as onion, garlic, celery and/or caramel powder, animal or vegetable fats, gelatine and spices are often used in meat flavour compositions The use of some yeast extracts, however, is limited owing to their undesirable ‘yeasty’ character Therefore, research groups have developed processes for manufacturing yeast hydrolysates with improved meat-like taste, in which the yeasty notes are absent De Rooij and Hakkaart (1992), for example, improved the meaty character of yeast hydrolysates prepared from several yeast species by combining the enzymatic degradation of yeast cells with an additional fermentation step, which was carried out using lactic acid-producing micro-organisms or additional yeasts Hyăoky et al 80 Food Flavour Technology (1996) developed a method for the production of yeast extracts in which undesirable bitter and yeasty flavour notes were removed The authors evaluated several non-ionic and slightly basic macroporous polymeric adsorbents as well as activated carbon for their ability to bind bitter and other undesirable flavouring substances of yeast hydrolysates, without binding yeast peptides, amino acids or nucleotides The best results were obtained with Amberlite XAD-16 and Amberlite XAD-765, which are a non-ionic styrene/divinylbenzene copolymer and a weakly basic phenolformaldehyde polymer, respectively 3.4.4 Sweet process flavours A few patents and articles on the generation of chocolate and caramel flavours are discussed here Rusoff (1958) prepared artificial chocolate flavours by heating partially hydrolysed proteins with sugars The Maillard reactions were performed between protein hydrolysates derived from casein, soy, wheat gluten or gelatine and mixtures of pentoses and hexoses The reaction medium contained 30% water and the reaction temperature ranged between 130 and 150◦ C These ‘base chocolate flavours’ were rounded off by adding ingredients such as caffeine, theobromine and tannins before or after the heating process The need for hydrolysed proteins to generate chocolate flavours through Maillard reaction was stressed by Răodel et al (1988) The authors based their investigations on precursor studies of Mohr et al (1971, 1976), who found that only mixtures of peptide and free amino acid fractions isolated from fermented raw cocoa beans developed cocoa aroma on thermal treatment The study of Răodel et al (1988) covered a complete range of parameters such as source of protein, rate of hydrolysis, source of enzyme, amount of sugar, water content as well as temperature and time The Maillard reaction flavours obtained were evaluated organoleptically and analytically The authors revealed that gelatine that is enzymatically hydrolysed by more than 20% is an appropriate protein source for producing cocoa flavours through Maillard reaction The quality of the cocoa aroma was further affected by water and sugar contents, whereas the source of enzyme had no influence A water content of at least 5% and a sugar content of 20 g per 100 g hydrolysed protein gave a positive effect on flavour In addition, temperature and time, which were the most sensitive parameters, were optimised at 144◦ C and 21 minutes Pittet and Seitz (1974) disclosed processes for the preparation of various flavours resembling chocolate, sweet corn, popcorn, bread, cracker and caramel toffee The flavours were prepared by heating cyclic enolones such as Furaneol, maltol or cyclotene with amino acids in propylene glycol or glycerol The temperatures ranged between 120 and 205◦ C Chocolate flavours, for example, were obtained when valine or leucine was reacted with maltol or Furaneol, whereas proline (in combination with Furaneol or maltol) yielded cracker-like, popcorn, sweet corn and bread aromas In addition, caramel toffee and burnt sugar flavours were established from proline and cyclotene or ethyl cyclotene Gilmore (1988) also developed a caramel butterscotch flavour by heating a mixture of sugar syrup and butter in the presence of ammonia at a pH of and a temperature of about 100◦ C 3.5 OUTLOOK Flavour formation is a minor but important pathway within the complex cascade of chemical reactions occurring during Maillard processes A strong focus on the key flavour compounds (aroma and taste-active molecules), their precursors and reaction routes is required for the Basic chemistry and process conditions for reaction flavours 81 optimisation of process flavours In addition, the formation of undesirable molecules needs to be taken into account when it comes to the optimisation of processing parameters The better understanding of the influence of various process parameters on the formation of both the key aroma compounds and their precursors is still a challenge for future research Experimental design and kinetic parameter estimations are good tools for limiting the amount of experiment required In addition, the evaluation of important taste compounds derived from the Maillard reaction, the elucidation of their formation pathways as well as the understanding of how the interaction of aroma and taste compounds affects the sensorial quality of the flavours are certainly research areas that are worth investigating REFERENCES Amrani-Hemaimi, M., Cerny, C and Fay, L.B (1995) Mechanisms of formation of alkylpyrazines in the Maillard reaction J Agric Food Chem 43, 2818–2822 Anet, E.F.L.J (1960) 3-Deoxyhexosones J Am Chem Soc 82, 1502–1504 Anonymous (1979) Foodstuff with flavor precursor Res Disclosure 177–188, 17918 Apriyantono, A and Ames, J.M (1990) Volatile compounds produced on heating lysine with xylose In: Flavour Science and Technology (eds Y Bessi`ere and A.F Thomas), Wiley, Chichester, UK, pp 117– 120 Beksan, E., Schieberle, P., Robert, F., Blank, I., Fay, L.B., Schlichtherle-Cerny, H and Hofmann, T (2003) Synthesis and sensory characterization of novel Umami-tasting glutamate glycoconjugates J Agric Food Chem 51, 5428–5436 Belitz, H.-D., Grosch, W and Schieberle, P (2008) Food Chemistry, 4th edn, Springer, Berlin Bel Rhild, B., Fleury, Y., Devaud, S., Fay, L.B., Blank, I and Juillerat, M.A (2002) Biogeneration of roasted notes based on 2-acetyl-2-thiazoline and its precursor 2-(1-hydroxyethyl)-4,5-dihydrothiazole In: Heteroatomic Aroma Compounds, ACS Symposium Series 826 (eds G.A Reineccius and T.A Reineccius), American Chemical Society, Washington, DC, pp 179–190 Blank, I., Devaud, S., Matthey-Doret, M., Pollien, P , Robert, F and Yeretzian, C (2003b) Formation of odour-active compounds in Maillard model systems based on proline In: Flavour Research at the Dawn of the Twenty-First Century, Proceedings of the 10th Weurman Flavour Research Symposium (eds J.-L Le Qu´er´e and P.X Eti´evant), Lavoisier, Intercept, London, pp 458–463 Blank, I., Devaud, S., Matthey-Doret, M and Robert, F (2003a) Formation of odorants in Maillard model systems based on L-proline as affected by pH J Agric Food Chem 51, 3643–3650 Blank, I., Sen, A and Grosch, W (1992) Aroma impact compounds of Arabica and Robusta coffee Colloq Sci Int Cafe 14, 117–129 Broderick, J.J and Linteris, L.L (1960) Flavoring agents and process for imparting a meat-like flavor to an edible composition US Patent 2955 041 Buttery, R.G., Haddon, W.F., Seifert, R.M and Turnbaugh, J.G (1984) Thiamin odor and bis(2-methyl-3furyl) disulfide J Agric Food Chem 32, 674–676 Buttery, R.G and Ling, L.C (1995) Volatile flavor components of corn tortillas and related products J Agric Food Chem 43, 1878–1882 Buttery, R.G., Ling, L.C and Juliano, B.O (1982) 2-Acetyl-1-pyrroline: an important aroma compound of cooked rice Chem Ind (London) 23, 958–959 Buttery, R.G., Ling, L.C., Juliano, B.O and Turnbaugh, J.G (1983) Cooked rice aroma and 2-acetyl-1pyrroline J Agric Food Chem 31, 823–826 Cerny, C (1995) Proc´ed´e de transformation du soufre e´ l´ementaire en sulfure European Patent 778350 Cerny, C (2003) Maltol generation from lactose in dry systems In: Flavour Research at the Dawn of the Twenty-First Century, Proceedings of the 10th Weurman Flavour Research Symposium (eds J.L Le Qu´er´e ´ evant), Lavoisier, Intercept, London, pp 544–547 and P.X Eti´ Cerny, C (2007) Process flavourings In: Flavourings Production, Composition, Applications, Regulations (ed H Ziegler), Wiley-VCH, Weinheim, Germany, pp 274–297 Cerny, C and Davidek, T (2003) Formation of aroma compounds from ribose and cysteine during Maillard reaction J Agric Food Chem 51, 2714–2721 82 Food Flavour Technology Cerny, C and Grosch, W (1993) Quantification of character-impact odor compounds of roasted beef Z Lebensm.-Unter Forsch 196, 417–422 Chan, F and Reineccius, G.A (1994) The reaction kinetics for the formation of isovaleraldehyde, 2-acetyl1-pyrroline, di(H)di(OH)-6-methylpyranone, phenylacet-aldehyde, 5-methyl-2-phenyl-2-hexenal, and 2acetylfuran in model systems In: Maillard Reactions in Chemistry, Food and Health (eds T.P Labuza, G.A Reineccius, V Monnier, J O’Brien and J Baynes), The Royal Society of Chemistry, Cambridge, pp 131–139 Chen, T.-K and Tandy, J.S (1988) Production of flavour European Patent 295509 Corn Products Company (1969) Process for the production of seasonings with a flavor similar to meat extract GB Patent 1148449 Cremer, D.R and Eichner, K (2000) The influence of the pH value on the formation of Strecker aldehydes in low moisture model systems and in plant powders Eur Food Res Technol 211, 247–251 Davidek, T., Clety, N., Aubin, S and Blank, I (2002) Degradation of the Amadori compound N-(1-deoxyD-fructos-1-yl)glycine in aqueous model systems J Agric Food Chem 50, 5472–5479 Davidek, T., Devaud, S., Robert, F and Blank, I (2006a) Sugar fragmentation in the Maillard reaction cascade: formation of acetic acid by a hydrolytic ␤-dicarbonyl cleavage mechanism J Agric Food Chem 54, 6667–6676 Davidek, T., Illmann, S., Gou´ezec, E., Rytz, A., Schuchmann, H.P and Blank, I (2009) Effect of reaction conditions on generation of 4-hydroxy-2,5-dimethyl-3(2H)-furanone from rhamnose In: 12th Weurman Flavor Research Symposium, Interlaken, Switzerland, July 1–4, 2008 Davidek, T., Robert, F., Devaud, S., Arce Vera, F and Blank, I (2006b) Sugar fragmentation in the Maillard reaction cascade: formation of short-chain carboxylic acids by a new oxidative ␣-dicarbonyl cleavage pathway J Agric Food Chem 54 6677–6684 Decnop, C., van Dort, J.M and de Hey, J.T (1990) Hydroxyfuranone preparation European Patent 398417 De Rooij, J.F.M and Hakkaart, M.J.J (1992) A process for the preparation of food flavours European Patent 191513 de Roos, K.B (1992) Meat flavor generation from cysteine and sugars In: Flavor Precursors: Thermal and Enzymatic Conversions (eds R Teranishi, G.R Takeoka and M Găuntert), American Chemical Society, Washington, DC, pp 203–216 Doornbos, T., Haring, P.G.M and Van Der Heijden, A (1991) Use of diketone in food products European Patent 293957 Doornbos, T., Van Den Ouweland, G.A.M and Tjan, S.B (1981) Amadori compounds, derived from 6-deoxy sugars, as flavour precursor In: Progress in Food and Nutrition Science, Maillard Reactions in Food (ed C Eriksson), Pergamon Press, Oxford, pp 57–63 Eichner, K., Reutter, M and Wittmann, R (1994) Detection of Amadori compounds in heated foods In: Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes (eds T.H Parliament, M.J Morello and R.J McGorrin), American Chemical Society, Washington, DC, pp 42–54 Fickert, B (1999) Untersuchungen zur Bildung von Aromastoffen bei der Măalzung von Getreide PhD thesis, Technical University of Munich, Garching, Germany Frank, O., Jezussek, M., Hofmann, T (2003) Sensory activity, chemical structure, and synthesis of Maillard generated bitter-tasting 1-oxo-2,3-dihydro-1H-indolizinium-6-olates J Agric Food Chem 51, 26932699 Gasser, U (1990) Flăuchtige Verbindungen aus gekochtem Rind- und Hăuhnerfleisch, Fleischaromen und Fleischextrakt PhD thesis, Technical University of Munich, Garching, Germany Gasser, U and Grosch, W (1988) Identification of volatile flavor compounds with high aroma values from cooked beef Z Lebensm.-Unters Forsch 186, 489–494 Gasser, U and Grosch, W (1990) Primary odorants of chicken broth A comparative study with meat broths from ox and cow Z Lebensm.-Unters Forsch 190, 3–8 Gasser, U and Grosch, W (1991) Aroma von gekochtem Schweinefleisch Lebensmittelchemie 45, 15–16 Gi, U.