New Food Engineering Research Trends

NEW FOOD ENGINEERING RESEARCH TRENDS

NEW FOOD ENGINEERING RESEARCH TRENDS

ALAN P. URWAYE EDITOR

Nova Science Publishers, Inc. New York

Copyright © 2008 by Nova Science Publishers, Inc.

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA New food engineering research trends / Alan P. Urwaye, editor. p. cm. Includes index. ISBN-13: 978-1-60692-828-8 1. Food industry and trade--Research. I. Urwaye, Alan P. TP370.8.N49 2007 664--dc22 2007028954

Published by Nova Science Publishers, Inc.

New York

CONTENTS Preface

vii

Chapter 1

Ionizing Irradiation of Foods Albert Ibarz

Chapter 2

Fruits and Vegetables Dehydration in Tray Dryers Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Chapter 3

Ultrasound in Fruit Processing Sueli Rodrigues and Fabiano A.N. Fernandes

Chapter 4

Optimisation of the Conversion of Ergosterol in Mushrooms to Vitamin D, and Its Bioavailability Conrad O. Perera and Viraj J. Jasinghe

Chapter 5

Chapter 6

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto The Development of the Processing of Yuba (Protein-Lipid Film) Li Zaigui, Shen Qun and Lin Qing

Chapter 7

Far-Infrared Heating in Paddy Drying Process Naret Meeso

Chapter 8

A Novel Two-Stage Dynamic Packaging for Respiring Produce: Concepts and Mathematics Tobias Thiele and Benno Kunz

Index

1 45 103

137

169

195 225

257 271

PREFACE This new book presents new research in the growing field of food engineering which refers to the engineering aspects of food production and processing. Food engineering includes, but is not limited to, the application of agricultural engineering and chemical engineering principles to food materials. Genetic engineering of plants and animals is not normally the work of a food engineer. Food engineering is a very wide field of activities. Among its domain of knowledge and action are: Design of machinery and processes to produce foods Design and implementation of food safety and preservation measures in the production of foods Biotechnological processes of food production Choice and design of food packaging materials Quality control of food production Chapter 1 - Irradiation, like other types of food treatments, is a method used to make food safer for the consumer and to increase its useful life in good conditions. In this chapter the interaction of ionizing radiation with matter and the sources of production of ionizing radiation are described. The biological effects caused by this type of radiation are also described. Likewise, the application of ionizing radiation in the food industry is described as well as the effects that it has on most food components. The inhibitory effect on microorganisms is described, as well as the effects on different kinds of foods such as meat, poultry, fish and shellfish, eggs and egg-derived products, tubers and bulbs, seeds, legumes, dry fruits, spices, seasonings and herbs, and for quarantine treatment. Finally, a short description of food treatment plants, dosimeters and certain current normative aspects of the ionizing radiation used are given. Chapter 2 - Dehydration involves simultaneous transfer of heat, mass and momentum in which heat penetrates into the product and moisture is removed by evaporation into an unsaturated gas phase. Owing to the complexity of the process, no generalized theory currently exists to explain the mechanism of internal moisture movement. In this Chapter, the investigation of momentum, heat and mass transfer phenomena, in both laboratory and large scale convective drying systems (suitable for dehydration of thermolabile products) by means of experimental measurements and numerical simulation are presented. The air flow inside complex geometry spaces, such as drying rooms containing hundreds of trays arranged in rows and columns, is analyzed by solution of 3-D momentum turbulent flow equations for different room configurations. Laboratory measurement data, concerning the space velocity distribution and the pressure field of the air flow over one tray, are provided and used for validation of turbulence models. The results of the flow investigation

viii

Alan P. Urwaye

lead to practical suggestions for the improvement of the air flow uniformity inside the drying space which is very important for the quality of the product. A novel numerical code, DrySAC, able to predict the unsteady-state processes taking place in a complex drying system, was developed. Unlike other attempts to predict drying processes, DrySAC takes into account not only the drying process itself, but also the behavior of the other system equipment and the interaction between them. Drying curves, evolution of the air state parameters in characteristic points of the system and product properties are predicted during the drying of various fruits and vegetables and. As a practical validation of the code, the predicted values compared with the measured data taken in-situ showed very good agreement. When a dryer configuration is given, the numerical DrySAC code can be used for optimization of the process parameters when a dryer configuration is given. For the most of the studied cases, an air recirculation ratio of around 75 % has proved to be the optimum, giving a minimum drying time. The code can be used both for evaluation of existing dryers and for optimum design of the new units with valuable impacts in increasing the efficiency of the systems and in reduction of energy consumption. Aiming to overcome the lack of experimental data in the open literature, a laboratory drying unit was constructed and is under operation for testing and monitoring the dehydration of agricultural products. Using this facility, experimental drying curves are set up for the drying of horticulture products under controlled conditions of the drying air parameters, which are gathered by means of a data acquisition system. The laboratory experimental results are useful for the validation of numerical models which further are an essential tool for optimization and increasing the efficiency of the drying process. Drying of agricultural products remains an open research field mainly because of their delicate and hard to be established, properties. Chapter 3 - Power ultrasound has been successfully employed in the chemical industry, polymer and plastic industry for many years and its use has been growing in the food industry. Power ultrasound can produce chemical, mechanical or physical effects on the processes or products where it is applied. Taking advantage of one of the effects or their combination, power ultrasound has been used in the food industry in drying, freezing, extraction processes and enzyme inactivation. The use of ultrasound in ambient fluids is well known to cause a number of physical effects (turbulence, particle agglomeration, microstreaming and biological cell rupture) as well as chemical effects (free radical formation). These effects arise mainly from the phenomenon known as cavitation. Herein a brief review of the use of ultrasound in the food industry is presented and the main applications are discussed. A comprehensive discussion on the effects of ultrasound in the tissue structure of fruits is presented along with photomicrographs of melons submitted to ultrasound. A detailed discussion is presented concerning the use of ultrasonic waves in drying, where an ultrasonic pre-treatment can be used prior to air-drying. The methods involved in the ultrasonic pre-treatment are presented along with the results obtained for several fruits such as melons, bananas, pineapples, papayas and other. Mathematical models that can be used to simulate the process are presented. Optimization of the drying process is also discussed for ultrasonic pre-treatment and ultrasound-assisted osmotic dehydration. Chapter 4 - The conversion of ergosterol in mushrooms to vitamin D2 by exposure to ultra violet (UV) light was studied under different UV lamps (UV-A, UV-B, and UV-C) and was found to be significantly different (p<0.05). Analysis of ergosterol content in different

Preface

ix

tissues of Shiitake mushrooms showed a significant difference (p < 0.01) in its distribution. Thus, the conversion of ergosterol in whole mushrooms to vitamin D2, by exposure to UV light was significantly affected (p < 0.01) by the orientation of the mushroom tissues to the UV source. The conversion of ergosterol to vitamin D2 was about four times higher when gills were exposed to UV light compared to when the outer caps were exposed to the same. The highest vitamin D2 content (184.22 ± 5.71 μg/g DM) was observed in Oyster mushrooms exposed to UV-B light at 35 oC and around 80% moisture. On the other hand, under the same conditions of UV-exposure, the lowest vitamin D2 content (22.90 ± 2.68 μg/g DM) was observed in Button mushrooms. The kinetics of conversion of ergosterol to vitamin D2 showed that Oyster mushrooms (Pleurotus ostreatus) had the highest conversion rate followed by Shiitake (Lentinula edodes) and Abalone (Pleurotus cystidus), whereas the lowest conversion rate was observed in Button mushrooms (Agaricus bisporus). Both initial moisture content and temperature of UV exposure influenced the conversion of ergosterol. The conversion of ergosterol to vitamin D2 followed zero-order kinetics, where the rate constant varied with temperature according to the Arrhenius equation (K0 = 7.32 s-1; Ea = 51.5 kJ mol-1). For the bioavailability of vitamin D, thirty male Wistar rats were fed for one week with a diet deficient in vitamin D. After the first week, six rats were randomly selected and sacrificed for analysis of initial Bone Mineral Density (BMD), and serum level of 25hydroxyvitamin D [(25(OH)D]. The remaining animals were divided into two groups of 12. One group received 1 μg of vitamin D2/day from UV-exposed mushrooms for a period of four weeks until sacrificed. The other group was fed the same amount of mushrooms that was not exposed to UV light, and was use as the control. At the end of four weeks, the mean serum 25(OH)D level of the experimental group was 129.42 ± 22.00 nmol/L, whereas, it was only 6.06 ± 1.09 nmol/L in the control group. The Femur BMD and the serum calcium concentration of the experimental group of animals were significantly higher (p < 0.01) than the control group. It may be concluded from the results that vitamin D2 from UV-exposed mushrooms is well absorbed and metabolised in this model animal system. Chapter 5 - Enzymatic hydrolysis allows to improve the functional, nutritional and immunological properties of proteins. For instance, protein hydrolysates are used in the food industry as ingredients in hypoallergenic formulae and clinical nutrition. Conventional batch hydrolysis of proteins has been traditionally used to obtain protein hydrolysates due its simple operation. However, there are several disadvantages associated to this method, the high enzyme consumption being the main one. Among the solutions assayed to enhance the yield of the process, enzyme immobilisation onto highly activated supports allows to work continuously and reuse the enzyme. Since there are loss of enzyme activity and constrains for the diffusion into the support, the feasibility of this technique is limited. Continuous reaction and simultaneous separation of products from the reaction mixture can be achieved in a continuous membrane recycle reactor. Here, the low molecular weight peptides generated permeate through an ultrafiltration module with the appropriate molecular weight cut-off, while the enzyme (which acts in soluble form) is continuously recycled to the reaction tank. As important drawbacks, permeate flux decline due to membrane fouling and frequent purges are required to eliminate non-reacting substrate which involves severe difficulties in the control process.

x

Alan P. Urwaye

In order to profit from the advantaged of both batch and continuous membrane recycle reactor, the objective of this research work was to design and optimise a reactor for the production of low antigenicity protein hydrolysates. The operation proposed comprises 3 consecutive steps: a) hydrolysis in a stirred tank reactor; b) ultrafiltration of the reaction mixture through a membrane with full retention of the enzyme; c) enzyme recycling, in which the retentate is returned to the tank reactor for a new hydrolysis. The materials employed were a whey protein concentrate as substrate, a subtilisin from Bacillus licheniformis as protease and an 8 kDa flat polyethersulfone membrane. The pH-stat method was used to monitor the hydrolysis reaction. The molecular weight profile and the antigenicity of the hydrolysates were determined by HPLC and ELISA, respectively. Regarding the process kinetics, zero-order for the substrate and second order for the enzyme deactivation were identified. Process optimisation involved the calculation of the optimum number of enzyme uses that minimised the enzyme consumption, subject to a required productivity. The performance of the system was compared at several temperatures to that of the conventional stirred tank reactor. Significant enzyme savings were achieved, which demonstrate the viability of this approach. Chapter 6 - Yuba is a kind of protein-lipid film formed from soymilk under continuous heating, so it is also called to be “Tofupi” or “Tofuyi”, which means “tofu sheets or tofu shirts”. Yuba is one of the most famous traditional foods which has a history of over 2000 years and is very popular now in China. The annual yield of yuba is over 200,000 ton in China. The film can be consumed directly as an ingredient of soups or be used as a sheet for wrapping and shaping meats or vegetables into various forms with different tastes. Now freezing yuba is also used as salad for sashimi in Japan and Korea. Yuba has first been found from the supernatant film of heated soymilk for Tofu making and was usually dried to 8% moisture content for storage. It can be stored for more than three months to six months. It contains about 55% protein and 25% lipid, which are important nutrients for monks to whom meat was prohibited. As its fine taste and unique texture, yuba spread fast in all of the country through the rede of people lived in temples. Chapter 7 - Far-infrared heating is applied in two paddy drying processes, namely, singlestage and multi-stage drying processes. The single-stage drying process is the combination of far-infrared radiation and hot-air convection in fluidized-bed drying, and the multi-stage drying process consists of hot-air convective fluidized-bed drying, far-infrared heating, tempering and ambient air ventilation. The effect of far-infrared heating in paddy drying process on moisture content, grain temperature and milling qualities (e.g. head rice yield and whiteness) is investigated together with the microstructure of rice kernels and the pasting behavior of rice flours. Moreover, the mathematical model of far-infrared heating, which is the set of coupled heat and mass transfer equations, is developed to describe the paddy drying. This model assumed that the absorbed infrared energy completely converts to heating within the superficial layer under the surface of paddy grain, and heat is transferred into the deeper layer via conduction. Validation of the developed models is made by comparing predicted and experimental data for the average moisture content and the grain temperature of paddy. Chapter 8 - Even when applying optimal storage conditions, off-flavors occur in packages with respiring commodities. In the case of fresh-cut produce mixes, compromises due to different optimal storage conditions have to be made and off-flavors are more likely to occur. Additionally unwanted temperature changes during the cold-chain can not always be

Preface

xi

avoided, which leads to excessive respiration and anaerobic conditions and off-flavors are formed due to fermentation. In this work a two-stage, dynamic packaging concept is presented, where the undesired off-flavors can evaporate at the end of the shelf life by shifting the gas exchange from permeation to diffusion. This can be realized by removing an adhesive film strip from the package by the customer. A mathematical model is developed to describe the gas composition inside this kind of two-stage packages by combination of model terms for permeation, diffusion and respiration. To estimate the effect of diffusion, the gas exchange of a model package with perforations is determined and a coefficient is calculated by nonlinear regression analysis. This coefficient can be used for the model to describe the gas exchange after opening the perforations. To prove the effectiveness of the two-stage, dynamic concept, a test series with fresh-cut chicory endive was carried out. The results showed a significant decrease of lactic acid (67% at 7°C, 36% at 20°C) and ethanol (56 % at 20°C) and the end of the shelf-life compared to the traditional concept of modified atmosphere packaging. By storage experiments with temperature changes from 7°C to 20°C for 4 and 8 hours it was shown that the gas composition changed for the remaining shelf life and ethanol was detected within the packages.

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 1-43

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 1

IONIZING IRRADIATION OF FOODS Albert Ibarz Food Technology Department University of Lleida (Spain)

ABSTRACT Irradiation, like other types of food treatments, is a method used to make food safer for the consumer and to increase its useful life in good conditions. In this chapter the interaction of ionizing radiation with matter and the sources of production of ionizing radiation are described. The biological effects caused by this type of radiation are also described. Likewise, the application of ionizing radiation in the food industry is described as well as the effects that it has on most food components. The inhibitory effect on microorganisms is described, as well as the effects on different kinds of foods such as meat, poultry, fish and shellfish, eggs and egg-derived products, tubers and bulbs, seeds, legumes, dry fruits, spices, seasonings and herbs, and for quarantine treatment. Finally, a short description of food treatment plants, dosimeters and certain current normative aspects of the ionizing radiation used are given.

1. INTRODUCTION Irradiation, like other types of food treatment, is a method used for treating foods in order to make them safer for the consumer and to increase the period for which they can be kept in good condition. In other words, it is used for making food safe and prolonging preservation times. It is a method that does not attempt to replace conventional treatments, but rather one that can be used as a complement to these treatments. The dictionary of the Real Academia Española (Spanish Royal Academy) defines the verb radiate as “production of radiation by means of waves or particles”, while irradiate is “emit rays of light, heat or other energy from a body, or subject something to radiation”. In other dictionaries (Larouse, van Nostrand) radiation is defined as “the emission and propagation of energy in the form of waves through space or a natural medium”. Common usage of the term radiation is to refer to the waves or rays in the electromagnetic spectrum.

2

Albert Ibarz

Irradiation from a body should not be confused with its being radioactive, as a body is radioactive when its atoms split spontaneously. The irradiation of foods is not a new treatment technology, as its roots can be traced back to the late 19th century; although it is not until the 1940’s that the term irradiation appears. Three clearly differentiated stages or periods in the history of food irradiation can be distinguished: the period from 1890 to 1940 represents the beginnings of the physics of irradiation and of the different sources used, all tied to the first treatments of food with radiation. The period from 1940 to 1970 corresponds to a stage of intensive research and development in the application of radiation in food treatment, and to the study of the healthiness of irradiated foods. Since 1970 a series of regulations for the safe control and application of irradiation have appeared. From a historical viewpoint, radiation can be dated back to 1895 with the discovery of Xrays by von Roentgen. In 1896 Becquerel discovered radioactivity, and in the same year there was an initial proposal for the application of ionizing radiation in the preservation of foods in Germany. In 1898 Thompson discovered the nature of cathode rays, observing that they were electrons, and in that same year the effects of radiation on micro-organisms were already observed. In 1902-1903 Rutherford and Soddy published a theory on radioactive decay, and Marie Curie published her thesis on the nature of alpha, beta and gamma radiation. It was in 1904 when studies were published on the bactericidal effect of ionizing radiation, while in the following year, when Einstein published his theory of relativity, in Great Britain a patent was issued for the use of ionizing radiation to eliminate bacteria from foods. At the same time, in the USA, another patent was issued for the mixing of foods with radioactive matter aimed at prolonging food preservation. The period between 1905 and 1920 corresponds to a stage of basic research into the nature and the physical, chemical and biological effects of ionizing radiation. In this period studies appeared on the processing of strawberries with radiation, and a USA patent was issued for the multi-tube processing of food with X-rays. In the years 192030, important developments were obtained in the design of electron accelerating machines, and studies were published on the lethal effects that X-rays exerted on Trichinella spiralis in raw pork. Additionally, publications on the effects of ionizing radiation on enzymes appeared for the first time, a bioassay being performed on rodents in order to determine the possible toxicity of the irradiated foods. In 1930 a French patent was issued for the use of ionizing radiation in food preservation. In the 1940’s in the Massachusetts Institute of Technology (MIT) the viability of the preservation of minced beef by treatment with X-rays was demonstrated. In the 1950’s food irradiation programs began in the USA and in the United Kingdom, while of special note was the declaration of “Atoms for Peace” (1953) which Eisenhower delivered in the General Assembly of the United Nations, for the pacific use of atomic energy, including its application in food preservation. At the end of the same decade the USSR approved the irradiation of potatoes and grain, while in Germany a licence was granted for the treatment of spices with radiation. Furthermore, in the USA, irradiation was classified as an “additive”. In the 1960’s in Canada the irradiation of potatoes was approved, but in the Federal Republic of Germany the irradiation of food was prohibited, while the Food and Drug Administration (FDA) of the USA approved the irradiation of wheat, flour, potatoes and bacon, although it was repealed for this latter foodstuff in 1968. In 1969 in Spain the irradiation of potatoes and animal feed was initiated, as was the irradiation of mushrooms and frozen meat in the Netherlands. In the 1960’s NASA adopted irradiation as a method of food sterilization for astronauts, in Japan the irradiation of potatoes was initiated on an industrial

Ionizing Irradiation of Foods

3

scale, while in Spain the irradiation of onions was initiated. In 1976 the Joint FAO/IAEA/WHO Expert Committee of the Food and Agriculture Organization (FAO), the World Health Organization (WHO) and the International Agency for Atomic Energy (IAEA) certified the wholesomeness of various irradiated foods (potatoes, wheat, chicken, papayas and strawberries), recommending that the irradiation of foods be classified as a physical process. In 1980 the Joint FAO/IAEA/WHO Expert Committee on the Wholesomeness of Irradiated Foods (JECFI) declared that the irradiation of foods at doses of up to 10 kGy did not constitute any danger, causing neither nutritional nor microbiological problems. It was in the 1980’s when the Codex Alimentarius Commission accepted as a worldwide norm the conclusions elaborated by the Committee of Experts, and this was also the decade in which an International Consultative Group for Food Irradiation (ICGFI) was established under the patronage of FAO/WHO/IAEA to evaluate the developments in food irradiation. In 1985 the final regulations of the USA and Canada on the irradiation of foods were established, and the FDA approved the irradiation of pork at low doses in order to control Trichinella. Likewise, the FDA approved irradiation to control insects, to delay the ripening of fruits and vegetables, spices and dehydrated enzymes. It was also in that decade when the European Community prepared the first draft to harmonize the legislation of the different member states with regard to irradiated foods. In the 1990’s the FDA approved the irradiation of poultry for the control of Salmonella, and in 1992 the WHO reaffirmed that irradiated foods were safe. In 1996 there were 40 countries authorizing the irradiation of at least one foodstuff, while 28 countries applied the irradiation of food on a commercial basis. In 1999 a European Union directive approved the irradiation of spices, herbs and seasonings, while the FDA presented a petition to unblock the irradiation of pre-prepared foods, due to lysteriosis epidemics that could be caused by this kind of food. In 2000 the FDA unblocked irradiation in order to be able to control Salmonella in eggshells and to decontaminate seeds. Because of the problems implied in the irradiation of foods and all of the regulations concerning it, the ICGFI has published different Good Practice Codes on irradiation (table 1). The irradiation of foods is a technique which has been given an unfortunate name, as the treatment of irradiation processing has been related to nuclear energy. This has meant that many times the irradiated food has been mistaken for being radioactive. However, these are completely different terms, as the former is treatment with radiation, while the latter refers to foods with radioactive potential. For this reason this kind of processing must demonstrate that food treated with irradiation is safe, much more so than with any other processing technique, even though there are numerous scientific tests that corroborate its healthiness. This kind of treatment has been attacked because it produces physical and chemical changes in the foods, but if one thinks of thermal treatments these produce very important alterations, and are still accepted by the consumer. Perhaps if human beings had not eaten cooked food and were at the dawn of thermal treatment this might have the same detractors as irradiation has, as the changes it produces in foods are much more intense than those produced by irradiation, in the form that this treatment is currently proposed.

4

Albert Ibarz Table 1. Codes of Good Irradiation Practice (ICGFI) Code of Good Irradiation Practice for Insect Disinfestation of Cereal Grains (ICGFI, Document 3), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Prepackaged Meat and Poultry (for control de pathogens and/or extended shelf-life)(ICGFI, Document 4), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Control of Pathogens and Other Microflora in Spices, Herbs and Others Vegetable Seasonings (ICGFI, Document 5), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Shelf-life Extension Bananas, Mangos and Papayas (ICGFI, Document 6), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Insect Disinfestation of Fresh Fruits (as a quarantine treatment) (ICGFI, Document 7), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Sprout Inhibition of Bulb y Tuber Crops (ICGFI, Document 8), IAEA, Vienna, 1991 Code of Good Irradiation Practice for Insect Disinfestatcion of Dried Fish and Salted and Dried Fish (ICGFI, Document 9), IAEA, Vienna, 1991 Code o f Good Irradiation Practice for Control of Microflora in Fish, Frog Legs and Shrimps (ICGFI, Document 10), IAEA, Vienna, 1991 Code o f Good Irradiation Practice for the Control of Pothogenic Microorganisms in Poultry Feed (ICGFI, Document 19), IAEA, Vienna, 1995 Code of Good Irradiation Practice for Insect Disinfestation of Driede Fruits and Tree Nuts (ICGFI, Document 20), IAEA, Vienna, 1995

Source: Molins, 2001.

Processed foods are those that reach the consumers, and it is the consumers who are the most concerned with their healthiness and safety. With regard to irradiated foods, as early as 1925 studies were initiated into their safety, and currently there are large numbers of publications which have been carried out on this subject. Thayer (1994) and Diehl and Josephson (1994) have performed reviews on the subject of healthiness and safety from radiological, microbiological and toxicological viewpoints and the nutritional suitability of irradiated foods. The Joint FAO/WHO/IAEA Expert Committee has examined 100 compounds of irradiated meat from cow, pig and chicken, declaring that these foods treated such are healthy and safe. Their declaration of 1980 states: “the irradiation of any food product at an average general dose of 10 kGy presents no toxicological risk; therefore it is not necessary to carry out more toxicological trials on the foods treated in this way” (WHO, 1981). Another problem that certain groups impute to the irradiation of foods is that it can present a potential genetic toxicity. In this respect studies on meat and poultry (Renner et al., 1982; Phillips et al., 1980) have been carried out which conclude that there is no genetic toxicity, the researchers having detected no abnormal effects on the X or Y chromosomes. The possible mutagenicity of meat and poultry treated with radiation has also been studied (Fruin et al., 1980), having found no mutagenic activity at all in any of the samples of meat studied. With regard to the unwillingness to accept irradiated foods, Satin (2000) presents what he calls the “Salmonella Russian roulette”. Currently it is possible to find on the market untreated fresh milk and pasteurised or sterilized milk, and if the consumer does not look at the label there is a certain possibility of acquiring untreated fresh milk, which may well contain the bacteria Salmonella, resulting in a high possibility of contracting salmonellosis.

Ionizing Irradiation of Foods

5

However, in this case there is the option of acquiring thermally treated milk, which overcomes this problem. However, when the consumer wishes to buy chicken, this is sold raw, and then it is certainly a game of Russian roulette whether or not the chicken contains the aforementioned bacterium, with the problems that it could represent. The irradiation of chicken eliminates this problem, as it directly affects the Salmonella, eliminating it from the food, and that is why the opportunity should exist on the market to choose between raw and irradiation treated chicken, as otherwise the acquisition of chicken is little more than a lottery. Irradiation is the food treatment technique that has been most studied for more than a hundred years, there being a huge number of documents which corroborate that irradiation is a safe, practical and beneficial process. Nevertheless, there is a current opposed to the treatment of foods by irradiation, denying the reality of studied facts, substituting this scientific reality for ungrounded suspicion. This rejection experimented by irradiation is not the only rejection in the history of food processing. Pasteurisation itself, which nowadays is a conventional treatment and one assimilated by the consumer, also needed some time before being generally applied, due to its detractors, not only in its early days, but rather in some cases it took more than a century before being applied. Even today it is not available in some countries. A noteworthy case of the application of pasteurisation is the treatment of milk, which nowadays is a general treatment in the majority of countries; nevertheless the case of Scotland should be pointed out, as this was one of the last European countries to publish legislation making pasteurisation obligatory, which occurred in 1983. In this country the incidence of salmonellosis was the highest in Europe, until the normative of obligatory pasteurisation was passed, and one year after its passing this incidence had become one of the lowest. Pasteurisation is a thermal treatment which is easily applied to liquid foods, as the heat is evenly distributed by mixing or shaking the fluid. However, in solid foods this is not possible, as suitable thermal treatment would imply that the least heated part of the solid would reach a suitable temperature to destroy the pathogenic micro-organism. That, however, would mean that the solid would be cooked, and this is not the intention. Another form of thermal treatment would be to make a mixture of the different parts of the solid, which could only be done by triturating it, but this would destroy the structure of the food and the result would be a completely different product. Therefore, some kind of treatment for solids should be sought which does not destroy their characteristics. This kind of treatment could be irradiation, but at the outset there were difficulties, as suitable sources of radiation were not available. At present this problem has been solved and suitable installations are available for carrying out the food irradiation treatments, aimed at obtaining foods which are healthy and safe. The milk industry presents an interesting parallelism between pasteurisation and irradiation. The difficulties associated with the introduction of milk pasteurisation have already been mentioned, although in the 1930’s the irradiation of milk with ultraviolet rays (UV) was already employed. This kind of treatment, besides affecting pathogenic microorganisms, presents an additional beneficial effect, as it increases Vitamin D content. The milk treated with UV rays was well accepted, and even the lots of the Red Cross for prisoners of war contained irradiated milk.

6

Albert Ibarz

2. IONIZING RADIATION The term ionizing radiation is given to the series of emissions of subatomic particles and electromagnetic radiation of nuclear or atomic origin, which on interacting with matter are capable of ionizing it. In other words, it is radiation which acts on matter, making it lose electrons, which leads to the production of ions. Radiation is a form of energy, and every person receives natural radiation from the sun and other natural components of the environment. In the same way as with other forms of radiant energy the radiation waves used to treat the foods form part of the electromagnetic spectrum (figure 1).

Figure 1. Radiation spectrum.

Irradiation has mostly been used in the field of medicine, as is the case of X-rays, and nuclear radiation in the detection and treatment of diseases, the sterilization of medical equipment, apparatus, pharmaceutical products and the production of sterilized foods for special hospital diets. The emission of ionizing radiation is a common characteristic of many unstable atoms. These atoms described as radioactive transform themselves to become stable atoms, which is achieved by freeing energy in the form of radiation. The kind of radiation freed by the radioactive atoms can be one of four different types: • • • •

Alpha particles (α): helium atoms, which contain two protons and two neutrons. Beta particles (β): electrons or positrons deriving from transformation in the nucleus. Gamma radiation (γ): electromagnetic radiation from the most energetic extreme of the radiation spectrum. Neutrons: chargeless particles.

Radioactive activity is the speed at which the transformations are produced in a radioactive substance, and measures the number of atoms which disintegrate in the unit of time. The unit of radioactive activity in the International System is the Becquerel (Bq), defined as a disintegration by second, although sometimes the Curie (Ci) is used, which is the activity existing in a gram of the 226Ra atom. 1 Bq = 1 disintegration/second 1 Ci = 3.7·1010 disintegrations/second

Ionizing Irradiation of Foods

7

The Ci represents a considerable activity, while the Bq is a small unit. Radioactive substances with an activity below 100 Bq/g, or natural solid substances with an activity below 500 Bq/g, are considered to be harmless. Any human being possesses an activity of approximately 4000 Bq due to 40K.

2.1. Interaction of Radiation with Matter Radiation which affects matter experiments different kinds of interactions, depending on the nature of the radiation and the type of matter. Radiation which affects matter can cause two kinds of phenomena, atomic excitation or ionization. The former produces a thermal effect, while the latter causes the formation of ions, which is why radiation is sometimes classified as thermal or ionizing, depending on the kind of interaction that it causes in the matter on which it acts. α radiation consists of particles with two protons and two neutrons, hence they are charged and are furthermore considered heavy particles. Therefore this is radiation with a limited power of penetration, of a few centimetres through the air or of a few microns in any tissue, and thus is not able to penetrate the skin. However, it can produce a high concentration of ions, which makes these particles very dangerous, as they can cause grave cellular damage. β radiation consists of electrons, which is a charged particle, and the interaction it experiments with matter responds to Coulomb’s law of electrical charges, in the same way as for the α particles. The power of penetration of electrons is greater than that of α particles, being a few metres through air, capable of traversing human skin, although not the subcutaneous tissue. γ radiation consists of high energy photons, which can be absorbed by matter according to three types of processes: the photoelectric effect, the Compton effect and the production of electron-positron pairs (e--e+). Depending on the energy of the incident radiation one type or other of interaction will occur, although the final effect of all of them is the production of charged particles. In the photoelectric effect (figure 2a) all the incident energy is absorbed by the atoms of the matter, and is used to expel an electron form the atom’s shell; this process requires energy absorption of up to 500 keV. The Compton effect (figure 2b) implies the incidence of the photon, producing an elastic collision with an electron from the atom’s shell, which provides it with enough kinetic energy to be separated from the atom, while the photon leaves in a different direction with less energy and a greater wavelength (lower frequency). If the energy of the incident photon is high enough, on interaction with the atoms a pair of electron-positron particles is produced (figure 2c). In order for the Compton Effect to occur the incident radiation must possess energy of between 500 keV and 10 MeV. In the process of electron-positron pair formation the incident energy is absorbed inside the electric field of the nucleus, the amount of incident energy required being greater than 1.02 MeV. The positron formed possesses a very short life and disappears (positron annihilation) with the appearance of two photons of 0.51 MeV of energy. γ radiation possesses a power of penetration estimated at various hundreds of metres in the air, and is capable of traversing the human body, metal sheets and up to several centimetres of lead.

8

Albert Ibarz

Figure 2. (a) Ionization process. (b) Compton Effect. (c) Pair formation.

Lastly, neutrons are chargeless particles, thus they can only interact with the atomic nuclei. Because the nuclei occupy a very small part of the total volume of matter the probability of the neutrons interacting with the matter is low, which means that these particles possess a large penetration capacity. Therefore, to summarize, the capacity for the penetration of matter of these types of radiation is different. In the case of α particles this capacity is limited as, for example, they can be retained by sheets of paper. β particles have somewhat more penetration, being able to traverse a sheet of paper, although the human body detains them. γ radiation is more penetrating, traversing the human body and sheets of different metals, although it is detained by lead sheets. As neutrons do not possess an electrical charge they are very penetrating, traversing all the materials traversed by the previously mentioned radiation types, but they can be retained by concrete walls of sufficient thickness. In any case, the ionizing radiation that interacts with matter undergoes a certain attenuation, which basically depends on two factors, one geometric and the other material. The geometric factor is that when increasing the distance between the source of radiation and the target matter the radiation becomes increasingly weak, with the attenuation inversely proportional to the square of the distance between the source and the object. The second factor is due to the fact that the attenuation depends on the type of radiation and its energy level, and on the type and composition of the matter. In this case the attenuation shows an exponential decay with the distance travelled by the radiation inside the matter.