-S and Baltes, W (1995) Model reactions on roast aroma formation, 15 Investigations on the formation of pyrido[3,4-d]imidazoles during the Maillard reaction J Agric Food Chem 43, 2226–2230 Giacino, C (1968a) Poultry flavor composition and process US Patent 3394017 Giacino, C (1968b) Product and process of reacting a proteinaceous substance with a sulfur-containing compound to provide a meat-like flavor US Patent 3394015 Giacino, C (1969) Meat flavor compositions GB Patent 1146337 Giacino, C (1970) Meat flavor compositions US Patent 3519437 Gilmore, L.T (1988) Process for preparing a caramel butterscotch flavor syrup US Patent 4753814 Godman, J.L and Osborne, D.R.D (1972) Edible products GB Patent 1285568 Basic chemistry and process conditions for reaction flavours 83 Grosch, W (1994) Determination of potent odourants in foods by aroma extract dilution analysis (AEDA) and calculation of odour activity values (OAVs) Flavour Frag J 9, 147–158 Grosch, W (2001) Flavour of chemistry III: Coffee and volatile compounds In: Coffee, Recent Developments (eds O Vitzthum and R.J Clarke), Blackwell Science, Oxford, pp 68–89 Grosch, W and Schieberle, P (1997) Flavor of cereal products – a review Cereal Chem 74(2), 91–97 Guadagni, D.G., Buttery, R.G and Turnbaugh, J.G (1972) Odor thresholds and similarity ratings of some potato chip components J Sci Food Agric 23, 14351444 Găuntert, M., Bertram, H.-J., Hopp, R., Silberzahn, W., Sommer, H and Werkhoff, P (1996) Thermal generation of flavor compounds from thiamin and various amino acids In: Recent Developments in Flavor and Fragrance Chemistry (eds R Hopp and K Mori), VCH-Verlagsgesellschaft, Weinheim, Germany, pp 215240 Găuntert, M., Brăuing, J., Emberger, R., Hopp, R., Kăopsel, M., Surburg, H and Werkhoff, P (1992) Thermally degraded thiamin: a potent source of interesting flavor compounds In: Flavor Precursors: Thermal and Enzymatic Conversions (eds R Teranishi, G.R Takeoka and M Găuntert), American Chemical Society, Washington, DC, pp 140–163 Gunther, R (1972) Flavoring compositions produced by reacting hydrogen sulfide with a pentose US Patent 3642497 Guth, H and Grosch, W (1993a) 12-Methyltridecanal, a species-specific odorant of stewed beef Lebensm Wissensch Technol 26, 171–177 Guth, H and Grosch, W (1993b) Identification of potent odorants in static headspace samples of green and black tea powders on the basis of aroma extract dilution analysis (AEDA) Flavour Frag J 8, 173–178 Guth, H and Grosch, W (1994) Identification of the character impact odorants of stewed beef juice by instrumental analysis and sensory studies J Agric Food Chem 42, 2862–2866 Hack, A.W and Konigsdorf, W (1969) Production of seasonings with a flavor similar to meat extract US Patent 3480447 Havela-Toledo, E., Naim, M., Zehavi, U and Rouseff, R.L (1997) 4-Hydroxy-2,5-dimethyl-3(2H)-furanone formation in buffers and model solutions of citrus juice J Agric Food Chem 45, 1314–1319 Havela-Toledo, E., Naim, M., Zehavi, U and Rouseff, R.L (1999) Effect of L-cysteine and n-acetylL-cysteine on 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol), 5-hydroxymethylfurfural and 5methylfurfural formation and browning in buffer solutions containing rhamnose or glucose and arginine J Agric Food Chem 47, 4140–4145 Henderson, S.K and Nawar, W.W (1981) Thermal interaction of linoleic acid and its esters with valine J Am Oil Chem Soc 58, 632–635 Herz, W.J and Schallenberger, R.S (1960) Some aromas produced by simple amino acid-sugar reaction Food Res 25, 491–494 Heyland, S (1977) Verfahren zur Herstellung eines Aromatisierungsmittels German Patent 2546035 Hidalgo, F.J., Zamora, R (2004) Strecker-type degradation produced by the lipid oxidation products 4,5epoxy-2-alkenals J Agric Food Chem 52, 7126–7131 Hidalgo, F.J and Zamora, R (2005) Interplay between the Maillard reaction and lipid peroxidation in biochemical systems Ann N Y Acad Sci 1043, 319–326 Hidalgo, F J and Zamora, R (2007) Conversion of phenylalanine into styrene by 2,4-decadienal in model systems J Agric Food Chem 55, 4902–4906 Hodge, J.E (1953) Chemistry of browning reactions in model systems J Agric Food Chem 1, 928–943 Hofmann, T (1995) Charakterisierung intensiver Geruchsstoffe in Kohlenhydrat/Cystein-Modellreaktionen und Klăarung von Bildungswegen PhD thesis, Technical University of Munich, Garching, Germany Hofmann, T (1999) Quantitative studies on the role of browning precursors in the Maillard reaction of pentoses and hexoses with L-alanine Eur Food Res Technol 209, 113–121 Hofmann, T (2005) Taste-active Maillard reaction products: the ‘tasty’ world of nonvolatile Maillard reaction products Ann N Y Acad Sci 1043, 20–29 Hofmann, T., Măunch, P and Schieberle, P (2000) Quantitative model studies on the formation of aroma-active aldehydes and acids by Strecker-type reactions J Agric Food Chem 48, 434–440 Hofmann, T., Ottinger, H., Frank, O., Soldo, T., Cerny, C., Robert, F and Blank I (2001) Use of alpha-keto enamine derivatives as cooling ingredients Eur Pat Appl., EP 1157617 A2 Hofmann, T and Schieberle, P (1995) Evaluation of the key odorants in a thermally-treated solution of ribose and cysteine by aroma extract dilution techniques J Agric Food Chem 43, 2187–2194 Hofmann, T and Schieberle, P (1997) Identification of potent aroma compounds in thermally-treated mixtures of glucose/cysteine and rhamnose/cysteine using aroma extract dilution techniques J Agric Food Chem 45, 898–906 84 Food Flavour Technology Hofmann, T and Schieberle, P (1998a) Quantitative model studies on the effectiveness of different precursor systems in the formation of the intense food odorants 2-furfurylthiol and 2-methyl-3-furanthiol J Agric Food Chem 46, 235–241 Hofmann, T and Schieberle, P (1998b) Identification of key aroma compounds generated from cysteine and carbohydrates under roasting conditions Z Lebensm.-Unters Forsch 207, 229–236 Hofmann, T and Schieberle, P (2000a) Formation of aroma-active Strecker-aldehydes by a direct oxidative degradation of Amadori compounds J Agric Food Chem 48, 4301–4305 Hofmann, T and Schieberle, P (2000b) Acetylformoin – an important progenitor of 4-hydroxy-2,5-dimethyl3(2H)-furanone and 2-acetyltetrahydropyridine during thermal food processing In: Flavour 2000 – Perception, Release, Evaluation, Formation, Acceptance, Nutrition/Health (ed M Rothe), Eigenverlag, Bergholz-Rehbrăucke, Germany, pp 311322 Hyăoky, G., Sarkki, M.-L and Tylli, M (1996) Method of producing a yeast extract usable in foodstuffs, where undesirable flavouring agents in the extract have been removed World Patent 96/38057 IFF, International Flavors & Fragrances, Inc (1965) Composition and process for preparation thereof GB Patent 1099711 IFF, International Flavors & Fragrances, Inc (1967) Flavoring compositions GB Patent 1069104 Illmann, S., Davidek, T., Gou´ezec, E., Rytz, A., Schuchmann, H.P and Blank, I (2009) Generation of 4hydroxy-2,5-dimethyl-3(2H)-furanone from rhamnose as affected by reaction parameters – experimental design approach J Agric Food Chem 57, 2889–2895 Ishizu, A., Lindberg, B and Theander, O (1967) 1-Deoxy-D-erythro-2,3-hexodiulose, an intermediate in the formation of D-glucosaccharinic acid Carbohydr Res 5, 329–334 Jaeggi, K (1973) Process for the manufacture of meat flavors US Patent 3761287 Katz, I and Evers, W.J (1973) Flavoring compositions and processes utilizing alpha-ketothiols US Patent 3773524 Kerler, J (1996) Objektivierung eines Aromafehlers (Warmed-over Flavour) bei aufgewăarmtem Hăuhner-und Rindfleisch PhD thesis, Technical University of Munich, Garching, Germany Kerler, J and Grosch, W (1996) Odorants contributing to warmed-over flavor (WOF) of refrigerated cooked beef J Food Sci 61(6), 1271–1274, 1284 Kerler, J and Grosch, W (1997) Character impact odorants of boiled chicken: changes during refrigerated storage and reheating Z Lebensm.-Unters Forsch 205, 232–238 Kerscher, R (2000) Objektivierung tierartspezifischer Aromaunterschiede bei erhitztem Fleisch PhD thesis, Technical University of Munich, Garching, Germany Kerscher, R and Grosch, W (1997) Comparative evaluation of potent odorants of boiled beef by aroma extract dilution and concentration analysis Z Lebensm.-Unters Forsch 204, 3–6 Kerscher, R and Grosch, W (1998) Quantification of 2-methyl-3-furanthiol, 2-furfurylthiol, 3-mercapto-2pentanone, and 2-mercapto-3-pentanone in heated meat J Agric Food Chem 46, 1954–1958 Keyhani, A and Yaylayan, V.A (1996) Elucidation of the mechanism of pyrazidone formation in glycine model systems using labeled sugars and amino acids J Agric Food Chem 44, 2511–2516 Kim, Y.-S., Hartman, T.G., Ho, C.-T (1996) Formation of 2-pentylpyridine from the thermal interaction of amino acids and 2,4-decadienal J Agric Food Chem 44, 3906–3908 Kroh, L.W (1994) Caramelisation in food and beverages Food Chem 51, 373–379 Lane, M.J and Nursten, H.E (1983) The variety of odors produced in Maillard model systems and how they are influenced by reaction conditions In: Maillard Reaction in Food and Nutrition (eds G.R Waller and M.S Feather), American Chemical Society, Washington, DC, pp 141–158 Leathy, M.M and Reineccius, G.A (1989a) Kinetics on the formation of alkylpyrazines: effect of pH and water activity In: Thermal Generation of Aromas (eds T.H Parliament, R.J McGorrin and C.-T Ho), American Chemical Society, Washington, DC, pp 196–208 Leathy, M.M and Reineccius, G.A (1989b) Kinetics on the formation of alkylpyrazines: effect of amino acid type and type of sugar In: Flavor Chemistry (eds R Teranishi, R.G Buttery and F Shahidi), American Chemical Society, Washington, DC, pp 76–91 Ledl, F and Schleicher, E (1990) New aspects of the Maillard reaction in foods and in the human body Angew Chem Int Ed Engl 29, 565–706 Lee, P.S (1995) Modeling the Maillard reaction – a computer-simulation and a discussion of its application to Maillard reaction analysis and design In: Flavor Technology: Physical Chemistry, Modification, and Process (eds C.-T Ho, C.T Tan and C.H Tong), American Chemical Society, Washington, DC, pp 74–97 Limacher, A., Kerler, J., Davidek, T., Schalzried, F and Blank, I (2008) Formation of furan and methylfuran by Maillard-type reactions in model systems and food J Agric Food Chem 56, 3639–3647 Basic chemistry and process conditions for reaction flavours 85 Lin, L.J (1995) Regulatory status of Maillard reaction flavors In: Savory Flavors (ed T.W Nagodawithana), Esteekay Associates, Wisconsin, pp 7–15 MacLeod, G and Ames, J (1986) 2-Methyl-3-(methylthio)furan: a meaty character impact compound from cooked beef Chem Ind 3, 175–177 Madruga, M.S and Mottram, D.S (1995) The effect of pH on the formation of Maillard-derived aroma volatiles using a cooked meat system J Sci Food Agric 68, 305–310 Madruga, M.S and Mottram, D.S (1998) Sensory analysis of a model system using -IMP and cysteine at different pH Ciˆenc Tecnol Aliment 18(4), 397–404 Manley, C.H (1995) Process flavors and precursor systems In: Savory Flavors (ed T.W Nagodawithana), Esteekay Associates, Wisconsin, pp 16–25 May, C.G (1961) Flavouring substances and their preparation GB Patent 858333 May, C.G and Morton, I.D (1960) Process for the preparation of a meat flavor US Patent 2934436 McKenna, A.B (1988) Influence of heating on the formation of caramel flavour, brown colour and hydroxymethylfurfural in reconstituted concentrated skim milk New Zeal J Dairy Sci Technol 23, 363–372 Meynier, A and Mottram, D.S (1995) The effect of pH on the formation of volatile compounds in meatrelated model systems Food Chem 52, 361–366 Milo, C and Grosch, W (1993) Changes in the odorants of boiled trout (Salmo fario) as affected by the storage of the raw material J Agric Food Chem 41, 2076–2081 Milo, C and Grosch, W (1995) Detection of odor defects in boiled cod and trout by gas chromatography–olfactometry of headspace samples J Agric Food Chem 43, 2076–2081 Milo, C and Grosch, W (1996) Changes in the odorants of boiled salmon and cod as affected by the storage of the raw material J Agric Food Chem 44, 459–462 Mohr, W., Landschreiber, E and Severin, T (1976) Zur Spezifităat des Kakaoaromas Fett Wiss Technol 78(2), 8895 ă Mohr, W., Răohrle, M and Severin, T (1971) Uber die Bildung des Kakaoaromas aus seinen Vorstufen Fett Wiss Technol 73(8), 515–521 Morton, I.D., Akroyd, P and May, C.G (1960) Flavouring substances and their preparation GB Patent 836694 Mottram, D.S., Ames, J.M., Mlotkicwiez, J.A., Copsey, J.P and Anderson, A (1998) Flavouring agents World Patent 98/42208 Mottram, D.S and Madruga, M.S (1994) Important sulfur-containing aroma volatiles in meat In: Sulfur Compounds in Foods (eds C.J Mussinan and M.E Keelan), American Chemical Society, Washington, DC, pp 180–187 Mottram, D.S and Nobrega, C.C (1998) Formation of volatile sulfur compounds in reaction mixtures containing cysteine and three different ribose compounds In: Food Flavors: Formation, Analysis and Packaging Influences (eds E.T Contis, C.