2.2. Absorbed Radiation Dose When matter receives radiation, the incident energy of the radiation can cause ionization and/or excitation of the matter’s atoms, although other effects may also appear such as different photoelectric effects, Compton effects, the formation of electron-positron pairs etc. Furthermore part of the incident radiation may not interact with the matter, traversing it

Ionizing Irradiation of Foods

9

without producing any effect. Thus it is necessary to measure the energy absorbed by the matter, as it is this which can cause ionization. The absorbed dose (D) is the amount of energy absorbed (E) per unit mass (m) of the matter during the time that it is exposed to the radiation:

D=

E m

(1)

The dose is controlled by the intensity of the radiation and by the time that the matter is exposed to this radiation. In the International System, the absorbed dose is measured in Gray (Gy), which is one joule of energy absorbed per kilogram of mass of the matter. Sometimes a historical unit known as the rad (“radiation absorbed dose”) is still used, which corresponds to the energy of 10-2 J absorbed by each kg of irradiated matter. 1 Gy = 1J/kg 1 Gy = 100 rad These units of measurement are very small and often multiples are used. Thus, for example, the kGy is normally used, and to give an idea of the energy level it possesses, 1 kGy is equal to the amount of energy needed for a kg of water to increase its temperature by 0.25ºC. Therefore, this type of process is a “cold” or “non-thermal” method of food treatment. Another important variable is the speed with which a body absorbs the radiation, known as absorbed radiation rate, which is defined as the variation of the absorbed dose with regard to the exposure time:

dD D& = dt

(2)

This is an important variable, as the final effect of the radiation does not only depend on the amount absorbed but also on the time that the matter is exposed to the radiation. In addition to the amount of radiation absorbed by the matter (absorbed dose), the type of radiation should also be taken into account and its potential for causing biological damage. For this the equivalent dose (H) is determined, its unit of measurement in the International System being the Sievert (Sv), which as with the Gray is the relation between the energy absorbed from one jule for each kilogram of mass, but taking into account the kind of radiation:

H = D · FR

(3)

where the equivalent dose is the absorbed dose multiplied by a weighting factor FR, which depends on the type of radiation. In order to have an idea of the effect produced by the different types of radiation, if a power of penetration of 1 is given to α rays, β rays would have a value of 100 and γ rays a value of 10,000. The weighting factor values for the different types of radiation are shown in table 2.

10

Albert Ibarz Table 2. Weighting Factor FR FR

Radiation type

1 5 5 a 20 > 20

X-rays; β rays; γ rays; electrons and positrons Protons Neutrons, according your energy α radiation; heavy nucleus

In the same way as with the absorbed dose, for the equivalent dose a historical unit known as the rem (“Roentgen equivalent man”) has been used, which is equivalent to 10-2 J for each kg of matter: 1 Sv = 1J/kg 1 Sv = 100 rem Another variable used in the measurement of radiations is the effective dose (E), which takes into account the risk of developing cancers or hereditary effects and is measured in Sv. This effective dose is a weighted sum of the average doses received by the different tissues and organs of the human body:

E = ∑ Fi · H i

(4)

i

where Hi is the equivalent dose for an organ i, while Fi is the weighting factor of this organ, its value depending on the organ considered. Furthermore, it is important to take into account the dose that a person can accumulate over time, and for that purpose the committed dose is determined, which is the dose accumulated over a certain period of time. To achieve a better understanding of the possible danger of radiation, it should be mentioned that a dose is lethal when its value exceeds 4 Sv, and that for doses of up to 0.25 Sv no harmful effects have been observed. By means of illustration, the radiation that can be received in an X-ray of the thorax shows a value of 0.02 mSv, for a CT head scan the value is 3 mSv, and the average annual dose per person in Spain is approximately 3.5 mSv, adding all the natural and artificial contributions, while the worldwide average is 2.5 mSv. The annual average dose received by the Spanish population due to the nuclear industry is in the order of 0.015 mSv, which is equivalent to what a person would receive when undertaking a 3-hour flight, due to cosmic radiation.

2.3. Sources of Ionizing Radiation The sources of ionizing radiation are not only artificial, as there are also natural sources. It is normal to find radioactivity and radiation in nature. The presence of ionizing radiation in our world and in the whole universe is normal, and as such a very important part of Earth’s natural radiation is due to cosmic radiation which is of extraneous origin. It is thought that

Ionizing Irradiation of Foods

11

every second in the order of 2x1018 particles of very high energy reach the Earth, most of which are protons (86%) and α particles (12%), in addition to neutrons and other particles. Cosmic radiation represents about 10% of the radiation that a person receives. The dose of radiation received in this way depends on the altitude, as the atmosphere absorbs part of it, which means that at higher altitudes the radiation is greater, hence when travelling by plane the radiation is more than that received at sea level. It is estimated that at an altitude of some 10,000 m, which is a normal altitude of transatlantic flights, a dose of some 5 mSv is received; at the summit of a mountain at 6,700m the dose is 1 mSv; in cities located at 2,000m the dose is 1 mSv; while at sea level it is usually in the order of 0.03 mSv (UNSCEAR, 1988). Moreover, this radiation is due to electrically charged particles and due to the Earth’s magnetic field they are diverted, so that in the equatorial zone less radiation is received than at the poles. Therefore, dose depends on both terrestrial latitude and longitude. In the case of Spain every hour people are traversed by 105 cosmic rays of neutrons and 4·105 secondary cosmic rays (CSN, 1992a,b). Furthermore, the cosmic radiation that reaches us from outer space can interact with different atmospheric components to produce radioactive substances. Cosmic radiation is, on average, about 10% of the total dose received. Most of the dose received is due to the radiation which comes from the Earth itself. This is due to the fact that in the subsoil there are large amounts of radioactive elements, such as uranium and thorium, among others. This radiation means that the whole planet is impregnated with radioactivity, including in the human body. It is estimated that every hour 2·108 of γ radiation is received from the soil. Due to this cause the average radioactive content in Spain of different materials is estimated at being 3,000 Bq for an adult human being, 1,000 Bq for 1 kilogram of coffee, and 25,000 Bq for 25kg of fertilizer. These figures are much higher for radioactive residues from medical and industrial applications. It is estimated that for 1 kg of low activity residues the activity is 106 Bq, for those of average activity it is 108 Bq, and for high activity residues it is 1013 Bq. Radiation which comes from the natural decay of uranium is important, as it provokes the appearance of the gas radon, which passes through cracks and pores in the soil, mixing with the air. The decay of radon produces radioactive compounds which remain attached to the particles of dust contained in the air and reach the lungs, with an estimated disintegration in every person every hour of some 30,000 atoms, with the emission of α and β particles and γ rays. Furthermore, it should also be borne in mind that natural radionuclides are also taken in with the ingestion of food. Noteworthy among these is 40K, which the human body contains to such an amount that it is estimated that every hour some 15·106 atoms disintegrate, emitting high energy β particles and in some cases producing γ rays (NRPB, 1986). Besides these natural sources there are also those which are due to the processes which man carries out in his medical and industrial applications, which could be called artificial sources. Of particular note is the ionizing radiation deriving from medical activity, due to its use in diagnosis and the treatment of diseases. This radiation covers the range from X-rays to nuclear radiation. But artificial radiation does not only come from medical applications, as in industry and in daily activity there are numerous examples, such as luminous watches, smoke detectors, radiation to define the structure of welding, and many other cases. The production of electrical energy in nuclear installations is another source of artificial ionizing radiation, although in thermal power stations the combustion of coal also produces

12

Albert Ibarz

natural radionuclides. In Spain the dose received due to this type of activity is less than 0.001 mSv, although the people who work in these power stations can receive greater doses which nevertheless do not reach 0.01 mSv per year. Finally, mention should also go to the radiation deriving from nuclear weapons testing and accidents like that of Chernobyl, which also contribute to the dose received by the human population. This is estimated at some 0.01 mSv per year.

3. BIOLOGICAL EFFECTS OF IONIZING RADIATION The biological effects that ionizing radiation produces depend on the type of interaction that occurs with the matter on which it acts. The absorption of radiation by living organisms depends on the kind and quantity of the radiation, as well as the structure and the kind of absorbing matter. Thus different kinds of effects may be shown, although in any case the incident radiation is an energy bearer, energy which is transferred to the absorbing medium either directly or indirectly, according to the mechanisms of excitation or ionization. When the absorbed radiation produces the effect of excitation of the matter’s atoms and molecules it can cause molecular changes if enough energy is absorbed, and if this is greater than that of the atomic bonds. If the ionization process is involved the effect is more important, as changes are always produced in the atoms, and it is capable of causing alterations in the structure of the molecules on which the radiation has fallen. The ionization induced in live tissues by exposure to radiation is usually quantified by the so-called lineal energy transfer (LET), which is the amount of energy yielded per unit distance travelled by the radiation in the tissue. Radiation is classified into two categories of high and low LET. α radiation and that of neutrons are considered to be of high LET, while X-rays and β and γ radiation are considered to be of low LET. Radiation produces different effects if the doses are high or low. For high doses of radiation the effects can be of two kinds, deterministic and stochastic. Deterministic effects are those that produce immediate effects, and show a minimum dose below which these effects do not occur, but appear immediately when the dose is greater than this minimum. Stochastic effects are those of delayed appearance and are probabilistic in nature, as is the case of cancer, which may develop some years after the exposure to the radiation. Also considered as stochastic are hereditary effects, due to genetic alterations, which appear in the descendents of the organisms which have received the dose of radiation. One thing to be borne in mind is that any organism is exposed to natural radiation, and so it is difficult to evaluate the effects of exposure to low doses of radiation, as these effects could be masked by the manifestation of conditions which could be considered normal, and which may not be due to the radiation received. For these low doses the possible effects can either be genetic or cancer. The biological effects of radiation can act at different levels, on cells, tissues or on whole organisms. The biological damage as well as the acceptable doses can vary greatly with each case. According to the molecular complexity of the living organisms, the biological effect is produced for different doses of radiation (Urbain, 1986). Thus, for mammals the lethal dose is ranged from 0.005 to 0.01 kGy, for humans this dose is in the order of 4 Gy; for insects it is from 10 to 1000 Gy, for plants it is 1 kGy. For the bacteria, the lethal dose depends on if they

Ionizing Irradiation of Foods

13

are in vegetative or sporulated form; thus, in their vegetative form the lethal dose is of 0.010.0 kGy, while for the sporulated forms it is ranged from 10 to 50 kGy. The most resistant organisms are the virus whose lethal dose is ranged from 10 to 200 kGy. This indicates that the greater the molecular complexity the smaller the dose required producing biological effects. Low overall doses are capable of killing a person, or else causing significant damage, while in order to completely destroy insects, larvae and eggs, the required doses are higher, and the doses needed to destroy bacteria, fungi and yeasts are much higher still. When a cell is irradiated the radiation may act directly on the genetic material or on the macromolecules, although it may also act indirectly on the water contained in the cell, or the ionized molecules may interact with the surrounding matter. There are considered to be three different stages in the overall process, a physical stage, a chemical stage and a biochemical stage. In the first physical stage the radiation interacts with the matter, which can excite or ionize its atoms, with characteristic times of 10-15 and 10-17 s, respectively. In the chemical stage free radicals are formed, with characteristic times in the order of 10-12 s. The last is a molecular or biochemical stage, in which the free radicals recombine and can form toxic molecules. The molecules formed by direct irradiation, or radicals, and those obtained indirectly, are known as radio-induced substances. These substances can be toxic or harmful to the cell. The presence of these substances causes the cell to set in motion the so-called cell repair mechanisms, giving rise to three different possible situations. If a very large amount of toxic substances has been produced the result is death. Otherwise the cell may survive although the harm caused to the genetic material is great and does not allow the cell to reproduce; this is called reproductive failure. If the amount of genetic damage is not excessive the cell can repair part of the genetic material, allowing it to reproduce, although in this case a delay in cell division is observed, enabling the transmission of mutations to subsequent generations. Therefore, in irradiated cells two types of damage are caused, the formation of toxic substances or genetic damage. From a survival viewpoint there may be important consequences, with the loss of tissue or organ functionality, the development of cancer, sterility problems, or the transmission of mutations to offspring. On penetrating tissues charged particles (α and β) lose energy by electric interaction with the electrons in the shells of the atoms which they strike. Due to indirect effects such as the photoelectric effect, the Compton effect and pair formation, when X and γ rays hit the tissues they end up freeing atomic electrons, which produces an end result of ionization. Due to the electric interaction of the charged particles an electron is split from the atom’s shell, which produces a positively charged atom. The separated electron can ionize other atoms. Generally, both the electron and the ionized atom are very unstable and react rapidly, giving rise to new molecules, some of which are highly reactive, being as they are free radicals. These radicals may react with each other and with other molecules, causing changes in molecules which are biologically important for cell functioning. Altogether the process lasts about one millionth of a second. The biological transformations which can occur in such a short interval may destroy or modify the cells, possibly giving rise to the appearance of genetic defects or cancer. In the case of food irradiation the problem is quite a different one, as it is necessary to evaluate the effects of irradiation from a food viewpoint. Foodstuffs are biological material and irradiation can cause different effects, such as the destruction of insects and microorganisms, the production of toxic substances, it may damage genetic material, or it may reduce the nutrient content of the foods. Of these four effects, on principle, the only desired

14

Albert Ibarz

one is the first. The appearance of toxic substances is not an effect which is unique to irradiation, as the appearance of this kind of substance has also been observed in chemical and thermal food treatments (Raventós, 2003; Satin, 2000; Molins, 2001). With regard to genetic damage, it should be pointed out that from the food viewpoint reproductive viability is not important, although care should be taken to avoid causing any problem for the consumer. The destruction of nutrients not only occurs in irradiation treatments, but is also observed in other kinds of food treatments, such as in thermal processes.

4. IONIZING RADIATION IN THE FOOD INDUSTRY Food irradiation is considered as the process of applying high energy to a food with the aim of pasteurising, sterilizing or prolonging its commercial life, eliminating micro-organisms and insects.

4.1. Types of Ionizing Radiation The sources of ionizing radiation which are applied in the food industry are X-rays, electron beams and γ radiation. Other kinds of radiation which have been used are ultraviolet rays (UV). UV rays are obtained using lamps which contain gases at different pressures, and are characterised by a low power of penetration, of only a few millimetres because their emission spectrum corresponds to rays with a wavelength considerably longer than X-rays. X-rays constitute a much more energetic electromagnetic radiation, for which their power of penetration is higher, showing a continuous spectrum of radiation with a maximum value of 5 MeV. X-rays are usually obtained by bombarding a metal plate with a high potential electronic beam (figure 3).

Figure 3. X Rays generation.

γ radiation is produced with radioactive isotopes, which in the food industry are normally the radioisotopes of 60Co and 137Cs. At present 60Co is the most commonly used for irradiating foods with γ radiation, as it is relatively straightforward to obtain and it produces

Ionizing Irradiation of Foods

15

radiation with a greater power of penetration than that of 137Cs. The energy spectrum of γ radiation is not continuous, but rather is discreet, and depends on the radioisotope used. The energy proceeding from 137Cs is 0.66 MeV, while that from 60Co is 1.17 and 1.33 MeV. 60Co is produced in a nuclear reactor bombarded with neutrons granules of 59Co highly refined, and in the decay process β and γ radiation is produced. 137Cs is obtained as a result of the fission of 235U, producing β and γ radiation (figure 4).

Figure 4. Radioactive disintegration of

60

Co and 137Cs.

The source of irradiation of the isotope 60Co is obtained from 59Co, which is compressed in cylindrical tablets which are placed in 50cm long steel tubes. These tubes containing 59Co are placed in a nuclear reactor where they are bombarded constantly with neutrons, which produce radioactive 60Co, which is capable of producing a controlled emission of γ rays. 137Cs is extracted from the bars of used combustible of the nuclear reactors. This reprocessing of nuclear waste has become very controversial and its possible use as a source of irradiation in food is very improbable. Thus it appears that 60Co offers better possibilities for food processing; moreover this type of source shows a greater degree of effectiveness, greater penetration of γ rays and greater environmental safety, as it is insoluble in water. The electron beam is a series of electrically charged particles of high energy, of up to 10 MeV. For the electrons to have a high energy level they are led to a linear accelerator which confers them with high voltages, thereby obtaining electrons with high speed, approaching that of light. The advantage of this compared to γ radiation is that the electronic beam is produced in an electric machine and can be turned on and off like a light bulb. Nevertheless, its power of penetration is low, from some 5 to 10 cm. Table 3 shows the advantages and disadvantages of the different sources of radiation used in the irradiation of foods. When foods are irradiated with γ rays, X-rays and electron beams, a certain degree of radioactivity can be induced in them. However, this is such a small amount that it is not distinguishable from the natural radiation possessed by the food. Hence the variation in radioactivity among different non-irradiated foods is greater than any difference existing between the same food when irradiated and non-irradiated (Stewart, 2001). From a food viewpoint, of all the energy transfer mechanisms of ionizing radiation due to the incidence of photons, the most important is the Compton Effect. For the photoelectric effect to be produced the energy of the incident photon would have to be lower than that

16

Albert Ibarz

provided in the normal intervals of food irradiation. On the other hand, for the formation of electronic pairs higher energy levels would be required (Stewart, 2001). Table 3. Advantages and disadvantages for the different irradiation sources Source type γ-Rays (60Co 137Cs)

Electron beam (10 MeV)

X-rays (5 MeV)

Advantages - Deep penetration - Reliability of irradiation source - Easy automation

- Electric source only work when switch on - Unit control possibility - High dose rate (several kGy/s) - Absence of environmental impact - Low cost for running - Electric source only work when switch on - Unit control possibility - High dose rate (several kGy/s) - Absence of environmental impact - Low cost for running

Disadvantages - First category radioactive facility - Transport and storage radioactive sources - Activity loss of storage radioactive source - Dose rate determined by source - Permanent irradiation emission - High cost for running and security - First category radioactive facility - Limited Penetration - Need a lot of handling staff - Need automated equips - First category radioactive facility

Source: Raventós (2003).

4.2. Irradiation Dose in Foods As the radiation used in food treatment is electromagnetic in nature (X-rays and γ rays) or else accelerated electrons, the weighting factor FR (table 2) of the equivalent dose is 1, with which the absorbed dose (D) and equivalent (H) coincide. Hence in this case the absorbed dose is usually employed. For every food product the permitted doses of radiation depend on its characteristics and the aim of the treatment. This means that the dose for the elimination of insects, for pasteurisation and sterilization will be different. Hence three irradiation categories are considered according to the dose employed, either low, medium or high doses. Low doses are those which do not exceed 1 kGy, and are used in the control of insects in grain, in the control of trichina in pork and can also inhibit the decomposition of fruits and vegetables. Average doses are those in the range from 1 to 10 kGy, and are applied in the control of pathogens in meat, poultry and fish and also retard the growth of moulds on strawberries and other fruits. High doses exceed 10 kGy, and are used to kill micro-organisms and insects in spices, and also when aiming to obtain commercially sterile foods. According to the dose of radiation the treatment usually receives different names. Thus the elimination of non-spore producing pathogenic microorganisms and parasites to an imperceptible level is called radicidation. The treatment of foods with ionizing radiation aimed at increasing their average life by reducing the number of modifying micro-organisms (pasteurisation) receives the name of radurisation, while the elimination of micro-organisms by irradiation to levels of sterilization is called radapertization. Table 4 shows the doses used to irradiate foods and the applications of each case.

Ionizing Irradiation of Foods

17

Table 4. Food irradiation, dose and applications Dose Low

< 1 kGy

Medium

1 a 10 kGy

High

10 a 50 kGy

Absorbed dose (kGy) 0,04 – 0.10 0,03 – 0,20 0,50 – 1,00 1–3

Application Sprout inhibition of tubers and bulbs Insects, grubs and eggs sterilization Fruit and vegetables ripening process control Insect death

1–7 2 - 10 15 – 50

Radicidation (pathogen elimination) Radurization (pasteurization) Radapertization (sterilization)

10 - 50

Spices and seasonings decontamination

4.3. Changes in Irradiated Foods Irradiated foods are treated at low levels of radiation, which means that only chemical changes are possible, and that changes which would make them radioactive do not occur. The large number of investigations carried out indicates that the changes produced in irradiated foods are similar to those produced by a conventional cooking treatment. The studies show that in irradiated foods toxic or mutagenic effects do not exist, and that irradiation does not produce chemical residues in the food. Irradiation is a cold process, which means that there is only a slight temperature rise of the food during processing. There is hardly any change in the physical appearance of the irradiated foods, which do not undergo the changes in texture and colour shown by foods treated by heat pasteurisation, or by tinned and frozen foods. Certain bad tastes in meat and excessive softening of fresh peaches and nectarines have been reported. In irradiated foods some changes do occur, although they are not as important as those that occur with conventional cooking methods. When the high energy particles hit the matter, electrons are released from the atoms, giving rise to ions. The radiolytic products formed in this way can interact to form new compounds. A few of these reactions may give rise to strange tastes. The FDA concluded that “very few of these radiolytic compounds are unique to irradiated foods; approximately 90% of radiolytic compounds are natural compounds of the food” (Web and Penner, 2000).

4.4. Irradiated Food Labelling Retailed irradiated foods must bear the symbol radura (figure 5) which identifies them as such. Furthermore the sentence “treated with irradiation” must also appear. Manufacturers are permitted to add the objective of the treatment; thus, for example, it may be labelled as “treated with radiation to control deterioration”. With non-packaged fruits and vegetables every piece must be labelled, and furthermore on the shelf containing the product and clearly visible to the consumer there must be a sign indicating that they have been treated with radiation.

18

Albert Ibarz

For unpackaged irradiated elements sold by weight, such as spices, it is also necessary to indicate that they have been irradiated. Nevertheless, if these foods are incorporated as ingredients in other foods, it is not necessary for the final products to be labelled as irradiated. Hence, if irradiated pepper is added to a processed meat, it is not necessary for it to be mentioned on the product that this irradiated spice has been used as an ingredient.

Figure 5. Radura symbol.

5. EFFECTS OF RADIATION ON THE FOOD’S COMPONENTS Any food is composed of different compounds, of which water is by far the commonest component, so that the possibility of the incident radiation affecting a water molecule is very high. Apart from water other majority components of foods are carbohydrates, fats and proteins, although other so-called minority components should also be considered, among which vitamins and minerals feature. Of all of these components contained in foods, minerals are not affected by irradiation of the food whatever the dose absorbed. However, irradiation produces changes in the other components, which are studied in detail in the following sections.

5.1. Water Radiolysis The interaction of radiation with a molecule of water can cause it to be ionized, due to the loss of an electron, resulting in H2O+, or else its dissociation into OH- and H+, due to the breakage of the link. Excitation may also occur, whereby the water molecule moves to a state of higher energy. The different ions that appear may react with water molecules to give rise to a whole series of compounds, among which are included atomic and molecular hydrogen,

Ionizing Irradiation of Foods

19

hydroxyl radicals, hydrogen peroxide, a solvatated proton and an aqueous (solvated or hydrated) electron (Stewart, 2001). The range of molecules produced is known as radioinduced products. Of all of these, only molecular hydrogen and hydrogen peroxide are stable molecules, although they are consumed in the reactions following on from irradiation. Among the range of radio-induced products the most reactive are H2O2 and the OH- radical, which can react with adjacent molecules to form new compounds. The hydroxyl radical is a powerful oxidizing agent, while hydrogen and the aqueous electron are reducing agents, which means that the water contained in the food may well undergo both oxidising and reducing reactions. The hydroxyl radical can remove the hydrogen atoms in the C-H links of the olefins or aromatic compounds. It should be noted that the presence of oxygen in the medium favours the formation of the most reactive products and the processes of oxidation. If oxygen does not exist in the medium the amount of peroxide formed is very low, which leads to a reduction in the oxidation reactions of the food. pH is another of the variables of the medium which may affect the final result of irradiation. The H+ ions can combine with a solvatated electron to produce atomic hydrogen radical, which can combine with a hydroxyl radical freeing the aqueous electron (Diehl, 1995). Therefore, an acid medium may favour the disappearance of the aqueous electron, while a basic medium favours its formation. Acid foods will favour reduction reactions during the irradiation process. Another factor affecting the whole process is the temperature of the product being irradiated. It should be mentioned that the initial stages of ionization and excitation and the reactions of the more reactivate species are temperature independents (Swallow, 1977). If the product is frozen the intermediate reactive components of water radiolysis are trapped and, as a result, cannot react with other components of the food (Urbain, 1986). Hence the effects of irradiation in frozen foods are fewer than when the irradiation is carried out on a product containing water in its liquid form.

5.2. Carbohydrates The chemistry of the radiation of carbohydrates is quite complex, as different radiolytic compounds can be formed. The main reaction involving carbohydrates when a food is irradiated is that which occurs with the hydroxyl radicals formed by water radiolysis. The end products of these reactions are ketones, aldehydes and acids. Depending on the dose they hydrolyse and oxidise, giving rise to simpler compounds, although they may also undergo a depolymerisation, making them more susceptible to the attack of hydrolytic enzymes. The action of the hydroxyl radical on the carbohydrates begins with the separation of a hydrogen atom linked with a carbon atom, giving rise to new radicals. These radicals may undergo new disproportioning, dimerisation and dehydration reactions. All of this can lead to the appearance of a large number of radiolytic products. The irradiation of solid sugars with a low molecular mass causes the reduction of their fusion point. In some cases this leads to the appearance of browning, as occurs with the irradiation of reducing sugars like glucose and fructose, and a mix of gases like hydrogen, carbon monoxide and carbon dioxide, among others (Dauphin and Saint-Lèbe, 1977). In the case of carbohydrates with a high molecular mass, like the polysaccharides (starch, pectin,

20

Albert Ibarz

cellulose), radiation causes their degradation. This degradation seems to be due to the division of the glycosidic bonds, giving rise to sugars of a lower molecular mass, although the degradation may continue to give other radiolytic products like formic acid, methanol, acetone, ethanol and ethyl formiate (Dauphin and Saint-Lèbe, 1977; Sokhey and Hanna, 1993; Stewart, 2001). The presence of other food compounds exercises a protecting effect on the carbohydrates. Therefore the end result of irradiation on the carbohydrates present in a food is not the same as that which occurs in the irradiation of pure sugar solutions. Moreover, carbohydrates generally occur in foods of plant origin, and the presence of the plant cell wall will exercise a protecting effect (Stewart, 2001), which means that in the end these foods are less susceptible to the effects of radiation and that the changes produced are insignificant.

5.3. Fats Fats are mostly insoluble in water, so that the effects of radiation on them will be different from those produced on carbohydrates. If the radiation dose is lower than 35 kGy, the changes in the physical and many chemical properties of the lipids are insignificant (Urbain, 1986). The radiation of fats causes the formation of cationic radicals and excited molecules which subsequently react to give rise to new compounds. The radicals formed may break up or dimerise, or a molecular disproportioning may be produced, although the adjacent free electrons may also be fixed (Nawar, 1977, 1978, Delincée, 1983a). The number of end products obtained is wide and diverse, as fatty acids, propanediol esters, aldehydes, ketones, diglycerides, alkanes, alkenes, methyl esters, hydrocarburates and short chain triglycerides may all be formed. However these compounds are also produced when fats are thermally treated, and the amount produced is even greater than in the irradiation treatments.

5.4. Proteins In the irradiation of proteins various types of reaction may be produced (Hanna and Shepherd, 1959; Delincée, 1983b; Simic, 1983). An important feature is the rupture of the protein bond, giving rise to peptides of shorter chains than the protein from which they originate. The amino acids that form part of the polypeptidic fraction may react with the free radicals produced by water radiolysis, which breaks the peptidic bonds. Irradiation can cause the denaturalisation of the proteins due to aggregation or disgregation of the polypeptides, or else due to the changes in their secondary and tertiary structures. However, it should be noted that the denaturalisation of proteins deriving from irradiation processes is less than that produced by thermal treatments. In order for the radiation to affect the amino acids doses of 40-50 kGy are required, and in this case certain organoleptic changes in the treated foods are brought about. Irradiation of proteins at doses up to 35 kGy causes no discernible reduction in amino acid content (Urbain, 1986). In the presence of fats, the irradiation of amino acids can give rise to aldehydes and ketones, which are products that cause bad odours in the treated food. It should be noted that irradiation has hardly any effect on enzymes. In order to inhibit their activity doses of up to 60 kGy would be needed. As the doses used in the irradiation of

Ionizing Irradiation of Foods

21

food do not approach this value, there exists the possibility of the occurrence of deteriorating enzymatic reactions in irradiated foods. If these foods are to be stored for long periods they should receive additional heat treatment for the enzymes to be inactivated.

5.5. Vitamins Vitamins are food micronutrients which are sensitive to irradiation. This sensitivity depends on the dose received by the food and the type of vitamin. It should be supposed that water-soluble vitamins are more sensitive to deterioration by radiation, due to the reactions of the hydroxyl radical produced in the radiolysis of water. The vitamins A, E, C, K and thiamine (B1) are relatively sensitive to radiation. In contrast, riboflavin, niacin, pyridoxine, pantothenic acid and vitamin D are much more stable (Diehl and Josephson, 1994). Of all of these thiamine is the most sensitive to irradiation, although its losses amount to no more than 2.5% after treatment. Generally, the loss of vitamins in irradiated foods is usually insignificant (Diehl ,1991;.Thayer et al., 1991). The sensitivity of water soluble vitamins to irradiation follows the order: vitamin B1 > vitamin C > vitamin B6 > vitamin B2 > niacin = folate > vitamin B12; and for fat soluble vitamins the order for their sensitivity is: vitamin E > carotene > vitamin A > vitamin D > vitamin K (Stewart, 2001). As occurs with some macronutrients, the presence of oxygen and the temperature affect the deterioration of vitamins caused by radiation. This deterioration of vitamins can be reduced by working at low temperatures in the absence of oxygen, with vacuum packaging or under an atmosphere of nitrogen. Table 5 shows the content of different vitamins in chicken meat after receiving different treatments. It can be seen that the irradiated chicken meat shows similar vitamin content to that treated thermally, and that for thiamine the results are even better. Table 5. Vitamin content in poultry with different treatments Vitamin Tiamine Riboflavin Piridoxine Ac. nicotinic Ac. pantotenic Biotine Folic ac. Vitamin A Vitamin D Vitamin K Vitamin B12

Source: Thayer (1990).

Vitamin concentration (mg/kg) (dry weight) Frozen Thermal γ−Irradiated 2.31 1.53 1.57 4.32 4.60 4.46 7.26 7.82 5.52 212.9 213.9 197.9 24 21.8 23.5 0.093 0.097 0.098 0.83 1.22 1.26 2716 2340 2270 375.1 342.8 354 1.29 1.01 0.81 0.008 0.016 0.014

β−Irradiated 1.98 4.90 6.70 208.2 24,9 0.100 1.47 2270 466.1 0.85 0.009

22

Albert Ibarz

6. INACTIVATION OF MICROORGANISMS BY RADIATION Irradiation is an efficient method of treatment for the destruction of parasites and both pathogenic and non-pathogenic bacteria, and to lesser degree viruses. The mechanisms of inactivation, as indicated in section 3, are due to the damage that irradiation produces in the genetic material, both directly and indirectly. Thus, ionizing radiation can collide directly with the cell’s genetic material and damage its DNA, although it can also act on an adjacent molecule which subsequently reacts with the genetic material. When ionizing radiation acts on a DNA molecule it does so at random, so that it may collide with a single or double strand of the molecule. When the collision occurs with a single strand of DNA the damage caused is not necessarily lethal and may give rise to mutations; although it is also possible that the damage caused exceeds the cell’s capacity for reparation, resulting in its death. If the damage is inflicted on a double strand it may slice the DNA molecule with lethal effects for the cell, which is incapable of repairing the damage caused by the incident radiation. If the radiation does not fall directly on the genetic material but rather on an adjacent molecule, the effects are more complex. As mentioned previously, the commonest molecule is that of water, and so there is a high probability of the radiation falling on one of these molecules. The resulting molecules, or radio-induced products, may react with the nucleic acids damaging the genetic material, although they may also give rise to a series of reactions with the different molecules they encounter, finally producing a range of molecules which are toxic for the cell, with possible lethal consequences. In addition to these effects on the genetic material, irradiation may affect other cell components, such as the membrane, the enzymes and the cytoplasm. This damage may not be directly lethal, but can damage the cell sub lethally, so that it can impede the survival of the damaged cell. The different organic compounds that form part of the cell show different sensitivity towards irradiation, proportional to their molecular mass. Therefore amino acids are less sensitive than enzymes, which in turn are less sensitive than DNA molecules (Pollard, 1966). Cells which have received irradiation are damaged by some of the mechanisms described, and the cell responds to this damage with its own repair and survival mechanisms. The recovery of surviving cells after irradiation depends on pH (Fielding, et al., 1994). Generally the inactivation of micro-organisms is caused by the damage done to the DNA, and thus the survival mechanisms are centred on the reparation of this molecule. Consequently, those micro-organisms which are efficient in the repair of DNA show the greatest probability of survival, and are therefore the most resistant to irradiation. In addition to this survival mechanism there are micro-organisms which are resistant to radiation owing to the fact that they possess mechanisms for the elimination of damaged genetic material, which aids their survival. One example of a bacterium resistant to radiation is Deinococcus radiodurans, which can be found in foods which have been irradiated with doses of up to 40 kGy (Dickson, 2001; Moseley, 1976). In thermal treatments the inactivation of micro-organisms at a certain temperature follows first-order kinetics. In a similar fashion, it has been observed that the number of micro-organisms present in an irradiated food decreases with the applied dose according to first-order kinetics:

Ionizing Irradiation of Foods

N = N 0 exp(− k D )

23 (5)

where N0 is the number of micro-organisms initially present in the food, N the number of micro-organisms which survive, D is the dose applied and k is the kinetic constant of microorganism destruction by irradiation. For pasteurisation doses of between 1 and 10 kGy are required, while for the sterilization of the product doses of between 15 and 50 kGy are needed. In food irradiation treatments it is very useful to apply the variable D10, known as the decimal reduction dose, which represents the dose applied in order to reduce the number of micro-organisms to one tenth of the original. In this case N = 0.1N0, so that:

⎛ N0 ln⎜⎜ ⎝ 0,1N 0

⎞ ⎟⎟ = ln (10 ) = k D10 ⎠

(6)

which is the same as:

D10 =

2,303 k

(7)

Table 6a shows the D10 values for some of the commonest bacteria in food. It can be seen that for the spore-producing forms the D10 values are higher, which indicates that they are more difficult to destroy. This difference compared to the vegetative forms can be explained by the fact that in the spore-producing forms the water content is much greater, which would minimize the secondary effects of the radiation, leading to an increase in the resistance to radiation. D10 data is also available for some parasites and certain pathogenic viruses (Dickson, 2001) (table 6b). In the case of viruses, the contamination of the food generally originates from infected food handlers which can transmit the disease if they have contaminated the food they have prepared. The D10 values for viruses are greater than those for vegetative bacteria, and are more similar to the spore-producing forms, which indicate greater resistance to radiation. Thus, the coxsackie virus shows values of 7-8 kGy, the polio virus a value of 3, while the hepatitis A virus shows a value of 2 (Sullivan et al., 1973; Heidelbaugh and Giron, 1969; Mallet et al., 1991; Dickson, 2001).

24

Albert Ibarz Table 6a. D10 values for selected gram-positive bacteria Bacteria Gram-positive Spore formers Bacillus cereus

Clostridium botulinum Clostridium perfringens Non-spore formers Listeria monocytogenes

Staphylococcus aureus Gram-negative Aeromonas hydrophila Campylobacter jejuni Escherichia coli O157:H7 Salmonella

Shigella

Vibrio Yersinia enterocolitica

Medium

Conditions

D10 (kGy)

Distilled water Mozzarella cheese Yogurt Beef stew Water

20 – 25ºC; aerobic - 78ºC, aerobic; spores - 78ºC 20 – 25ºC; type E 20 – 25ºC

1.6 3.6 4.0 1.4 1.2 – 1.3

Chicken Chicken Ground beef Trypticase soy broth Phosphate buffer Ice cream Poultry Meat

2 – 4ºC 12ºC 12ºC 0ºC 0ºC - 78ºC 10ºC ---

0.77 0,49 0.5 – 0.9 0.21 0.18 2.0 0.42 0.86

Ground fish Ground fish Ground turkey Ground beef Ground beef Salsa Roast beef Ground beef Deboned chicken Deboned chicken Deboned chicken Deboned chicken Liquid whole egg Liquid whole egg Oysters Crabmeat Oysters Crabmeat Oysters Crabmeat Prawns Shrimps Ground beef Ground beef Minced meat

2ºC - 15ºC 0 – 5ºC, vacuum - 17ºC 2 – 5ºC 3ºC; S. typhimurium 3ºC; S. typhimurium 20ºC; S. typhimurium - 40ºC; air; S. typhimurium - 40ºC; air; S. enteritidis - 40ºC; air; S. newport - 40ºC; air; S. anatum Frozen; S. seftenberg Frozen; S. gallinarum S. dysenteriae S. dysenteriae S. flexneri S. flexneri S. sonnei S. sonnei Frozen; V. cholerae Frozen;V. parahaemolyticus 25ºC - 30ºC ---

0.16 0.274 0.19 0.307 0.241 0.416 0.567 0.55 0.533 0.534 0.436 0.542 0.47 0.57 0.40 0.35 0.26 0.22 0.25 0.27 0.11 0.1 0.2 0.39 0.10 – 0.21

Source: Dickson (2001).

Table 6b. D10 values for selected virus Virus Coxsackie Polio Echovirus Hepatitis A Rotavirus SA11

Source: Dickson (2001).

Medium Raw and cooked beef Fish MEM medium Oysters Oysters

Conditions -90 – 16ºC 0ºC -------

D10 (kGy) 6.8 – 8.1 3 4.3 – 5.5 2 2.4

Ionizing Irradiation of Foods

25

7. EFFECT OF IRRADIATION ON FOODS The effects produced by irradiation depend on the type of food which is being treated, as well as the characteristics of the medium. These effects depend on radio-induced processes, as they are favoured by the presence of oxygen, by an increase in pH, by an increase in temperature and by the water content. In dry and dehydrated food products direct irradiation is the most effective, as there is less water available and therefore the formation of free radicals is reduced. It is also recommendable to irradiate at freezing temperatures, as the water is not present in liquid form, which eliminates the process of radiolysis. Furthermore, it is better to irradiate vacuumpacked foods or those in a modified atmosphere, in the absence of oxygen or with very low levels of this gas. With regard to pH, the irradiation of foods with a value of less than 4.5 is safer, this requiring lower radiation doses than a higher pH. It is important to underline that irradiation, like other treatments, is only effective with foods in a healthy and hygienic state. Irradiation will never improve already degraded foods. The food’s components will vary depending on the food under consideration. Thus carbohydrates will be the majority components of foods of plant origin, while proteins and fats are predominant in foods of animal origin. A brief description of the effects of irradiation for different types of food and irradiation processes according to their objective is given below.

7.1. Irradiation of Meat and Meat Products From the sacrifice of the animal to the retail of the meat, meats and poultry are subjected to different stages of handling and processing, which could lead to their becoming contaminated with pathogenic micro-organisms such as Salmonella, Campylobacter, Listeria, Yersinia and Escherichia coli. Much intoxication by Escherichia coli have been caused by contaminated minced meats. In chicken and other poultry the most problematic intoxications are due to Salmonella, either because the food has not received suitable treatment or it has not been handled correctly. Several studies have demonstrated that the irradiation can be an effective treatment in the destruction of different pathogen microorganisms in meat (Farkas, 1987; Lee et al., 1996) Irradiation at doses of between 10 and 50 kGy eliminates the previously mentioned microorganisms, and even destroys Trichinella in raw beef and pork (Kotula, 1983; Molins, 2001). The Food and Drug Administration (FDA) in the United Stated in 1985 approved irradiation of fresh pork at 0.3 to 1.0 kGy in order to inactivate Trichinella spiralis. One of the most dangerous micro-organisms is Clostridium botulinum, due to the botulinic toxin that it produces. In order to eliminate Clostridium botulinum and the spore-producing forms of other micro-organisms it is necessary to apply higher doses of radiation. For this microorganism it is necessary a 12D value of 38 to 48 kGy in a radiation treatment in pork, ham or chicken (Kreiger et al., 1983). In order to obtain high quality meats and meat products a recommended procedure is an initial stage of packaging, followed by gentle thermal treatment to inactivate enzymes, then a stage of freezing to end with the irradiation of the product. One advantage obtained by

26

Albert Ibarz

irradiation in comparison to other preservation treatments is that it is possible to eliminate the preservatives which are added to many processed meat products.

7.2. Irradiation of Fish and Shellfish Products The processing of fish and shellfish using irradiation can control pathogenic and harmful micro-organisms, as well as prolong the product’s commercial life. Irradiation does not cause sensorial changes in the products. Several authors have reviewed different aspects of the irradiation on fish and shellfish products (Kilgen, 2001; Nickerson et al., 1983). Pathogenic micro-organisms present in fish and shellfish come from the water in which they are submerged. The micro-organisms which usually appear are Clostridium perfringens, Clostridium botulinum, Vibrio parahemolyticus, Vibrio cholerae, and Aeromona hydrophila. In addition to these pathogens, owing to deficient handling of the fish and shellfish products, there may also be Salmonella, Shigella and Staphylococcus aureus. Generally this type of food is sold fresh and by weight, exposed on ice, for which processing with irradiation effects the initial flora that it contains, but in subsequent handling it can become contaminated again. Therefore, although a safe product is obtained initially, if the subsequent handling is unsuitable, the global treatment will be ineffective. Shellfish, such as crustaceans, feed by filtering the surrounding water and absorbing the food that it contains. If the water is contaminated this contamination also passes on to the shellfish. A dangerous case could be contamination by bacteria like Escherichia coli and viruses found in waste waters. The doses normally applied in the irradiation of fresh and cooked fish and shellfish is from 0.75 to 1.5 kGy, while for frozen products this dose is higher, of between 2 and 5 kGy (Raventós, 2003; Kilgen, 2001). These doses are not high enough to eliminate the spores of bacteria like Clostridium botulinum, or the toxins that the product may contain. For this reason and other, the fish and shellfish products should be processed and stored in cold below 3ºC, in ice or frozen. This way, after the irradiation treatment, if the product is conserved under appropriate conditions it can increase the time of storage from 1 to 3 weeks in the case of fresh or cooked product, and if it is frozen it can duplicate the conservation time (Raventós, 2003).

7.3. Processing of Eggs and Egg Products by Irradiation In eggs and their derivatives, the most problematic pathogenic micro-organism is Salmonella enteriditis. Thermal pasteurisation treatments eliminate this pathogen, but the resulting product is different from the original, as the heat causes the denaturalisation of the proteins. As mentioned in the previous section irradiation is considered to be a “cold” thermal treatment, hence by irradiating this type of food the pathogenic micro-organism can be eliminated, while the food conserves all of its organoleptic properties. The dose for eliminating Salmonella is 2.5 kGy. With this dose whole eggs and derivatives (mayonnaise, creams, sauces, etc) can be irradiated, producing products which have no perceivable differences compared to their non-irradiated equivalents.

Ionizing Irradiation of Foods

27

The irradiation process has been applied to inactivate the Salmonella contamination in dried egg products, and a dose of 2 kGy has been recommended (Narvaiz et al., 1992; Farkas, 2001).

7.4. Irradiation of Fresh Fruits and Vegetables Diverse revision works exist on the fruits and vegetables processed with radiation (Akamine and Moy, 1983; Willemot et al., 1996; Thayer and Rajowsky, 1999; Thomas, 2001a). Once harvested fruits and vegetables start deteriorating, and for this reason postharvest treatments are normally applied to these products, based on refrigeration processes. This type of product possesses high carbohydrate content, although its water content is between 80% and 95%, and what’s more the intercellular spaces contain oxygen. Hence the irradiation of fruits and vegetables is usually carried out at relatively low doses, in order to avoid the previously mentioned problems associated with water radiolysis and subsequent oxidation of the carbohydrates due to the presence of oxygen. Irradiation at low doses can delay the ripening of some fruits and vegetables, leading to the prolongation of their useful life. With doses of 0.3 to 1 kGy a significant increase in their commercial life can be obtained (Thomas, 2001a). In the case of mangos and bananas a prolongation of their useful life by one and two weeks, respectively, has been achieved. Additionally, the irradiation of mushrooms and asparagus at doses of 1-1.5 kGy has slowed their maturation (Satin, 2000). Generally, doses of up to 0.25 kGy applied to most fruits and vegetables do not affect their organoleptic properties. Doses of between 0.25 and 1 kGy can cause significant modifications and may even cause a reduction in the vitamin content. Doses of more than 1.75 kGy enable the control of storage-related diseases. If a dose of between 1 and 3 kGy is applied it may accelerate the softening of the product, and may lead to the development of undesirable organoleptic properties. In the case of irradiation at doses of more than 3 kGy an excessive ripening is produced, with the appearance of certain disagreeable tastes (Raventós, 2003).

7.5. Irradiation of Tubers and Bulbs Tubers and bulbs are vegetables which are widely used ain human food. The most important tuber crops are potato, yam, sweet potato and ginger; while the most important bulbs are onion, garlic and shallot. Of these the potato and the onion are the products for which most information on irradiation treatments exists. Irradiation is applied to these products with the aim of reducing the losses which occur after harvesting. These losses are due to four main factors (Thomas, 2001b): physical, physiological, microbiological and entomological factors. Included among the physical factors is mechanical damage done during harvesting and handling. Physiological factors are losses caused by the formation of sprouts, the appearance of rot, and the transpiration and respiration of the product. These last two factors involve the loss of water during storage, which decreases if the product has a well-formed skin or dry outer scales. In stored potatoes greening of the skin can occur as a as a result of the formation of chlorophyll in artificially lit

28

Albert Ibarz

warehouses, which is accompanied by the production of solanine, a glycoalkaloid which confers a bitter taste and which can be poisonous. In some cases during storage the starch may be converted into sugars, the presence of which causes browning problems when elaborating crisps. With regard to microbiological factors the growth of bacteria and fungi is favoured by a warm and humid environment, and can be controlled by storage at low temperatures. In addition to these factors another significant cause of losses in tubers is that owing to the attack of the moth Phthorimaea operculella (Zeller), which is very common in hot tropical and subtropical regions. Of all of these factors, one which causes the most concern is the appearance of sprouts due to germination, and in order to reduce this kind of deterioration different techniques have been applied. Thus germination is inhibited by storing potatoes at 4ºC, while with bulbs the temperature should be 0ºC. Another way of inhibiting sprout formation is by the use of chemical products like maleic hydrazide (HM; 1,2-dihidropiridacina-3-6-diona), applied before the bulbs are picked (del Rivero and Cornejo, 1971; Isenberg, 1956; Thomas, 2001b). Other chemical inhibitors applied to potatoes are Chlorpropham [CIPC; isopropyl-N-(3chlorophenyl) carbamate], Propham [IPC; isopropyl-N-phenyl carbamate] and Tecnazene [TCNB; tetrachloro nitrobenzene] applied after harvesting, and they are proved effective to extend stored life of these products at 7-10ºC (Booth and Proctor, 1972; Smith, 1977). The application of these chemical inhibitors has been questioned for the possible risks they represent for human health. The application of irradiation can prevent the germination of this type of food. As early as in 1936 it was observed that the application of X-rays could inhibit sprout formation in vegetables (Metlitsky, 1936). The application of X-rays at a dose of 4.5 Gy inhibits the germination of potatoes (Sparrow and Christensen, 1950). The works carried out on the inhibition of germination in tubers and bulbs is very extensive, including studies on the effects of the applied dose rate, the interval between harvesting and irradiation, sprout inhibition mechanisms, the influence of temperature and humidity during storage, sensorial characteristics, susceptibility of the irradiated products to rotting, and methods of irradiation detection (Thomas, 2001b). Irradiation appears to produce changes in the levels of endogenous growth hormones and affects the metabolism of nucleic acids. It has also been observed that irradiation is more effective if applied immediately after harvesting, when they are in their period of lethargy, as the dose applied to prevent germination must be increased as the time between harvesting and irradiation increases. For tubers normal treatment doses are from 70 to 150 Gy, which not only prevent germination but also wither the sprouts already formed. In the case of onions and other bulbs doses of between 20 and 90 Gy are usually applied (Thomas, 2001b), which bring about the inhibition of germination if applied within 2-4 weeks after harvesting. In general the period of lethargy is prolonged if the onions are stored at low temperatures (º0C), which enables suitable inhibition of germination if an irradiation treatment is applied during this time. Before treating with irradiation it should be ensured that the tubers are clean, dry and soil-free. In the case of bulbs it is essential that they are well hardened, with dry outer scales and neck. Furthermore, if these products are washed it is essential that they are dry when treated with irradiation.

Ionizing Irradiation of Foods

29

7.6. Disinfestation of Cereals, Seeds, Legumes, Dried Fruits, Nuts and other Dried Foods Many insects can cause serious problems in stored products, particularly in grains and cereals. The FAO estimates that during storage between 5% and 10% of cereals can be lost due to infestation by insects. To avoid these losses control techniques of temperature and controlled atmosphere have been used, although the use of pesticides for the control of insects in stored foods is very common. Currently a widely used product in this control is methyl bromide, as it presents a wide fumigation spectrum, although it has been put under very strict control by the developed countries. This fumigant possesses an ozone reduction potential of more than 0.2, which is the limit marked by the Montreal Protocol (Ahmed, 2001), which recommends its gradual removal and total prohibition before the year 2010. Therefore there is widespread concern to find alternative treatments to fumigation with pesticides. Disinfestation by radiation began with studies on the sensitivity to radiation of the rice weevil (Hunter, 1912). Since then there have been various studies on the irradiation of insects aimed at disinfestation (Lorenz, 1975; Ahmed, 2001). The Codex Alimentarius Commission recommends a dose of 1 kGy for the disinfestation of all agricultural foods and products (CAC, 1984), although with most dry foods this can be achieved at lower doses. In insects, as in other organisms, the effects of radiation are due to the effects that are produced in their cells. Sensitivity to radiation is directly proportional to the cells’ reproductive activity and inversely proportional to their degree of differentiation. Hence, dividing cells are more sensitive than those at the adult stage. Eggs are more sensitive to radiation, and even at sublethal doses subsequent malformations and adult sterility are observed. Insect larvae in their period of lethargy are more resistant to radiation, although these larvae die after the pupal stage. Male pupae are more resistant to radiation than females. It appears that sensitivity to radiation varies from one type of insect to another, so that, for example, the Coleoptera (beetles) are more sensitive to radiation, while the Lepidoptera (moths) are the most resistant group. The sterilization doses differ from one type to another, so that for beetles a dose of 50 Gy is sufficient, while with moths doses of some 1000 Gy are necessary. However, in order to kill the insect a dose of between 3 and 5 kGy will be needed (Ahmed, 2001). To perform a satisfactory irradiation treatment the type of insect that is to be targeted should be known, and its development stage which is most sensitive to radiation identified. Any food product treated with doses up to 10 kGy does not present a toxicological risk and such doses can be applied; nevertheless in 1983 the Codex Alimentarius Commission (table 1) recommended that in the disinfestation of cocoa beans, dates, legumes, rice, wheat and derivatives the dose of 1 kGy should not be exceeded. To summarise, the disinfestation doses applied will depend on the type of insect and product although, in general, it is considered that this objective is achieved with doses in the interval of 0.5 kGy to 3 kGy. However, it should be borne in mind that for doses of more than 1 kGy some products may be affected in their organoleptic qualities, in their vitamin content and in their starch properties. Table 7 shows the radiation doses most commonly used to control insects present in stored foods. Irradiation treatment for the disinfestation of cereals, legumes and seeds does not cause any harmful effects in them, although it does not prevent the products from being reinfested. Although there is no harmful effect on the quality of these irradiated products, it should be

30

Albert Ibarz

borne in mind that irradiation at disinfestation doses may prevent the seeds and cereals from germinating, which in some cases may be more of an inconvenience than an advantage. Disinfestation treatment by radiation has not only been applied to cereals, legumes and seeds, but also has been successfully applied to other dehydrated products such as salted, smoked and dried fish, nuts, dehydrated fruits, coffee and cocoa beans, etc. Table 7. Irradiation dose used to insect control in stored products Species Coleoptera Sitophilus oryzae S. granarius S. zeamais Tribolium castaneum T. confusum T. cadeus Rhyzopertha dominica Latheticus oryzae Oryzaephilus surinamensis O. mercator Callosobruchus analis C. chinensis C. maculatus Bruchus rufimanus Bruchidius incarnatus Trogoderma granarium Dermestes maculatus Lasioderma serricorne Nerobia rufipes Araecerus faciculatus Lepidoptera Anagastus kuehniella Plodia interpunctella Cadra cautella Sitrotoga cerealella Nemapogon granellus

Stage

Dose (kGy)

All All All All All All Larvae Adults All All All All All All All All All All All All

0.16 0.16 0.16 0.20 0.20 0.20 0.25 0.20 0.20 0.20 0.20 0.20 0.20 0.40 0.40 0.25 0.50 0.50 0.30 0.75

Larvae, pupae Larvae Larvae, pupae All All

0.60 0.45 0.45 0.60 0.50

Source: Ahmed (2001).

7.7. Irradiation of Spices, Herbs, Seasonings and other Dry Food Ingredients Among the different possibilities for the application of irradiation, perhaps that which presents an immediate application is the treatment by irradiation of dry food ingredients, as is the case of spices, herbs and plant-based seasonings (Farkas, 2001). These kinds of products are appreciated because they confer special flavours and aromas to many cooked dishes. The main problem is that they possess high levels of contamination due to bacteria, fungi and yeasts, which can contaminate the food to which they are added. Therefore it is necessary for them to receive suitable treatment to reduce this contamination.

Ionizing Irradiation of Foods

31

Until fairly recently, the destruction of micro-organisms was achieved by means of fumigation with ethylene oxide, and to a lesser degree with propylene oxide. However, this type of fumigation can cause toxicity problems, as it is carcinogenic, and subsequently its use on food has been prohibited in many European countries. Hence the best alternative for treating this kind of product is irradiation, as it does not present this problem of toxicity, achieving a degree of decontamination much greater than any other kind of treatment. High dose irradiation treatment does not cause sensorial changes or affect the functional properties of spices. Furthermore, the process of microbial reduction continues during subsequent storage. Spices and food ingredients possess low water content, enabling the use of high doses in the irradiation treatment, and in fact doses of more than 5 kGy are usually employed. For doses of between 5 and 10 kGy the bacterial population is reduced to between 1% and 1 per thousand. Doses of between 7.5 and 15 kGy do not affect the sensorial properties of spices, although they may affect herbs (Sugimoto et al., 1986; Onyenekwe et al., 1997; Farkas, 2001). Many of these products are sold already packaged, which makes it essential to identify packaging materials which are not affected by radiation. The most common treatment is the irradiation of the food once it has been packaged, which enables the destruction of the microbial load while at the same time conserving the characteristic aromas of these dry food ingredients.

7.8. Irradiation of Milk Products As mentioned earlier, irradiating milk with ultraviolet radiation, besides destroying micro-organisms also produces an increase in vitamin D, although the content of other vitamins such as vitamin B and C is reduced. This irradiation may produce bad odours, which can be prevented if the treatment is performed in an atmosphere of nitrogen. The most important contamination that can occur in cheeses is due to Listeria monocytogenes, which may be present even in pasteurised cheese. These bacteria can proliferate at low temperatures. Table 8. Irradiation of different cheese types Cheese type

Irradiation source

Purpose

Brinsen

60

Increase stored time

Camembert

60

Cottage/Camembert

60

Increase shelf-life Destruction of Listeria and Salmonella Bacterial population destruction

Cheddar

Electrons

Rind decontamination

Fresco

Electrons

Listeria elimination

Gouda

60

Organoleptic changes

Kashar

60

Increase shelf-life

Mozzarela

60

Listeria elimination

Ras

60

Bacteria elimination

Source: Calderón (2000).

Co Co Co

Co Co Co Co

32

Albert Ibarz

In France the irradiation of Camembert cheeses elaborated with non-pasteurised milk has been permitted. In general, irradiation used in the elaboration of cheeses enables the use of non-pasteurised milk, prolonging its useful life, while at the same time ensuring the elimination of pathogenic bacteria (Raventós, 2003). Table 8 shows different cases of the application of radiation in the elaboration of different kinds of cheeses.

7.9. Irradiation of Wines and Liqueurs With the aim of preventing the development of bacteria in wines chemical and physical agents are added. The chemical agents used are, among others, sulphurous anhydride, ascorbic acid, caproic and caprilic acids; while physical treatments are in the form of microwaves and ultrasounds; although antibiotics have also been used (nisine and piramicine) and lactic enzymes (lisozimes and cimolases). Nevertheless, studies are being carried out into the possible application of ionizing radiation in the treatment of wines. There are various objectives of radiation, one of which is to use it as an alternative to treatment with carbon dioxide to prevent the development of bacteria and viruses in bottled wines. Irradiation is effective in preventing the souring of wines and enables the prolongation of their commercial life. The ionising irradiation can be applied to improve organoleptic properties such as colour, odour and flavour. Table 9 shows the applications of irradiation in different alcoholic products. Table 9. Irradiation of different alcoholic products Product Brandy (from sweet potato) Bottled beer

Irradiation source γ−Rays

γ−Rays

Target Improvement of organoleptic quality Pathogen reduction Microbial load reduction

Wine

γ−Rays < 0.8 kGy γ−Rays

Prevent bacterial and virus growth in bottled wines Speed up aging process

γ−Rays < 0.6 kGy

Microorganisms elimination Organoleptic properties variation

γ−Rays (2.4 kGy) (70ºC, 10 min) γ−Rays

Increase of shelf-life

γ−Rays Electrons

Microbial load reduction Sour stopping

Wine (Madeira, Rakia) Wine (Romania)

wine (from rice) Grapefruit and pulp

Cork

Source: Calderón (2000).

Sterilization

Effect Bitter taste elimination Taste improvement Dark color appearance Unpleasant taste Possible organoleptic changes Improvement of organoleptic characteristics Discoloration Decrease of pigments and tannins Decrease of SO2 and permanganate content No detection in organoleptic changes Favorable development of organoleptic characteristics s Pulp irradiation produces low quality wines Prevent to unpleasant taste formation

Ionizing Irradiation of Foods

33

7.10. Irradiation as a Quarantine Treatment Due to the great demand for food generally no country is self-sufficient and must resort to importation. Imported foods more often than not can contain a high microbial load which can be hazardous. Furthermore, they may contain entomological species which do not belong to the importing country, and if appropriate measures are not taken these can reach plague proportions in a short period of time. The insects contained in foods develop in equilibrium in their natural environment, but when they are transported to another country where their natural predators do not exist they can reproduce without control and cause serious problems. In the face of the possibility of the outbreaks of plagues and their grave consequences, the laws of every country impose obligatory quarantine measures. The difference between a disinfestation treatment and a quarantine treatment should be noted. Disinfestation kills, eliminates and/or inactivates pests in a product at any level, while a quarantine treatment can achieve the disinfestation of a pest put in quarantine at a predetermined level. In some cases an exemption from quarantine can be conceded if there is the guarantee that the appropriate measures have been applied in order to achieve the effective disinfestation of the food. These treatments are normally performed in the country of origin, during transport or at the final destination. Table 10. Quarantine treatment dose by irradiation for several pests Quarantine pest

Geographic distribution

Mexican fruit fly, Anastrepha ludens West Indian fruit fly, A. obliqua Zapote fruit fly, A. serpentina Caribbean fruit fly, A. suspensa Melon fly, Bactrocera cucurbitae Oriental fruit fly, B. dorsalis

Extreme southern Texas to Guatemala Caribbean islands, Mexico to Brazil Mexico to Argentina Florida, Caribbean islands Asia, parts of eastern Africa, Hawaii India to southern China, Hawaii, N. Mariana Islands Northern Australia India to China, Laos, Singapore, Hawaii Australia, New Guinea, New Caledonia, Austral Islands, Society Islands Mediterranean, Africa, Central America and South America, Middle East, Hawaii, Western Australia, North Mariana Islands United States and Canada east of Rocky Mountains, Utah Sub-Saharan Africa, India, Asia, Australia, Oceania,, much of tropical and subtropical Americas

Jarvis fruit fly, B. jarvisi Malaysian fruit fly, B. latifrons Queensland fruit fly, B. tryoni Mediterranean fruit fly, Ceratitis capitata Plum Curculio, Conotrachelus nenuphar Sweet potato weevil, Cylas formicarius elegantulus

Dose (kGy) 0.07 0.1 0.1 0.1 0.21 0.25 0.075 0.15 0.075 0.225

0.092 0.165

Source: Hallman (2001).

One of the most widely used quarantine treatments has been ethylene dibromide, although in 1987 it was prohibited in the USA, owing to its carcinogenic potential (Ruckelhaus, 1984), and in subsequent years other countries have also prohibited it. This provided the opportunity to apply radiation as a quarantine treatment (Hallman 1999, 2000,

34

Albert Ibarz

2001; Johnson and Marcotte, 1999). Hence papayas treated with irradiation were initially sent from Hawaii to California, and subsequently in 1995 they were also sent from Hawaii to Chicago treated with a dose of 0.25 kGy, while in August of the year 2000 different tropical fruits were sent from Hawaii and treated in irradiation centres of New Jersey and Illinois (Hallman, 2001). Currently Hawaii possesses irradiation centres for fruit treatment, which enables importation from these islands. In Texas and California guavas irradiated with a dose of 0.15 kGy have been received, with the aim of eliminating the Caribbean fruit fly, while Florida has received approval to send irradiated fruit, and sweet potatoes have been treated with a dose of 0.165 kGy to control the sweetpotato weevil. Various studies have been carried out on different pests at different radiation doses, which are shown in table 10, while table 11 shows the absorbed doses required in order achieving quarantine safety for various groups of pests (Hallman, 2001). Table 11. Absorbed doses that might achieve quarantine security Pest Grup Aphids and whiteflies Seed weevils (Bruchidae) Scarab beetles Fruit flies (Tephritidae) Weevils (Curculionidae) Borers (Lepidoptera) Thrips Borers (Lepidoptera) Spider mites Stored product beetles Stored product moths Nematodes

Objective Sterilize actively reproducing adult Sterilize actively reproducing adult Sterilize actively reproducing adult Prevent adult emergence from third instar Sterilize actively reproducing adult Prevent adult growth from late larva Sterilize actively reproducing adult Sterilize late pupa Sterilize actively reproducing adult Sterilize actively reproducing adult Sterilize actively reproducing adult Sterilize actively reproducing adult

Dose (Gy) 50–100 70–100 50–150 50–150 80–165 100-280 150–250 200–350 200–350 50–400 100–1000 ~ 4000

Source: Hallman (2001).

8. FOOD IRRADIATION PLANTS Food irradiation plants, whatever the process they apply, consist of different elements which are common to all of them. Hence the differential stage is usually that of the treatment applied. It is important to bear in mind that the zone where the products awaiting treatment are stored is separated from the zone where the already-treated products are stored; otherwise there may be cross-contamination. In any treatment plant the following elements can be distinguished. a) Storage zone of the products awaiting treatment. Generally this is located near the loading area of the treatment devices. b) Loading zone. In this zone the products awaiting treatment are loaded in crates or on suitable supports, which are then placed on a conveyor belt. c) Conveyor belt. This is used to transport the products awaiting treatment from the storage zone to the treatment point. This belt also conveys the irradiated products to the treated products storage zone. The belt’s movement should allow the products to receive the adequate dose.

Ionizing Irradiation of Foods

35

d) Irradiation zone. This is the main element of any plant, and as it is potentially the most dangerous area it must be isolated from the plant operation personnel. For this reason the treatment chamber is covered with 2 metre thick concrete walls which protect the operators. Inside the chamber there is the source of radiation. Because water is one of the best protections against radiation deriving from the decay of 60Co or 137Cs, the radioactive source is submerged in a pool, and is raised by remote control once the food to be treated is inside the treatment chamber. In the case of the electron accelerator the machine comes on when the product enters the chamber. The trajectory followed by the product should ensure that it receives the dose calculated for the treatment’s objectives; hence the travelling speed of the conveyor belt should be such that the time spent by the product inside the treatment chamber is that required. When the product is large or very dense, it is turned over and irradiated again in order to ensure that it receives adequate treatment. e) For plants which use an electron accelerator it is necessary for them to be equipped with a chamber for the refrigeration circuit. These accelerators are usually compact, and there are different models on the market, with a differentiated range of potentials. One such model has a potential of 35 kW, which emits electrons with energies of between 3 and 10 MeV, although there are also models with potentials of 80 and 150 kW. As for the sources of 60Co o 137Cs, there should be a pool to store these sources, which consist of bars containing the capsules of radioactive material. These bars are kept in deionised water in a covered pool buried at a depth of some 4 metres underground, in order to guarantee operator safety. When the product is treated, these bars rise vertically out of the pool and once the treatment is over they are re-submerged. f) Treated product loading zone. In a part of the plant well away from the loading zone, where the treated product is received. g) Storage zone. Before definitive storage, the product must be suitably labelled and the received dose measured. Then it is taken to the warehouse where it is stored under suitable conditions, so that there is no recontamination of the product. For this reason this zone should be at some distance from the zone where the non-treated product is stored. h) Control laboratory. This is where the dosimetry in all parts of the plant is controlled. It is necessary to measure the dose received by the products, although it is also essential to measure the radiation received by the personnel manning the plant. The two basic types of ionizing radiation treatment plants differ of the source used to provide this radiation. In anyone of them the product is taken to the irradiation area by means of conveyor belt, although they can differ in the form in that the foods are transported; this way, plants based on different concepts of food transport exist. The food to be treated is packed in boxes or pallets that are transported by a conveyor until the irradiation zone; thus, depending on the product-handling concepts there are several models of plants (Kunstadt, 2001). In addition to all of these elements the plant should have a control room, from where all the operations of the plant are controlled, including the speed of the conveyor belt, the systems for raising and lowering the bars of the radioactive source, and all the safety systems of the plant. It should be underlined that this type of plant is completely automated, and that

36

Albert Ibarz

all of the mechanisms can be controlled from the control room. Other elements which are found in irradiation plants are offices and auxiliary service rooms.

Figure 6. Tote box irradiator (Courtesy of MDS Nordion).

Figure 7. Pallet conveyor irradiator (Courtesy of MDS Nordion).

Ionizing Irradiation of Foods

37

TROLLEY RAIL (RAIL SUPPORTING STRUCTURE NOT SHOWN FOR CLARITY)

REPRESENTATIVE HOIST MECHANISM

RETAINING MECHANISM POOL BELL SURGE TANK

FLOOR LEVEL UNDERGROUND PORTION

EARTH

SOURCE PLENUM

BELL LIMITER PLENUM GUIDE

GRAY*STAR Genesis Irradiator TM

Figure 8. Genesis Irradiator™ (Courtesy of GRAY*STAR, Inc.)).

Figure 9. Electron beam for food irradiation device.

Figure 6 shows a typical plant for a tote box irradiator, while figure 7 shows pallet irradiator. Also, there are irradiation plants where the cobalt-60 radiation source never leaves the shielded pool (figure 8). Product is moved into the pool by special product containers

38

Albert Ibarz

(bells) via an overhead hoist and trolley system. At the bottom of the pool, the product is irradiated in a stationary position on two sides of a fixed dry plenum filled with inert helium that contains the source of radiation. Consequently, the irradiator is inherently safe. Instead of lifting the source out of the pool into a shielded chamber, the product is lowered into the pool adjacent to the source. To accomplish this product must obviously be kept dry, and a solution to the problem of lost efficiency that is normally associated with underwater irradiation, had to be found. Figure 9 shows a diagram of a high voltage generator of electron beams applied to the irradiation of food.