-T Ho and C.J Mussinan), Elsevier Science, Amsterdam, pp 483–492 Măunch, P and Schieberle, P (1998) Quantitative studies on the formation of key odorants in thermally-treated yeast extracts using stable isotope dilution assays J Agric Food Chem 46, 4695–4701 Nedvidek, W., Ledl, F and Fischer, P (1992) Detection of 5-hydroxymethyl-2-methyl-3(2H)-furanone and of ␣-dicarbonyl compounds in reaction mixtures of hexoses and pentoses with different amines Z Lebens.-Unters Forsch 194, 222–228 Nestl´e (1966) Flavouring agent GB Patent 1032334 Ottinger, H., Bareth, A and Hofmann, T (2001b) Characterization of natural ‘cooling’ compounds formed from glucose and L-proline in dark malt by application of taste dilution analysis J Agric Food Chem 49, 1336–1344 Ottinger, H., Hofmann, T (2003) Identification of the taste enhancer alapyridaine in beef broth and evaluation of its sensory impact by taste reconstitution experiments J Agric Food Chem 51, 6791–6796 Ottinger, H., Soldo, T., Hofmann, T (2001a) Systematic studies on structure and physiological activity of cyclic ␣-keto enamines, a novel class of ‘cooling’ compounds J Agric Food Chem 49, 5383–5390 Parliment, T.H., McGorrin, R.J and Ho, C.-T (eds) (1989) Thermal Generation of Aromas, American Chemical Society, Washington, DC Parliment, T.H., Morello, M.J and McGorrin, R.J (eds) (1994) Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes, American Chemical Society, Washington, DC Pippen, E.L and Mecchi, E.P (1969) Hydrogen sulfide, a direct and potentially indirect contributor to cooked chicken aroma J Food Sci 34, 443–446 Pittet, A.O., Rittersbacher, P and Muralidhara, R (1970) Flavor properties of compounds related to maltol and isomaltol J Agric Food Chem 18, 929–932 86 Food Flavour Technology Pittet, A.O and Seitz, E.W (1974) Flavoring compositions and processes US Patent 3782973 Reineccius, G.A (1990) The influence of Maillard reactions on the sensory properties of foods In: The Maillard Reaction in Food Processing, Human Nutrition and Physiology (eds P.A Finot, H.U Aeschbacher, R.F Hurrell and R Liardon), Birkhăauser Verlag, Basel, Switzerland, pp 157–170 Reineccius, G.A (1998) Kinetics of flavor formation during Maillard browning In: Flavor Chemistry: Thirty Years of Progress (eds R Teranishi, E.L Wick and I Hornstein), Kluwer Academic/Plenum Publishers, New York, pp 345–352 Roberts, D.D and Acree, T.E (1994) Gas chromatography-olfactometry of glucose-proline Maillard reaction products In: Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes (eds T.H Parliment, M.J Morello and R.J McGorrin), American Chemical Society, Washington, DC, pp 7179 Răodel, W., Habisch, D and Ruttloff, H (1988) Formation of cocoa flavour by Maillard reaction Charact Prod Appl Food Flavours 2, 301309 Rolli, K., Răoschli, D and Sihver, J.J (1988) Process for preparing a flavouring ingredient European Patent 286838 Rosing, E.A.E and Turksma, H (1997) Process for the preparation of a savoury flavour European Patent 784936 Rusoff, I.I (1958) Flavor US Patent 2835592 Rychlik, M and Grosch, W (1996) Identification and quantification of potent odorants formed by toasting of wheat bread Food Sci Technol (London) 29, 515–525 Schieberle, P (1990) The role of free amino acids present in yeast as precursors of the odorants 2-acetyl1-pyrroline and 2-acetyltetrahydropyrridine in wheat bread crust Z Lebensm.-Unters Forsch 191, 206–209 Schieberle, P (1991a) Primary odorants in popcorn J Agric Food Chem 39, 1141–1144 Schieberle, P (1991b) Primary odorants of pale lager beer Differences to other beers and changes during storage Z Lebensm.-Unters Forsch 193, 558565 Schieberle, P (1992a) Bildung wichtiger Răostaromastoffe in Lebensmitteln aus Getreide Getreide, Mehl und Brot 46, 338–342 Schieberle, P (1992b) Formation of Furaneol in heat-processed foods In: Flavour Precursors – Thermal and Enzymatic Conversion (eds R Teranishi, G.R Takeoka and M Găuntert), ACS Symposium Series 490, Washington, DC, pp 164–174 Schieberle, P (1993a) Studies on the flavour of roasted white sesame seeds In: Progress in Flavour Precursor Studies – Analysis, Generation and Biotechnology (eds P Schreier and P Winterhalter), Allured Publishing, Carol Stream, IL, pp 343–360 Schieberle, P (1993b) Untersuchungen zum Aromabeitrag und zur Bildung von 4-Hydroxy-2,5-dimethyl3(2H)-furanon in thermisch behandelten Lebensmittel Lebensmittelchemie 47, 15–16 Schieberle, P (1995) Quantification of important roast-smelling odorants in popcorn by stable isotope dilution assays and model studies on flavor formation during popping J Agric Food Chem 43, 2442– 2448 Schieberle, P (1996) Odor-active compounds in moderately roasted sesame Food Chem 55, 145–152 Schieberle, P (2005) The carbon module labeling (CAMOLA) technique: a useful tool for identifying transient intermediates in the formation of Maillard-type target molecules Ann N Y Acad Sci 1043, 236–248 Schieberle, P., Fischer, R and Hofmann T (2003) The carbohydrate module labeling technique: a useful tool to clarify formation pathways of aroma compounds formed in Maillard-type reactions In: Flavour Research at the Dawn of the Twenty-First Century, Proceedings of the 10th Weurman Flavour Research Symposium (eds J.-L Le Qu´er´e and P.X Eti´evant), Lavoisier, Intercept, London, 447–452 Schieberle, P and Grosch, W (1987) Evaluation of the flavor of wheat and rye bread crusts by aroma extract dilution analysis Z Lebensm.-Unters Forsch 185, 111–113 Schieberle, P and Grosch, W (1994) Potent odorants of rye bread crust Differences from the crumb and wheat bread crust Z Lebensm.-Unters Forsch 198, 292–296 Schieberle, P and Hofmann, T (1998) Characterization of key odorants in dry-heated cysteine-carbohydrate mixtures: comparison with aqueous reaction systems In: Flavor Analysis (eds C.J Mussinan and M.J Morello), American Chemical Society, Washington, DC, pp 320–330 Schieberle, P and Hofmann, T (2002) New results on the formation of important Maillard aroma compounds In: Advances in Flavours and Fragrances: From the Sensation to the Synthesis (ed K.A Swift), Royal Society of Chemistry, Cambridge, pp 163–177 Basic chemistry and process conditions for reaction flavours 87 Schieberle, P., Hofmann, T and Munch, P (2000) Studies on potent aroma compounds generated in Maillardtype reactions using the odor-activity-value concept ACS Symp Ser 756, 133–150 Schlichtherle-Cerny, H., Affolter, M., Blank, I., Cerny, C., Robert, F., Beksan, E., Hofmann, T and Schieberle, P (2002) Amadori and Heyns rearrangement products as flavoring compounds for imparting umami taste to food products Eur Pat Appl EP 1252825 A1 Schnermann, P and Schieberle, P (1997) Evaluation of key odorants in milk chocolate and cocoa mass by aroma extract dilution analysis J Agric Food Chem 45, 867–872 Schăonberg, A and Moubacher, R (1952) The Strecker degradation of a-amino acids Chem Rev 50, 261–277 Semmelroch, P and Grosch, W (1995) Analysis of roasted coffee powders and brews by gas chromatography–olfactometry of headspace samples Food Sci Technol (London) 28, 310–313 Semmelroch, P., Laskawy, G., Blank, I and Grosch, W (1995) Determination of potent odorants in roasted coffee by stable isotope dilution assays Flavour Frag J 10, 1–7 Shibamoto, T (1983) Heterocyclic compounds in browning and browning/nitrite model systems In: Instrumental Analysis of Foods, Vol (eds G Charalambous and G Inglett), Academic Press, New York, pp 229–277 Shu, C.-K and Ho, C.-T (1989) Parameter effects on the thermal reaction of cystine and 2,5-dimethyl-4hydroxy-3(2H)-furanone In: Thermal Generation of Aromas (eds T.H Parliment, R.J McGorrin and C.-T Ho), American Chemical Society, Washington, DC, pp 229–241 Soldo, T., Blank, I., Hofmann, T (2003) (+)-(S)-Alapyridaine – a general taste enhancer? Chem Senses 28, 371–379 Stahl, H.D and Parliment, T.H (1994) Formation of Maillard products in the proline-glucose model system – high-temperature short-time kinetics In: Thermally Generated Flavors, Maillard, Microwave and Extrusion Processes (eds T.H Parliment, M.J Morello and R.J Mc Gorrin), American Chemical Society, Washington, DC, pp 251–262 Tai, C.-Y and Ho, C.-T (1997) Influence of cysteine oxidation on thermal formation of Maillard aromas J Agric Food Chem 45, 3586–3589 Tandy, J.S (1985) Chicken flavorants and the processes for preparing them GB Patent 2157538 Teranishi, R., Buttery, R.G and Guadagni, D.G (1975) In: Geruchs- und Geschmacks-stoffe (ed F Drawert), Verlag Hans Carl, Năurnberg, Germany, pp 177–186 Theron, P.P.A., von Fintel, R.G.H., Saisselin, A.L and Vissers, A.M (1975) Flavouring agents GB Patent 1382335 Tressl, R (1989) Formation of flavor compounds in roasted coffee In: Thermal Generation of Aromas (eds T.H Parliment, R.J McGorrin and C.T Ho), American Chemical Society, Washington, DC, pp 285–301 Tressl, R., Helak, B., Kersten, E and Rewicki, D (1993) Formation of flavor compounds by Maillard reaction In: Recent Developments in Flavor and Fragrance Chemistry (eds R Hopp and K Mori), Verlag Chemie, Weinheim, Germany, pp 167–181 Tressl, R., Nittka, C and Kersten, E (1995) Formation of isoleucine-specific Maillard products from [113 C]-d-glucose and [1-13 C]-d-fructose J Agric Food Chem 43, 1163–1169 Tressl, R and Rewicki, D (1999) Heat generated flavors and precursors In: Flavor Chemistry: Thirty Years of Progress (eds R Teranishi, E.L Wick and I Hornstein), Kluwer Academic/Plenum Publishers, New York, pp 305–325 Turksma, H (1993) Process for the preparation of savoury flavours European Patent 571031 Ullrich, F and Grosch, W (1987) Identification of the most intense volatile flavor compounds formed during autoxidation of linoleic acid Z Lebensm.-Unters Forsch 184, 277–282 Van den Ouweland, G.A.M., Demole, E.P and Enggist, P (1989) Process meat flavor development and the Maillard reaction In: Thermal Generation of Aromas (eds T.H Parliment, M.J Morello, R.J McGorrin and C.T Ho), American Chemical Society, Washington, DC, pp 433–441 Van den Ouweland, G.A.M and Peer, H.G (1968) Mercapto furane and mercapto thiophene derivatives GB Patent 1283912 Van den Ouweland, G.A.M and Peer, H.G (1970) Synthesis of 3,5-dihydroxy-2-methyl-5,6-dihydropyran4-one from aldohexoses and secondary amine salts Recl Trav Chim Pay B 89, 750–754 Van den Ouweland, G.A.M and Peer, H.G (1975) Components contributing to beef flavor Volatile compounds produced by the reaction of 4-hydroxy-5-methyl-3(2H)-furanone and its thio analog with hydrogen sulfide J Agric Food Chem 23, 501–505 Van Pottelsberghe de la Potterie, P.J (1972) Verfahren zur Herstellung von Geschmacksstoffen mit Rinderbratenaroma German Patent 2149682 Van Pottelsberghe de la Potterie, P.J (1973) Beef flavor US Patent 3716380 88 Food Flavour Technology Vauthey, S., Leser, M and Milo, C (1998) An aroma product comprising saturated C16 and C18 monoglycerides European Patent 1008305 Wagner, R and Grosch, W (1997) Evaluation of potent odorants of French fries Food Sci Technol (London) 30, 164–169 Weenen, H (1998) Reactive intermediates and carbohydrate fragmentation in Maillard chemistry Food Chem 62, 393–401 Weenen, H and Apeldoorn, W (1996) Carbohydrate cleavage in the Maillard reaction In: Flavour Science: Recent Developments (eds A.J Taylor and D.S Mottram), The Royal Society of Chemistry, Cambridge, pp 211–216 Weenen, H., Kerler, J and van der Ven, J (1997) The Maillard reaction in flavour formation In: Flavours and Fragrances (ed K.A.D Swift), The Royal Society of Chemistry, Cambridge, pp 153–171 Weenen, H and Tjan, S.B (1992) Analysis, structure, and reactivity of 3-deoxyglucosone In: Flavor Precursors: Thermal and Enzymatic Conversions (eds R Teranishi, G.R Takeoka and M Găuntert), American Chemical Society, Washington, DC, pp 217231 Weenen, H and Tjan, S.B (1994) 3-Deoxyglucosone as flavour precursor In: Trends in Flavour Research (eds H Maarse and D.G van der Heij), Elsevier Science, Amsterdam, pp 327–337 Weenen, H and van der Ven, J.G.M (1999) Formation of Strecker aldehydes In: Book of Abstracts, 218th ASC National Meeting, New Orleans, American Chemical Society, Washington, D.C Weenen, H., van der Ven, J.G.M., van der Linde, L.M., van Duynhoven, J and Groenewegen, A (1998) C4, C5, and C6 3-deoxyosones: structures and reactivity In: The Maillard Reaction in Foods and Medicine (eds J O’Brien, H.E Nursten, M.J.C Crabbe and J.M Ames), The Royal Society of Chemistry, Cambridge, pp 57–64 Werkhoff, P., Bretschneider, W., Emberger, R., Găuntert, M., Hopp, R and Kăopsel, M (1991) Recent developments in the sulfur flavour chemistry of yeast extracts Chem Mikrobiol Technol Lebensm 13, 30–57 Whitfield, F.B (1992) Volatiles from interactions of Maillard reactions and lipids Crit Rev Food Sci Nutr 31, 1–58 Whitfield, F.B and Mottram, D.S (1999) Investigation of the reaction between 4-hydroxy-5-methyl-3(2H)furanone and cysteine or hydrogen sulfide at pH 4.5 J Agric Food Chem 47, 1626–1634 Widder, S., Sabater Lăuntzel, C., Dittner, T and Pickenhagen, W (2000) 3-Mercapto-2-methylpentan-1-ol, a new powerful aroma compound J Agric Food Chem 48, 418–423 Yaylayan, V.A., Forage, N.G and Mandeville, S (1994) Microwave and thermally induced Maillard reactions In: Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes (eds T.H Parliament, M.J Morello and R.J McGorrin), American Chemical Society, Washington, DC, pp 449–456 Yaylayan, V.A and Huyghues-Despointes, A (1994) Chemistry of Amadori rearrangement products: analysis, synthesis, kinetics, reactions and spectroscopic properties Crit Rev Food Sci Nutr 34, 321–369 Yaylayan, V.A and Keyhani, A (1999) Origin of 2,3-pentanedione and 2,3-butanedione in d-glucose/lalanine Maillard model systems J Agric Food Chem 47, 3280–3284 Yaylayan, V.A and Keyhani, A (2000) Origin of carbohydrate degradation products in l-alanine/d[13 C]glucose model systems J Agric Food Chem 48, 2415–2419 Yaylayan, V and Sporns, P (1987) Novel mechanisms for the decomposition of 1-(amino acid)-1-deoxy-dfructoses (Amadori compounds): a mass spectrometric approach Food Chem 26, 283–305 Yeretzian, C., Blank, I and Palzer, S (2007) Process flavourings In: Flavourings Production, Composition, Applications, Regulations (ed H Ziegler), Wiley-VCH, Weinheim, Germany, pp 549–572 Zamora, R and Hidalgo, F.J (2005) Coordinate contribution of lipid oxidation and Maillard reaction to the nonenzymatic food browning Crit Rev Food Sci Nutr 45, 49–59 Zhang, Y and Ho, C.-T (1989) Volatile compounds formed from thermal interaction of 2,4-decadienal with cysteine and glutathione J Agric Food Chem 37, 1016–1020 Zhang, Y and Ho, C.-T (1991) Formation of meatlike aroma compounds from thermal reaction of inosine -monophosphate with cysteine and glutathione J Agric Food Chem 39, 1145–1148 Zheng, Y., Brown, S., Ledig, W.O., Mussinan, C and Ho, C.-T (1997) Formation of sulfur-containing flavor compounds from reactions of Furaneol and cysteine, glutathione, hydrogen sulfide, and alanine/hydrogen sulfide J Agric Food Chem 45, 894–897 4 Biotechnological flavour generation Ralf G Berger, Ulrich Krings and Holger Zorn 4.1 INTRODUCTION Aromas and flavours possess antimicrobial, medicinal and signalling properties, as well as acting as food preservatives They were even used to preserve human corpses in ancient Egypt Above all, it is their alluring sensory properties that promised financial profits high enough to prompt the Phoenicians, Arabs, and, later, the Portuguese, Dutch, Spanish and Venetians to discover and conquer entire countries Today’s captains are scientists, their ocean is the metabolic flow, their ships and weapons are gene shuttles and biochemical knowledge, and their targets are no longer leaves, fruits and seeds, but cells, enzymes and genes This chapter describes the following: r r r r r How micro-organisms transform flavour precursors directly into flavour molecules (biotransformation) or produce flavours along multi-step processes (bioconversion and de novo synthesis) How enzymes catalyse hydrolytic or other chemical reactions leading to flavours How single plant cell in sterile culture may replace field-grown flavour producers How the dissemination of methods of genetic engineering fertilises bioflavour research How laboratory developments have been successfully transferred into industrial applications No mention is made of flavours derived from traditional and genetically engineered starter cultures, because informative reviews exist, e.g on fermented meat flavours (Tjener and Stahnke, 2007), on dairy flavours (Smit et al., 2004), on bread flavours (Hansen and Schieberle, 2005) and on wine flavours (Swiegers et al., 2008) 4.2 NATURAL FLAVOURS: MARKET SITUATION AND DRIVING FORCES Beverages are the largest market segment for added flavours, accounting for about one-third of the worldwide flavour sales Around 90% of the flavours added to beverages in the EC are natural, and about 80% in the US; savoury products contain about 80% naturals in both the EC and the US, while the percentages for flavoured dairy products are 50% in the EC and 75% in the US (the remaining proportions are nature-identical and artificial flavours) These numbers clearly reflect a strong consumer preference for naturalness This scientifically 90 Food Flavour Technology unfounded, vague chemophobia has also been extended to furniture, clothing and cosmetics As a result, the attribute ‘natural’ is an excellent marketing point, and the use of ‘natural’ flavour sources is the last bastion for food manufacturers to suggest the naturalness of their product explicitly According to the definitions of the Code of Federal Regulations in the US (CFR, 1993) and to the mandatory guidelines of the Council of the European Communities (88/388/EWG of 22 June 1988; 91/71/EWG and 91/72/EWG of 16 January 1991), aromas generated by biotechnology are classified as natural if the starting materials used were ‘natural’ The revised flavour regulation, scheduled for coming into effect in autumn 2010, will no longer distinguish between nature-identical and artificial flavours; they will be termed ‘flavouring substances’ (EC 1334/2008 of 16 December 2008) Article 3.2(c) defines a ‘natural flavouring substance’ as ‘a flavouring substance obtained by appropriate physical, enzymatic or microbiological processes from material of vegetable, animal or microbiological origin either in the raw state or after processing for human consumption by one or more of the traditional food preparation processes listed in Annex II Natural flavouring substances correspond to the substances that are naturally present and have been identified in nature’ (www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+TA+P6-TA-20080331+0+DOC+XML+V0//EN) The flavour industry, accounting for some legal uncertainties, has adopted a productoriented point of view on the basis of the GRAS (generally recognised as safe; Waddell et al., 2007) procedure and has set its own standards The guidelines of the International Organisation of the Flavour Industry demand that the bioprocessing of flavours is in accordance with good manufacturing practice and with the general principles of hygiene of the Codex Alimentarius Bioflavours shall comply with national legislation, and safety must be ‘adequately established’ Doubts about contamination with microbial toxins or the like have been resolved as the volatile nature of aroma compounds facilitates efficient downstream processing and exclusion of recombinant nucleic acids or other undesired non-volatiles by distillation or lipophilic extraction This is thought to sufficiently provide protection of the consumer’s health The ‘all-natural’ mania of the consumer and the favourable legal situation are, however, only two of the forces driving the development of bioflavours A bundle of driving forces that may be subdivided into ‘business pull’ and ‘technical push’ can be listed Among the business factors listed are the decline of availability of some traditional raw materials, an increasing market size, modern processes and ingredient formulations, and the growing importance of functional, ethnic and exotic products (Bauer, 2000; Cheetham, 2004) Technical factors include again and again improved analytical methods (Steinhart et al., 2000), improved bioprocessing techniques, genetic engineering, and an improving understanding of structure–activity relationships of flavours and fragrances (for example, ‘olfactophore models’; Kraft et al., 2000) 4.3 ADVANTAGES OF BIOCATALYSIS If a chemosynthetic route results in a mixture of products or isomers, a subsequent separation may be more expensive than a comparable but more selective bioprocess Trace impurities of a chemosynthetic compound may adulterate the sensory character of a product, and sensory activity is often dependent on an exact stereochemical structure, as the sense of smell is chiral Biocatalysts not only provide high regiospecificity and stereospecificity but show Biotechnological flavour generation 91 high reaction velocity even at low molar fraction Although biocatalysts are not free from inherent drawbacks, such as operational instability (isolated enzymes), processing cost and sometimes occurrence of side reactions, the advantages of biocatalysts, such as ecological compatibility or multi-step synthesis, cannot be matched by a chiral chemical catalyst While the closing of mass cycles and sustainable production have gained high priority in the chemical industries, nature itself sets the example Biocatalysis will not provide immediate solutions for all fine chemicals, but highly prized flavours and fragrances appear to present a particularly suitable playground for biotechnologists Many authors have discussed one class or various groups of aroma compounds under quite different aspects For an introduction to the literature until 1997 and a description of some ´ evant and Schreier processes already operating, the reader is referred to Berger (1995), Eti´ (1995) and Berger (1997) More recent biotechnological advances, which will broaden and facilitate the use of biocatalysts for industrial production of volatile chemicals, are described in this chapter 4.4 MICRO-ORGANISMS Intact microbial cells possess active transport systems, enzyme arrangements optimised by evolution, and they regenerate their biocatalytic molecules together with the co-factors required; costs arise for the equipment needed to maintain a suitable biochemical environment, but costs for isolation and stabilisation of the biocatalyst are inapplicable Bacterial pathways to volatile flavours are often based on strong hydrolytic properties Incomplete substrate oxidation and Strecker degradation of amino acids yield a spectrum of products (Rizzi, 2008) Yeasts, as unicellular non-filamentous fungi, are more complex eukaryotic cells that show a much more diverse biochemistry including a broad range of volatile metabolites, such as esters, lactones, aldehydes and phenolics (Debourg, 2000) The most developed fungal species, the class of basidiomycetes, show a complicated sexual cycle, pseudo-tissue formation, and the distinct ability to degrade native cellulose or lignin aerobically Volatile flavours from all chemical classes were found in basidiomycete fruit bodies and cell cultures, with a particular emphasis on volatile phenylpropanoic and phenolic compounds (Berger and Zorn, 2004; Wu et al., 2007) 4.4.1 Biotransformation and bioconversion of monoterpenes Many essential oils are dominated by monoterpene hydrocarbons Because of their low sensory activity, low water solubility and tendency to autoxidise and polymerise, they are usually rectified from the oil and regarded as processing waste These properties in conjunction with their role as physiological precursors of high-valued oxyfunctionalised terpenoids turn terpene hydrocarbons, such as limonenes, pinenes and terpinenes, into ideal starting materials for microbial transformations (Berger et al., 1999b; de Carvalho and da Fonseca, 2006) The amount of R-(+)-limonene separated from cold-pressed citrus peel oil was