9. DOSIMETRY With irradiation treatments it is necessary to be able to determine the dose received by the food, in order to ensure that the received doses are within the limits dictated by the current legislation. A simple dosimetry system capable of covering the whole range of irradiation levels received by the food does not exist. The choice of dosimetry system depends on the requirements of the type of food and the aim of the application. For such purposes the so-called dosimeters are used, which vary according to the value of the dose they must measure. All measuring systems possess a device which shows an effect produced by the radiation, as well as a counter which measures this effect. The commonest effects caused by radiation are optical changes produced in solutions or in solid materials, and these are measured by means of spectrophotometers. For reliable measurement there must be a pre-calibration according to pre-established norms, and this must be appropriate for the situation in which it is going to be used. The products most frequently used in dose measurement are alanine, amino acids, radiometric films, cellulose triacetate, colorants, iron sulphate, K/Ag dichromate, among others, depending on the range of radiation that is to be measured (Ehlermann, 2001). Some dosimeters are routinely used to check whether the food has received radiation; generally these are substances which change colour when irradiated and which enable the user to see at a glance whether the food has been subjected to irradiation for the appropriate period. The advantage of this type of dosimeter is that it can be placed inside the packaging and can tell if the product has been irradiated. The commonest routine dosimeters are radiometric colorants and those that indicate the decomposition of plastics, the former measure the range of 0.1 to 50 kGy, while the latter measure from 5 to 50 kGy.

10. NORMATIVE Throughout the history of food irradiation the process of approval by different countries has been gradual and unequal. Hence, in the former Soviet Union the irradiation of potatoes and grains was unblocked in the years 1958/59, in Canada the irradiation of potatoes was permitted in 1960, while in the USA in the years 1963/64 it was already permitted to irradiate wheat, flour, bacon and potatoes. The Joint FAO/IAEA/WHO Expert Committee on the Wholesomeness of Irradiated Foods (JECFI) declared in 1980 that irradiation with doses of up to 10 kGy does not present toxicological problems, and does not cause nutritional and microbiological problems. However, it wasn’t until after 1984, when the General

Ionizing Irradiation of Foods

39

Standardised Codex for Irradiated Foods was published, that many countries started establishing regulations on the irradiation of food. The International Consultative Group on Food Irradiation (ICGFI) was established in the same year under the patronage of the FAO/IAEA/WHO, with one of its aims being that of advising on the application of irradiation, providing the necessary information. Table 1 shows the good practice codes of irradiation for different kinds of food, while table 12 shows the technological limits of the recommended doses for good irradiation practice, published by the consultative group ICGFI. Table 12. Advisory technological dose for good irradiation practice Food type

Purpose

Type 1: bulbs, roots and tubers

To inhibit sprouting during storage To delay ripening Insect disinfestation Shelf-life extension Quarantine control

Type 2: fresh fruits and vegetables (different to type 1)

Type 3: cereals and their milled products, nuts, oilseeds, pulses, dried fruits Type 4: fish, seafood, and their products (fresh or frozen)

Type 5: raw poultry and meat and their products (fresh or frozen)

Type 6: dry vegetables, spices, condiments, animal feed, dry herbs, and herbal teas Type 7: dry food of animal origin Type 8: miscellaneous foods, including by not limited to honey, space foods, hospital foods, military rations, spices, liquid egg, thickeners

Dose Maxima (kGy) 0.2 1.0 1.0 2.5 1.0

Insect disinfestation Microbial load reduction Reduction of pathogenic microorganisms Shelf-life extension Control of infection by parasites Reduction of pathogenic microorganisms Shelf-life extension Control of infection by parasites Reduction of pathogenic microorganisms Insect disinfestation Insect disinfestation Control of molds

1.0 5.0

Microbial load reduction Sterilization Quarantine control

> 10 > 10 > 10

5.0 3.0 2.0

7.0 3.0 2.0

10.0 1.0 1.0 3.0

Source: IAEA (1998), Molins (2001).

In 1999 some 45 countries possessed authorization for the irradiation of one or more food products (ICGFI, 1999). Unfortunately, national regulations on the irradiation of food are different from one country to the next, and furthermore there is no international agreement on packaging materials used in the irradiation of food. In the USA and the UK regulations on irradiation have been adopted based on types of food, something which has not been done in most countries. For several years now efforts have been made to promote the harmonization of the regulations existing in different countries which have approved the irradiation of food. For this purpose work is taking place in the development of a model based on the General

40

Albert Ibarz

Standardised Codex for Irradiated Foods, although it also incorporates diverse concepts contained in the good practice codes of irradiation (table 1); however, due to the specific needs of each country or region there are four variations of a basic model, one applicable for the Asian/Pacific region, one for Africa, a third for Latin America and the Caribbean and finally a fourth for the Near East (Molins, 2001). In a similar fashion as to what occurs on a world scale the situation in the European Union varies according to the country, and there is a heated debate about the irradiation of food. However, in 1999 directives were approved in order to permit the irradiation of different food products. Likewise, the conditions under which the irradiated products should be labelled are established as well as the types of ionizing irradiation sources permitted and the methodology for measuring the doses received by the food. Furthermore, the member states are obliged to set in motion the legal, regulatory and administrative dispositions that must be fulfilled. In August 2001 a communication of the Commission of the European communities was published regarding the foods and ingredients authorized for treatment with ionizing radiation. All of the existing regulations on irradiated foods in the European Union can be found on the corresponding web (European Commission; European Food Safety Authority-EFSA). Likewise, there are also the Directives 1999/2/CE and 1999/2/CE of the European Parliament and the Council of the European Union, published in the Official Bulletin of the European Communities (13th March 1999), relating to the laws that govern foods and food ingredients treated with ionizing radiation, including dry aromatic herbs, spices and plant-based seasonings. Moreover, in the Official Bulletin of the European Union of 11th March 2003 a list is provided of the foods and food ingredients which the member states have authorized for treatment with ionizing radiation.

REFERENCES Akamine, E.K.; Moy, J.H. In: Preservation of Food by Ionizing Radiation; Vol. 3. Josephson, E.S. and Peterson, M.S.; Eds.; CRC Press: Boca Raton, FL, 1983; pp 129-158 Ahmed, M. (2001). In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; Wiley-Interscience: New York, NY, 2001; pp: 77-112 Booth, R.H.; Proctor, F.J. PANS, 1972, 18, 409-430 CAC. Codex General Standard for Irradiated Foods, Codex Alimentarius Comission, CAC/, 1984; Vol. XV, E-1, Codex Stan 106-1983, Joint FAO/WHO Food Standards Programme, FAO Rome Calderón, T. La Irradiación de Alimentos. Principios, Realidades y Perspectivas de Futuro; McGraw-Hill, Madrid, Spain, 2000 CSN (1992a). Consejo de Seguridad Nuclear. Dosis de Radiación, Madrid, Spain [http://www.csn.es/] CSN (1992b). Consejo de Seguridad Nuclear. Protección Radiológica, Madrid, Spain [http://www.csn.es/] Dauphin, J.F.; Saint-Lèbe, L.R. (1977). In: Radiation Chemistry of Major Food Components; Elias, P.S. and Cohen, A.J.; Eds.; Elsevier Scientific: Amsterdam, 1977; pp: 131-220 del Rivero, J.M.; Cornejo, J. Bol. Patol. Veg. Entomol. Agric., 1971, 31, 71-75

Ionizing Irradiation of Foods

41

Delincée, H. In: Recent Advances in Food Irradiation; Elias, P.S. and Cohen, A.J.; Eds.; Elsevier Biomedical Press: Amsterdam, 1983a; pp: 89-114 Delincée, H. In: Recent Advances in Food Irradiation; Elias, P.S. and Cohen, A.J.; Eds.; Elsevier Biomedical Press: Amsterdam, 1983b; pp: 129-147 Dickson, J.S. 2001). In: Food Irradiation. Principles and Applications; Molins, R.; Ed. Wiley-Interscience: New York, NY, 2001, pp: 23-36 Diehl, J.F; Josephson, E.S. Acta Alimentaria 1994, 23(2): 195-214 Diehl, J.F. Food Control 1991, 2 (1): 20-25 Diehl, J.F (1995). In: Safety of Irradiated Foods; 2nd ed., Marcel Dekker: New York, NY; pp: 43-88 Farkas, J. Acta Alimentaria 1987, 16, 351-384 Farkas, J. (2001). In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 291-312 Ehlermann, D. In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 387-414 Fielding, L.M.; Cook, P.E.; Grandison, A.S. J. Appl. Bacteriol. 1994, 76, 412-416 Fruin, J.T.; Kuzdas, C.D.; Guthertz, L.S. Proc. Eur. Meeting of Meat Research Workers, 26, 1980; Vol. I, E-20, p. 241 Hallman, G.J. Postharvest Biol. Technol. 1999, 16, 93-106 Hallman, G.J. J. Agric. Forest Entomol. 2000, 2, 1-11 Hallman, G.J. In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 113-130 Hannan, R.S.; Shepherd, H.J. J. Sci. Food Agric., 1959, 10, 286-295 Heildelbaugh, N.D.; Giron, D.J. J. Food Sci. 1969, 34, 239-241 Hunter, W.D. J. Econ. Entomol., 1912, 5(2): 188 IAEA. Report of Joint AAEA/FAO/IAEA Regional Workshop on Present Status and Guidelines for Preparing Harmonized Legislation on Food Irradiation in the Near East, Tunis, Tunisia, Oct. 12-16. IAEA, Vienna, 1998 ICGFI. Database of Food Irradiation Clearances, 1999, (in http://www.iaea.org/icgfi) Isenberg, F.M. Proc. Am. Soc. Hortic. Sci. 1956, 63, 343-348 Johnson, J.; Marcotte, M. Food Technol., 1999, 53(6), 46-51 Kilgen, M.B. In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 193-211 Kotula, A.W. Food Technology 1983, 37(3), 91-94 Kreiger, R.A.; Snyder, O.P.; Pflug, I.J. J. Food Sci. 1983, 48, 141-145 Kunstadt, P. In: Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 415-442 Lee, M.; Sebranek, J.G.; Olson, D.G.; Dickson, J.S. J. Food Prot. 1996, 59, 62-72 Lorenz, K. CRC Crit. Rev. Food Scie. Nutr., 1975, 6, 317-382 Mallet,J.C.; Beghian, L.E.; Metcalf, T.G.; Kaylor, J.D. J. Food Safety, 1991, 11, 231-245 Metlitsky, L.V. Cited by P. Thomas. In Food Irradiation. Principles and Applications; Molins, R.; Ed.; Wiley-Interscience: New York, NY, 2001; pp: 241-272 Molins, R. In Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 1-21 Moseley, B.E.B.. Photchem. Photbiol. Rev., 1976, 7, 223-274 Narvaiz, P.; Lescano, G.; Kariyama, E.; Kaupert, N. J. Food Safety 1992, 12, 263-282

42

Albert Ibarz

Nawar, W.W. Radiation Chemistry of Major Food Components; Elias, P.S. and Cohen, A.J. Eds.; Elsevier Scientific: Amsterdam, 1977; pp: 21-61 Nawar, W.W. J. Agric. Food Chem. 1978, 26(1), 21-25 Nickerson, J.F.R., Licciardello, J.J., and Ronsivalli, L.J. In Preservation of Food by Ionizing Radiation; Josephson, E.S. and Peterson, M.S. Eds.; CRC Press: Boca Raton, FL, 1983; pp: 12-82 NRPB Living with Radiation. National Radiological Protection Board, Her Majesty’s Stationary Office, London, UK; 1986 Onyenekwe, P.C.; Ogbadu, G.H.; Hasimoto, S. Postharvest Biol. Technol. 1997, 10, 161-167 Phillips, B.J.; Kranz, E.; Elias, P.S. Food Cosmetic Toxicol. 1980, 18, 471-475 Pollard, E.C. (1966). In Enciclopedia of Medical Radiology; Zuppinger, A. Ed.; SpringerVerlag: New York, NY, 1966; Vol. 2 Raventós, M. Indústria Alimentària. Tecnologies Emergents; Edicions UPC: Barcelona, Spain, 2003; pp: 101-127 Renner, H.W.; Graf, U.; Wurgler, F.E.; Altmann, H.; Asquith, J.; Elias, P.S. Food Chem. Toxicol., 1982, 20, 867-878 Ruckelhaus, W.D. Fed. Reg., 1984, 49(70), 14182-14185 Simic, M.G. In Preservation of Food by Ionizing Radiation; Josephson, E.S. and Peterson, M.S. Eds.; CRC Press: Boca Raton, 1983; Vol. II , pp: 1-73 Smith, O. Potato: Production, Storage and Processing; AVI Publishing: Wesport, CO; 1977 Sokhey, A.S.; Hanna, M.A. Food Struct. 1993, 12, 397-410 Sparrow, A.H.; Christensen, E. Am. J. Bot. 1950, 37, 667-371 Stewart, E.M. In Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001; pp: 37-76 Sugimoto, B.M.; Cherry, W.B.; Dodd, D.J. Appl. Environ. Microbiol. 1986, 34, 602-603 Sullivan, R.; Scarpino, P.V.; Fassolitis, A.C.; Larkin, E.P.; Peeler, J.T. Appl. Microbiology 1973, 22, 61-65 Swallow, A.J. In Radiation Chemistry of Major Food Components; Elias, P.S. and Cohen, A.J.; Eds.; Elsevier Scientific: Amsterdam, 1977; pp: 5-20 Thayer, D.W. J. Food Quality, 1990, 13, 147-169 Thayer, D.W.; Fox Jr., J.B.; Lakritz, L. In: Food Irradiation; Thorpe, S. Ed.; Applied Science: London, 1991; pp: 285-325 Thayer, D.W. Food Technology In Food Irradiation. Principles and Applications, Molins, R.; Ed.; Wiley-Interscience: New York, NY, 2001;, 1994, 48(5): 132-135 Thayer, D.W.; Rajowky, K.T. Food Technology 1996, 53(11), 62-65 Thomas, P. In Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001a; pp: 213-240 Thomas, P. In Food Irradiation. Principles and Applications; Molins, R.; Ed.; WileyInterscience: New York, NY, 2001b; pp: 241-272 UNSCEAR United Nations Scientific Committee on the Effects of Atomic Radiation. Sources, Effects and Risks of Ionising Radiation. Report to the General Assembly, United Nations. New York, 1998 Urbain, W.M. Food Irradiation, Academic Press, London, 1986 Web, M; Penner, K.P. Food irradiation. MF-246, Kansas State University, Kansas, ,2000 WHO WHO, Wholesomeness of Irradiated Foods. Technical Report Series 659, Geneva, 1981

Ionizing Irradiation of Foods

43

Willemot, C.; Marcott, M.; Deschenes, L. In: Processing Fruits: Science and Technology; Somogy, L.P. and Ramaswamy, H.P. Eds.; Technomic Publishing: Lancaster, 1996; Vol. 1, pp: 221-260

Electronic Media Resources Codex Alimentarius (FAO/OMS): http://www.codexalimentarius.net/ European Commission: http://www.europa.eu.int/comm/ Consejo de Seguridad Nuclear (CSN)-Spain: http://www.csn.es/ European Food Safety Authority (EFSA): http://www.efsa.eu.int/ Foundation for Food Irradiation Education (FFIE): http://www.food-irradiation.com/ International Consultative Group on Food Irradiation (ICGFI): http://www.iaea.orat/icgfi/ Nuclear Energy Institute: http://www.nei.org/ Ministerio de Sanidad y Consumo (Spain): http://www.msc.es/

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 45-101

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 2

FRUITS AND VEGETABLES DEHYDRATION IN TRAY DRYERS Dionissios P. Margaris*1 and Adrian-Gabriel Ghiaus*2 1

Fluid Mechanics Lab., Mechanical Engineering and Aeronautics Dept., University of Patras, GR-265 00, Patras, Greece 2 Technical University of Civil Engineering – Bucharest Bd. P. Protopopescu nr. 66, RO- 021414, Bucuresti, ROMANIA

ABSTRACT Dehydration involves simultaneous transfer of heat, mass and momentum in which heat penetrates into the product and moisture is removed by evaporation into an unsaturated gas phase. Owing to the complexity of the process, no generalized theory currently exists to explain the mechanism of internal moisture movement. In this Chapter, the investigation of momentum, heat and mass transfer phenomena, in both laboratory and large scale convective drying systems (suitable for dehydration of thermolabile products) by means of experimental measurements and numerical simulation are presented. The air flow inside complex geometry spaces, such as drying rooms containing hundreds of trays arranged in rows and columns, is analyzed by solution of 3-D momentum turbulent flow equations for different room configurations. Laboratory measurement data, concerning the space velocity distribution and the pressure field of the air flow over one tray, are provided and used for validation of turbulence models. The results of the flow investigation lead to practical suggestions for the improvement of the air flow uniformity inside the drying space which is very important for the quality of the product. A novel numerical code, DrySAC, able to predict the unsteady-state processes taking place in a complex drying system, was developed. Unlike other attempts to predict drying processes, DrySAC takes into account not only the drying process itself, but also the behavior of the other system equipment and the interaction between them. Drying curves, evolution of the air state parameters in characteristic points of the system and product * *

Dionissios P. Margaris: Tel.: +30.2610.997193, Fax: +30.2610.997202, E-mail: [email protected] Adrian-Gabriel Ghiaus: Tel.: +40.21.2524280, Fax: +40.21.2526880, E-mail: [email protected]

46

Dionissios P. Margaris and Adrian-Gabriel Ghiaus properties are predicted during the drying of various fruits and vegetables and. As a practical validation of the code, the predicted values compared with the measured data taken in-situ showed very good agreement. When a dryer configuration is given, the numerical DrySAC code can be used for optimization of the process parameters when a dryer configuration is given. For the most of the studied cases, an air recirculation ratio of around 75 % has proved to be the optimum, giving a minimum drying time. The code can be used both for evaluation of existing dryers and for optimum design of the new units with valuable impacts in increasing the efficiency of the systems and in reduction of energy consumption. Aiming to overcome the lack of experimental data in the open literature, a laboratory drying unit was constructed and is under operation for testing and monitoring the dehydration of agricultural products. Using this facility, experimental drying curves are set up for the drying of horticulture products under controlled conditions of the drying air parameters, which are gathered by means of a data acquisition system. The laboratory experimental results are useful for the validation of numerical models which further are an essential tool for optimization and increasing the efficiency of the drying process. Drying of agricultural products remains an open research field mainly because of their delicate and hard to be established, properties.

1. INTRODUCTION In an extremely varied range of types and industries, drying processes are used to obtain dry products, the performance of which is often determined by morphological aspects created in the drying process. In process optimization and drier design the strong non-linearity of the equilibrium relationship and the internal resistance of water diffusion prohibit the use of effective overall mass-transfer coefficients, as it is common used in chemical engineering. The complexity of materials, in composition as well as in (multi-phase) structure, strongly limits the linking of diffusion coefficients and sorption isotherms to materials on the basis of first principles, in contrast to the data bank and group estimation methods in chemically welldefined mixtures. The exchange of mass, heat and momentum between air and product demands specific models for each drier, as it depends on geometry. Morphological properties are dependent on the material (deformation) properties which change in the transient process. Drying is generally considered an intensive energy process, and because of this it is often mentioned in national and international research programs.

1.1. Need for Drying The reasons for drying are as diverse as the materials which may be dried. However, most of the materials investigated in drying processes are agricultural and food products. A dried product must be suitable for either subsequent processing or sale, [Keey, 1982]. Although drying is considered a unit operation, it covers a rather diverse field and many configurations for drying equipment exist. Compared to other classical unit operations, this diversity in technology has been and remains an intrinsic obstacle in the development of scientific understanding of the drying process. From an economical point of view, there is only one good reason for drying, namely making profit. This requires that the drying process:

Fruits and Vegetables Dehydration in Tray Dryers • • •

47

adds value to the product (quality, specifications), is performed at minimum cost (design, control, optimization, equipment, technology, energy, other resources), and is achieved within the criteria, imposed by society (safety, environmental protection, loss prevention).

These three factors are essentially the main driving forces for research and development. Nevertheless, the personal motives of individual researchers should not be ignored. With well-defined physical properties, it is possible to predict the heat and mass transfer in food processing and to solve the heat and mass transfer equations. For geometrically simple systems, mass and heat transfer equations can be solved analytically, but for complex systems, a computer program is required to solve the equations numerically. Furthermore, the porous structure and the changes taking place during the drying process will give rise to further problems in the calculations. Sophisticated models for mass and heat transfer in drying materials are very important for developing a better understanding of the drying process at the microscopic scale. These rigorous models incorporate a maximum of physical relevance and are not easy to handle. Most model parameters depend on moisture content and temperature and are laborious to establish experimentally. In most cases drying kinetics is characterized by means of drying curves, which represent an overall drying behavior of the sample and from which it is rather difficult to derive reliable intrinsic material properties. Taking into account: • • • •

the large amount of dryer configurations and the many, varied materials to be dried, the modeling at different scales, from micro- (pores, molecules), meso- (particle) to macro- scale (incremental or total dryer), and and the balance between physically meaningful and practically manageable models, one can explain the great diversity in theoretical approaches of drying processes.

Modeling flow patterns and particle trajectories has shown increased interest in using Computation Fluid Dynamics (CFD) due to the availability of excellent commercial software in this field. It is without any doubt that CFD shows great potential in the practical design of dryers. The combination of CFD modeling and sophisticated drying kinetics seems to be a future objective. Stable, accurate and fast computer programs for these rigorous approaches are still hard to develop and accordingly, the need for simplified models will persist in the future. It is generally felt that technology management in industry is rather conservative. There is often a lack of qualified personnel with drying expertise, and it may be that drying is just one of their many duties, therefore not taking highest priority. Furthermore, one has to admit that in many cases the trial and error approach leads more quickly to a working solution than scientific routes. Especially for food products, enhancement of product sales can be better achieved via advertisements, commercials and attractive packaging rather than improving the product quality via more risky research efforts. On the other hand, the academic world produces a great diversity of drying models which do not always show a drive for putting theory into real industrial practice. A possible solution for bringing industry and academia closer together should take into consideration that there is a need for:

48

Dionissios P. Margaris and Adrian-Gabriel Ghiaus • •

good, reliable software development within industry, and practical laboratory methods for establishment of the relevant material properties.

The field of dehydration is full of puzzles, inconsistencies, disputed observations and conflicting interpretation. An enormous amount of well designed experimental work needs to be done, leading into many lines of theoretical inquiry and technological advance. Advances in understanding the thermodynamics of moist materials and better solutions to the equations of change would obviate the need for empiricism in selecting optimum drying conditions, [Luikov, 1970]. More detailed knowledge of the mechanism of moisture transport would undoubtedly be needed if drying is to become more of a science than an art.

1.2. Historical Features Humans have benefited from dried foods since the Cro-Magnon era. Sun and wind fostered the first dried foods. Early man copied the drying process he observed in nature. Dried grasses, seeds, fruits and nuts were gathered and stored. Samples of dried foods dating to 4000 years before present have been found in both ancient settlements in Jericho and Egyptian tombs. The first patent for dehydration was granted in England in 1780 to J. Grafer, who scalded vegetables in boiling, salt water and then dried them in a heated room. Later, the Royal Navy expedition that searched for the British arctic explorer Sir John Franklin in 1852 carried carrots and potatoes that had been dried in hot air for 20 to 30 hours, [Borgstrom, 1971]. During the Crimean War (1853-1856), an attempt was made to combat scurvy by using potatoes dried by Edwards’ patented process, which was to boil them, press them through small holes to form fine spaghetti-like threads, and dry them on steam-heated plates. Many air dried vegetables were produced during World War I, [Borgstrom, 1971].

1.3. Drying of Foods Throughout history, man has learned that removal of water increases the period of usefulness of perishable products. The lower storage and transportation costs associated with the reduction of weight and volume due to water removal have provided additional economic incentives for widespread use of dehydration processes, [Rizvi, 1986], features especially important for developing countries and in military feeding and space food formulation, [Jayaraman, 1992]. From a theoretical point of view, dehydration of porous materials, such as foods, is a rather complex process. It involves interactions not only between heat and mass transfer processes occurring within the food itself, but also between the food and the drying medium circulating around the solid matter. Successful outcome of these competing phenomena requires simultaneous solution of the separate differential equations for heat, mass and momentum transport within the food system being dried and in the external drying medium. Coupling of the two processes at the surface of the solid for general theoretical solutions to the overall drying problem, along with the dependence of the transport coefficients on the values of driving forces, presents serious computational problems. Further complications arise

Fruits and Vegetables Dehydration in Tray Dryers

49

as a result of the lack of thermo-physical property data for real foods. Clearly, much work is needed toward obtaining numerical or analytical solutions of the basic differential equations for the dehydration of real foods under practical drying conditions. The delicate characteristics of foods require the skilful operation and design of dehydration systems, and this requires an understanding of the principles of dehydration. The conditions under which a product should be dried (drying time, drying temperature, amount of heat to be supplied, amount of water vapor to be removed, etc.) vary greatly because the products to be dried are very different from each other in nature and in properties such as shape and dimensions, moisture content and temperature sensitivity, [Leniger, 1975]. Besides providing crude fiber and bulk, fruits and vegetables are indispensable sources of essential dietary nutrients, vitamins and minerals. However, due to their high moisture content (above 80 %) they are highly perishable, [Jayaraman, 1992]. Total world production of dried grapes averages about 700,000 tons per year. The major producers are: U.S.A. (35 %), Turkey (31 %) and Greece (15 %). About 40 % of total production reaches the international market and is mainly directed to the industrialized countries of Europe, where dried grapes are used as ingredients by the confectionery industry, [Riva, 1986]. The preservation of fruits and vegetables by dehydration offers a unique challenge. To achieve the desired results for dehydration of fruits and vegetables, the process must provide the optimum heat and mass transfer within the product. Only through analysis and understanding of these processes can maximum efficiency and optimum quality be achieved, [Singh, 1993]. Due to changing lifestyles, especially in the developed world, there is currently great demand for a wide variety of dried products with emphasis on high quality and freshness as well as convenience. This calls for sustained basic research on drying conditions and equipment and their influence on food qualities, [Jayaraman, 1992].

1.4. Food Quality and Safety Scientific research and development in agricultural engineering fosters optimum utilization of available human, physical and financial resources. Improved technologies for drying fruits and vegetables have been introduced to reduce losses arising from seasonal gluts. In most developing countries, less than 20 percent of agricultural output undergoes industrial processing compared with 80 or more percent in developed countries. The safety of food is essential for the health and well-being of man, and its quality for his satisfaction. Bacteria, yeasts, moulds, insects and rodents are in constant competition with man for his food supply. Foods are also subject to destruction by almost every variable in the natural environment. Heat and cold, light, oxygen, moisture, dryness and natural enzymes within the food, all tend to cause deterioration. Scientific advances and better knowledge through research have supported older and newer technologies alike in their ability to ensure the safety and quality of the processed food supply. Without knowledge of the drying mechanism there is no sound way to predict methods for increasing the drying rate or for improving product quality or retention of nutritional value.

50

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

1.5. Objectives of the Research While computer models have been increasingly successful in simulating an ever widening range of engineering problems, it is nevertheless essential that advances in these models are validated and verified by experimentation. Experimental measurements are themselves conditioned to the requirements of the computational models. Accordingly, it is important that scientific work on experimental facilities must be correlated with research in developing computer codes as well as with monitoring and measurements on full-scale prototypes. The orderly and progressive concurrent development of all these fields is essential for the progress of engineering sciences. This chapter presents research carried out by numerical and experimental investigation on the operation of convective drying systems used for dehydration of agricultural products. The objectives of the research have been: to investigate the air flow field inside large capacity tray drying rooms, to develop a numerical code suitable for simulating the operation of a drying system, to predict the drying time and the drying process parameters, to evaluate experimentally the drying parameters for specific products, to improve the uniformity of air distribution inside the drying room, to minimize the drying time, to increase the efficiency of the drying system by optimizing the process parameters, and finally to reduce the energy consumption.

2. FUNDAMENTALS OF DRYING PROCESSES Dehydration involves the simultaneous transfer of heat, mass and momentum in which heat penetrates into the product and moisture is removed by evaporation into an unsaturated gas phase. Owing to the complexity of the process, no generalized theory currently exists to explain the mechanism of internal moisture movement. Although it is now accepted that in most practical situations of air drying of foods the principal rate-determining step is internal mass transfer, there is no agreement on the mechanism of internal moisture movement, [Chirife, 1983]. In the case of capillary-porous materials such as fruits and vegetables, interstitial spaces, capillaries and gas-filled cavities exist within the food matrix and water transport takes place via several possible mechanisms acting in various combinations. The possible mechanisms proposed by many workers include: • • • • • • •

liquid diffusion caused by concentration gradients, liquid transport due to capillary forces, vapor diffusion due to shrinkage and partial vapor-pressure gradients (Stefan’s law), liquid or vapor transport due to the difference in total pressure caused by external pressure and temperature (Poiseuille’s law), evaporation and condensation effects caused by differences in temperature, surface diffusion in liquid layers at the solid interface due to surface concentration gradient, liquid transport due to gravity.

Fruits and Vegetables Dehydration in Tray Dryers

51

Additionally, moisture may also be transported inside a material if a suitable temperature gradient exists (thermo-gradient effect), because of thermodynamic coupling of heat and mass transport processes. Most foods are classified as capillary porous rigid or capillary porous colloids, [Bruin, 1980]. Therefore, it is often proposed that a combination of capillary flow and vapor diffusion mechanisms should be used to describe internal mass transfer. Water activity, rather than moisture content, influences biological reactions. In the regions of water adsorption on polar sites or when a mono-molecular layer exists, there is little enzyme activity. Enzyme activity begins only above the region of mono-molecular adsorption. When the moisture content of a substrate is reduced below 10 %, microorganisms are no longer active. It is necessary however to reduce the moisture content to below 5 % in order to preserve nutrition and flavor, [Charm, 1978].

2.1. Dehydration Principles In air-drying processes, two drying periods are usually observed: an initial constant-rate period in which drying occurs as if pure water was being evaporated and a falling-rate period where moisture movement is controlled by internal resistance. Figure 1a illustrates this by showing the moisture content as a function of time, where segment A-B represents the initial unsteady-state warming-up period, B-C the constant rate period and C-D-E the falling-rate period.

Figure 1. Drying curves showing: a) moisture content vs. drying time, b) drying rate vs. drying time, and c) drying rate vs. moisture content.

Figure 1b shows the drying rate as a function of time, whereas in figure 1c the drying rate is plotted against moisture content. The drying rate during the constant-rate period may be computed using either the heat transfer or mass transfer equation. As the surface of the material maintains a saturated condition and its temperature is the wet-bulb temperature of the drying air, the drying rate is given as:

52

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

&v= −m

h t ⋅ A ⋅ ( Tdb − Twb ) = k p ⋅ A ⋅ (pw − pv ) h lg

heat transfer eq.

(2.1)

mass transfer eq.

& v -drying rate, h t -global heat transfer coefficient, A-area of the where the symbols are: m drying surface, Tdb -dry-bulb temperature of moist air, Twb -wet-bulb temperature of moist air, h lg -latent heat of vaporization, k p -mass transfer coefficient pressure-basis, p w -partial pressure of water vapor at saturation, and p v -partial pressure of water vapor of the drying air. The drying rate is actually the mass flux of water vapor evaporated from the product. As moisture content of the material decreases during the drying process, the drying rate has negative value. The negative sign of the drying rate in Eq.(2.1) is a consequence of the fluxes’ direction convention (positive from the drying air to the material).

(

)

The temperature difference Tdb − Twb is often defined as the wet-bulb depression. The driving force for vaporizing water from the surface is the difference between the vapor pressure of water at the temperature of the surface and the partial pressure of water in the air. Eq.(2.1) can be used to calculate the drying rate in the case of flat surfaces and can only be applied to plates, sheets, etc. In other cases, it may be better to relate the drying rate to the weight of the material rather than the drying surface area. In practice, the heat transfer equation gives a more reliable estimate of the drying rate than the mass transfer equation because its parameters can be measured directly. The first falling-rate period is the period of unsaturated surface dehydration. During this period, increasingly larger proportions of dry areas appear on the surface as drying progresses. In many food materials, the migration of moisture occurs through the mechanism of diffusion. In practice the diffusion coefficient is dependent, to some extent, on the moisture content. To estimate the average drying time during the first falling-rate period, Fick’s second law of diffusion is widely used. Assuming an idealized system with a constant diffusion coefficient, the partial differential equation for one-dimensional diffusion is given as:

∂X = D eff ∂t

⎛ ∂ 2 X C ∂X ⎞ ⋅ ⎜⎜ 2 + ⋅ ⎟⎟ z ∂z ⎠ ⎝ ∂z

(2.2)

where X is the moisture content dry-basis of the material, t is the time, D eff is the effective diffusion coefficient, z the distance, and C a constant of 0 for planar, 1 for cylindrical, and 2 for spherical geometry. Assuming a uniform initial moisture distribution and in the absence of any external resistance, the analytical solutions of Fick’s law for a slab of material are given in the form of infinite series, [Chirife, 1983]:

Fruits and Vegetables Dehydration in Tray Dryers

X − Xe 8 ∞ 1 = 2⋅∑ ⋅ exp X i − X e π n = 0 ( 2n + 1) 2

⎡ ⎤ ⎢− ( 2n + 1) 2 ⋅ π 2 ⋅ D eff ⋅ t ⎥ ⎢ z2 ⎥ ⎣ ⎦

53

(2,3)

where X is the average, X e the equilibrium and X i the initial moisture content dry-basis. For long drying times and for an un-accomplished moisture ratio less than 0.6,

[ (X − X e ) / (X i − X e )] < 0.6

(2.4)

generally only the first term of the infinite series (Equation 2.3) is used to estimate the drying rate:

X − Xe 8 = 2 ⋅ exp ( − K ⋅ t ) Xi − Xe π

(2.5)

where

K = π2 ⋅

D eff z2

(2.6)

−1 is termed the drying constant (dehydration constant) and has the unit of measurement sec . The drying analysis presented above is based on the assumption that the heat transfer effects can be neglected and drying can be treated as a purely diffusion controlled mass transport phenomenon with a constant, effective diffusion, coefficient. This approach is based on several experimental studies, which indicate the existence of very small internal temperature gradients within foods during drying, [Chirife, 1983]. When the foodstuff is a porous solid, the mass flux involved is one of vapor through the intercellular spaces and hence, there is a need to include porosity in the mass transport equation, [Lozano, 1980]:

ε⋅

&v ∂ρ v 1 ∂m ∂X + ⋅ = −(1 − ε ) ⋅ ρ ds ⋅ ∂t A ∂z ∂t

(2.7)

where ε is the porosity, ρ v the density of water vapor, t the time, A the drying area, z the thickness of the material, and ρ ds the density of the dry product. During the second falling-rate period, drying occurs at a moisture content where the equilibrium relative humidity is below saturation and heat transfer should be considered along with mass transfer because desorption of moisture requires consumption of substantial amounts of heat. The diffusion coefficient is moisture content dependent and Fick’s second law of diffusion for an infinite slab can be written as:

54

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

∂X ∂ ⎛ ∂X ⎞ = ⎜ D eff ⋅ ⎟ ∂t ∂z ⎝ ∂z ⎠

(2.8)

Diffusion-like theories of drying cannot describe the complete spectrum of moisture transport mechanisms that occur during the drying of granular porous medium and therefore gas-phase momentum should be included in any comprehensive theory. This incorporates the liquid- and vapor-phase continuity equations, combines the liquid-, solid- and vapor-phase thermal energy equations into a single temperature equation and makes use of Darcy’s law for the liquid phase to account for moisture transport due to capillary action, [Whitakar, 1984]. If the rate of heat transfer to the material is sufficiently high, vaporization takes place within the material and the rate of drying is determined by the heat transfer rate into the material. Among the most important parameters in the evaluation, design and specification of drying systems are the energy requirements and drying times. With few exceptions, existing drying models have taken the heat of vaporization to be constant and invariant with moisture content. This is in fact true in most drying situations. For most foods the enthalpy of evaporation does not differ significantly from the latent heat of vaporization of pure water until moisture contents of 0.1 kg w/kg ds or less are reached. Most commercial operations do not dry materials to this degree. The general case of food material dehydration involves energy inputs to meet the following energy requirements: • • •

Removal of free water through evaporation, Removal of water associated with the food matrix, Superheating of water vapor evaporated as it passes through the food.