estimated at 36 000 tons per year, and the steam distillation of pine oils delivered 160 000 tons of ␣-pinene and 26 000 tons of ␤-pinene (Ohloff, 1994; Nonino, 1997) The first systematic studies on terpene hydrocarbon transformation date back to the early 1960s (Bhattacharyya et al., 1960; Prema and Bhattacharyya, 1962), and since then a vast number of publications on the subject has appeared 92 Food Flavour Technology Fig 4.1 Products of allylic oxidations: (I) carveol; (II) perillyl alcohol; (III) p-mentha-1,8-diene-4-ol; (IV) p-mentha-2,8-diene-1-ol; (V) isopiperitenol; (VI) verbenol; (VII) myrtenol 4.4.1.1 Allylic hydroxylation From a biotechnological point of view, the allylic hydroxylation is one of the most important transformation reactions leading to compounds with direct or indirect economic importance (carveol, borneol, verbenol, nootkatol) Concurrent rearrangements of double bonds are explained either by radical transition states, as they occur at the end of the cytochrome P450 mono-oxygenase cycle, or by radical-cationic intermediates, typical of peroxidase-mediated reactions Transformation organisms range from bacteria, such as Pseudomonas through deuteromycetes, such as Penicillium or Aspergillus to higher fungi, such as the basidiomycetes A number of frequently occurring transformation products of limonene and ␣-pinene are compiled in Fig 4.1 High regioselectivities have been observed While some strains preferentially attacked ring carbons, other strains hydroxylated exocyclic allylic positions Because of its abundance, limonene has been the most popular transformation substrate in recent years Its regiospecific hydroxylation by a strain of Pseudomonas putida yielded up to g of perillic acid per litre within days (Speelmans et al., 1998) The success of this transformation was based on the selection of a limonene-tolerant micro-organism, biphasic operation, and careful optimisation of reaction conditions including the co-substrate glycerol The same strategy of pre-screening microbial strains, using substrate-enriched nutrient media, succeeded in the case of Pseudomonas alcaligenes, which transformed (+)limonene into the hydration product ␣-terpineol (Teunissen and de Bont, 1995) However, when 120 Gram-positive bacterial strains were selected by their growth on carvone as the sole source of carbon, none of them transformed limonene efficiently to carvone (Van der Werf and De Bont, 1998) Other regiospecific transformations of (+)-limonene were observed for the basidiomycete Pleurotus sapidus, which produced mainly carveols and carvone (Onken and Berger, 1999a), and for a non-conventional Hormonema black yeast to yield Biotechnological flavour generation 93 trans-isopiperitenol (Van Rensburg et al., 1997) In the latter case, a varying morphological appearance was associated with unstable product yields The enantiospecific transformation of racemic limonene to (+)-␣-terpineol by Penicillium digitatum (Tan et al., 1998) encouraged these researchers to immobilise the cells in calcium alginate and to perform an airlift reactor study (Tan and Day, 1998) While good molar conversion was achieved, the absolute yields remained in the lower milligram per litre range Marostica and Pastore (2007) have covered recent work on limonene biotransformation Starting with a report on the microbial hydroxylation of ␣-pinene (Bhattacharyya et al., 1960), the history of pinene transformation now spans more than four decades Work on the transformation of the pinenes, as well as other bicyclic monoterpenes, such as cineoles, camphor and carene, is summarised by Trudgill (1994) Recently, classical (ultra-violet) UVmutagenesis preceded the selection of strains of Aspergillus, which converted ␣-pinene either to verbenol (Agrawal et al., 1999) or to verbenone (Agrawal and Joseph, 2000) This group has recently studied (+)-␣-pinene transformation by a Pseudomonas strain and found high yields of several volatiles, among them a novel flavour, dihydrocarveol acetate (Divyashree et al., 2006) As long as the public concern about genetic engineering applications in the food sector persists, random mutagenesis will continue to claim scientific interest The transformation of monoterpenols, such as citronellol, geraniol and nerol, using Botrytis cinerea followed the same general routes (Bock et al., 1988) Cystoderma carcharias, a basidiomycete, transformed citronellol to 3,7-dimethyl-1,6,7-octanetriol and several minor products, among them the flavour-impact compounds E/Z-rose oxide (Onken and Berger, 1999b) Nerol or citral was transformed by isolates of Penicillium and Aspergillus to a number of other monoterpenes and to the fruity smelling C2 -shortened 6-methyl-5-hepten-2-one (Demyttenaere and De Pooter, 1998; Demyttenaere and De Kimpe, 2000) A microcultivation method in conjunction with solid-phase micro-extraction was used to speed up the process monitoring on a laboratory scale Another study confirmed the minor role of yeasts in the formation of terpenes of wine and beer flavour (King and Dickinson, 2000) Neither Saccharomyces cerevisiae nor Kluyveromyces lactis or Torulospora delbrueckii accumulated significant amounts of the transformation products – linalool, ␣-terpineol and terpin hydrate Low yields and the lack of stereoselectivity for linalool and ␣-terpineol suggest that chemical transformation overlapped with the biocatalysed routes 4.4.1.2 Oxidation of non-activated carbons This reaction may be called the Holy Grail of biotechnology However, there will be a low chance of incidence with oligo-isoprenoid compounds, as most of them carry carbon–carbon double bonds, where attack will be favoured Aliphatic bicyclic compounds, such as fenchol, borneol and 1,8-cineole, were shown by the group of Kieslich to be selectively converted to mono- and diols by a Zygomycete, Diplodia bisporus and by Bacillus cereus (Abraham et al., 1988) A chemically related reaction is the terminal hydroxylation of saturated fatty acids catalysed by Torulopsis yeasts, a reaction permitting access to macrocyclic musks (Williams, 1999) 4.4.1.3 Epoxidation Epoxidations are typical cytochrome P450-mediated reactions The reaction products were either isolated as stable products or considered to be intermediates, on the basis of the 94 Food Flavour Technology Fig 4.2 Degradation pathway of ␣-pinene in Pseudomonas fluorescens vicinal diols identified Obviously, the pH of the incubation medium governs the extent of hydrolysis Asymmetric dihydroxylation of alkene moieties was obtained when strains of P putida were exposed to isoprene or to related dienes (Boyd et al., 2000) Further, diol oxidation was successfully controlled by addition of propylene glycol as an inhibitor Substrates with high ring tension, such as ␣-pinene epoxide, were often transformed by fungal strains to a number of products, among them campholenic aldehyde, trans-sobrerol, and E/Zcarveols In such cases, biocatalysis appears at first glance not to have any advantage over a chemical transformation The same epoxide substrate, however, was efficiently degraded by Pseudomonas fluorescens and Nocardia strains to a small set of products belonging to one specific pathway (Best et al., 1987; Griffiths et al 1987; Fig 4.2) An unusual (4R,8R)-limonene-8,9-epoxide was the only transformation product when a Xanthobacter isolate selected on cyclohexane as the sole carbon source was exposed to (4R)-limonene (Van der Werf et al., 2000) Many other transformation pathways of limonene were compiled by the same authors (Van der Werf et al., 1999) The conversion of ␤-myrcene by the edible fungus Pleurotus ostreatus was investigated using trideutero-labelled ␤-myrcene or presumed intermediates as the substrates Myrcene diols were formed from the cleavage of several myrcene epoxides identified as the immediate reaction products (Krings et al., 2008) The diverse substrates and reactions catalysed by cytochrome P450 isoforms were summarised in a recent review (Bernhardt, 2006) 4.4.1.4 Oxidation of alcohols Interest in the widespread oxidation of primary and secondary terpenols rests on the fact that many terpene aldehydes and ketones are more volatile and more powerful (in flavour terms) than the corresponding alcohols Yeasts and some higher fungi may deliver the target compounds in yields close to 100% Further oxidation to carboxylic acids is rarely observed (Fig 4.2 shows one of the few exceptions), because the free acyl moieties are quickly transformed to activated conjugates for further metabolism through ␤-oxidation and related pathways Another pathway along which terpene acids are formed is the Baeyer–Villiger oxidation: P putida strains were able to insert an oxygen atom next to a carbonyl function, and the subsequent ring opening yielded a degradable acid intermediate (Abraham et al., 1988) 4.4.1.5 Hydration Hydration of a double bond or of a ring bridge may proceed as an acid- or enzyme-catalysed reaction As always, a cell-free control experiment is indispensable to assess the extent of chemical side reactions As a rule of thumb, troublesome chemical hydration occurs at ambient temperature only at a pH Ͻ Useful information on this aspect is sometimes hidden in reports on off-flavour formation in acidic food matrices, such as citrus juices (Haleva-Toleo Biotechnological flavour generation 95 et al., 1999) The required biocatalytic activity appears to be located mainly in bacteria, such as Escherichia coli, Bacillus stearothermophilus and Pseudomonos gladioli, but also in P digitatum Candida tropicalis transformed ␣-pinene to ␣-terpineol in a previously unmatched yield of 77% (Chatterjee et al., 1999a) 4.4.2 Bioconversion of C13 -norisoprenoids and sesquiterpenes The same fundamental transformation steps as for monoterpenes apply for larger oligoisoprenoids and related compounds The increased structural complexity, however, allows more substructures of the molecules to fit into the active sites of enzymes, resulting in a mixture of products with a broad range of concentrations C13 -Norisoprenoids of the ionone/damascone group have received particular attention, because they belong to the small group of highly potent character-impact components with pleasant floral fruity flavours In tomatoes and maize, a carotenoid cleavage dioxygenase was found to asymmetrically cleave numerous carotenoids to yield, for example, 6-methyl-5-hepten-2-one from lycopene (Vogel et al., 2008) Similar activities were reported from higher fungi A concerted biogeneration of ␤-ionone and some related compounds has become known using submerged grown fungal mycelium (Zorn et al., 2003a) Several of these fungal enzymes, all so far belonging to the large peroxidase family, were isolated from the source basidiomycete, sequenced, and cloned, and one of them is now commercial (de Boer et al., 2005) A summary on this subject is available (Rodriguez-Bustamante and Sanchez, 2007) A broad screening of Streptomyces strains showed that some strains transformed ␣-ionone regioselectively to 3-hydroxyionone without many side products (Lutz-Wahl et al., 1998) Racemic ␣-ionone was selectively transformed to trans-(3R,6R) and (3S,6S) alcohols In contrast, there was limited transformation of ␤-ionone to the 4-hydroxy product with no formation of the 3-hydroxy product Strains of Aspergillus niger metabolised ␤-ionone much better and yielded almost 100% conversion to 3- and 4-hydroxy products in a fedbatch process (Larroche et al., 1995) An advanced mathematical model of the mass transfer rates in this two-phase system was presented (Grivel et al., 1999) Because of the observation that the growth inhibition of B cinerea by patchoulol faded with time, the metabolic fate of the fungicide was followed (Aleu et al., 1999) Numerous hydroxy compounds were found, indicating first steps of a detoxification pathway As these hydroxy compounds are substrates for further degradation, only transient accumulation of volatile oxidation products can be expected The group of Miyazawa has issued a series of papers looking, for example, at the conversion of substrates, such as (+)-cedrol, (+)-aromadendrene, (−)-␣-bisabolol, ␤-selinene, (−)-globulol, ␥ -gurjunene and farnesol isomers (Miyazawa et al., 1997, 1998; Nankai et al., 1998, and references