2.2. Batch Drying Systems Batch drying systems are also called tray or compartment dryers and consist of an enclosed, insulated housing in which solids are placed upon tiers of trays in the case of particulate solids, or stacked in piles or upon shelves in the case of large objects. The drying material bed is in a static condition in which each particle rests upon another at essentially the settled bulk density of the solids phase (figure 2). During drying, there is a permanent & v ) between the air stream and the bed surface. exchange of heat ( q& ) and water vapor ( m Specifically, there is no relative motion among solid particles. Satisfactory operation of tray-type dryers depends on maintaining a constant temperature and uniform air velocity over all the material being dried. Circulation of air at velocities of 1 to 10 m/s is desirable to improve the surface heat transfer coefficient and to eliminate stagnant air pockets. Proper air flow in tray dryers depends on sufficient fan capacity, the design of ductwork to modify sudden changes in direction, and on properly placed baffles. Non-uniform air flow is one of the most serious problems in the operation of tray dryers.

Fruits and Vegetables Dehydration in Tray Dryers

55

Figure 2. Parallel air flow over a static bed of solids.

Tray dryers may be of the tray-truck or the stationary-tray type. In the former, the trays are loaded on trucks which are pushed into the dryer; in the latter, trays are loaded directly into stationary racks within the dryer. Trays may be square or rectangular, with 0.5 to 1 m2 per tray, and may be fabricated from any material compatible with corrosion and temperature conditions. When the trays are stacked in the truck, there should be a clearance of not less than 4 cm between the material in one tray and the bottom of the tray immediately above. When material characteristics and handling permit, the trays should have screen bottoms for additional drying area. Batch dryers are the most popular family of convective industrial dryers used in the food industry undertaking the processing of all materials that need to be dried for an extensive time period (usually days). Performance of these dryers depends on several factors related to dryer design, product type and on conditions of the drying air. Over a time interval from tn to tn+1, the fall in moisture content of the drying material is related to the increase in air humidity over the tray from the inlet value x in to the outlet value x out :

& a⋅ − m ds ⋅ ( X n +1 − X n ) = m

t n +1

∫ ( x out − x in ) ⋅ dt

(2.9)

tn

where m ds is the mass of dry solid, X n +1 and X n are the moisture content dry-basis at

& a the mass flow rate of air. time n+1 and n, respectively, and m

(

)

For an infinitesimal time interval t n +1 − t n → dt , the drying rate is:

& v = − m ds ⋅ m

dX & a ⋅ ( x out − x in ) =m dt

(2.10)

56

Dionissios P. Margaris and Adrian-Gabriel Ghiaus So, the outlet air humidity can be calculated as:

& m x out = x in + v &a m

(2.11)

Over a given batch, x out rises rapidly from x in as drying begins to reach a maximum value, then falls as the drying rate declines towards the end of the process. Under quasi-steady conditions, when the accumulation of heat by the moist material is & a becomes: zero, the energy balance for an intake of mass flow rate of drying air, m

& F +Q & H − h out ⋅ m &L =0 & a +Q & a −Q h in ⋅ m

(2.12)

& is the fan where h in and h out are the specific enthalpy at inlet and outlet, respectively, Q F & the heat input from the heater and Q & the heat loss from the drying room to the work, Q H L outside. The work performed by the fan is normally small. The resistance of airflow in a simple drying chamber is about 20 Pa for an air velocity of 2 m/s over the tray, [Keey, 1982]. If the fan efficiency is 80 %, the work done per unit cross-section of the dryer is 20 x 2 / 0.8 = 50 W/m2. Heat losses are more significant even at fairly low drying air temperatures. Installing suitable lagging would eliminate about two-thirds of this loss. As air passes over the wet material, the air temperature is reduced and its humidity is increased as described by the adiabatic cooling curve of the psychrometric chart. The temperature of the air leaving the tray is given by:

Tout = Tsurf + ( Tin − Tsurf ) ⋅ e − N t

(2.13)

where Tsurf is the temperature of the drying surface, and N t the number of transfer units calculated from the following equation:

Nt =

h t ⋅ L tr ρ a ⋅ w ⋅ c p ⋅ z fs

(2.14)

where h t is the global heat transfer coefficient, L tr the tray length, ρ a the air density, w the air velocity, c p the specific heat at constant pressure and z fs the free space between trays. During the constant-rate period and in the absence of radiant heat transfer, Tsurf assumes the adiabatic saturation temperature of the air. In order to maintain an economic drying operation, the warm, moist air leaving a dryer is frequently mixed with dryer, fresh air in order to reduce the heating cost of the incoming air. Air recirculation is generally in the order of 80 to 95 percent except during the initial drying stage of rapid evaporation [Perry, 1987].

Fruits and Vegetables Dehydration in Tray Dryers

57

In a batch through-circulation dryer, heated air passes through a stationary permeable bed of wet material placed on removable screen-bottom trays suitably supported in the dryer. This dryer is similar to a standard tray dryer except that hot air passes through the wet solid instead of across it. The air pressure loss in a through-circulation dryer is related to the bed characteristics by:

Δp = z bed ⋅ (1 − ε ) ⋅ ( ρ s − ρ a ) ⋅ g

(2.15)

where Δp is the pressure drop, z bed the depth of the material bed, ε the porosity, ρ s and ρ a the densities of material and air, respectively, and g the acceleration of gravity. Batch through-circulation dryers are restricted in application to granular materials which permit free flow-through circulation of air. In these cases drying times are usually much shorter than those of parallel-flow tray dryers.

2.3. Mathematical Modeling of Drying Processes The simulation of various product drying systems involves solving a set of heat and mass transfer equations which describe: a) heat and moisture exchange between product and air, b) adsorption and desorption rates of heat and moisture transfer, c) equilibrium relations between product and air and d) psychrometric properties of moist air, [Jayas, 1991]. The most rigorous methods of describing the drying process are derived from the concepts of irreversible thermodynamics in which the various fluxes are taken to be directly proportional to the appropriate “potential” [Ghiaus, 1997]. The mass balance inside the product can be written as:

∂( ρ s ⋅ X) ∂t

∂X ⎞ ⎛ = div ⎜ ρ s ⋅ D eff ⋅ ⎟ ⎝ ∂z ⎠

(2.16)

and the heat-energy balance can be set down as:

ρs ⋅ c p ⋅

∂T ⎛ ∂T⎞ = div ⎜ λ ⋅ ⎟ ⎝ ∂z ⎠ ∂t

(2.17)

The following initial and boundary conditions apply for the system:

T( z,0) = Ti −λ⋅

∂T( n, t ) & v =0 + h c ⋅ [Tdb − T( n, t )] − h lg ⋅ m ∂z

(2.18)

(2.19)

58

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

∂T(1, t ) =0 ∂z

(2.20)

where T(n, t) is the temperature of the slab layer ‘n’ in contact with the drying air, and T(1, t) the temperature of bottom slab layer. No general theory or equation would be valid for the prediction of drying kinetics of fruits and vegetables, and actual experimentation using specific material is essential for describing its drying behavior. Although correlation for the calculation of the heat and mass transfer coefficients has been proposed in the literature, few data are available to allow the constants in this correlation to be fixed with certainty. For a particulate bed, and air-flow parallel to the product surface, the following equation to calculate heat transfer coefficient can be used [Jayas, 1991]:

h c = 0.00327 ⋅ ρ a ⋅ w ⋅ Re −0.65 ⋅ Pr −0.66

(2.21)

In the case of through-circulated beds, where the product is particulate (e.g. grapes), for a laminar flow with Re < 350 (the characteristic length is taken as the particle diameter), the heat transfer coefficient can be calculated as:

h c = 182 . ⋅ ρ a ⋅ w ⋅ c p ⋅ Re −0.51

(2.22)

For Re > 350, the flow through the bed becomes turbulent and the following alternative correlation applies:

h c = 0.989 ⋅ ρa ⋅ w ⋅ c p ⋅ Re −0.41

(2.23)

The mass transfer coefficient (kp) is difficult to measure and is usually estimated from the Chilton and Colburn analogy, [Bird, 1960]. A widely used relationship in drying calculation, that correlates heat and mass transfer coefficients with the latent heat of water vaporization hlg, is the following, [Sereno, 1990]:

hc = 64.7 h lg ⋅ k p

(2.24)

2.4. Modeling of Agricultural Product Properties Variability in composition and physical characteristics is typical for all food products. Most thermal property models are empirical rather than theoretical, i.e. they are based on statistical curve fitting rather than theoretical derivations involving heat transfer analysis. The specific heat indicates how much heat is required to change the temperature of a material. The ratio of the heat supplied to the corresponding temperature rise is defined as the heat capacity of a body. Specific heat is given by the mass heat capacity equation:

Fruits and Vegetables Dehydration in Tray Dryers

c=

1 dQ ⋅ m dT

59

(2.25)

where m is the mass, Q the heat, and T the temperature of the body. Specific heat of a product is influenced by the product components, moisture content, temperature and pressure. For products whose composition is known, the following equation may be used:

c = 4187 . ⋅ m w + 1549 . ⋅ m p + 1675 . ⋅ m f + 1424 . ⋅ m c + 0.837 ⋅ m a

(2.26)

where m is the mass fraction and the subscripts are: w-water; p-protein; f-fat; c-carbohydrate; a-ash. Some composition values of selected foods are given in table 1. Table 1. Composition values of selected foods Food Apples, fresh Garlic Peaches Peas, raw Pineapple, raw Potatoes, raw Rice, white Spinach Tomatoes

Water ,% 84.4 61.3 89.1 78.0 85.3 79.8 12.0 90.7 93.5

Protein, % 0.2 6.2 0.6 6.3 0.4 2.1 6.7 3.2 1.1

Fat, % 0.6 0.2 0.1 0.4 0.2 0.1 0.4 0.3 0.2

Carbohydrate, % 14.5 30.8 9.7 14.4 13.7 17.1 80.4 4.3 4.7

Ash, % 0.3 1.5 0.5 0.9 0.4 0.9 0.5 1.5 0.5

Specific heat of agricultural products has been most commonly modeled with equations of the form [Sweat, 1986]:

c = C1 + C 2 ⋅ X

(2.27)

with the constants C 1 and C 2 varying with product. For moist foodstuffs, the specific heat is equal to the sum of the specific heat of water and that of the solid material, [Mohsenin, 1980]. This can be mathematically expressed as follows, [Abalone, 1994]:

c = W ⋅ c p , w + (1 − W ) ⋅ c ds

(2.28)

where W is the moisture content wet-basis, c p, w the specific heat at constant pressure for water, and c ds the specific heat of the dry product. Table 2 presents the values of specific heat for several fruits and vegetables correlated with their water content and temperature. The enthalpy of a dry solid is defined by the product of the heat capacity or specific heat c ds and the temperature excess:

h ds = c ds ⋅ ΔT

(2.29)

60

Dionissios P. Margaris and Adrian-Gabriel Ghiaus Table 2. Specific heat for some fruits and vegetables Foodstuff Apples Apples Apricots Berries Carrots Figs Grapes Peaches Pears Plums Potatoes

Water content W, % 84.1 75-85 85.4 75-85 88.2 78 81.8 86.9 83.5 85.7 77.8

Temperature T, °C 0-100 0-100 0-100 -

Specific heat c, J/kg K 3600 3730 3680 3730-4100 3770 3430 3600 3770 3600 3680 3430

Reference ASHRAE, 1986 Ordinanz, 1946 ASHRAE, 1986 Ordinanz, 1946 ASHRAE, 1986 ASHRAE, 1986 ASHRAE, 1986 ASHRAE, 1986 ASHRAE, 1986 ASHRAE, 1986 ASHRAE, 1986

In foods, thermal conductivity depends mostly on the individual components. For products whose composition is known, the following equation may be used:

λ = 0.58 ⋅ m w + 0155 . ⋅ m p + 016 . ⋅ m f + 0.25 ⋅ m c + 0135 . ⋅ ma

(2.30)

Thermal conductivity of most high moisture foods has values closer to the thermal conductivity of water. For products that are predominantly water, a model of the form:

λ = C1 + C 2 ⋅ W

(2.31)

is commonly used. For fruits and vegetables with a water content greater than 60 %, the coefficients C 1 and C 2 are 0.148 and 0.493, respectively [Sweat, 1974]. To cover widely varying temperatures, a model should account for variation in temperature. For greater accuracy, this would probably include a T term and a T2 term, because the thermal conductivity of water varies as temperature is squared. For potatoes, [Abalone, 1994], thermal conductivity can be calculated as:

λ = C1 + C 2 ⋅ T + C3 ⋅ T2

(2.32)

with C1 = 1.05 W/m K, C 2 = -1.96 102 W/m K2, and C 3 = 1.90 10-4 W/m K3. Temperature and water content are the major factors affecting thermal diffusivity. A multiple regression analysis on 246 published values, gave the following equation for the calculation of thermal diffusivity, [Singh, 1982]:

α = ( 0.057363 ⋅ W + 0.000288 ⋅ T) ⋅ 10 −6

(2.33)

Effective mass diffusion coefficient is related to various physical properties of the food such as thermal conductivity, bulk density, enthalpy, etc., as well as to environmental conditions [King, 1968]. The equations derived may be written as:

Fruits and Vegetables Dehydration in Tray Dryers

D eff =

b ⎛ ∂a w ⎞ a ⋅⎜ ⎟ ⋅ pw ⋅ ρs ⎝ ∂X ⎠ T 1+ a

61

(2.34)

where b is the vapor-space permeability, ρs the solid density, a w the water activity, X the moisture content dry-basis, p w the partial pressure of water vapor at saturation, and a is calculated from the following equation:

(

)(

2 a = λ ⋅ R w ⋅ T 2 / b ⋅ a w ⋅ p w ⋅ h st

)

(2.35)

where λ is the thermal conductivity, R w the gas constant for water vapor, T the temperature, and h st the heat of sorption. The term a / (1+ a ) in Eq.(2.34) determines the degree of mass or heat transfer control. If a >> 1 , the process is totally mass transfer controlled; if a << 1 , the process is totally heat transfer controlled. This equation is based on an analysis of the ways in which heat and mass transfer processes interact to govern desorption (or adsorption) of water vapor in dried food. It also represents a link between moisture and vapor pressure gradients as the driving forces of water diffusion during dehydration. The effect of temperature on the effective diffusion coefficient is usually expressed by an Arrhenius-type equation:

D eff = C ⋅ exp( − E a / R ⋅ Tdb )

(2.36)

where E a is the activation energy, R the universal gas constant, Tbd the dry-bulb temperature and C a constant. The values of activation energy range between 14.2 to 39.8 kJ/mol, for the first falling-rate period, and between 33.5 to 58.6 kJ/mol, for the second falling-rate period of drying. For grapes, C = 3·10-6 m2/s and E a = 22 kJ/mol, [Belessiotis, 1995]. For potatoes, Abalone (1994) gives C = 2.78 10-8 m2/s and E a = 50 kJ/mol. The ratio between the water vapor pressure at the surface of a foodstuff and the vapor pressure of pure water at the same temperature is called the water activity, a w .

aw =

p surf pw

(2.37)

Water activity also represents the relative humidity of air with which the foodstuff is in equilibrium and the expression ‘equilibrium relative humidity’ (e.r.h.) is sometimes used. For pure water a w = 1; when water is more or less bound to the product, a w < 1. The water activity depends on the water content and the temperature of the product. Within a product, water can be present in “free” and “bound” form. Free water is bound by such minute forces that its vapor pressure is equal to the vapor pressure of pure water. Thus a product containing free water has a water activity of 1.00 and behaves, during drying, like a so-called wet body. Free water can evaporate in unsaturated air of arbitrarily high relative humidity, and may be present on the surface or in cavities and wide capillaries. In the drying of food, free water can

62

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

only be expected to be present when the products have been washed or blanched prior to drying. The amount of free water is small and can easily be reduced even further by mechanical separation methods used before drying. Water activity is one of the most important factors in the drying of products because it determines the drying rate. When graphically expressed, the relationship between total moisture content and the corresponding water activity of a substance over a range of values at a constant temperature yields a moisture sorption isotherm (MSI). Moisture sorption isotherms give an impression of how strongly water is bound to a product when the moisture content is at a particular level. The water activity is a measure of this. The amount of water that can theoretically be evaporated when drying with air of a certain relative humidity is shown schematically in figure 3.

Figure 3. Amount of water evaporated at certain relative humidity.

Of the large number of MSI models, the three-parameter Guggenheim-Anderson-deBoer (GAB) equation has been suggested to be the most versatile sorption model available in the literature and has consequently been adopted by West European food researchers [Rizvi, 1986]. Fundamentally, it represents a refined extension of the Langmuir and BrunauerEmmett-Tetter (BET) theories, with three parameters having physical meanings. For sorption of water vapors, it is mathematically expressed as:

X C1 ⋅ C 2 ⋅ a w = X o (1 − C 2 ⋅ a w ) ⋅ (1 − C 2 ⋅ a w + C1 ⋅ C 2 ⋅ a w )

(2.38)

where C1 is the Guggenheim constant and C 2 is a factor correcting properties of multi-layer with respect to the bulk liquid, both being related to the energies of interaction. The GAB model is basically similar to the BET equation in its assumptions of localized physical adsorption in multi-layers with no lateral interactions. The major advantages of the GAB model are that (1) it has a viable theoretical background; (2) it describes sorption behavior of nearly all foods from zero to 0.9 water activity; (3) it has a simple mathematical form with only three parameters, which makes it very amenable to engineering calculations; (4) its

Fruits and Vegetables Dehydration in Tray Dryers

63

parameters have physical meaning in terms of the sorption processes; (5) it is able to describe some temperature effects on isotherms by means of Arrhenius-type equations. An empirical equation describing moisture sorption isotherms for Corinthian grapes was developed by Belessiotis (1995):

X = ( C1 ⋅ C 2 ⋅ C 3 ⋅ a w ) / [(1 − C 3 ⋅ a w ) ⋅ (1 − C 3 ⋅ a w + C 2 ⋅ C 3 ⋅ a w )]

(2.39)

where the moisture content dry-basis, X is in percent and coefficients C1, C2 and C3 are functions of temperature as follows:

C1 = −59.5 + 818889 . ⋅ T − 0.237111⋅ T2 + 19901 . ⋅10 −3 ⋅ T3

(2.40)

C 2 = 5.22 − 0.447111⋅ T + 0.0127111⋅ T2 − 104691 . ⋅10 −4 ⋅ T3

(2.41)

C 3 = 1194 . − 0.0348778 ⋅ T + 0.0112444 ⋅ T 2 − 9.72839 ⋅ 10 −6 ⋅ T 3

(2.42)

Critical moisture content at the end of the constant-rate period has been found to vary from 3.5 to 5 kg water/kg dry matter in vegetables, and from 5.5 to 7.7 kg water/kg dry matter in fruits, [Saravacos, 1962]. These values are close to the initial moisture content, so the importance of the constant-rate period in food dehydration is diminished. The moisture content remaining in a dry material, when drying rate drops to zero at specific conditions of the drying medium, is called the equilibrium moisture content. The equilibrium moisture values predicted by the static and dynamic moisture sorption (obtained from equilibrium isotherms and drying experiments, respectively), do not always agree over the entire relative humidity range of the drying air. When drying takes place with air of relative humidity less than that corresponding to the equilibrium value with the mono-layer moisture content, the equilibrium conditions no longer hold and the static and dynamic moisture sorption values show differences. There are three types of densities when dealing with foods, namely solid density ( ρ s ),

particle density ( ρ p ) and bulk density ( ρ b ). The values of these different densities depend

on how the pore spaces present in a food material are considered. Solid density is the density measured when the material has been thoroughly broken into pieces small enough to guarantee that no pores remain, [Lozano, 1980]. Particle density accounts for the presence of internal pores in the food particle which has not been structurally modified. This density is defined as a ratio of the actual mass of a particle to its actual volume. Bulk density is defined as the mass of particles occupied by a unit volume of bed and accounts for the void space between the particles. Bulk and particle densities of Corinthian grapes were measured by Belessiotis (1995) for different moisture content values and were expressed in the form of 3degree polynomial functions [Ghiaus, 1997] having the moisture content dry-basis as argument:

ρ b = 775.988 - 228.086 X + 133.573 X2 - 22.1892 X3

(2.43)

64

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

ρ p = 1480.37 - 382.184 X + 131.754 X2 - 15.4767 X3

(2.44)

For several fruits and vegetables, the particle density can be correlated with an equation of the following type [Lozano, 1983]:

⎛ X X⎞ ρ p = C1 + C 2 ⋅ + C 3 ⋅ exp⎜ − C 4 ⋅ ⎟ Xi Xi ⎠ ⎝

(2.45)

where the constants C1 , C 2 , C 3 and C 4 are given for different products. On the basis of the density distinctions, the following porosities result: particle porosity, which is the ratio of the volume of pores to the total volume of the particle, bulk porosity which is the ratio of inter-particle volume to the total volume of the bed and total porosity which is the ratio of air space volume to the total volume of the bed. A generalized correlation for the modeling of particle porosity is given by [Lozano, 1983]:

ε p = 1 − sp ⋅

ρ p,i ρp

⋅

1+ X 1+ Xi

(2.46)

where s p is the particle shrinkage coefficient, ρ p, i the initial particle density, and X i the initial moisture content dry-basis. Shrinkage is the process in which a material undergoing dehydration contracts or reduces in size due to water loss. It is convenient to define two shrinkage coefficients: the bulk shrinkage coefficient ( s b ) and the particle shrinkage coefficient ( s p ). Their mathematical expressions are as follows:

s b = Vb Vb,i

(2.47)

s p = Vp Vp,i

(2.48)

where Vb and Vp are the bulk and particle volumes at one moment and Vb , i and Vp , i are the initial bulk and particle volumes. A generalized correlation for the particle shrinkage coefficient, applied to fruits and vegetables, is [Lozano, 1983]:

s p = 0.37 + 0.607 ⋅

⎛ X X⎞ ⎛ 0.018 ⎞ + 0.022 ⋅ exp⎜ ⎟ ⎟ − s p,f ⋅ ⎜ 1 − ⎝ X + 0.025⎠ Xi Xi ⎠ ⎝

(2.49)

where s p, f is the final value of the particle shrinkage coefficient at X = 0, and is given by:

Fruits and Vegetables Dehydration in Tray Dryers

s p,f =

0.966 X i + 0.796

65

(2.50)

During the early stages of drying, the shrinkage in volume of vegetables equals the volume of water lost by evaporation. In later stages, the volume shrinkage is less and no substantial further decrease in volume occurs as the moisture content drops below 15-20 %.

3. EXPERIMENTAL ANALYSIS 3.1. Laboratory Drying Model The lack of thermo-physical properties for agricultural products is the main obstacle in extensive analysis of the dehydration process by numerical simulation. A laboratory drying unit, equipped with an appropriate set of measuring instruments connected to a data acquisition system, was constructed in order to evaluate the main drying characteristics of various fruits and vegetables. The analysis and interpretation of the experimental results are used to validate specific drying models which, implemented as sub-routines, will enlarge the applicability of the numerical codes. The experimental laboratory drying facility is presented schematically in figure 4. The main components of the facility are: drying room (1), centrifugal fan (2), electrical heater (3), drying tray (4), data acquisition system (5), personal computer (6), dampers (7) and orifice-plate devices (8). An overall view of the facility is presented in the photograph taken at the Fluid Mechanics Laboratory - University of Patras, Greece, where the drying unit is under operation, figure 5.

Figure 4. Schematic representation of the laboratory drying facility.

The drying room has a rectangular shape with the dimensions 117 x 25 x 17 cm, and is constructed from 5 mm thick Plexiglas plates. The room is thermally insulated with polystyrene of 10 cm thickness at the base, and 5 cm thickness for the sides and top. The two ends of the room are connected with two conversion pieces (from rectangular to circular in cross-section), each one including a pair of thermocouples for measuring the dry- and wetbulb air temperature. Inside the drying room is placed the tray having the dimensions 75 x 24 x 2 cm, constructed from an aluminum frame and having an iron wire mesh at the bottom.

66

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

The drying capacity inside the drying room is up to 8 kg of fresh product, depending on the bulk density and the thickness of the product bed.

Figure 5. Overall view of the laboratory drying facility.

A system of three load cells, type 8523-100 N manufactured by “burster”, measures the weight of the tray containing the product on-line with inaccuracy less than ±0.5 % F.S. and sensitivity tolerance ±0.5 %. The load cells operate by the approved strain gauge method, [Pavese, 1992]. The force to be measured must be centered. Calibration of the load cell was done for different excitation voltages and for loading weights ranging from 0 to 2 kg. The centrifugal fan works with a power supply of 12 V DC and can deliver the air into the system at flow rates ranging from 0 to 55 m3/h according to the supplied voltage that modifies the RPM of the motor. The fan has two admission ways: one for fresh air, and the other for the amount of air that is re-circulated in the system. The ratio between fresh and recirculated air is adjusted from 0 to 100 % according to the baffle position. At the exit, two thermocouples (with dry and wet bulb) measure the two temperatures of the air in order to establish the state parameters of the air at this point. The connection between the fan and the air pipes is made through flexible tubes in order to decrease the vibration transmission from the fan to the system. The electrical resistance, with a maximum power of 2 kW and connected to a power supply of 220 V AC, is inserted in the middle of two coaxial galvanized iron tubes and can heat the drying air up to 80 °C. The outer tube is 40 cm long and has a diameter of 10 cm. The inner tube is a radiation screen with 5 cm diameter and 10 cm length. The power of the electrical resistance can be adjusted in five steps from 0 to 2 kW, in this way obtaining the desired drying air temperature. The operation of the fan and the electrical heater is conducted from an electrical panel.

Fruits and Vegetables Dehydration in Tray Dryers

67

The links between the drying room, centrifugal fan and electrical heater, are made with Plexiglas pipes of 100 ID. The air mass flow rate is measured by means of standard orificeplate devices according with ISO 5167 - 1980. Drying air parameters are evaluated by direct measurements of the dry and wet bulb temperatures using type J thermocouples with a tolerance value, expressed as a deviation in degrees Celsius, of ±1.5 °C under tolerance class 1. The approximate working temperature range of the measuring junction is from +20 to +700 °C, with an approximate generated EMF change per degree Celsius, with reference junction at 0 °C, of 46 μV/°C at 100 °C. The points of air temperature measurement are: environmental conditions, exit of the fan, inlet and outlet of the drying room. The wet-bulb temperature was measured by covering the thermocouples with a wick immersed in distilled water. The thermocouples and the load cells are connected to analog input modules type “ADAM-4018” of a PC-based “ADVANTECH” data acquisition system. The accuracy of the modules is ±0.05 % or better. “Advantech Genie” was chosen as a comprehensive, flexible data acquisition and control software package designed to run in the Microsoft Windows environment. During the drying process, the real time, temperatures and tray weight are displayed on-line on the monitor and are also stored in a .log text file.

3.2. Measurements of Carrot Drying In order to evaluate the potential of the electrical heater and the heat losses from the drying room, a test measurement was carried out with the drying room empty and no recirculation of the air in the system, for a period of two hours. The mass flow rate of the air & a = 64 kg/h and the mean environmental temperature, Tenv = 18.8 °C. The evolution of was m the dry-bulb temperatures taken at characteristic points of the system is presented in figure 6. After approximately one hour, the temperatures reached steady-state conditions with the following average values: Tfan 20 °C at the exit of the fan, Tin 59.4 °C at the entrance of the drying room and Tout 54.5 °C at the exit of the drying room.

& eh , transmitted by the electrical heater to the drying air can be The heat flux, Q calculated with the following expression: & eh = m & a ⋅ c p ,a ⋅ (Tin − Tfan ) Q

(3.1)

& a is the mass flow rate of the drying air, and c p the specific heat of the air at where m & eh is equal constant pressure. During the stationary regime, and for the above conditions, Q to 705 W.

68

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

60

Temperature, oC

50

Measuring points: inlet drying room outlet drying room fan exit environment

40

30

20 0

10

20

30

40

50

60

70

80

90

100

110

120

Time, min Figure 6. Dry-bulb temperature evolution for empty drying room.

A heat transfer balance for the drying room gives the heat loss:

⎛ Tin − Tout ⎞ & loss = m & a ⋅ c p ,a ⋅ (Tin − Tout ) = h c ⋅ A ⋅ ⎜ Q − Tenv ⎟ = 87.46 W ⎝ ⎠ 2

(3.2)

where A = 1.05 m2 is the total wall area of the drying room. From Eq.(3.2) it can be calculated the mean convective heat transfer coefficient, h c = 218 . W m2 ⋅ K . Two series of measurements concerning the drying of fresh carrots were taken: one with uncut jasmine-like carrots, and the other with 15 mm thick sliced carrots of the same variety. 2.6 kg of uncut carrots, having an average length of 180 mm and an average diameter of 25 mm, were uniformly placed on the drying tray. The average initial moisture content of the carrots was 5.42 kgw/kgdm. After a drying period of 37 hours under a hot stream of air with a mass flow rate of 59 kg/h, the weight of the carrots reached 0.405 kg which corresponds to a final moisture content of 0.05 kgw/kgdm. The evolution of the dry-bulb temperatures during the drying process is given in figure 7. The environmental conditions are constant for the first 20 hours, after which the temperatures begins to decrease gradually. The first hour is the warm-up period, after this the inlet drying air parameters can be considered constants.

Fruits and Vegetables Dehydration in Tray Dryers

69

60

Temperature, oC

50

Measuring points: inlet drying room outlet drying room fan exit environment

40

30

20

0

5

10

15

20

25

30

35

Time, h Figure 7. Dry-bulb temperatures during the drying of uncut carrots.

The drying curve of uncut carrots, representing the decrease of the product weight due to the dehydration process is represented in figure 8.

2.5

Weight, kg

2.0

1.5

1.0

0.5 0

5

10

15

20

Time, h Figure 8. Drying curve for uncut carrots.

25

30

35

70

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Scatterable values Polynomial interpolation

Drying rate, g/min

2.0

1.5

1.0

0.5

0.0 0

5

10

15

20

25

30

35

Time, h Figure 9. Drying rate for uncut carrots.

Evolution of the drying rate during the process, expressed in grams of evaporated water per minute is presented in figure 9. It is taken as a global variable for the entire quantity of product. The scattered points are values of measured drying rate and the continuous line is the polynomial regression of the type:

& v = A + B1 ⋅ t + B 2 ⋅ t 2 + B 3 ⋅ t 3 m

(3.3)

& v is the drying rate in g/min, t is the drying time in hours, and the coefficients have where m the values: A = 2.26011; B1 = -0.07174; B2 = -9.88078·10-4 and B3 = 3.8169·10-5. The second series of measurements was taken during the drying of 1.3 kg sliced carrots with an initial moisture content of 4.65 kgw/kgdm. The thickness of each carrot slice was 15 mm and the average diameter 25 mm. The slices were dried for 15 hours with a hot stream of air having 59 kg/h mass flow rate. After this period, the carrots reached a final weight of 0.24 kg and moisture content of 0.05 kgw/kgdm. The evolution of the dry-bulb temperature, during the drying process is given in figure 10, and the drying curve, representing the decrease of the product weight (in kg) during the process is shown in figure 11.

Fruits and Vegetables Dehydration in Tray Dryers

71

70

Temperature, oC

60

Measuring points: inlet drying room outlet drying room fan exit environment

50 40 30 20

0

1

2

3

4

5

6

7

8

9

10

11

12

10

11

12

13

14

15

Time, h Figure 10. Dry-bulb temperatures during the drying of sliced carrots.

1.4 1.2

Weight, kg

1.0 0.8 0.6 0.4 0.2 0

1

2

3

4

5

6

7

8

9

13

14

15

Time, h Figure 11. Drying curve for sliced carrots.

Drying rate is represented in figure 12. The scattered points are values of measured drying rate and the continuous line is the polynomial regression of the type:

& v = A + B1 ⋅ t + B 2 ⋅ t 2 + B 3 ⋅ t 3 m

(3.4)

& v is the drying rate in g/min, t is the drying time, expressed in hours, and the where m coefficients have the values: A = 2.21515; B1 = 0.14375; B2 = -0.05077 and B3 = 0.00212.

72

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

2.5

Scatterable values Polynomial interpolation

Drying rate, g/min

2.0

1.5

1.0

0.5

0.0 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Time, h Figure 12. Drying rate for sliced carrots.

3.3. Measurements of Grape Drying Another series of measurements were taken on the drying of 3.282 kg Sultana grapes, [Margaris, 2007]. The average initial moisture content of the grapes was 3.0 kgw/kgdm. After a drying period of 72 hours under a hot air stream with a mass flow rate of 59 kg/h, the weight of the grapes reaches 0.988 kg which corresponds to a final moisture content of 0.18 kgw/kgdm. The evolution of the drying air dry-bulb temperatures at specific points during the drying process is given in figure 13, and the wet-bulb temperatures in figure 14.

70

60

Temperature, oC

Meas. Simul. inlet drying room outlet drying room fan exit environment

50

40

30

20 0

10

20

30

40

Time, h Figure 13. Dry-bulb temperatures during the drying of Sultana grapes.

50

60

70

Fruits and Vegetables Dehydration in Tray Dryers

73

Temperature, oC

30

Meas. Simul.

25

inlet drying room outlet drying room environment

20

15 0

10

20

30

40

50

60

70

Time, h Figure 14. Wet-bulb temperatures during the drying of Sultana grapes.

The drying curves of Sultana grapes, representing the decrease of the product weight and the moisture content due to the dehydration process, are represented in figures 15 and 16, respectively. 3.5

Measurements Simulation

Weight, kg

3.0

2.5

2.0

1.5

1.0 0

10

20

30

40

50

60

70

Time, h Figure 15. Drying curve (weight vs. drying time) for Sultana grapes.

Evolution of the drying rate (as a global variable) during the process, expressed in grams of evaporated water per minute, is presented in figure 17. The process, under the same drying condition, was simulated with DrySAC code and the predicted parameters are plotted in the same diagrams with the measurement representation. There are very good agreements between measurements and simulation.

74

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Moisture content, kg/kg-db

3.0

Measurements Simulation

2.5 2.0 1.5 1.0 0.5 0.0 0

10

20

30

40

50

60

70

Time, h Figure 16. Drying curve (moisture content vs. time) for Sultana grapes. 1.2

Measurements Simulation

Drying rate, g/min

1.0

0.8

0.6 0.4

0.2

0

10

20

30

40

50

60

70

Time, h Figure 17. Drying rate for Sultana grapes.

Experimental drying curves were built for un-cut and sliced carrots. The drying constant, K, was evaluated to 0.0675 s-1 in the case of un-cut carrots and 0.172 s-1 in the case of sliced carrots. Knowing the drying constant, it is possible to estimate the moisture content as a function of time, or the drying time after which the product reaches a certain moisture content. The analytical expression is given by the approximate solution of the Fick’s law:

X Xe 8 = 2 exp (- K t ) Xi Xe π

(3.5)

Fruits and Vegetables Dehydration in Tray Dryers

75

Measurements Simulation

0.8

i

0.6

e

e

(X-X )/(X -X )

1.0

0.4

0.2 0.0 0

10

20

30

40

50

60

70

Time, h Figure 18. Relative drying rate vs. drying time for Sultana grapes.

In the case of Sultana grapes, the drying constant, K was evaluated to 0.02852 s-1. The relative drying rate curve is graphically represented in figure 18 both from measured values and simulation with DrySAC. The facility is under operation for further product investigation.

4. THERMAL ANALYSIS OF THE TOTAL DRYING SYSTEM Tray drying systems used for dehydrating agricultural products are great energy consumers for which the maintenance and operation control are made mainly heuristically (based on the operator’s experience). To increase of process efficiency, reduce of energy consumption and optimize of process parameters, it is convenient to simulate the total system behavior with a numerical code and to analyze the evolution of the significant parameters in different operating conditions.

4.1. Description of the Self Numerical Code - Drysac A dynamic mathematical model, based on physical and transport properties and mass and energy balances, was developed for the simulation of unsteady convective drying of agricultural products (fruits and vegetables) in static bed conditions. The local material-averaged drying rate and the heat flux depend on local air humidity and temperature, as well as local mass and heat transfer coefficients in interaction with the moisture and temperature distribution inside the material. The model utilizes water sorption isotherm equations and the change in solid density due to the shrinkage phenomenon. The unsteady-state differential equations for temperature and moisture profiles inside the product were numerically solved using a central finite difference scheme, [Welt, 1997].

76

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

The numerical code DrySAC (Drying System Analysis Code) was set up to simulate the dehydration of grapes in a complex drying system, consisting of the main drying room and an adjoining room comprising the equipment of the system (a heat exchanger with a primary heat source being the combustion gases of a burner, a suction fan which introduce hot air inside the drying space, two economizers, two exhaust fans, and a system of dampers which controls the rate of air flow). A schematic diagram for the computational model is given in figure 19. Fresh air (0) passes through two economizer heat exchangers of the cross flow air-to-air type (EC) where it is preheated. The air then enters (via passage 9) the air mixing room where, on mixing with the amount of air (5) that is re-circulated from inside the system, both the temperature and humidity content are changed reaching the state point (1). From this state, the air is heated by contact with the outside surface of a multi-tubular heat exchanger (HE). The primary source of heat is given by a burner (BR) that can operate with different liquid fuels. The heated air (2) is delivered inside the drying room (DR) by a suction fan (SF) and is assumed to be distributed uniformly between the static trays containing the product to be dried. After heat and mass (water) exchange, mainly with the product to be dried, the air (4) leaves the drying space and is divided, by suitable arrangement of baffles (DS), into two courses: one (5) is re-circulated into the system and the other (6) passes through the economizer heat exchangers (EC), where it is cooled until state (7) and is emitted into the environment by the exhaust fans (EF). The algorithm flow chart is given in figure 20.

Figure 19. Schematic diagram for the computation model of the drying system.

Fruits and Vegetables Dehydration in Tray Dryers INPUT PARAMETERS Air, Product, Geometry

t = t + Δt

77

t=0

AIR PREHEATING

NO

YES

MIXING OF PREHEATED AND RECYCLED AIR

MAXIMUM DRYING TIME EXCEEDED

AIR MAIN HEATING

DRYING PROCESS

NO FINAL MOISTURE CONTENT ACHIEVED

HEAT RECOVERY

YES OUTPUT

END

Figure 20. Code flow chart for drying process simulation.

The DrySAC code has three main parts: a) the input data which include the initial values for the numerical solution, ambient conditions, room geometry, equipment characteristics, air flow and product parameters, b) the computing algorithm comprising the numerical modeling of the processes in each system component, and c) the output results grouped in seven output files (one text- and six data-files). After the initial parameters are set out, the programme starts the computing algorithm and calculates the state parameters of the drying air at each passage numbered in figure 19 from 0 to 9. For each apparatus, the three outlet air parameters (pressure, temperature and humidity ratio) at the next time step, are computed according to the inlet parameters at the current time and the specific process of that apparatus. The time step is calculated for each iteration at the upper limit that ensures solution convergence. This is done by putting the condition of Fourier number Fo≤0.25, both for thermal and mass diffusion inside the product bed. A restrictive condition prohibits the time step of numerical integration to be greater than the output data time step. The process inside the drying room involves both heat and mass (water) transfer between the drying air and the product (grapes). The drying air will lose heat and gain humidity. For each iteration, the air velocity above the bed is recalculated as it varies due to the modification of the free space between the trays (as a result of the shrinkage effect of the product). The system of partial differential equations, Eq.(4.1) and (4.2), describing heat and mass transfer within the product bed, is integrated by using an implicit, central finite-differences algorithm, (Welt, 1997).

ρb ⋅ c ⋅

∂T ⎛ ∂T ⎞ = div ⎜ λ ⋅ ⎟ ⎝ ∂z ⎠ ∂t

(4.1)

78

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

∂(ρ b ⋅ X) ∂X ⎞ ⎛ = div ⎜ ρ b ⋅ D eff ⋅ ⎟ ⎝ ∂t ∂z ⎠

(4.2)

The algorithm gives the temperature and the moisture content for each product layer at the next time step ( t + Δt ), and calculates the new thickness of the bed as the sum of each layer thickness:

Tj, t + Δt = Tj, t +

λ ⋅ Δt

( )

c ⋅ ρ b ⋅ Δz j

2

D eff ⋅ Δt

(

X j, t + Δt = X j, t +

( ) Δz j

2

m ds n ⎛ 1 ⎞ ⎟ z bed = ⋅∑ ⎜ n ⋅ A j=1 ⎝ ρ b,db ⎠

(

⋅ Tj+1, t − 2 ⋅ Tj, t + Tj−1, t

⋅ X j+1, t − 2 ⋅ X j, t + X j−1, t

)

)

(4.3)

(4.4)

(4.5) j

where j refers to a current layer, Δz j is the thickness of the layer j, m ds is the mass of the dry product, n the number of layers, A the drying surface area, and ρ b, db the bulk density dry-basis which is calculated as:

ρ b, db =

ρb 1+ X

(4.6)

The programme terminates when the average moisture content of the product is lower than the requested one, or when the maximum drying time is reached. The final main results, which represent the predicted values of computed parameters, are recorded in a text file. These parameters are: drying time, final parameters of the product (weight, temperature, average moisture content on wet and dry basis), final porosity and thickness of the product bed, final air velocity between the trays, and the total quantity of evaporated water from the product. Specific energy consumption follows: thermal energy used by the main heat exchanger to heat the drying air in order to evaporate one kilogram of water from the product, the thermal energy needed to decrease the moisture content of one kilogram of product from the initial value to the requested one, and the total energy consumption (thermal and electrical) for evaporating one kilogram of water. The efficiency of the drying process is defined as the ratio between the thermal energy needed exclusively for evaporating the water from the product and the thermal energy consumed in the main heat exchanger for heating the drying air. Finally, the total energy consumption is recorded by categories: thermal for heating the drying air, thermal needed for the drying process, thermal recovered in the two economizers, electrical consumed by the suction fan, and electrical consumed by the two exhaust fans.

Fruits and Vegetables Dehydration in Tray Dryers

79

The unsteady-state parameters are recorded in six data-files for each output data time step chosen at the beginning of the programme. The first column of each data file is the drying time. The first data file stores the temperatures and the moisture contents of the product bed (mean, basal and surface temperature, mean, basal and surface moisture content in dry-basis and mean moisture content in wet-basis), the drying rate, the thickness of the product bed, and the bulk weight of the product. The second data file contains the time step of numerical integration, the heat and mass transfer coefficients, Reynolds number, drying air velocity between the trays, thermal power of the main heat exchanger, thermal energy flux needed for drying, energy rate recovered in the economizer heat exchangers and the specific energy consumed to evaporate one kilogram of water from the product. The temperature, moisture content, relative humidity and specific enthalpy of the drying air are recorded in the rest four data files, for all characteristic passages of the system.

4.2. Numerical Simulation of the Drying System with Drysac Prediction of the drying process, correlated with the operation of the system equipment, was carried out by numerical simulation of the designed operation conditions for the drying of Corinthian grapes [Ghiaus, 1998]. The input parameters for the studied case, as specified in the text file, are as follows: initial mass of wet product, 5000 kg; initial temperature of product, 26 °C; initial product humidity wet basis, 75 %; required product humidity wet basis, 15 %; max. air temperature, 70 °C; drying-air volume rate, 24000 m3/h; recirculation ratio, 75 %; number of trays, 400 pcs; ambient temperature, 26 °C; ambient relative humidity, 50 %. Using the DrySAC code, the predicted drying time was calculated to be 38 hours and 21 minutes during which the grapes are dried from 75 % moisture content - wet basis to 15 % moisture content - wet basis. Figure 21 shows graphically the evolution of grape temperature during the drying process, at the surface and bottom of the bed and as an average. Differences of temperature between the base and surface of the bed appear only during the first period of drying, approx. the first 5 hours. After this the bed temperature remains uniform until the end of the process. During the first 5 hours the temperature gradient is high and corresponds to the so-called warm-up period of drying. During the next 25 hours the temperature remains practically constant, and during the last period it starts to increase again. At the end of the process, the temperature of the grapes reaches 42.5 °C. One of the most important drying parameters is the drying rate (figure 22) which represents the rate of evaporated water from one square meter of drying product. At the beginning of the process the drying rate increases from 0.06 g/s m2 to 0.12 g/s m2 and then has a very small decreasing slope. At the end of the process the drying rate decreases rapidly. Drying air parameters are predicted at characteristic points of the system: inlet of the fresh air into the economizers (ambient air), outlet of the economizers onto the fresh air path (preheated fresh air), inlet of the main heat exchanger (the mixing between preheated and recycled air), inlet and outlet of the drying room, and the outlet of the economizers onto the exhaust air path.

80

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Temperature, oC

40

35

30

surface of grape bed average bottom of grape bed

25 0

5

10

15

20

25

30

35

30

35

Time, h Figure 21. Evolution of grape temperature during drying process. 0.14 0.13

Drying rate, g/s m2

0.12 0.11 0.10 0.09 0.08 0.07 0.06 0

5

10

15

20

25

Time, h Figure 22. Drying rate vs. drying time for grape dehydration.

The ambient air parameters are given as input data and are considered constant. Evolution of drying air temperature is given in figure 23 and the relative humidity of the drying air in figure 24.

Fruits and Vegetables Dehydration in Tray Dryers

70

inlet drying room outlet drying room inlet heat exchanger

65 60

Temperature, oC

81

55 50 45 40

preheated exhaust ambient

35 30 25 0

5

10

15

20

25

30

35

Time, h Figure 23. Drying air temperature at different locations vs. drying time.

exhaust ambient outlet drying room

Relative humidity, %

70 60 50 40

inlet heat exchanger preheated inlet drying room

30 20 10 0

5

10

15

20

25

30

35

Time, h Figure 24. Relative humidity of drying air at different locations vs. drying time.

4.3. In-Situ Monitoring of a Full Scale Drying System The prototype drying system was manufactured by the Greek industry Vencon - Varsos S.A., has the overall dimensions (L x W x H) 8.1 x 3.11 x 2.81 in m, 3,500 kg in weight, and consists of two main parts: the drying room and the equipment room (figure 25). The equipment room contains the following: a suction fan, the main heat exchanger, a LPG burner, two economizer heat exchangers and two exhaust fans. A set of adjustable baffles permits the control of inlet, outlet and recycled air within the system.

82

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Figure 25. Schematic representation of drying system and air flow.

Measurements were taken in-situ in Kimi, Evia Island, Greece, during the drying of Corinthian grapes. The initial weight of fresh grapes to be dried was 2,958 kg with average initial moisture content of 76 % - wet basis. The grapes were distributed uniformly in 308 trays, arranged in 22 columns of 14 trays each, and each tray containing between 8 and 10 kg of grapes. The maximum temperature of the drying air (with an inlet volume flow rate of 23,000 m3/h) was set-up at 65 °C. During the drying period, the ambient temperature varied between 16 and 32 °C and the ambient relative humidity between 40 and 90 %. At the entrance of the drying room, the drying air parameters were evaluated by measuring the drybulb temperature and the relative humidity. At the exit of the drying room, the dry- and wet-bulb temperatures were measured and recorded. The temperature of the exhaust drying air exiting of the economizer heat exchanger was also measured. The initial weight of the fresh grapes and the final weight of the dried grapes contained by each tray were weighed at the beginning and at the end of the process, respectively. After a drying time of 44 hours, the grapes reached a final weight of 1,031 kg and a final average moisture content of 31 %.

5. FLOW ANALYSIS OF THE TOTAL DRYING SYSTEM Aerodynamics of drying spaces for batch drying systems is the most important factor which ensures reasonable design results and high quality of the final dried product. Experimental analysis of air motion inside drying rooms in a real life situation is difficult and sometimes impossible to be carried out, expensive and time consuming. In such situations, it is more convenient to use techniques of computation fluid dynamics (CFD).

Fruits and Vegetables Dehydration in Tray Dryers

83

5.1. Local Flow Analysis - Flow Over one Tray The Navier-Stokes (N-S) equations together with the continuity equation, comprise a closed set of equations, the solution of which provides a valid description of laminar and turbulent flows. The turbulent motion has a wide spectrum of eddy sizes, and large and small eddies can coexist in the same volume of fluid. The transport equation, in a general form, can be written as follows:

∂ r ρ ⋅ Φ) + div(ρ ⋅ w ⋅ Φ) = div (ΓΦ grad( Φ) ) + SΦ ( ∂t

(5.1)

transient convection diffusion source

where Φ is the potential, ΓΦ diffusivity, and SΦ the source term, and are given in table 3. Table 3. Parameters of the general transport equation Equation

Φ

ΓΦ

SΦ

Continuity u- momentum

1 u

0

0

v- momentum

v

μ

w- momentum

w

μ

Kinetic energy

k

Dissipation rate

ε

μ

ρ⋅

νt σk

ρ⋅

νt σε

−

∂ ∂w ∂p ∂ ∂u ∂ ∂v ) + (μ ) + (μ ) + (μ ∂x ∂x ∂x ∂y ∂x ∂z ∂x

−

∂p ∂ ∂u ∂ ∂v ∂ ∂w + (μ ) + (μ ) + (μ ) − ρ⋅ g ∂y ∂x ∂y ∂y ∂y ∂z ∂y

−

∂w ∂p ∂ ∂u ∂ ∂v ∂ ) + (μ ) + (μ ) + (μ ∂z ∂x ∂z ∂y ∂z ∂z ∂z

ρ ⋅ ( Pk + G b − ε )

ρ⋅

ε ⋅ (C1 ⋅ Pk + C3 ⋅ G b − C 2 ⋅ ε ) k

Engineers are not concerned with all the details of turbulent motion, but rather with its effects on the gross properties of the flow. Consequently, there is no need to solve the instantaneous variables if averaged variables are all that is required. Most turbulence models use the eddy-viscosity concept ( μ t = ρ ⋅ ν t ) which is a property of the local state of the turbulence. The simplest turbulence model is one which uses a constant value for the eddy viscosity. For dimensional reasons, the effective kinematic viscosity made by the local turbulence is proportional to typical velocity and characteristic length:

νt = C ⋅ w ⋅ lc

(5.2)

84

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

where C is a constant, w the typical velocity, and l c the characteristic length. The k-ε two-equation turbulence model has proved the most popular, mainly because it does not require a near-wall correction term. The standard form of the k-ε turbulence model employs the following turbulence transport equation: kinetic-energy equation:

⎛ ν ⎞ ∂ r ρ ⋅ k ) + div(ρ ⋅ w ⋅ k ) = div ⎜ ρ ⋅ t grad( k )⎟ + ρ ⋅ ( Pk + G b − ε ) ( ∂t ⎝ σk ⎠

(5.3)

dissipation equation:

⎛ ν ⎞ ε ∂ r ρ ⋅ ε ) + div(ρ ⋅ w ⋅ ε ) = div ⎜ ρ ⋅ t grad(ε )⎟ + ρ ⋅ ⋅ (C1 ⋅ Pk + C3 ⋅ G b − C 2 ⋅ ε ) ( ∂t k ⎝ σε ⎠

(5.4)

where ε is the dissipation rate, σ k and σ ε are the effective Prandtl and Schmidt numbers, Pk is the volumetric production rate of turbulent kinetic energy by shear forces, G b is the volumetric production rate of turbulent kinetic energy by gravitational forces interacting with density gradients (buoyancy production), and C1 , C 2 and C 3 are constants [Bradshaw, 1981]. The kinematic turbulent viscosity is defined by:

k2 ν t = Cμ ⋅ Cd ⋅ ε

(5.5)

Two-equation models account for transport effects of velocity and characteristic length, and the distribution of l c is determined by the model. The characteristic length may be recovered from:

k1.5 lc = Cd ⋅ ε

(5.6)

The model constants are: C μ = 0.5478; C d = 0.1643; σ k = 1.0; σ ε = 1.314; C1 = 1.44;

C 2 = 1.92 and C 3 =1.0. The constant C 3 has been found to depend on the flow situation. It should be close to zero for stably-stratified flow, and close to 1.0 for unstably-stratified flow. The volumetric production rate of turbulent kinetic energy by shear forces, Pk , is defined as:

∂w i Pk = ν t ⋅ ∂z j

⎛ ∂w i ∂w j ⎞ ⎟ ⋅ ⎜⎜ + ⎟ ⎝ ∂z j ∂z i ⎠

(5.7)

Fruits and Vegetables Dehydration in Tray Dryers

85

and the volumetric production rate of turbulent kinetic energy by gravitational forces interacting with density gradients, G b (buoyancy production), is defined as:

G b = −νt ⋅

g i dρ ⋅ ρ ⋅ Pr dz i

(5.8)

G b is negative for stably-stratified (dense below light) layers, so that k is reduced and turbulence damped. G b is positive for unstably-stratified (dense above light) layers, in which therefore, k increases at the expense of gravitational potential energy. This model forms a good compromise between generality and economy of use for many CFD problems. In order to evaluate the influence of different turbulence models on the flow pattern of geometric arrangements specific for drying rooms, a simplified laboratory model was constructed. This laboratory model consists of one empty tray, having the dimensions length x width x height as 50 x 100 x 9 cm respectively, and placed into a rectangular channel with a cross-section of 100 x 19 cm2. The numerical simulation was carried out using the PHOENICS CFD commercial code. Due to the fact that only two velocity components are significant and side effects are negligible, the computational domain was chosen in 2-D, representing a vertical plane, parallel with the main direction of flow and passing through the longitudinal axis of the tray. The tray, placed in the middle of the domain, was simulated by two vertical walls having 0.5 m distance between them, 9 cm height and 1 cm width, each. The domain was divided into 3,040 cells using a 160 x 19 Cartesian grid. The inlet opening boundary condition is the air velocity, which is supposed to be uniform and equal to 2.87 m/s. The exit boundary condition is the ambient pressure, 101.325 kPa. No-slip boundary condition and appropriate wall functions apply on the solid surfaces. Figure 26 shows the velocity vector distribution for the simulation using k-ε two-equation turbulence model.

Figure 26. Simulation of velocity vectors distribution using two-equation k-ε model.

An experimental arrangement, similar to the one investigated numerically, was set up to carry out air flow measurements in the laboratory. The model consists of a wood tray, having the dimensions: length 0.5 m, width 1.0 m and height 0.1 m, which was inserted in the middle of the test section of a low-speed wind tunnel. The test compartment of the wind tunnel has a cross-section of 1.0 x 0.2 m2 and is 1.5 m in length. The air flow, delivered by a centrifugal blower, passes through a settle room (1.0 x 1.0 m2 cross-section, 1.2 m in length) having a series of wire sieves, and enters the test channel

86

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

through a convergent nozzle. At the entrance of the test section, the air velocity was measured and found to be uniform and constant having the value 2.87 m/s. The flow field characteristics were measured with a five-hole tube, with a spherical head 8 mm in diameter and a bent shaft, type Schiltknecht f.881, [Ghiaus, 1994]. The measured values of the mean velocity vector field, was plotted using the SURFER graphic commercial code and is presented in figure 27.

Figure 27. Distribution of velocity experimentally measured.

The availability of measurement techniques permitting the description of the spatial structure of the flow allows cross-validation of the results obtained and better suitability of the proposed numerical simulation models. Validation of zero-equation turbulence models showed the closest simulation results with the experimental measurements for a viscosity ratio ν t ν l = 100 .

18

18

16

16

14

14

Channel height, cm

Channel height, cm

The two-equation turbulence models did not differ significantly from each another and compared with the experiments. However, the best approximation was given by the basic k-ε model. Figure 28 shows the profiles of the horizontal components of velocity vectors obtained from numerical simulation with different turbulence models compared to actual measured values, in three cross-sections of the test domain.

12 10 8 6

12 10 8 6

4

4

Measured ν = 100 ν

2

2

k-ε model

t

-2

0

2

4

6

8

10

Velocity, m/s

a) 15 cm before tray

-2

0

2

4

6

Velocity, m/s

b) middle of tray

Figure 28 - Continued on next page

l

8

10

Fruits and Vegetables Dehydration in Tray Dryers

87

18

Channel height, cm

16 14 12 10 8 6 4 2 -2

0

2

4

6

8

10

Velocity, m/s

c) 15 cm after tray Figure 28. Profiles of velocity horizontal components for constant viscosity and k-ε turbulence models.

We can conclude that the best simulation results for this type of problem can be obtained by using the k-ε turbulence model. It is useful practice, however, to start the numerical simulation with a simple model (constant effective viscosity) and after the configuration problems are solved, the convergence of the solution can be achieved using a more complicated model (k-ε for instance).

5.2. Global Flow Analysis - Flow Inside Drying Room The analyzed drying space has the overall dimensions 6 x 3 x 2 m (length, width and height). Two rows of trays are placed symmetrically from the longitudinal vertical plane of the room. The length of one tray is 75 cm, and the width can be either 50 or 75 cm. Therefore, using trays with 50 cm width, each row will have 12 columns; and with 75 cm, 8 columns. The inlet and outlet openings for the drying air are on the back wall: the two inlets at the centre-upper part (each of 44 x 36 cm2), and the two outlets at the lateral-bottom sides (each of 200 x 30 cm2). Above the drying room, in the direction of the longitudinal axis, there is a wedge-shape distribution channel. Next to the drying space is the equipment room containing fans, heat exchangers, a burner and the control panel, figure 29 and 30. Numerical simulation of air movement inside the drying room was carried out using the PHOENICS CFD code, [Margaris, 2006]. Due to the longitudinal plane of symmetry of the drying room, the computational domain chosen was half of the room. The grid was generated using non-uniform Cartesian grids, with number of cells and grid steps according to each studied case. Many different geometry configurations were studied including: room with and without trays, use of guiding wings for air distribution, separation of the columns by vertical walls, different size of the inlet openings as well as some specific test cases. Analysis of the flow inside the drying room containing trays requires a great number of cells.

88

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Figure 29. Outline of the industrial 5-ton dryer and prototype photo during the execution phase.

side view

front view Figure 30. Side and front view of the industrial dryer with 14 trays per column.

Fruits and Vegetables Dehydration in Tray Dryers

89

The airflow inside this complicated geometry space was simulated with the PHOENICS CFD code. The computational domain (half of the space due to its longitudinal plane of symmetry) has the overall dimensions 1.5 x 2.7 x 6 m and was divided in 22320 cells as follows: 10 cells in x direction, 62 cells in y direction from which 50 cells for the drying room containing 25 trays and 12 cells for the distribution channel, and 36 cells in z direction. Therefore, each cell has 15 cm and 16.5 cm for x and z direction respectively, whereas in y direction has 24 mm for those cells corresponding to the tray thickness, 16 mm between the trays and 58 mm inside the distribution channel. The grid for the computational domain is given in figure 31.

Figure 31. Grid of the computational domain for 25 trays per column.

Because the flow between the trays is horizontal, the velocity has only two components (in x and z directions) and the grid has only one cell between two trays. Each tray is considered to be a blockage for the flow. No-slip boundary conditions and appropriate wall functions apply on the solid surfaces. The selected turbulence model was the standard 2equation k-ε. For an inlet flow rate of 3.33 m3/s, which corresponds to a mean inlet velocity of 21 m/s and a mean velocity over the trays of 2.86 m/s, a typical example of 3-D representation of the velocity vector distribution inside the drying room is given in figure 32. It can be observed a non-uniform flow inside the drying space associated with re-circulation regions. In the drying space, the velocity vectors are plotted in several vertical planes, and in

90

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

24 levels representing the spaces between the trays. The most intense circulation of the air was found to be at the bottom of the room.

Figure 32. Overall velocity distribution inside drying room with 25 trays.

Graphical representations of the flow parameters for different geometrical configuration and their interpretation led to the optimization of the tray arrangement, size and position of the inlet opening, opportunity of using separation walls between rows of trays and directing wings inside the distribution channel and central corridor. For example, the optimization of the distribution channel geometry led to a uniform inlet air for the drying space, thus more uniform flow and no re-circulation are achieved between the trays as can be seen in figure 33, where 3-D velocity distribution in vertical and horizontal planes after optimization is presented. The proposed improvements were implemented into practice on an industrial dryer that operates in Kimi, Evia Island of Greece. The drying system was monitored in-situ during the drying of currants and grapes. Air velocity was measured at the inlet and outlet openings of the drying room and at some specific points of the distribution channel, inlet and outlet corridors and between the trays, and the experimental results were every time compared to the corresponding numerical results. The design improvements resulted from the numerical simulation led to an increment of 23% in the uniformity of the humidity inside the final product, when 3 tons of Corinthian grapes were dried during the monitoring period.

Fruits and Vegetables Dehydration in Tray Dryers

91

Figure 33. 3-D velocity distribution in vertical and horizontal planes after optimization.

The evaluation of the airflow parameters inside drying units is a very important task. In some cases, the air has to be removed, while in others, it just helps the separation of the other two phases. Experimental investigations give valuable information for the understanding of a specific phenomenon and help to validate general CFD codes that are afterward used to extensively analyze the phenomenon by numerical simulation. Improvement solutions by flow manipulation techniques using CFD codes can be given to both, dryers that are already in operation and new units under design.

92

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

6. OPTIMIZATION OF DRYER DESIGN AND OPERATION When sufficient information exists regarding the initial moisture content of the product to be dried and the final moisture content required, it is necessary to decide on the most suitable drying conditions from the point of view of quality and energy savings. When designing dryers, it is very important to ensure that drying will be highly uniform, heat consumption will be low, and that the construction is simple and easy to operate.

6.1. Improvement of Drying Room Geometry The distribution of air over the drying space inlet cross-section determines the uniformity of drying in the plane perpendicular to the direction of air flow. The longer the distance over which the air has been in contact with the product, the lower the drying capacity of the air, so that drying along the line of air flow is never uniform. As discussed in the previous paragraph, the flow field inside the drying room of the 5-ton dryer analyzed is not uniform. This is due mainly to the fact that the drying room has very large dimensions and the air distribution system is a very simple one. One possibility to improve air distribution uniformity is to decrease the velocity of the inlet drying air without modifying the flow rate. This can be done by increasing the cross-sectional area of the inlet opening.

Figure 34. Initial position of the suction fan - side view.

As the system analyzed has already been constructed and is in operation, any modifications to its design should not involve substantial changes. Taking this into account, a very good solution would be to shift the suction fan 70 cm backward from its initial position (figure 34), and to insert a diffuser between the exit of the fan and the inlet opening of the drying room. In this way, the air velocity at the inlet of the drying room will decrease from 21 m/s to 13.7 m/s according to the continuity equation.

Fruits and Vegetables Dehydration in Tray Dryers

93

Figure 35. Shift and rotation of the suction fan.

In order to achieve better distribution of the drying air, two adjustable dampers were inserted at the top of the distribution corridor. Each one of these guides the drying air to a two meter section of the drying room. Another possibility to improve the uniformity of the air distribution is, besides shifting the suction fan, to rotate it 10 ° from its initial vertical position (figure 35). In this way, the air jet of the centrifugal fan will have the direction of the inclined distribution channel. Alternatively, instead of adding two adjustable dampers, five dampers can be added, which will adjust the air flow for each meter of the drying room (figure 35).

6.2. Minimization of Drying Time An important question is how the heat efficiency of a batch drying system can be improved. One effective method to do this would be to re-use some of the exhaust air. This is called recirculation of the drying air. Another parameter that affects the process efficiency is the inlet flow rate. Both the air recirculation ratio and the inlet flow rate can be optimized for a given drying unit configuration in order to minimize drying time and energy consumption.

6.2.1. Optimum Air Recirculation Ratio A specified percentage of the exhaust drying air is mixed with the pre-heated fresh air in order to increase the thermal potential of the heated inlet drying air. In this way the air that enters in the main heat exchanger has a higher temperature and hence the thermal energy required to bring the inlet air to the requested operation temperature is lower. At the same time, together with the calorific contribution, the recycled air brings an amount of humidity which increases the relative humidity of the inlet drying air and, therefore, the driving force for drying will be lower. This leads to an optimum amount of air to be re-circulated into the system for which, when the configuration system is specified, the drying time will be a minimum. Using the repetitive analysis technique, the drying process was simulated with the “DrySAC” code for the nominal conditions specified in Paragraph 4, and for different values of air recirculation ratios ranging from 0 % to 98 %. A minimum drying time of 38 hours and 49 minutes was calculated for a recirculation ratio of 80 %. The drying time plotted vs. air

94

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

recirculation ratio is presented in figure 36. When no air is re-circulated in the system, the drying time is about 50 hours. As the air recirculation ratio increases, the drying time decreases until it reaches the minimum value at 80 % recirculation ratio. After this point, further increase of the air recirculation ratio, leads to a very sharp increase of the drying time. 70 65

Drying time, h

60 55 50 45 40 35 0

10

20

30

40

50

60

70

80

90

100

Air recirculation ratio, % Figure 36. Minimization of drying time by means of air recirculation ratio.