therein) A plant pathogenic fungus, Glomerella cingulata, was the strain of choice, and the problem of low substrate solubility was partially circumvented by using monoalcohols instead of the hydrocarbon compounds As a general outline, exomethylene carbons and isopropyl substituents were oxidised non-stereoselectively, while the oxidation at ring positions proceeded stereoselectively, indicating a more restricted conformational situation at the active site of the enzyme Some evidence is accumulating that ␤-ionone and other terpenes (De-Oliveira et al., 1999) and sesquiterpenes (Sime, 2000) are quite efficient inhibitors of cytochrome enzymes This would mean that low substrate solubility was responsible for the not only low product 96 Food Flavour Technology yields reported in the past but also the inhibition of the first step of substrate detoxification, particularly if the substrate was administered to the cells all at once 4.4.3 Generation of oxygen heterocycles A number of oxygen-containing heterocycles, such as the well-known flavour compounds maltol and Furaneol (and some alkanolides), have received much attention because, besides possessing flavour themselves, they may also affect the smell, taste and umami impressions of other flavour compounds Furaneol (2,5-dimethyl-4-hydroxy-2H-furan-3-one) was obtained by converting 6-deoxyhexoses along a Maillard-imitating soft chemistry route (Whitehead, 1998) This approach suffers, however, from the lack of economic sources of the precursor 6-deoxy sugars Furaneol and the related 2- (or 5-) ethyl-5- (or 2-) methyl-4-hydroxy2H-furan-3-one have been frequently found in oriental food fermented in the presence of certain yeasts (Hayashida et al., 1998) Concentrations of the two compounds (in the milligram per litre range) accumulated after supplementing Zygosaccharomyces rouxii in simple fermentation media with an autoclaved mixture of a single amino acid and reducing sugars (Hayashida et al., 1999) Hexoses favoured the formation of Furaneol, while pentoses, such as ribose, stimulated formation of the ethyl-substituted furanone with a nonlinear correlation Glutamate was the most efficient source of nitrogen The authors, well aware of the competing chemical pathway of formation, described the formation of furanones in aging miso and simple fermentation media as a yeast-dependent reaction and suggested that ‘many amino acids form compounds able to act as Furaneol precursors’ The same experiments did not result in the formation of significant amounts of furanones, when the heating of the precursor combination was replaced by a filter sterilisation (0.22 ␮m) protocol Even higher concentrations of 2- (or 5-) ethyl-5- (or 2-) methyl-4-hydroxy-2H-furan-3-one (74.3 mg/L in the presence of l-alanine) were reported subsequently (Sugawara and Sakurai, 1999) The occurrence of volatile lactones, such as 4,5-dimethyl-3-hydroxy-5H-furan-2-one (sotolon), in higher fungi (Liz´arraga-Guerra et al., 1997) underscores the potential of fungal cells as catalysts in bioprocesses The most common pathway is ␤-oxidative degradation of hydroxy fatty acids followed by intramolecular esterification to the respective 4- and 5-alkanolides with fruity and fatty odour notes, but a number of different pathways have also been reviewed (Gatfield, 1997) 6-Pentyl-␣-pyrone is a lactone with coconut-like flavour, and its formation by Trichoderma harzianum, Trichoderma viride and other fungi is well documented Various surface and submerged fermentations (Kalyani et al., 2000) and solid-state fermentation on sugarcane bagasse (Sarhy-Bagnon et al., 2000) have yielded up to almost g of volatile product per litre The lactone inhibits the growth of some phytopathogenic fungi and may be formed as a part of the chemical self-defence system of the Trichoderma fungi T harzianum was also able to convert castor oil into 4-decanolide (Serrano-Carr´eon et al., 1997) Using a 14-L bioreactor, rheological studies on the shear-sensitive mycelial cells were performed A lactone yield of 5.3 g/kg of castor oil was calculated according to the principle of extractive fermentation Indeed, the age of industrial bioflavours started with a patented process to convert ricinoleic acid (12-hydroxyoleic acid) to 4-decanolide using Candida yeast (Farbood and Willis, 1983) Initially, the market price was US$20 000/kg, which created a lot of enthusiasm among biotechnologists Process development was facilitated by access to an inexpensive precursor substrate, by the widespread occurrence of the pathway in fungi and by relatively simple Biotechnological flavour generation 97 bioprocessing conditions An alternative access was opened by the transformation of 3-decen4-olide to the reduced lactone using baker’s yeast (Gatfield and Sommer, 1997) Among the more recently described producers were Sporidiobolus yeasts (Haffner and Tressl, 1998; Dufosse et al., 2000) Today, 4-decanolide commands a market price 40-fold lower than initially A survey of the pathways to 4-decanolide was presented (Krings and Berger, 1998) Less work has been devoted to sulfur-containing heterocycles A patent (Bel Rhlid et al., 1997a,b) claims a bioprocess on the basis of the baker’s yeast that converted cysteamine or cysteine and hydroxy or oxopropanoic acid (or derivatives) to a flavour mixture containing 2methyl-3-furanthiol, mercaptopentanone and 2-acetyl-2-thiazoline The composition is used to intensify meat flavour of food and pet food 4.4.4 Generation of vanillin, benzaldehyde and benzoic compounds Vanillin is the world’s number one flavour chemical Vested with a unique flavour characteristic, the compound also exhibits antioxidant, flavour-enhancing and bitterness-masking properties Three orders of magnitude of difference of the market prices of chemosynthetic (around US$12/kg) and Vanilla pod-derived vanillin (up to US$16 000/kg) have again triggered a lot of research Vanilla planifolia is an orchid and fixes CO2 according to the rather uncommon CAM (Crassulacean acid metabolism) pathway Multi-component stable isotope analysis of vanillin will therefore be able to discriminate various ‘natural’ vanillins according to the isotope patterns of the vanillin precursor substrates (Hener et al., 1998) Among the vanillin precursors listed (Rabenhorst, 2000; Xu et al., 2007) are ferulic acid, eugenol and isoeugenol, vanillylamine, methoxytyrosine, coniferyl aldehyde, coumaric acid, and others More than two dozen microbial strains, mainly soil bacteria and higher fungi, and at least two different enzymes, converted these substrates to vanillin The most efficient process (Ͼ11 g/L in 30 hours) is based on an actinomycete of the Amycolatopsis family and a batch-fed dosage of ferulic acid (Rabenhorst, 2000) This retroClaisen-type cleavage also resulted in metabolite overflow, subsequent secretion, and accumulation of amounts in gram of vanillin by Streptomyces setonii (Muheim and Lerch, 1999) A P putida strain was cultivated by the same authors Although ferulic acid catabolism was fast, vanillin did not accumulate, as it was oxidised faster than ferulic acid Accumulation of vanillin also failed in growing cultures of a Nocardia strain (Li and Rosazza, 2000) A purified carboxylic acid reductase from the same strain, however, quantitatively reduced vanillic acid to vanillin Isoeugenol from various essential oils is another inexpensive precursor substrate Owing to its considerable cytotoxicity, the development of a high-yielding bioprocess towards vanillin was prevented for a long time With the aid of enrichment techniques, isoeugenol-tolerant strains have now been selected Cell-free extracts of a Bacillus strain yielded 0.9 g vanillin per litre (Shimoni et al., 2000), and Rhodococcus rhodochrous tolerated 15 g of isoeugenol per litre, which was converted with a molar yield of about 60% to vanillin (Chatterjee et al., 1999b) The phenylpropanoid pool provides precursors for benzoic volatiles in flavour extracts from both cell cultures (Venkateshwarlu et al., 2000) and fruiting bodies (Răosecke and Kăonig, 2000) Benzaldehyde, mandelates, phenylacetates and their derivatives occur regularly Multiple pathways of the degradation of phenylpropanoid compounds appear to exist Although less likely to occur than with terpenes, mere chemical conversion cannot be ruled out: a cell-free extract from Lactobacillus plantarum deaminated phenylalanine to phenylpyruvic 98 Food Flavour Technology acid, which was then chemically oxidised to benzaldehyde and other compounds at pH if certain metal cations were present (Nierop Groot and de Bont, 1998) A P fluorescens strain, able to grow on ferulic acid as the sole carbon source, contained high levels of an inducible feruloyl-CoA ligase and a vanillin dehydrogenase (Narbad and Gasson, 1998) Acetyl-CoA was verified as a cleavage product by 13 C NMR, but the mechanism of the cleavage steps remained open The mandelate pathway was demonstrated in the related P putida by feeding benzoylformate (Simmonds and Robinson, 1998), and also in the higher fungus Gloeophyllum odoratum by analysing volatile and non-volatile (trimethylsilylated) constituents of a hot-water extract from the fruiting bodies (Răosecke and Kăonig, 2000) Flavour chemists and natural products chemists often stick to their selective non-polar and medium-polar extraction solvents, respectively, and are thus prone to overlook the respective non-soluble intermediates of a given pathway, which might impede biogenetic understanding Originating from the Strecker degradation of phenylalanine, the rose-like compound 2phenylethanol is a common volatile in yeast fermentation flavours (Fabre et al., 1998) Metabolites of labelled l-phenylalanine in the higher fungus Bjerkandera adusta were the non-volatile (E)-cinnamic, phenylpyruvic, phenylacetic, mandelic and benzoylformic acids, and volatiles such as benzaldehyde and benzyl alcohol (Lapadatescu et al., 2000) The direct ␤-oxidation of cinnamic acid to benzoic acid was concluded from the occurrence of acetophenone, the product of the spontaneous decarboxylation of the supposed intermediate ␤-oxophenylpropanoic acid Pycnoporus cinnabarinus, another white-rot fungus, was the subject of optimisation studies resulting in a high-density culture yielding >1.5 g of vanillin per litre (Oddou et al., 1999; Stentelaire et al., 2000) Vanillin formation was favoured by reduced concentration of dissolved oxygen, high carbon dioxide, gentle agitation, low specific growth rate and the application of a non-selective adsorbent On the basis of this well-explored model, 5-2 H-labelled ferulic acid was fed to the fungus (Krings et al., 2001; Fig 4.3) The major labelled phenolic compounds identified were four lignans: the methyl esters of ferulic and vanillic acid, (E)-coniferyl aldehyde and alcohol, vanillic acid, vanillin, and vanillyl alcohol Labelled 4-hydroxy-3-methoxyacetophenone occurred, suggesting the decarboxylation of free 4-hydroxy-3-methoxybenzoylacetic acid Detailed mass spectrometric examination revealed traces of 4-hydroxy-3-methoxybenzoylacetic acid methyl ester and 3-hydroxy-(4-hydroxy-3-methoxyphenyl) propanoic acid methyl ester in the culture medium Hence, the fungal degradation of the phenylpropenoic side chain, a principal key step of lignin decomposition, should proceed by analogy to the oxidation of fatty acids Raspberry ketone, 4-(4-hydroxyphenyl)butan-2-one, is one of the character-impact components of raspberry flavour If the compound were isolated from raspberry fruit, natural raspberry ketone would cost million dollars per kg (cost of fruit only), compared to $US8–10/kg for the synthetic product An enzymatic pathway was proposed involving the ␤-glucosidase-catalysed hydrolysis of the naturally occurring betuloside from the European white birch (Betula alba) to release betuligenol, which is transformed by an Acetobacter alcohol dehydrogenase (ADH) into the ketone Plant cell culture or biomimetic acyl anion transfer reactions using aldolases were also suggested, but none of the known routes appear to have met industrial processing requirements (Whitehead, 1998) The basidiomycete Nidula niveo-tomentosa synthesised traces of the ketone and its corresponding alcohol de novo starting from simple nutrients, such as glucose and amino acids