As the magnitude of the air flow rate is difficult to measure and therefore the exact air recirculation ratio is difficult to establish, it is better to set up the air recirculation ratio at a value less than the optimum. For the operation of the analyzed dryer, an air recirculation ratio of 75 % is recommended.

0.14

Air recirculation ratio 75 % 90 % 0% 97 %

Drying rate, g/s m2

0.12 0.10 0.08 0.06 0.04 0.02 0.00 0

10

20

30

40

Time, h Figure 37. Drying rate for several air recirculation ratios.

50

60

70

Fruits and Vegetables Dehydration in Tray Dryers

95

Influence of air recirculation ratio on the drying rate, which is one of the most important process parameters, is shown graphically in figure 37. During the warm-up period of the drying process, which is approximately one hour, it is preferable not to have recirculation of air. After this period, the air recirculation ratio can be set up at the optimum value. Air recirculation ratio also has great influence on the product temperature. The evolution of grape temperature during the drying period for several air recirculation ratios is given in figure 38. Using a higher air recirculation ratio increases the temperature of the product. For an air recirculation ratio of 75 %, the final product temperature will not be higher than 43 °C, which also ensures a high quality of the dried product.

55

Temperature, oC

50 45 40

Air recirculation ratio 97 % 90 % 75 % 0%

35 30 25 0

10

20

30

40

50

60

70

Time, h Figure 38. Influence of air recirculation ratio on grape mean temperature.

Moisture content, kg/kg-db

3.0

Air recirculation ratio 97 % 0% 90 % 75 %

2.5 2.0 1.5 1.0 0.5 0.0 0

10

20

30

40

Time, h Figure 39. Influence of air recirculation ratio on grape moisture content.

50

60

70

96

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Drying time, h

70

60

50

40 0

10

20

30

40

50

60

70

80

90

100

Air recirculation ratio, % Figure 40. Minimization of drying time by means of air recirculation ratio for three tons of drying grapes.

The decrease of the product moisture content, in dry-basis, during the drying period for different air recirculation ratios is given in figure 39. In each case, the initial product moisture content was 3 kg/kg db (or 75 % in wet-basis) and the requested moisture content to be achieved at the end of the process was 0.18 kg/kg db (or 15 % in wet-basis). Air recirculation ratio affects also the drying air parameters at the inlet of the drying room. The optimum air recirculation ratio is not a constant parameter. It depends on the configuration of the system, the characteristics of the product to be dried, and the state parameters of the drying air. It depends also, for the same dryer configuration and drying conditions, on the quantity of the product to be dried. The optimum air recirculation ratio when 3 tons of grapes were dried, instead of 5 tons, was found to be near 60 % (figure 40).

6.2.2. Optimum Inlet Air Flow Rate For a given dryer configuration, the inlet air flow rate affects the drying time in the following way: when the air flow rate is increased, the air velocity above the product bed will also increase and the heat and mass processes will be intensified. At the same time, because the main heat exchanger and the burner have an upper limit operation condition, the inlet air temperature will decrease and negatively affects the drying time. These two tendencies lead to a specific value of the inlet air flow rate, for which the drying time is minimum. For the design operation conditions of the 5-ton dryer, the predicted drying time was found to be a minimum and was 38 hours and 20 minutes for 24,000 m3/h inlet air flow rate (figure 41). For values of inlet air flow rate less than this optimum, the drying time increases very rapidly, whereas for higher values, the drying time remains almost constant.

Fruits and Vegetables Dehydration in Tray Dryers

97

50

Drying time, h

48 46 44 42 40 38 12

14

16

18

20

22

24

26

28

30

Inlet flow rate 10-3, m3/h Figure 41. Minimization of drying time by means of inlet air flow rate (5 tons).

The minimum drying time is achieved with an inlet air flow rate of 27,000 m3/h. The analysis was carried out for a 60 % air recirculation ratio which was found to be the optimum value for this configuration. If the nominal air recirculation ratio, i.e. 75 %, is considered, the optimum value of the inlet air flow rate becomes 33,000 m3/h.

6.3. Reduction of Energy Consumption Energy efficiency is of major concern in today’s environmentally conscious society. Industrial processes are increasingly required to meet specified emission standards. Energy use and energy efficiency are two important areas for the assessment of legal operational requirements. A reduction in industrial energy consumption is not only of benefit to the environment; the adoption of energy saving schemes can facilitate the lowering of production costs and help maintain competitiveness. A typical air drying unit will consume approximately 5 times its capital cost in energy during its life time [Gilmour, 1998] and as such, the energy efficiency of a drying system should be an important consideration in its purchasing and operating decisions. However, the importance of drying as a consumer of valuable energy resources has been given little attention, the majority of research in drying focusing on material applications and drying methods. The conditions under which a drying system operates affect not only the drying time but also the energy consumption. As the air recirculation ratio increases the total consumption of thermal energy needed for heating is less. At the same time, increasing the amount of air recycled in the system decreases the air flow rate that will pass through the economizer heat exchangers, and so the energy saved by the economizers will be less. It is common practice to express energy consumption as specific values, related to the amount of evaporated water or to the quantity

98

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

of product to be dried. As these functions have a gradually decreasing slope, the main criterion of optimizing the air recirculation ratio remains the drying time minimization. 90

Drying efficiency, %

85 80 75 70 65 60 55 0

10

20

30

40

50

60

70

80

90

100

Air recirculation ratio, % Figure 42. Drying efficiency vs. air recirculation ratio.

Drying efficiency, expressed as the ratio between the thermal energy actually needed for drying and the thermal energy consumption for heating the drying air, is given, as a function of air recirculation ratio, in figure 42. To ensure the best use of thermal energy for the drying process, it is profitable to recycle large quantities of drying air.

7. CONCLUSION The air flow inside complex geometry spaces, such as drying rooms containing hundreds of trays arranged in rows and columns, was analyzed by solution of 3-D momentum turbulent flow equations for different room configurations. Laboratory measurement data, concerning the space velocity distribution and the pressure field of the air flow over one tray, are provided and used for validation of turbulence models. The results of the flow investigation lead to practical suggestions for the improvement of the air flow uniformity inside the drying space which is very important for the quality of the product. A novel numerical code, DrySAC, able to predict the unsteady-state processes taking place in a complex drying system, was developed. Unlike other attempts to predict drying processes, DrySAC takes into account not only the drying process itself, but also the behavior of the other system equipment and the interaction between them. Drying curves, evolution of the air state parameters in characteristic points of the system and product properties were predicted during the drying of Sultana grapes and, to the best of author’s knowledge, appear for the first time in literature. For a practical validation of the code, the predicted values compared with the measured data taken in-situ when 3 tons of Corinthian grapes were dried in the full-scale prototype dryer in Kimi, Evia Island, Greece, showed good agreement. The numerical DrySAC code can be used for optimization of the process parameters when a dryer

Fruits and Vegetables Dehydration in Tray Dryers

99

configuration is given. For the studied case, an air recirculation ratio of 75 % and an inlet flow rate of 6.5 m3/s have proved to be the optimum, for which a minimum drying time of 38 hours is achieved. The code can be used both for evaluation of existing dryers and for optimum design of the new units with beneficial impacts in increasing the efficiency of the systems and in reduction of energy consumption. Aiming to overcome the lack of experimental data in the open literature, a laboratory drying unit was specially constructed for testing and monitoring the dehydration of agricultural products. Using this facility, experimental drying curves were built for the drying of carrots and Sultana grapes under controlled conditions of the drying air parameters, which were gathered by means of a data acquisition system. The laboratory experimental results are useful for the validation of numerical models which are further, an essential tool for optimization and increasing the efficiency of the drying process. The research in the field of drying of agricultural products remains open mainly because of their delicate, and hard to be established, properties. Further investigation must be done for evaluation of moisture sorption isotherms, shrinkage coefficient and drying curves of particular species of fruits and vegetables. Industry and academia should be brought closer together by development of reliable good software not only for design of the new dryers but also for evaluation of existing ones. DrySAC is such a code and can solve many practical industrial difficulties.

REFERENCES Abalone R.M., Lara M.A., Gaspar R. and Piacentini R.D., 1994: Drying of biological products with significant volume variation. Experimental and modeling results for potato drying, Drying Technology. 12, pp. 629-647 ASHRAE, 1997: ASHRAE Handbook 1997 Fundamentals, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, pp. 5.1-5.10 Belessiotis V., 1995: Unpublished data. National Centre for Scientific Research “DEMOKRITOS”, Greece Bird R.B., Stewart W.E. and Lightfoot E.N., 1960: Transport phenomena, John Wiley and Sons, New York Borgstrom G., 1971: Principles of food science, Vol. I - Food technology, Collier-MacMillan Canada, Ltd., Toronto Bradshaw P., Cebeci T. and Whitelaw J.H., 1981: Engineering calculation methods for turbulent flow, Academic Press, London Bruin S. and Luyben K.Ch.A.M., 1980: Drying of food materials: A review of recent developments, in Advances in drying, vol. 1, A.S. Mujumdar (Ed), Hemisphere Publishing Corporation, Washington, pp. 155-215 Charm S.E., 1978: The fundamentals of food engineering, The AVI Publishing Company, Inc., Westport, CT Chirife J., Boquet R., Ferro-Fontan C. and Iglesias H.A., 1983: A new model for describing the water sorption isotherm of foods, Journal of Food Science. 48, pp. 1382-1383

100

Dionissios P. Margaris and Adrian-Gabriel Ghiaus

Chirife J., 1983: Fundamentals of the drying mechanism during air dehydration of foods, Ch. 3 in Advances in drying, A.S. Mujumdar (Ed), Hemisphere Publishing Corporation, Washington, pp. 73-102 Ghiaus A.G., Margaris D.P., Papanikas D.G. and. Fertis D.K., 1994: Experimental and theoretical analysis of gas-liquid separation process in baffle-type units, Second European Fluid Mechanics Conference, 20-24 September 1994, Warsaw Ghiaus A.G., Margaris D.P. and Papanikas D.G., 1996:, Energy analysis and Optimization of Tray Drying Systems, Proceedings of the T.I.E.E.S. - 96, Vol. 1, Edited by: T. Ayhan, I. Dincer, H. Olgun, S. Dost and B. Cuhadaroglu, Trabzon, pp. 455-460 Ghiaus A.G., Margaris D.P. and Papanikas D.G., 1997: Mathematical modeling of the convective drying of fruits and vegetables, Journal of Food Science. 62, pp. 1154-1158 Ghiaus A.G., Margaris D.P. and Papanikas D.G., 1998: Evaluation of fixed bed drying of agricultural produce by simulation and laboratory investigation, Drying. ’98, Ziti Editions, Thessaloniki, pp. 357-364 Gilmour J.E., Oliver T.N. and Jay S., 1998: Energy use for drying processes: the potential benefits of airless drying, Drying. ’98, Ziti Editions, Thessaloniki, pp. 573-580 Jayaraman K.S. and Das Gupta D.K., 1992: Dehydration of fruits and vegetables - recent development in principles and techniques, Drying Technology. 10, pp. 1-50 Jayas D.S., Cenkowski S., Pabis S. and Muir W.E., 1991: Review of thin-layer drying and wetting equations, Drying Technology. 9, pp. 551-588 Keey R.B., 1982: Introduction to industrial drying operations, Pergamon Press, Oxford King C.J., 1968: Rate of moisture sorption and desorption in porous, dried foodstuffs, Food Technology. 22, pp. 509-514 Leniger H.A. and Beverloo W.A., 1975: Food process engineering, D. Reidel Publishing Company, Dordrecht Lozano J.E., Rotstein E. and Urbicain M.J., 1980: Total porosity and open-pore porosity in the drying of fruits, Journal of Food Science. 45, pp. 1403-1407 Lozano J.E., Rotstein E. and Urbicain M.J., 1983: Shrinkage, porosity and bulk density of foodstuffs at changing moisture content, Journal of Food Science. 48, pp. 1497-1502 Luikov A.V., 1970: A prognosis of the development of science of drying capillary-porous colloidal materials, International Chemical Engineering. 10, pp. 599-604 Margaris D. P., Ghiaus A. G., 2006: Dried product quality improvement by air flow manipulation in tray dryers, Journal of Food Engineering, 75, pp. 542-550 Margaris D. P., Ghiaus A. G., 2007: Experimental study of hot air dehydration of Sultana grapes, Journal of Food Engineering, 79, pp. 1115-1121 Mohsenin N.N., 1980: Thermal properties of foods and agricultural materials, Gordon and Breach Sci. Publ., New York Pavese F. and Molinar G., 1992: Modern gas-based temperature and pressure measurements (International cryogenics monograph series), Book News Inc., Portland Perry R.H. and Green D.W., 1987: Perry’s chemical engineers’ handbook, 6th Ed., Mc. GrawHill Book Company, Singapore Riva M. and Peri C., 1986: Kinetics of sun and air drying of different varieties of seedless grapes, Journal of Food Technology. 21, pp. 199-208 Rizvi S.S.H., 1986: Thermodynamic properties of foods in dehydration, Ch. 4 in Engineering Properties of Foods, M.A. Rao and S.S.H. Rizvi (Ed), Marcel Dekker, Inc., New York, pp. 133-214

Fruits and Vegetables Dehydration in Tray Dryers

101

Saravacos G.D. and Charm S.E., 1962: A study of the mechanism of fruits and vegetable dehydration, Food Technology. 16, pp. 78-81 Sereno A.M. and Medeiros G.L., 1990: A simplified model for prediction of drying rates for foods, Journal of Food Engineering. 12, pp. 1-11 Singh R.P, 1982: Thermal diffusivity in food processing, Food Technol. 2, pp. 87-91 Singh R.P. and Heldman D.R., 1993: Introduction to food engineering, Academic Press, Inc., San Diego Sweat V.E., 1974: Experimental values of thermal conductivity of selected fruits and vegetables, Journal of Food Science. 39, pp. 1080-1083 Sweat V.E., 1986: Thermal properties of foods, Ch. 2 in Engineering Properties of Foods, M.A. Rao and S.S.H. Rizvi (Ed), Marcel Dekker, Inc., New York, pp. 49-87 Welt B.A., Teixeira A.A., Chau K.V., Balaban M.O. and Hintenlang D.E., 1997: Explicit finite difference method for heat transfer simulation and thermal process design, Journal of Food Science. 62, pp. 230-236 Whitakar S. and Chou W.T.H., 1984: Drying granular porous media - theory and experiments, Drying Technology. 1, pp. 3-33

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 103-135

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 3

ULTRASOUND IN FRUIT PROCESSING Sueli Rodrigues*1 and Fabiano A.N. Fernandes*2 1

Universidade Federal do Ceara, Departamento de Tecnologia dos Alimentos, Campus do Pici, Bloco 858, Caixa Postal 12168, 60455-760 Fortaleza – CE, BRAZIL 2 Universidade Federal do Ceara, Departamento de Engenharia Quimica, Campus do Pici, Bloco 709, 60455-760 Fortaleza – CE, BRAZIL

ABSTRACT Power ultrasound has been successfully employed in the chemical industry, polymer and plastic industry for many years and its use has been growing in the food industry. Power ultrasound can produce chemical, mechanical or physical effects on the processes or products where it is applied. Taking advantage of one of the effects or their combination, power ultrasound has been used in the food industry in drying, freezing, extraction processes and enzyme inactivation. The use of ultrasound in ambient fluids is well known to cause a number of physical effects (turbulence, particle agglomeration, microstreaming and biological cell rupture) as well as chemical effects (free radical formation). These effects arise mainly from the phenomenon known as cavitation. Herein a brief review of the use of ultrasound in the food industry is presented and the main applications are discussed. A comprehensive discussion on the effects of ultrasound in the tissue structure of fruits is presented along with photomicrographs of melons submitted to ultrasound. A detailed discussion is presented concerning the use of ultrasonic waves in drying, where an ultrasonic pre-treatment can be used prior to air-drying. The methods involved in the ultrasonic pre-treatment are presented along with the results obtained for several fruits such as melons, bananas, pineapples, papayas and other. Mathematical models that can be used to simulate the process are presented. Optimization of the drying process is also discussed for ultrasonic pre-treatment and ultrasound-assisted osmotic dehydration.

Keywords: Ultrasound, Fruit, Drying, Osmotic dehydration, Extraction.

* *

Sueli Rodrigues: Phone: 55-85-33669656, E-mail: [email protected] Fabiano A.N. Fernandes: Phone: 55-85-33669611, E-mail: [email protected]

104

Sueli Rodrigues and Fabiano A.N. Fernandes

1. INTRODUCTION The use of ultrasound within the food industry has been growing for many years. Up to few years ago most applications of ultrasound in the food industry were related to noninvasive monitoring of the process or food quality. In recent years, many applications of ultrasonic energy have been investigated concerning the use of ultrasound to directly affect the process or product. Virtually, power ultrasound is successfully employed to chemical industry, polymer industry, plastic welding, etc. Power ultrasound can produce chemical, mechanical or physical effects on the processes or products where it is applied (Mason, 1998). Taking advantage of one of the effects or their combination, power ultrasound has been used to improve cleaning of surface, catalyse chemical reaction, enhance drying and filtration, and accelerate the extraction of separation process. Although the benefits from the use of ultrasound in the food industry have attracted more and more attention in recent years, application of ultrasound to directly improve processes and products is less popular in food manufacturing than in other industries. The sound ranges employed in the food industry can be divided into high and low frequency. High frequency ultrasound in the MHz range has low energy and is used mainly in monitoring food quality. Low frequency ultrasound in the kHz range has high energy and can be used in food processing, and is often called low frequency power ultrasound. High intensity ultrasound uses power in the range of 10-1000 W/cm2 and can cause physical disruption of the material to which it is applied (Mason, 1998). When low frequency power ultrasound is employed, the ultrasonic waves travel through the solid medium causing a rapid series of alternative compressions and expansions, in a similar way to a sponge when it is squeezed and released repeatedly (sponge effect) (FuenteBlanco et al., 2006). In liquid medium the sonication causes cavitation which consists in the formation of bubbles in the liquid, which can explosively collapse and generate localized pressure (Wan et al., 1992; Simal et al., 1998). The rate of cavitation or alternate compressions and expansions depends on the ultrasound frequency. Extreme physical phenomena (1000 K and 500MPa) at micro-scale take place when the bubbles collapse (Suslick, 1988). The stable cavitating bubbles interact with the acoustic field generating strong micro-streaming and high shear. Cavitation studies have been restricted to the areas of heat transport, liquid tensile strength, and superheating and boiling phenomena (Apfel, 1981; Rooney, 1981). The application of power ultrasound in food enzyme inactivation has been explored in recent years and was reported by Raviyan, Zhang and Feng (2005) as a tool for inactivation of tomato pectinmethylestearase. Ultrasonication in combination with other treatment(s) is more effective in enhancing the inactivation efficacy. Ordoñez, Sanz, Hernández, and LópezLorenzo(1984) explored the effect of combining heat with power ultrasound (thermosonication) and found that the microbial inactivation of thermosonication was greater than the sum of the inactivating effects of heat and ultrasound when acting independently. De Gennaro, Caavella, Romano, and Masi (1999) also used thermosonication as a means for peroxidase inactivation. The concept of combination treatment has been further explored by introducing elevated static pressure in an ultrasound treatment chamber in a process called manothermosonication. Manothermosonication has been used to deactivate lipoxygenase

Ultrasound in Fruit Processing

105

(López and Burgos, 1995a), peroxidase (López and Burgos, 1995b), lipase and protease (Vercet, López, and Burgos, 1997), and tomato or orange pectinmethylestearase (Kuldiloke, 2002; López, Vercet, Sánchez, and Burgos, 1998; Vercet, López, and Burgos, 1999; Vercet, Sánchez, Burgos, Montanes, and López-Buesa, 2002), all with an increased inactivation. For example, López et al. (1998) reported that the D-values of tomato pectinmethylestearase at 62.5oC were reduced 53-fold, from 45 min in thermal treatments to 0.85 min by manothermosonication. The inactivation effect of ultrasound is attributable mainly to the cavitation promoted by the sonication. Raviayan, Zhang and Feng (2005) carried out ultrasonic inactivation tests with tomato pectinmethylesterase at 50, 61, and 72oC. Thermal only tests at 50, 61, and 72oC were also conducted to delineate possible additive or synergistic effects. In thermal inactivation tests, the reduction in pectinmethylestearase residual activity at 50oC was negligible while D-values at 61 and 72oC were 299.0 and 25.3 min. D-values varied from 24.0 to 240.6 min in sonication tests but were reduced to 0.3min in the thermosonication test at 72oC. Compared to the enzyme inactivation test at 61oC, thermosonication at the same temperature increased the inactivation by 39 to 374-fold, while at 72oC the increase was 36–84-fold, depending upon cavitation intensity. The authors reposted a strong synergistic effect in the thermosonication tests. The increase in the inactivation was more pronounced at low temperatures. Sonication can also be used to destroy microorganisms in foods. Many authors have reported that cavitation caused by sonication is responsible for dispersion of clumps of microorganisms, modification of the cellular activity, puncturing of the cell wall and increasing sensitivity to heat. However, the lethal effect of ultrasound is not the same for all microorganisms. Generally the ultrasound treatment is not much effective on small round cells (Allinger, 1975). For example, Gram-positive bacteria, such as Staphylococcus aureus and enterococci (Ordonez, Sanz, Hernandez and Lopez-Lorenzo, 1984) are quite resistant to ultrasound application. Bacterial spores are even more resistant (Ahmed and Russell, 1975; Boucher and Lechowich, 1979). The combined effect of ultrasonic waves and heat treatment applied simultaneously appears to be very effective (Ciccolini, Taillandier, Wilhem, Delmas and Strehaiano, 1997) and even more effective combining heat and ultrasound under pressure (Raso, Condon and Sala Trepat, 1994). The same combination of effects has been demonstrated to have a quite effective deactivating action on various enzymes of interest in food technology (López et al., 1994; López and Burgos, 1995a). Although the possibility of deactivating enzymes or destroying microorganisms by ultrasound waves, alone or in combination with other physical treatments, has been widely used for laboratory applications in microbiology, immunology and enzymology, the same is not true for industrial applications. De Gennaro, Caavella, Romano and Masi (1999) studied the effect of sonication on peroxidase inactivation. The combined effect of high power ultrasound and temperature on the activity of peroxidase type VI from horseradish suspended in water was studied. The tests were performed at 80°C using power ultrasounds having frequencies of 20, 40, and 60 kHz. Ultrasonic powers varied in the range from 0 to 120 W. Combined treatments were carried out by using a laboratory scale plant operating in continuous and in batch mode. In continuous experiments, 46 mL of suspension was circulated in the plant having a sanitation chamber of 20 mL. In the other case, two sets of experiments were carried out by using a sanitation chamber of 100 mL treating 40 or 80 mL of suspensions, respectively. It was found that the decimal reduction time of peroxidase at 80°C, D80, reduces from 65 to 10 min ca.

106

Sueli Rodrigues and Fabiano A.N. Fernandes

when ultrasounds are applied. In particular, D80 varies with ultrasonic power, sonotrode geometry and volume of suspensions submitted to the treatment. Freezing is an excellent method for long preservation of food products and has gained widespread attention. Generally speaking, freezing preservation provides food products with better taste, texture and nutritional value than any other preservation methods. In recent years, increased ownership of domestic freezers and microwave ovens has boosted rapid increases in sales of frozen food (Fellows, 2000). Current challenges lie on production efficiency and high product quality. However, the drawback of the presence of large ice crystals within the food which results in reduction in eating quality inherently exists and has stalled the development of freezing preservation. More and more research (Bustabad, 1999; Martino, Otero, Sanz, and Zaritzky, 1998; Ngapo, Babare, Reynolds, and Mawson, 1999) manifested that the quality of frozen food is closely related to the size and location of ice crystals formed, which in turn largely depends on the rate of freezing and final temperature in freezing process. The freezing rate governs the formation of ice crystals within frozen products. Rapid freezing produces small intracellular ice crystals, while slow freezing forms large ice crystals. Large ice crystals would cause damages to food quality including appearance, sensory properties, textual attributes and nutritional value. Furthermore, the cellular structure of biologically originated foodstuffs would be ruptured with the result of drip loss on thawing. Extensive research has been carried out to reduce the size of ice crystals on frozen products. This can be achieved by very high freezing rate, forming small and even ice crystals throughout the foodstuffs. For this purpose, freezing methods capable of ensuring frozen food of the maximum final quality have been developed. Air freezing, contact freezing, immersion freezing, cryogenic freezing and their combinations are the most common methods used in freezing industry to obtain optimum freezing rate for food products. New methods are being proposed and developed mainly in laboratory scale, e.g. the high-pressure-assisted freezing (Kalichevsky, Knorr, and Lillford, 1995; Otero, Martino, Zaritzky, Solas, and Sanz, 2000; Sanz, Paklacios, Lopez, and Ordonez, 1985), or dehydrofreezing (Spiazzi, Raggio, Bignone, and Mascheroni, 1998). According to Li and Sun (2002) power ultrasound has proved to be useful in controlling crystallisation process. It plays effective roles in the initiation of nuclei and subsequent crystal growth. Under the influence of power ultrasound, much more rapid and even nucleation occurs. In freezing, this would lead to fine ice crystals and shortening the time between the onset of crystallisation and the complete formation of ice, thus reducing damages to cellular structure. This may be mostly due to acoustic cavitation, which consists of the formation, growth and violent collapse of small bubbles or voids in liquids (Simal, Benekito, Sanchez, and Rossello, 1998). The cavitation bubbles can act as nuclei for crystal growth, and/or the nuclei already present would be fragmented into smaller ones caused by the strong forces originating from the collapse of cavitation bubbles (Mason, 1998). Furthermore, other studies (Lima and Sastry, 1990; Sastry, Shen, and Blaisdell, 1989) have shown that ultrasound would effectively improve convective heat transfer coefficient. Enhancing the heat transfer coefficient between the food products being frozen and the refrigerating medium is one of the most useful measures for rapid removal sensible heat and latent heat of fusion so as to intensify the freezing process. Power ultrasound is promising to be applied in freezing with its direct effects on heat transfer and crystallization process. Li and Sun (2002) studied the effect of power ultrasound on freezing rate of potatoes. The effect of power ultrasound on freezing rate was influenced by ultrasound power, exposure time and the freezing phase to which ultrasound was applied. However, the ultrasound power

Ultrasound in Fruit Processing

107

and exposure time should be chosen with the consideration of thermal effect of ultrasound. In this study, the freezing rate was improved greatly when ultrasound (15.85 W) was applied for 2 min. When ultrasound was applied to the phase change period during freezing process, the freezing rate was increased significantly. The mechanical and physical effects of sound can be used to enhance many processes where mass transfer takes place. Thus, this mechanism can be of great relevance to drying, dewatering and extraction processes. Ultrasound application may also change the viscosity, surface tension and deform porous solid materials (Tarleton, 1992).

2. EFFECT OF ULTRASOUND IN THE FRUIT TISSUE The effect of ultrasound application on the cellular structure can be studied by light microscopy analysis. To show the effect of ultrasound over the cellular structure and the changes that occur with time will be shown and compared with the effects produced by osmotic dehydration (other popular pre-treatment).

Light Microscopic Analysis At the end the pre-treatment a cube sample has to be carefully cut (0.005 m average side). The sample cubes are fixed with 4% solution of paraformaldehyde in 0.1M phosphate buffer, pH 7.2 and 1% glutaraldehyde for 24h at ambient temperature (Karnovisk, 1965). The material is then dehydrated in a graded ethanol series (from 10% to 100%) and embedded in Historesin embedding kit (Jung). To prepare the laminas the tissue blocks should be sectioned at 5 to 10 µm on a microtome. Sectioning the blocks at 5 μm may result in section too thin that may be difficult to see under the microscope. Sectioning the blocks at a thickness over 10 μm can result in superposition of cells. Histochemical reaction can be carried out with Toluidine Blue (TB) at pH 4.0 as metachromatic stain to detect polyanionic pectins (Vidal 1977). The photomicrographs of the cell structure shown herein are for melons submitted to ultrasound and osmotic dehydration and were taken using an Olympus BX51 light microscope with digital image capture system and the blocks were sectioned at 8 μm.

Effect of Ultrasound in Fruit Tissue The microscopic image analysis of fresh melon showed typical thin-walled cells with normal morphology and some intercellular spaces (figure 1). Very few cells had undulated walls.

108

Sueli Rodrigues and Fabiano A.N. Fernandes

Figure 1. Photomicrographs of melon cubes before processing (raw fruit). Magnification of 380x.

At the end of the ultrasonic pre-treatment little change was observed in the fruit moisture content when submitted to ultrasound in a distilled water bath. The fruit incorporated water after 20 minutes submitted to ultrasound increasing by 8.7% its water content. Microscopic image analysis showed two distinct regions, one region where the cells became bloated as a result of water gain (figure 2A) and a second region where the cells became much smaller and needle shaped (figure 2B).

Figure 2. Photomicrographs of melon cubes after 20 minutes of ultrasound pre-treatment. A. Region with bloated cells and B. Region with microscopic channels. Magnification of 380x.

Figure 2B shows that microscopic channels are formed by the elongation and flattening of cells in some regions of the sample. These needle shaped cells were not observed in the fresh fruit and were observed in all samples submitted to ultrasonic pre-treatment. Some

Ultrasound in Fruit Processing

109

researchers have reported that ultrasonic waves may create microscopic channels (Tarleton, 1992; Tarleton and Wakeman, 1998; Fuente-Blanco, Sarabia, Acosta-Aparicio, BlancoBlanco, and Gallego-Juárez, 2006), but did not show any microscopic image. After 30 minutes of ultrasonic treatment the microscopic channels became broader (figure 3B) and some new smaller microscopic channels appeared. The region with bloated cells was still observed (figure 3A). The broadening of the microscopic channels explain the further increase in the effective water diffusivity of the fruit observed during air-drying process and the increase in the effective diffusivity of pigments and soluble solids observed in extraction processes. There is no indication of loss of strength of the cell wall and during knife cutting the cells at the border did not fail under stress as happens when the fruit is submitted to osmotic dehydration. No cell breakdown was observed in the samples submitted to low frequency ultrasound.

Figure 3. Photomicrographs of melon cubes after 30 minutes of ultrasound pre-treatment. A. Region with bloated cells and B. Region with microscopic channels. Magnification of 380x.

The effective diffusivity in fruits is dependent on tissue structure and on the fruit porosity. The cell walls act as a semi-permeable membrane inferring in a resistance to diffusion of water and solids. An increase in porosity increases the void space for water to diffuse and decreases resistance for diffusion. The use of ultrasound increased the effective water diffusivity of melons, bananas and sapota during the air-drying process reducing the time required for drying, confirming the observations of Fuente-Blanco et al. (2006) that the ultrasonic pre-treatment affects the fruit tissue making easier for the water to diffuse during air-drying and showed that the microscopic channels may contribute with the higher effective water diffusivity. Other pre-treatment used in drying is the osmotic dehydration, which presents a completely different effect over the fruit tissue. After 30 minutes under osmotic treatment several differences are observed in the cell structure due to the high water loss. During osmotic dehydration the cell walls become distorted and smaller throughout of the samples. In some regions the junctions between adjacent cells are still present but with reduced intercellular spaces (figure 4A) while in other regions disruption of the cells are observed causing an increase of intercellular spaces which occur due to the solubilizing of chelatorsoluble pectin of the middle lamella (figure 4B) (Prinzivalli et al., 2006). Chelator-soluble pectin is the substance that most contributes to cell adhesion and firmness and according to the microscopic images the chelator-soluble pectin partially solubilize at the early stages of

110

Sueli Rodrigues and Fabiano A.N. Fernandes

osmotic dehydration. During ultrasound application, no increase of intercellular spaces was observed or has been reported in the literature.

Figure 4. Photomicrographs of melon cubes after 30 minutes of osmotic dehydration. A. Region with distorted cells and B. Region with disrupted cells. Magnification of 380x.

After 1 hour of osmotic treatment most intercellular spaces between the cells disappear and the cell walls become even more distorted (figure 5A). In some regions of the samples cell wall breakdown starts to appear (figure 5B) although still rare. The cell wall strength diminishes, due to solubilizing of pectin, combined with high osmotic pressure and flow of sugar molecules into the fruit that contribute to the breakdown of cell walls. Pectin solubilizing during osmotic dehydration was shown by Prinzivalli et al. (2006).

Figure 5. Photomicrographs of melon cubes after 1 hour of osmotic dehydration. A. Region with distorted cells and B. Region with cells in initial stage of cell breakdown. Magnification of 380x.

A greater change is observed after 2 hours of osmotic treatment when most cell walls are broken-down and the few intact cells have severely distorted walls (figure 6). Solubilizing of pectin can be visually noted by decreasing cell wall strength.

Ultrasound in Fruit Processing

111

Figure 6. Photomicrographs of melon cubes after 2 hours of osmotic dehydration showing cells in advanced stage of cell breakdown. Magnification of 380x.

The water diffusivity in the air-drying process may also increase with the use of the osmotic dehydration pre-treatment, for some fruits and vegetable, when carried out for more than one hour. This increase is explained by the collapse of cell wall membranes, which makes easier for water to diffusive within the vegetable. In this case a good agreement was obtained between cellular structure changes and water diffusivity (Rodrigues and Fernandes, 2007a). Water diffusivity may decrease in the beginning of the osmotic dehydration process due to the incorporation of sugar, but increase when the osmotic dehydration is carried out for more than one hour due to the breakdown of cells which lowers the resistance to water diffusion. Ultrasound also induces changes on the cell structure, but different from osmotic dehydration, no cell breakdown is observed and the increase in diffusivity is attained by the formation of microscopic channels in the cell structure which also offer lower resistance to diffusion of water, pigments and soluble solids.