A systematic attempt was made to improve the productivity of this fungus in submerged culture (Băoker et al., 2001) Variation of the composition of the nutrient medium supported by a factorial experimental design yielded a 50-fold increase in metabolite concentrations This allowed for a Biotechnological flavour generation Fig 4.3 99 ␤-Oxidation-like degradation of ferulic acid by Pycnoporus cinnabarinus follow-up labelling study using phenylpropanoid precursor substrates, and it turned out that l-phenylalanine was degraded to a benzylic intermediate and then side chain elongated in two subsequent steps (Zorn et al., 2003b) Exposure to UV light stimulated the growth of the basidiomycete and the synthesis of raspberry alcohol and ketone As this is one of the very rare examples of a UV light-induced formation of a fungal volatile, differentially expressed proteins were identified by means of 2D-electrophoresis and ab initio sequenced by ESIMS/MS spectrometry The encoding nucleotide sequences were cloned Several stress and growth-related enzymes were up-regulated as a response to irradiation, but clear evidence for the involvement of polyketide synthases is still missing (Taupp et al., 2008) 4.4.5 Generation of miscellaneous compounds Short-chain methyl-branched fatty acids are not only important to cheese and other fermentation flavours, but are also added, for example, to strawberry flavours Their generation from fusel oil constituents by strongly oxidising bacteria is feasible Gluconobacter species preferentially produced (S)-2-methylbutanoic acid from 2-methylbutanols of known enantiomeric composition (Schumacher et al., 1998) Mechanistic aspects were studied in detail (Gatfield et al., 2000) Esterification of such fatty acids with ethanol using Geotrichum 100 Food Flavour Technology yeast removes the necessity to substitute inactive lipase during reverse-hydrolytic enzymebased processes (Daigle et al., 1999) Work on the production of 1-octen-3-ol, one of the mushroom flavour-impact compounds generated through the lipoxygenase/hydroperoxide lyase pathway, is continuing (Assaf et al., 1997) Enantioselective reduction of 2-octanone by Lactobacillus fermentum and other organisms produced (S)-2-octanol with high enantiomeric excess (Molinari et al., 1997) An extended screening of food-grade yeast species showed that some of them accumulated elevated concentrations of ␣-hydroxy ketones (acyloins) Zygosaccharomyces bisporus cells, for example, accepted a wide variety of aldehyde substrates and linked them to pyruvate to form ␣-hydroxy ketones (Neuser et al., 2000a) The enantiopurity of the products depended strongly on the substrate structure Odour qualities and threshold values of 34 acyloins were evaluated using a GC–olfactometry dilution technique, and 23 of them possessed novel and pronounced flavour properties (Neuser et al., 2000b) This catalytic property is known to rely on a pyruvate decarboxylase, and the respective enzyme was isolated from Z bisporus and compared with crude preparations from S cerevisiae, K lactis and Kluyveromyces marxianus Conversion rates of more than 50% showed that the potential of this type of enzyme to catalyse the formation of aliphatic acyloins has been underestimated An 1856 bp cDNA coding for the monomeric unit of the enzyme was amplified and sequenced (Neuser et al., 2000c) Aliphatics with sulfur functions often possess low-odour detection threshold values Of interest for cheese flavour, for example, is methanethiol Species commonly present in ageing cheese, such as Lactococcus, Lactobacillus, and Brevibacterium, degraded l-methionine through methionine aminotransferase or ␥ -lyase activities to methanethiol (Dias and Weimer, 1998) The enzyme activities were quantitatively assessed, but no attempts at further purification were made Particularly active formers of sulfur compounds are strains of Geotrichum candidum, which develop early in cheese ripening (Berger et al., 1999a; Demarigny et al., 2000) Methanethiol, as well as di- and trisulfides, and several thio fatty acids with straight and branched chains were identified From results obtained with l-[(S)-methyl-2 H]methionine, two different pathways for the generation of the sulfur compounds were suggested An extended range of microbial strains from surface-ripened cheese is currently under study (Deetae et al., 2007) Whether an odorous molecule is perceived by humans as a positive or negative (offflavour) sensation depends on the actual concentration, the composition of the matrix and the presence of other complementing volatiles Hence, off-flavours attract the same industrial attention as pleasant volatile flavours How the production of volatiles by yeast can be affected by bioprocess conditions has become evident during the development of continuous brewing processes (Van Iersel et al., 2000) Immobilised yeast showed slower growth and decreased amino acid metabolism, and, thus, released increased concentrations of oxo acids into the fermentation medium High levels of undesirable aldehydes may occur due to the activity of pyruvate decarboxylase, a problem present in the dealcoholisation of beer As these oxo acids are the precursors of the above-mentioned acyloins, this problem may be turned to advantage if the generation of acyloins is aimed at Strains of Penicillium, Trichoderma, Aspergillus, Mucor, Monilia and one of Streptomycetes were isolated from cork (Caldentey et al., 1998) Growth on cork resulted in the formation of off-flavours, but growth on malt extract medium did not Many aliphatic alcohols and carbonyls, sesquiterpenes and some halogenated aromatics were identified, but no structural assignments of the off-notes were reported The phenolic off-flavour of fermented soy products is now clearly attributable to 4-ethyl- and 4-vinylguaiacol (Suezawa et al., 1998; Karmakar et al., 2000) Bacteria, Biotechnological flavour generation 101 mainly strains of Bacillus, Pseudomonas and Staphylococcus that degrade phenylpropanoid precursors, are responsible for this problem 4.5 ENZYME TECHNOLOGY If a single chemical step is to be biocatalysed, it appears to be a waste to maintain all the thousands of concurrent metabolic steps of a living cell that are not wanted A single enzyme, whether in free form or included in gel or microcapsules or bound onto solid supports, could perform the desired reaction just as well Some problems associated with the use of enzymes may include location of the activity needed, isolation and purification, maintenance of activity during preparation and use, and, for kinases, methyl transferases and oxidoreductases, inevitable co-factor requirements A review on enzymes in flavour biotechnology by Menzel and Schreier (2007) shows how some of these disadvantages can be overcome Most enzymes currently used in the food industry stem from only about 25 micro-organisms Many possible sources of enzymes, for example marine organisms (Chandrasekaran, 1997), have remained almost unexplored so far With the rapid progress in basic enzymological and genetic knowledge even NAD+ -dependent redox enzymes gain interest, particularly if a synthetic electron sink, such as dichlorophenol indophenol, is accepted and high stereoselectivity is observed (Van der Werf et al., 1999) 4.5.1 Liberation of volatiles from bound precursors The application of glycosidases and, less frequently, of lyases, for the liberation of preformed flavours from non-volatile precursors (Winterhalter and Skouroumounis, 1997) should not be confused with the maceration of recalcitrant plant materials, when enzymes serve merely as processing aids (for example, Sakho et al., 1998) The majority of reports deal with wine flavour (Table 4.1) ␤-d-Glucopyranosidase, ␣-l-arabinofuranosidase, ␤-apiosidase, ␣-lrhamnosidase and a cysteine conjugate ␤-lyase were shown to enhance the level of volatiles in musts, wines and fruit juices While strains of Aspergillus and yeasts are typical sources of these enzymes, the lyase was from Eubacterium limosum (Tominaga et al., 1998) Flavour enhancement in common fruits and in vanilla bean (Pu et al., 1998) was often performed using fungal ␤-glucosidases Tropical fruits and black tea leaves, however, also contain less abundant glycosides, such as primeverosides, vicianosides and rutinosides, which may not be sufficiently hydrolysed by side activities of the standard glycosidases The long history of enzyme-modified cheese is discussed in reviews (Kilcawley et al., 1998; Klein and Lortal, 1999) The proteolysis of minced fish tissue to produce a ‘seafood flavour’ (Imm and Lee, 1999) was likewise adopted from traditional food biotechnology The reverse reaction may also be useful: lyophilised cells of Xanthomonas campestris and of Stenotrophomonas maltophilia synthesised l-menthyl ␣-d-glucopyranoside anomer-selectively from l-menthol (Nakagawa et al., 2000) An impressive molar yield of more than 99% in 48 hours was reported The glucoside obtained is slowly hydrolysed in the oral cavity, thereby acting as a flavour depot 4.5.2 Biotransformations The interfacial enzymology of lipid hydrolases is useful in flavour generation, and there are, as with all carbonyl reactions, two options for shifting the reaction equilibrium: first, towards 102 Food Flavour Technology Table 4.1 Hydrolases of oenological relevance Flavour precursor Enzymes References Glycoconjugated aroma compounds Endo-, exogenous glycosidases Winterhalter and Skouroumounis (1997) Mono- and diglycosides of terpenols (linalool, ␣-terpineol, citronellol, nerol, geraniol) ␣-L-Arabinofuranosidase, ␤-D-glucopyranosidase Spagna et al (1998) Monoglycosides of terpenols ␤-D-Glucosidase Yanai and Sato (1999) Apiofuranosylglucosides of geraniol and linalool ␤-Apiosidase Guo et al (1999) ␣-L-Rhamnopyranoside ␣-L-Rhamnosidase Orejas et al (1999) Glycosides of nerol, geraniol, linalool, ␥ -terpinene, 2-phenylethanol, benzyl alcohol ␤-Glucosidase Gueguen et al (1997) S-Cysteine conjugates of volatile thiols (4-mercapto-4-methylpentane-2-one, 4-mercapto-4-methylpentan-2-ol, 3-mercaptohexan-1-ol) Cysteine conjugate ␤-lyase Tominaga et al (1998) Glycosides of fruits, particularly Vitis vinifera Exogenous and endogenous glycosidases Sarry and Gunata (2004) Terpenyl glycosides Bacterial, yeast, plant glycosidases Maicas and Mateo (2005) Glycosides of juices Glycosidases Pogorzelski and Wilkowska (2007) the liberation of odorous fatty acids from triacylglycerols and related esters, and second, towards the synthesis of a variety of volatile esters from different acyl and alkyl precursor moieties in a microaqueous environment A broad range of reaction conditions including organic media is tolerated and can be varied to achieve the desired selectivities (Saxena et al., 1999) The lipase from Candida antarctica, which is a stable, versatile enzyme with broad substrate specificity, has been characterised in detail and become a classical catalyst Some of the more than 170 lipases have been analysed by X-ray crystallography down to the 1.5 Å level of resolution; thus, the position of the helical lid that buries the active triad and conformational changes exerted on this by bipolar chemicals have been well established (http://www.rcsb.org/pdb/) The acetates of common acyclic monoterpenols are used as flavours and fragrances These volatiles can be produced either by direct esterification or by transesterification using the C antarctica lipase SP435 in microaqueous n-hexane as reaction solvent (Claon and Akoh, 1994a,b) Whole cells of Hansenula saturnus or Pichia species in an interface reactor and n-decane (Oda et al., 1995; Oda and Ohta, 1997) and dry mycelium of Rhizopus delemar in n-heptane (Molinari et al., 1998a,b) were successfully used for the same purpose The reverse hydrolytic process runs on an industrial scale and constitutes one of the fundamentals of modern flavour biotechnology Lipases of C antarctica and of Rhizomucor miehei formed esters in high yields under solvolytic conditions and vacuum, where the substrate itself acted as a solvent (Chatterjee and Bhattacharyya, 1998a,b) The reaction solvent also affects the enantioselectivity The synthesis of (−)-citronellyl oleate from racemic citronellol was achieved using a Candida cylindracea lipase in supercritical carbon dioxide, but only in Biotechnological flavour generation 103 a narrow window of pressure and temperature (Ikushima et al., 1996) Some contradictory results suggest, however, that a generalisation of these findings should be treated with caution (Michor et al., 1996) The synthesis of (3Z)-hexenyl butanoate in n-hexane or in solvent-free medium was likewise achieved using the Mucor or Candida lipases (Bourg-Garros et al., 1997; KimJungbae et al., 1998) Aliphatic alcohols from fusel oil, a side-product of the distillation of spirits, were reacted with dodecanoic acid (De Castro et al., 1999) Yields dropped with decreasing chain length of the aliphatic acyl moiety Recent work also modified the structure of the alkyl moiety Phenylethanol was transformed to its acetate, a Koji-style wine flavour impact, using supercritical carbon dioxide and lipase PS (Wen et al., 1999), and thioethyl, thiobutyl and thiohexyl propanoate, butanoate and valerate were produced using different immobilised lipases (Cavaille-Lefebvre and Combes, 1997) An example of a successful redox process was the oxidation of vanillylamine to vanillin by an amine oxidase from A niger, or by a monoamine oxidase from E coli (Yoshida et al., 1997) 4.5.3 Kinetic resolution of racemates The preferred cleavage of one enantiomer of a racemic (usually ester) mixture is one possible route to enantiopure products, an approach patented for the generation of the bulk flavour chemical l-menthol in the early 1970s Many subsequent papers and patents confirm the competitiveness and industrial usefulness of this mature bioprocess This classical approach was followed to resolve (R)- and (S)-karahanaenol produced from ring enlargement of limonene epoxide (Roy, 1999) The racemic monoterpenol acetate was resolved using a Pseudomonas cepacia lipase, and the 4R-isomer alcohol was preferentially obtained through alcoholysis (Fig 4.4) Kinetic resolution in the course of the esterification of menthol was a novel aspect arising from work with lipase from Candida rugosa, originally focusing on the formation of methyl esters as a flavour depot (Shimada et al., 1999) When a menthol racemate was esterified with oleate in an emulsion containing 30% water, 96% esterification and an enantiomeric excess of 88% of the l-enantiomer menthyl ester were found The operational stability of some lipases is remarkable In a study undertaken to separate racemates of ibuprofen and 1-phenylethanol, the commercial enzyme remained partly active even after 14 hours at 140◦ C and 15 MPa (Overmeyer et al., 1999) These examples together with Table 4.2 show the broad range of substrates amenable to hydrolysis with various lipases The theoretical modelling of the energy differences between the transition states of the diastereomeric enzyme–substrate complexes and of substrate inhibition will aid in further refinement of kinetic resolution processes (Berendsen et al., 2006) There remains the drawback that the Fig 4.4 Enantiospecific alcoholysis of karahanaenol acetate by Pseudomonas cepacia lipase: (I) terpinolene oxide; (II) (R)-karahanaenol; (III) (S)-karahanaenol acetate (adapted from Roy, 1999) 104 Food Flavour Technology Table 4.2 Kinetic resolution of racemates: a mature technique in flavour enzymology Flavour compounds Enzymes References Lactones and epoxides 1-Phenylethanol 1-Phenylethanol 2-Phenyl-1-propanol Ibuprofen, 1-phenylethanol ␥ -, ␦-Lactones 2-Methylbutanoic acid methyl ester Reductase Immobilised lipase (Mucor miehei) Lipase (Pseudomonas sp.) Lipases Commercial lipase (Novozym R ) Lipase (Pseudomonas sp.) Lipase (R miehei) Danchet et al (1998) Frings et al (1999) Ceynowa and Koter (1997) Goto et al (2000) Overmeyer et al (1999) Enzelberger et al (1997) Kwon et al (2000) resolved substrates are usually derived from chemosynthesis; hence, the produced flavours are not naturals in a legal sense Tailor-made enzymes are now at hand With the invention of directed evolution in combination with high-throughput screening systems, literally any enzyme can be modified according to changes of substrate or reaction specificity Micro-organisms that cannot be cultured in vitro deliver their enzymes through gene partial sequences found in soils, deepsea sources or geysers Progress in sequence-based biocatalyst discovery allows to explore the bewildering catalytic diversity of nature Techniques for the stabilisation and immobilisation of enzymes have reached a high degree of perfection Further impetus comes from the spreading idea of white biotechnology, an industrial approach focusing on renewable substrates and environmentally friendly production processes 4.6 PLANT CATALYSTS Most of the natural flavours currently processed by the flavour and fragrance industry are obtained by extraction or distillation of parts of field-grown plants Their huge biosynthetic potential represents an attractive starting point for the in vitro culture of plant cells for aroma biotechnology (Fu, 1999) At the same time, the high degree of structural and biochemical organisation of a plant cell impedes the biotechnological approach: specialised plant cells tend to dedifferentiate in vitro, they demand complex nutrient media and low-shear bioreactors, and usually not accumulate and excrete elevated levels of volatile flavours The enzymes of a phototrophic metabolism are not easily inducible by carbon substrates This explains why plant cells in sterile culture are not as responsive to precursor substrates as are most micro-organisms Numerous attempts at the production of flavour and aroma compounds by plant cell, tissue and organ cultures have been described (Berger, 1995; Dăorneburg and Knorr, 1996; Kumar et al., 1998) A collection of promising recent results obtained on the laboratory scale is presented in the following sections 4.6.1 Plant cell, tissue and organ cultures Plant tissues, excised from differentiated and surface-sterilised materials and placed on phytoeffector-containing agar medium, start to divide again and form non-differentiated cell clusters (‘callus’) After transfer to liquid medium, a fine suspension of single cells and smaller cell aggregates develops This propagation of plant cells is based entirely on mitotic events; the full genetic potential to form flavours is retained in each cultured cell However, Biotechnological flavour generation 105 these cells dedifferentiate during subculturing and not accumulate flavours, even if isolated from a flavour-bearing source tissue 4.6.2 Callus and suspension cultures Exceptionally high amounts of odorous mono- and sesquiterpenes (0.34% total oil content in cells plus medium) were recovered from callus cultures of the Brazilian snapdragon (Otacanthus coeruleus), an ornamental pot plant from east Brazil (Ronse et al., 1998) The amount of essential oil extracted from the nutrient media was higher than the intracellular amount High sucrose treatments (Ͼ40 g/L) increased the quantities of oil found in the medium due to the high osmotic stress In cell cultures of two genotypes of rosemary, the concentrations of calcium ions, sucrose and plant growth regulators significantly affected the yields of camphene, 1,8-cineole, linalool, camphor, borneol and bornyl acetate (Tawfik et al., 1998) Photomixotrophic callus cells of grapefruit (Citrus paradisi Macf.), lemon (Citrus limon (L.) Burm.) and lime (Citrus aurantiifolia (Christm et Panz.) Swingle) generated monoterpenes, although no cytodifferentiation was found by electron microscopy (Reil and Berger, 1996a) Chlorophyll content and the formation of oligo-isoprenoid volatiles were positively correlated with high light intensities An optimisation of the growth medium and the light regime resulted in the identification of more than 40 mono- and sesquiterpenes and aliphatic aldehydes in grapefruit callus The maximum yield was 186 mg of volatiles per kg callus accumulated in weeks, representing about 5% of the volatiles accumulated in flavedo tissue in about 13 months A similar correlation existed for white diosma (Coleonema album Thunb.) photomixotrophic cell cultures (Reil and Berger, 1997) An extended photoperiod and certain concentrations of phytoeffectors were the conditions for the formation of volatiles including limonene and phellandrenes The light conditions also affected the production of compounds by vanilla (V planifolia) cultures, particularly of 4-hydroxy-3-methoxybenzyl alcohol (vanillyl alcohol) (Havkin-Frenkel et al., 1996) Microbial infections, whether caused by an endogenous infection of the source tissue or by a secondary infection, present a serious experimental problem because most micro-organisms simply overgrow the slower plant cells Otherwise, microbial infections are often accompanied by the formation of elicitors, components involved in plant chemical communication and defence As volatile flavours may be part of a plant’s response to a microbial attack, flavour biogenesis can be elicited A persistent contamination with Pseudomonas mallei was reported for cell cultures of rosemary (Rosmarinus officinalis L.), producing cineole and ␣-pinene (Shervington et al., 1997) An (unintended) elicitation may explain this observation A combination of photomixotrophy and elicitation led to the excretion of volatiles by cell cultures of parsley (Petroselinum crispum (Mill.) Nym.) Treatment with autoclaved homogenate of the wood-destroying basidiomycetes Polyporus umbellatus or Tyromyces sambuceus elicited a spicy odour, imparted by elemicin (5-allyl-1,2,3-trimethoxybenzene), 3-n-butylphthalide, (Z)- and (E)-butylidenephthalide, sedanenolide and (Z)-ligustilide (Reil and Berger, 1996b) (Fig 4.5) 4.6.3 Organ cultures There is abundant experimental evidence for the correlation of cytodifferentiation and the formation of secondary plant products Visible differential gene activity and biochemical 106 Food Flavour Technology n Fig 4.5 Z Z Flavours elicited in cell cultures of parsley specialisation occur with morphological differentiation Non-embryogenic cell lines derived from immature juice vesicles of sweet orange (Citrus sinensis (L.) Osbeck) failed to form the characteristic flavour constituents, but organised embryogenic cells emitted a fruity aroma and produced 420 mg of volatiles per kg tissue (Niedz et al., 1997) Some hairy root cultures, derived from the transformation of aseptic plantlets with Agrobacterium rhizogenes, were shown to be capable of forming volatile flavours The essential oils of hairy root cultures of anise (Pimpinella anisum L.) differed significantly from those of the fruits (Santos et al., 1998; Andarwulan and Shetty, 1999) While the major components of the essential oil from the hairy root cultures were the anethole precursor (E)-epoxypseudoisoeugenyl 2-methylbutanoate, zingiberene, ␤-bisabolene, geijerene and pregeijerene, the terpene spectrum of the fruits was dominated by (E)-anethole (Fig 4.6) The maintenance of morphological stability with no dedifferentiation or greening remains an experimental challenge (Santos et al., 1999) Matsuda et al (2000) generated mutant hairy roots of musk melon (Cucumis melo L.) by means of t-DNA insertion Of more than 6500 clones, fragrant hairy root clones were selected The volatile compounds were identified as (Z)-3-hexenol, (E)-2-hexenal, 1-nonanol and (Z)-6-nonenol, which also determine the flavour of the fruits Aroma production was stable for more than years The yield of aroma compounds was about 6.5-fold higher than in ripe melon fruit Fig 4.6 Secondary metabolites produced by hairy root cultures of anise ... Water-soluble liquid flavours 1.5.2 Clear water-soluble liquid flavours 1.5.3 Oil-soluble liquid flavours 1.5.4 Emulsion-based flavours 1.5.5 Dispersed flavours 1.5.6 Spray-dried flavours 1.6 Production... transfer 7.4.2 Liquid food products 7.4.3 Semi-solid food products 7.4.4 Solid food products 7.5 Delivery systems: food technology applications 7.6 Conclusions References Modelling flavour release... Food flavour technology / edited by Andrew J Taylor and Robert S.T Linforth – 2nd ed p cm Includes bibliographical references and index ISBN 978-1-4051-8543-1 (hardback : alk paper) Flavour Flavouring
- Xem thêm -

Xem thêm: Food flavour technology, Food flavour technology

Mục lục

Xem thêm

Gợi ý tài liệu liên quan cho bạn