3. APPLICATION OF ULTRASOUND IN DRYING PROCESSES Several fruits and vegetables are produced in large quantities, especially in tropical countries, aiming mainly exportation (banana, pineapple, orange, melon, mango and several others). Some fruits, like oranges and lemons, are cropped mainly to produce fruit juices, other are cropped aiming direct consumption or to be used in foodstuffs (cakes, jams, yogurt, and several others). Most fruit are extremely perishable and do not allow the use of freezing for their conservation. Drying can be applied to reduce post-harvest loss of fruits and also to produce dehydrated fruit used in foodstuffs formulations, providing an extension of shelf-life, lighter weight for transportation and less space for storage. Many vegetables are used in

112

Sueli Rodrigues and Fabiano A.N. Fernandes

dehydrated foodstuffs (sauces, food mixtures, soups and others) and benefit from drying to extend their shelf-life. Drying is the most common method of food preservation. Conventional air-drying is a simultaneous heat and mass transfer process, accompanied by phase change (Barbanti, Mastrocola and Severine, 1994). Drying fruits and vegetables through air-drying is done at mild temperatures (between 40ºC to 70ºC) to reduce the degradation of the fruit or vegetable. The process takes from 8 to 24 hours depending on the vegetable that is being processed, on its initial moisture content, on the desired final moisture content and on the temperature used in the process. Temperature is maintained by direct heating of the air passing through the samples making the air-drying process heat intensive, thus expensive.

Air-Drying Air-drying comprehends two periods: a heat transfer-controlled mass transfer period (or constant rate period) and a diffusion-controlled mass transfer period (or falling rate period). In the heat transfer-controlled mass transfer period the water removal rate is at its maxima, but in many vegetable this period is generally very short compared to the diffusion-controlled mass transfer period. Air-drying is generally modeled assuming diffusion-controlled mass transfer with liquid flowing within the fruit conforming to Fick’s second law of diffusion. The equation used to model the falling-rate period is a simplification of Fick’s second law considering long drying times (Perry and Green, 1999).

dH 2π = − 2 ⋅ D ⋅ (H − H eq ) dt δ

(1)

According to the equation, reduction of air-drying time can be achieved by reducing the initial moisture content (H0) of the fruit sent to dry or by increasing the effective water diffusivity (D). Pre-treatments can be used to reduce the initial water content or to modify the fruit tissue structure in a way that the effective water diffusivity increases (Cao, Zhang, Mujumdar, Du, and Sun, 2006; Pan, Zhao, Zhang, Chen, and Mujumdar, 2003; Madamba, and Lopez, 2002)

3.1. Ultrasound Pre-Treatment The forces involved by the sponge effect caused by ultrasonic waves can be higher than surface tension which maintains the moisture inside the capillaries of the fruit creating microscopic channels which may ease moisture removal. These microscopic channels can be used by water molecules as a preferential pathway to diffuse towards the surface of the fruit or vegetable increasing its effective water diffusivity. The ultrasonic waves also reduce the diffusion boundary layer and increase the convective mass transfer in the sample (Tarleton, 1992; Tarleton and Wakeman, 1998; Fuente-Blanco, Sarabia, Acosta-Aparicio, Blanco-

Ultrasound in Fruit Processing

113

Blanco, and Gallego-Juárez, 2006). In addition, the cavitations produced by ultrasound are beneficial for the removal of moisture strongly attached. The pre-treatment technique is simple and involves the immersion of the fruit or vegetable in water to which ultrasound is applied. The treatment can be carried out at ambient temperature and no heating is required, reducing the probability of food degradation (Mason, 1998).

Preparation of Samples Fruit or vegetable samples should be cut to obtain cubes, cylinders or any other kind of slices. To calculate water and sugar loss, the moisture content of the sample has to be determined by heating in a drying oven at 105oC for 48h, according to AOAC method (AOAC, 1990). The initial soluble solids content of the fruit (ºBrix) can be determined by refractometry.

Ultrasound Pre-Treatment The samples should be immersed in distilled water and submitted to ultrasonic waves during a period of time. Recent studies (Fernandes and Rodrigues, 2007; Rodrigues and Fernandes, 2007b) showed that the application of ultrasound should be carried out for at least 10 minutes. After 20 minutes the changes on the effective water diffusivity become slight and after 30 minutes the changes are insignificant. The process can be carried out under ambient temperature, since higher temperatures do not show to enhance the effect of ultrasound (Simal, Benedito, Sánchez, and Rosselló, 1998). The process can be carried out placing the samples in an ultrasonic bath or placing the samples in a vessel and using an ultrasonic tip to induce ultrasonic waves inside the vessel. No mechanical agitation is required. The ultrasound frequency should be between 25 and 100 kHz and intensity above 4000 W/m2. Temperature increase is slight not exceeding 3ºC during 30 minutes run (Fernandes and Rodrigues, 2007; Rodrigues and Fernandes, 2007b). The water to fruit ratio should be maintained low. Water to fruit ratios from 3:1 to 4:1 (weight basis) is preferred because the area of the required vessel (or ultrasonic bath) can be small. Higher water to fruit ratios requires larger vessels and more powerful ultrasound to generate an ultrasonic intensity above 4000 W/m2, increasing power consumption. Fernandes and Rodrigues (2007) showed that at this water to fruit ratio the increase in soluble solids content of the liquid medium was slight (less than 2.0 g solids/L for melons, bananas and pineapples). To calculate water loss during the ultrasonic pre-treatment the moisture content of the fruit has to be determined by heating in a drying oven at 105oC for 48h, according to AOAC method (AOAC, 1990). To calculate the sugar loss during the process two options can be used. The soluble solids content of the fruit (ºBrix) can to be directly determined by refractometry. Or the sugar content of the liquid medium can be determined using colorimetric methods or HPLC methods and then the soluble solids of the fruit can be calculated by mass balance:

114

Sueli Rodrigues and Fabiano A.N. Fernandes

SSFR = SS0FR − SLM

(2)

where, SLM is the sugar content of the liquid medium, SSFR is the soluble solids content of the fruit, and SSFR0 is the initial soluble solids content of the fruit.

Air-Drying At the end of the pre-treatment the dehydrated samples are drained, blotted with absorbent paper to remove the excess solution and transferred to an air-dryer, which can be a forced circulating air-drying oven, a fixed-bed air-dryer or a solar dryer. The air temperature should be as high as possible but not so high as to spoil the vegetable. Temperatures between 40 to 60ºC are usually applied in air-drying of fruits and vegetables.

Weight Reduction, Water Loss and Sugar Gain During the Pre-Treatment The performance of drying pre-treatments is commonly evaluated by the parameters: weight reduction (WR), water loss (WL) and sugar gain (SG). Water loss is the main parameter and indicates the amount of water removed during the pre-treatment in relation to the initial weight of the samples. Sugar gain is an important parameter that indicates the amount of soluble solids that are incorporated by the sample during the pre-treatment. High sugar gain values may alter the sweetness of the vegetable, also increasing the calories associated with the product. Negative sugar gain indicates that the pre-treatment was able to remove soluble solids from the vegetable. Generally the performance of the pre-treatment will be satisfactory when high water loss is associated with low sugar gain. Weight, moisture content and soluble solids content of the fruit samples are used to calculate weight reduction (WR), water loss (WL) and solid gain (SG), according to the following equations:

WR(%) =

wi − wf ⋅ 100 wi

WL(%) =

( w i ⋅ Xi − w f ⋅ X f ) ⋅ 100 wi

SG (%) =

[w f ⋅ (1 − X sf ) − w i ⋅ (1 − X si )] ⋅ 100 wi

(3)

(4)

(5)

where, Xi is the initial fruit moisture on wet basis [g water/g], Xf is the final fruit moisture on wet basis [g water/g], Xsi is the initial fruit soluble solid content [g solid/g], Xsf is the final fruit soluble solid content [g solid/g], wi is the initial fruit mass [g], and wf is the final fruit mass [g].

Ultrasound in Fruit Processing

115

The calculation of sugar gain based on the initial weight of the samples may be misleading when analyzing the process because it generally results in a small value (up to 10% in most cases). A fast analysis of the sugar gain value may be interpreted that a reduced amount of sugar was incorporated by the vegetable, whereas in some cases the amount of sugar in the sample doubles or even triples, having a direct impact in the sensory characteristics of the vegetable and on the calories associated to it. To proper evaluate the incorporation of soluble solids into the sample, the sugar gain can be calculated based on the initial soluble solids content rather than on the initial weight of the sample:

SG (%) =

[w f ⋅ (1 − X sf ) − w i ⋅ (1 − X si )] ⋅ 100 w i ⋅ (1 − X si )

(6)

At the end of the ultrasonic pre-treatment little change is observed in the fruit moisture content and the effect of the ultrasonic pre-treatment on drying is mainly observed during the air-drying stage where a significant increase in the effective water diffusivity is found. Different from the use of osmotic dehydration where an expressive water loss is observed, when the ultrasonic pre-treatment is applied each fruit display a different behavior, gaining or losing water during the pre-treatment. Bananas have gained water during the treatment. For ultrasound treatments lasting 20 minutes, water gain was 11.1%. This value decreased to 7.2% when the fruit was submitted to ultrasound for 30 minutes (table 1) (Fernandes and Rodrigues, 2007). The same behavior was observed with sapotas (Achras sapota L.) which also gained water during the ultrasonic treatment. Melons, papayas, pineapples and jenipapo (Genipa americana L.) lost a small amount of water during the pretreatment. For ultrasound treatments lasting 20 minutes the water loss observed in pre-treated melons was 1.3% increasing to 5.5% when the fruit was submitted to ultrasound for 30 minutes (table 1) (Rodrigues and Fernandes, 2007b). At the end of the ultrasonic pretreatment, little change was observed in the moisture content of pineapples which lost between 2.1 to 3.2% of its initial water (table 1). All fruit submitted to ultrasonic pre-treatment lost soluble solids to the liquid medium and the amount of soluble solids lost depended on the fruit characteristics (table 1). The loss of sugars occurs because of the different sugar concentration (osmotic pressure) between the fruit and the liquid medium, which favors a mass transfer of sugar from the fruit to the liquid medium and a mass transfer of water from the liquid medium to the fruit. The amount of sugars lost from bananas after 30 minutes in ultrasonic bath was 21.3% of the reducing sugars. The loss of glucose was 11.0% after 30 minutes in ultrasonic bath. Figure 7 shows the sugar loss as a function of the time spent in ultrasonic bath. In fruits with high initial soluble solids content, a reduction in water gain may be observed after 10 or 20 minutes. For bananas a reduction in water gain was observed after 20 minutes under ultrasound application compared to the results found for water gain after 10 minutes (Fernandes and Rodrigues, 2007). This phenomenon is explained by the higher sugar loss from the fruit to the liquid medium found after 20 minutes, which diminishes the solid concentration gradient between the fruit and the liquid medium and as a consequence less water enters the fruit to compensate for the osmotic pressure gradient. As a consequence of the water gain and the loss of soluble solids, the moisture content of the fruit after the ultrasonic bath increases (figure 8).

116

Sueli Rodrigues and Fabiano A.N. Fernandes

Table 1. Water loss and sugar gain of fruits submitted to ultrasound pre-treatment Operating condition

Sugar Gain

Water Loss

Banana (Musa spp.) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 11.0% ± 2.2 - 12.1% ± 0.2 - 21.3% ± 1.2

- 4.1% ± 0.9 - 11.1% ± 0.5 - 7.2% ± 0.9

Melon (Curcumis melo L.) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 21.1 ± 0.13 - 44.7 ± 2.0 - 52.2 ± 1.0

+ 1.3 ± 0.2 + 0.8 ± 0.6 + 5.5 ± 2.3

Pineapple (Ananas comosus) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 15.3% ± 2.2 - 15.8% ± 3.9 - 18.9% ± 1.8

+ 3.2% ± 0.6 + 2.1% ± 0.6 + 3.1% ± 0.8

Sapota (Achras sapota L.) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 1.6% ± 0.2 - 1.6% ± 0.3 - 3.8% ± 0.2

- 7.4% ± 0.8 - 3.3% ± 0.4 - 6.6% ± 0.3

Papaya (Carica papaya) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 6.7% ± 1.6 - 2.4% ± 0.3 - 1.5% ± 0.5

- 3.0% ± 0.7 - 8.4% ± 1.3 - 5.7% ± 1.3

Jenipapo (Genipa americana L.) After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment

- 12.2% ± 2.8 - 16.6% ± 3.6 - 8.2% ± 0.9

- 11.2% ± 2.9 - 9.5% ± 1.2 - 14.9% ± 1.4

Melons showed the highest loss of soluble solids to the liquid medium. The amount of sugars transferred to the liquid medium during the process was 52% after 30 minutes in ultrasonic bath. The amount of glucose lost was 24% after 30 minutes in ultrasonic bath (Rodrigues and Fernandes, 2007b). The lowest reduction in soluble solids content to the liquid medium was observed for sapota (Achras sapota L.) which lost only 3.8% of its initial sugar content. The result observed for sapota may be explained by the high starch content of this fruit (up to 34.6% of the total carbohydrate content of the fruit).

Ultrasound in Fruit Processing

117

Figure 7. Reducing sugars and glucose loss as a function of time in ultrasound bath for bananas. Error bars represent the standard deviation.

Figure 8. Moisture content of the fruit as a function of time in ultrasound bath for bananas. Error bars represent the standard deviation.

118

Sueli Rodrigues and Fabiano A.N. Fernandes

Studies on osmotic dehydration showed that melons gained up to 185% of sugar after 180 minutes in osmotic solution (55ºBrix, 65ºC), pineapples gained up to 120% of sugar after 180 minutes in osmotic solution (70ºBrix, 30ºC) and bananas gained up to 220% of sugar after 240 minutes in osmotic solution (70ºBrix, 60ºC) (Parkojo, Rahman, Buckle and Perera, 1996; Rastogi, and Raghavarao, 1997; Fernandes, Rodrigues, Gasparetto, and Oliveira, 2006; Teles et al., 2006; Rodrigues, and Fernandes, 2007a). As such the ultrasonic treatment is an interesting process to produce low sugar dried fruits, since during ultrasound pre-treatment the fruit losses sugar whereas under osmotic dehydration the incorporation of sugar is very high. The sugar loss showed to be related to the porosity of the samples. Melons have high porosity and showed the highest sugar loss to the liquid medium, while fruit with lower porosity showed lower loss of sugar. Other interesting finding is that fruits with high starch content such as sapotas and bananas showed to incorporate water during the ultrasonic pretreatment while fruits with low starch content showed to lose water to the liquid medium.

Effective Water Diffusivity During Air-Drying The effect of the ultrasonic pre-treatment on drying is mainly observed during the airdrying stage where a significant increase in effective water diffusivity is observed. Experimental data is required to estimate the effective diffusion coefficient of the airdrying process. The effective water diffusivity can be adjusted using Equation (4) with a parameter estimation procedure based on the minimization of the error sum of squares, where the mathematical model is solved by numerical integration using the Runge-Kutta method. Effective water diffusivity was affected by ultrasound application in some fruits, but no effect was observed for pineapple (table 2). As a consequence of the higher effective water diffusivity the fruit submitted ultrasound pre-treatment will dry faster during the air-drying stage if compared to the fresh fruit with no pre-treatment (figure 9). These results corroborates with the observations of Fuente-Blanco et al. (2006) and Fernandes, Gallão and Rodrigues (2007) that the ultrasonic pre-treatment affects the fruit tissue making easier for water to diffuse during air-drying, most probably due to the formation of microscopic channels in the fruit.

Ultrasound in Fruit Processing

119

Figure 9. Development of the moisture content during air-drying as a function of processing time for fresh bananas (no pre-treatment) and for bananas submitted to 30 minutes of ultrasound pre-treatment.

Table 2. Effective water diffusivity of fruits submitted to ultrasound pre-treatment Operating condition Banana (Musa spp.) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment Melon (Curcumis melo L.) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment Pineapple (Ananas comosus) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment Sapota (Achras sapota L.) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment Papaya (Carica papaya) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment After 30 minutes of ultrasound treatment Jenipapo (Genipa americana L.) Fresh fruit After 10 minutes of ultrasound treatment After 20 minutes of ultrasound treatment

Effective Diffusivity [m2/s]

Regression R2

1.28 .10-9 1.08 .10-9 1.47 .10-9 1.41 .10-9

0.98 0.99 0.99 0.97

5.00 .10-9 6.50 .10-9 6.42 .10-9 6.97 .10-9

0.995 0.989 0.996 0.998

4.97.10-10 4.92.10-10 5.00.10-10 4.89.10-10

0.999 0.999 0.999 0.994

7.31.10-9 7.09.10-9 7.43.10-9 7.12.10-9

0.998 0.997 0.994 0.995

5.90.10-9 5.66.10-9 5.72.10-9 5.76.10-9

0.997 0.995 0.996 0.996

1.41.10-8 1.49.10-8 1.56.10-8

0.999 0.997 0.994

120

Sueli Rodrigues and Fabiano A.N. Fernandes

Greater changes on effective water diffusivity during the air-drying process was found for melons pre-treated for 30 minutes under ultrasonic waves (6.97 .10-9 m2.s-1). The observed effective water diffusivity was 39.4% higher than the diffusivity of the fresh fruit (5.00 .10-9 m2.s-1) and 35.9% higher than the diffusivity of osmo-dehydrated melons (5.13 .10-9 m2.s-1). This higher effective water diffusivity observed for melons may be related to the fruit tissue structure and the high moisture content of the fruit which may facilitate the formation of micro-channels. Bananas also showed to be influenced by the ultrasonic treatment and the effective water diffusivity during the air-drying process was found to be higher when the bananas where pretreated for 20 minutes under ultrasonic waves (1.47 .10-9 m2.s-1). Pre-treating the bananas using osmotic dehydration resulted in a water diffusivity of 1.37 .10-9 m2.s-1. Both diffusivities were higher than the diffusivity obtained for the fresh fruit (1.28 .10-9 m2.s-1) during air-drying (Fernandes, Rodrigues, Gasparetto, and Oliveira, 2006). The observed effective water diffusivity after application of ultrasound was 14.8% higher than the diffusivity of the fresh fruit and 7.3% higher than the diffusivity of osmo-dehydrated bananas. This enhancement observed in the effective water diffusivity for bananas was lower than for melons and may be related to the higher density and the lower moisture content of bananas which may result in fewer micro-channels or in micro-channels only near the surface of the fruit. This theory still has to be checked by microphotograph analysis. The ultrasound pre-treatment affected slightly the effective water diffusivity of pineapples during the air-drying process (table 2). These differences show that the effect of ultrasound may be dependent on the structure of the fruit cell tissue. Further studies with light microscopy will be carried out with these fruits to evaluate the effect of ultrasound in the cell structure and why some fruit are more affected than others regarding water loss, sugar gain and effective water diffusivity. There still no phenomenological mathematical model to describe the changes in effective water diffusivity as a function of the time spent in ultrasound and only an empirical correlation can be found. For bananas the effective water diffusivity can be correlated by Eq. (7):

(

)

D = 7.69 − 0.0085 ⋅ t U + 0.0073 ⋅ t U 2 − 0.000205 ⋅ t U 3 ⋅10 −8

(7)

where, D is the effective water diffusivity [m2/s] and tU is the time spent under ultrasonic pretreatment [min].

Processing Time (Pre-Treatment + Air-Drying) The main objective of a pre-treatment is to reduce the total processing time, specially the air-drying time. Table 3 shows that when ultrasonic pre-treatment was used the total processing time (air-drying + ultrasound) was lower than when the fresh fruit was air-dried without any pre-treatment, showing that the ultrasound pre-treatment is efficient in reducing the total drying time, except for pineapples where no effect was observed. When ultrasound was applied to bananas during 20 minutes the total processing time reduced by 86 minutes (a 10.3% reduction) for bananas dehydrated to a final moisture content

Ultrasound in Fruit Processing

121

of 0.05 g water/g dry solids (Fernandes and Rodrigues, 2007). For melons the reduction in total processing time was even higher and when ultrasound is applied for 30 minutes, the total processing time to dehydrate melons to a final moisture content of 0.5 gwater/gdry solids will be reduced by 200 minutes (a 24.6% reduction). These reductions are considerable and able to reduce the process cost. The use of ultrasound with pineapples, however, showed an increase in total processing time and therefore this pre-treatment is not suitable for this fruit. Table 3. Total processing time (pre-treatment + air-drying) to achieve a moisture content of 0.05 g of water per gram of dry solids for bananas and melons and 1.25 g of water per gram of dry solids for pineapples Total Pre-treatment Time Time [min] [min] Bananas – Final moisture content = 0.05 g of water / g of dry solids No pre-treatment (air-drying only) 839 --10 minutes of ultrasound (distilled water) 803 10 20 minutes of ultrasound (distilled water) 753 20 30 minutes of ultrasound (distilled water) 791 30 Melons – Final moisture content = 0.05 g of water / g of dry solids No pre-treatment (air-drying only) 814 --10 minutes of ultrasound (distilled water) 636 10 20 minutes of ultrasound (distilled water) 654 20 30 minutes of ultrasound (distilled water) 614 30 Pineapples – Final moisture content = 1.25 g of water / g of dry solids No pre-treatment (air-drying only) 473 --10 minutes of ultrasound (distilled water) 473 10 20 minutes of ultrasound (distilled water) 513 20 30 minutes of ultrasound (distilled water) 550 30

Air-Drying Time [min] 839 793 733 761 814 626 634 584 473 463 493 520

Optimization The drying process of fruits and vegetables considered herein comprehends the pretreatment process followed by air-drying. Total processing time can be optimized to reduce the drying process to a minimum, which can reduce costs and increase the overall productivity. To optimize the process consisting of ultrasonic treatment and air-drying, the ultrasonic treatment should be carried out while the increase in water diffusivity it provokes leading to a continuous reduction of total processing time (figure 10).

122

Sueli Rodrigues and Fabiano A.N. Fernandes

Figure 10. Illustrative scheme of the development of water diffusivity of the fruit and total processing time as a function of time submitted to ultrasound. The arrow indicates the optimum period that the bananas should be submitted to ultrasonic bath prior to air-drying to minimize the total processing time.

The optimization can be carried out using the method of Levenberg-Marquardt or equivalent optimization method, setting the objective function to minimize the total processing time. It is important to validate the mathematical models used in the optimization procedure by using the F-test as a criterion to validate the models. The level of confidence of the model is established comparing the listed F-values and the calculated F-values for each operating condition. Levels of confidence above 90% are desired. After validation, the model can be used to optimize the total processing time to dry the fruit by ultrasound followed by air-drying. Table 4 shows the best operating conditions that should be used in the processing of bananas and melons when submitted to the ultrasonic pre-treatment using distilled water as the liquid medium. Table 4. Optimal processing times (pre-treatment + air-drying) for banana and melon processing Total Pre-treatment Time Time [min] [min] Bananas – Final moisture content = 0.05 g of water / g of dry solids 748 22 Melons – Final moisture content = 0.05 g of water / g of dry solids 614 30

Air-Drying Time [min] 726 584

3.2. Ultrasound-Assisted Osmotic Dehydration Osmotic dehydration is the most reported pre-treatment used prior to air-drying. The technique consists in immersing the fruit in hypertonic solution to partially remove water from the fruit. The driving force for water removal is the difference in osmotic pressure between the fruit and the solution where the complex cellular structure of the fruit acts as a

Ultrasound in Fruit Processing

123

semi-permeable membrane (Raoult, Lafont, Rios, and Guilbert, 1989; Torreggiani, 1993; Fito, 1994; Raoult-Wack, 1994; Prokharkar, Prasad, and Das, 1997; Rastogi and Raghavarao, 1997; Rastogi and Niranjan, 1998; Simal, Benedito, Sánchez, and Rosello, 1998; Rastogi, Eshtisghi, and Knorr, 1999; Corzo and Gómez, 2004). Ultrasound can be used to enhance water removal during osmotic dehydration, benefiting from the microscopic channels created by the application of ultrasound which may ease moisture removal.

Preparation of Samples The preparation of fruit or vegetable samples should be done using the same procedure as described earlier for ultrasound pre-treatment (Section 3.1).

Osmotic Solution The osmotic solution should be prepared by mixing food grade sucrose with distilled water to give the desired sugar concentration. Other sugars such as mannitol, fructose, glucose and others can also be used in substitution to sucrose. Most studies and industrial processes use osmotic solutions ranging from 30 to 70ºBrix. As the sugar concentration increases the osmotic pressure gradient between the fruit and the osmotic solution increases. Above 50ºBrix the dissolution of sugar in the osmotic solution becomes difficult and fast dissolution is only achieved through heating and vigorous stirring of the osmotic solution. Ternary osmotic solutions consisting of a sugar and salt (generally sodium chloride or potassium chloride) can also be used and in this case the salt content should not exceed 5% of the total solution weight otherwise the dried fruit becomes salty.

Ultrasound-Assisted Osmotic Dehydration The samples should be immersed in an osmotic solution and submitted to ultrasonic waves during a period of time. The process can be carried out under ambient temperature or at higher temperatures, since higher temperatures may enhance the mass transfer of water and soluble solids between the fruit or vegetable and the osmotic solution. The process can be carried out placing the samples in an ultrasonic bath or placing the samples in a vessel using an ultrasonic tip to induce the ultrasonic waves inside the vessel. No mechanical agitation is required. The ultrasound frequency should be between 25 and 100 kHz and intensity above 4000 W/m2. Most reports on osmotic dehydration employ an osmotic solution to fruit ratio from 4:1 to 20:1 (weight basis). In ultrasound-assisted osmotic dehydration the osmotic solution to fruit weight ratio should remain as low as possible, otherwise the required ultrasound power increases to output an ultrasound intensity between 4000 and 5000 W/m2. After removal from the solution, the dehydrated samples from each group should be drained and blotted with absorbent paper to remove the excess solution. To calculate water

124

Sueli Rodrigues and Fabiano A.N. Fernandes

loss during the ultrasonic-assisted osmotic dehydration the moisture content of the fruit has to be determined by heating in a drying oven at 105oC for 48h, according to AOAC method (AOAC, 1990). To calculate the sugar loss of the fruit and the sugar concentration of the osmotic solution during the process the soluble solids content of the fruit and of the osmotic solution (ºBrix) can to be directly determined by refractometry.

Air-Drying Air-drying of the samples should be carried out using the same procedure as described earlier for the ultrasound pre-treatment (Section 3.1).

Weight Reduction, Water Loss and Sugar Gain During the Pre-Treatment When an osmotic solution is used as the liquid medium (ultrasound-assisted osmotic dehydration) a higher water loss is observed if compared with the ultrasound pre-treatment using distilled water as the liquid medium. Studies with ultrasound-assisted osmotic dehydration of pineapples showed that pineapples lost between 5.1 to 8.3% of its initial water when an osmotic solution of 35ºBrix was used and between 9.8 to 14.2% of its initial water when an osmotic solution of 70ºBrix was used (table 5). Rastogi and Niranjan (1998), Rastogi and Raghavarao (2004) and Parkojo, Rahman, Buckle and Perera (1996) have studied the osmotic dehydration of pineapples without application of ultrasound and their results showed that after 30 minutes the pineapples lost 14% of water at 40ºBrix and 30ºC and 32% of water using an osmotic solution at 70ºBrix and 30ºC. Table 5. Sugar gain and water loss for pineapples submitted to ultrasound-assisted osmotic dehydration Operating condition 10 minutes of ultrasound (35ºBrix) 20 minutes of ultrasound (35ºBrix) 30 minutes of ultrasound (35ºBrix) 10 minutes of ultrasound (70ºBrix) 20 minutes of ultrasound (70ºBrix) 30 minutes of ultrasound (70ºBrix)

Sugar Gain 2.2% ± 1.1 2.7% ± 0.3 13.5% ± 1.2 12.3% ± 1.8 32.9% ± 3.8 34.1% ± 4.7

Water Loss 6.8% ± 0.7 5.1% ± 0.9 8.3% ± 0.2 9.8% ± 1.4 12.7% ± 0.9 14.1% ± 2.2

Using the osmotic solution the fruit gained soluble solids and an increase by 13.5% of sugar in pineapples was observed when an osmotic solution of 35ºBrix was used and an increase by 34.1% was observed when an osmotic solution of 70ºBrix was used. The gain of sugars occurred because of the different sugar concentration (osmotic pressure) between the fruit and the osmotic solution, which favored a mass transfer of sugar from the more sugar concentrated osmotic solution to the fruit and a mass transfer of water from the fruit to the osmotic solution. The studies on osmotic dehydration reported by Rastogi and Niranjan (1998), Rastogi and Raghavarao (2004) and Parkojo, Rahman, Buckle and Perera (1996)

Ultrasound in Fruit Processing

125

showed that pineapples gained 10% of soluble solids at 40ºBrix and 30ºC, after 30 minutes under osmotic dehydration, and gained 35% of soluble solids using an osmotic solution at 70ºBrix and 30ºC. The water loss observed when using ultrasound-assisted osmotic dehydration was slightly lower at 35ºBrix, but was 50% lower when a 70ºBrix osmotic solution was used. The lower water loss observed at 70ºBrix was influenced by the higher incorporation of sugar by the fruit which has decreased significantly the concentration gradient between the osmotic solution and the fruit. According to Fito (1994) and Shi and Fito (1994) when the osmotic dehydration is carried out under vacuum the gas that occupies the pores of fruits and vegetables is removed by the application of low pressure. This reduction in pressure causes the expansion and escape of the gas occluded in the pores. When pressure is restored, the pores can be occupied by the osmotic solution increasing the soluble solids content of the sample. The cavitations provoked by the application of ultrasound may also cause the expansion and escape of the gas occluded in the pores which later on are occupied with the osmotic solution. In addition, the micro-channels formed during ultrasound application may also facilitate the mass transfer of soluble solids into the sample. Studies on ultrasound-assisted osmotic dehydration of apples (var. Golden) carried out by Simal, Benedito, Sánchez and Rosselló (1998) showed that sugar gain was not dependent on temperature while water loss was. The high porosity of apples favored the incorporation of sugars when ultrasound was applied and after 30 minutes the sugar gain was 145% (table 6), which was much higher than the sugar gain observed for osmotic dehydration without application of ultrasound (88% at 40ºC and 128% at 60ºC). The use of ultrasound became an advantage for water loss only after 1 hour of application when the water loss observed for ultrasound-assisted osmotic dehydration was 8.7% higher than the water loss observed for traditional osmotic dehydration (table 6). Table 6. Sugar gain and water loss for apples submitted to osmotic dehydration and ultrasound-assisted osmotic dehydration using an osmotic solution of 70ºBrix (Simal, Benedito, Sánchez and Rosselló, 1998) Operating condition Ultrasound-assisted osmotic dehydration 10 minutes (40ºC) 30 minutes (40ºC) 60 minutes (40ºC) 10 minutes (60ºC) 30 minutes (60ºC) 60 minutes (60ºC) Osmotic dehydration 30 minutes (40ºC) 60 minutes (40ºC) 30 minutes (60ºC) 60 minutes (60ºC)

* based on equation (6).

Sugar Gain* [%]

Water Loss [%]

89 145 193 89 145 193

19.2 27.0 36.1 22.9 34.4 46.3

88 107 128 160

29.6 33.2 36.0 40.0

126

Sueli Rodrigues and Fabiano A.N. Fernandes

Modeling of the Mass Transfer Phenomena Three mathematical models for osmotic dehydration pre-treatment are presented in the literature and can be used to model ultrasound-assisted osmotic dehydration: Fickian, Peleg’s and Mass transfer model. The Fickian model (Fito and Chiralt, 1997) assumes that the main mass transfer mechanism during osmotic dehydration is diffusional in nature and sample deformation and shrinkage during drying can be considered negligible. The diffusional water and solute coefficients are considered effective parameters, as a result of all the transport phenomena that could take place during the process. The model assumes that the diffusivity coefficients are effective parameters, constant and uniform, and also assumes an isotropic behavior of the solid with regard to water and solute diffusion. As such the mass transports can be described by Fick’s law in an unsteady state mass balance. The Fickian model can be solved analytically assuming that the sample volume remains constant throughout the process and that the initial soluble solids content of the samples is also negligible. For a parallelepipedic shape sample: 2 ⎡ H − He 8 ∞ 1 2 3π ⋅ D w ⋅ t ⎤ ( ) exp 2 1 = 2⋅∑ ⋅ − ⋅ ν + ⋅ ⎢ ⎥ H 0 − H e π ν=0 (2 ⋅ ν + 1)2 4 ⋅ L2 ⎦ ⎣

(8)

2 ⎡ C − Ce 8 ∞ 1 2 3π ⋅ D s ⋅ t ⎤ = ⋅∑ ⋅ exp ⎢− (2 ⋅ ν + 1) ⋅ ⎥ C 0 − C e π 2 ν =0 (2 ⋅ ν + 1)2 4 ⋅ L2 ⎦ ⎣

(9)

where, C is the soluble solids concentration [g/m3]; C0 is the initial soluble solids concentration [g/m3]; Ce is the equilibrium soluble solids concentration [g/m3]; DS is the effective diffusivity of the solids [m2/s]; DW is the effective water diffusivity [m2/s]; H is the moisture content of the sample, H0 is the initial moisture content; He is the equilibrium moisture content; L is the height of the sample [m]; and t is the time [s]. Peleg’s model (1988) is used to describe the osmotic dehydration curves in some published papers (Azoubel and Murr, 2004; El-Aouar, Azoubel, and Murr, 2003; Park et al., 2002). Peleg’s equation presents a satisfactory fitting for water loss. This two parameter model describes most of the published curves which approach equilibrium asymptotically (Palou, Lopez-Malo, Argaiz, and Welti, 1993):

MC = MC 0 ±

t k1 + k 2 ⋅ t

(10)

where, MC is the water or solids’ amount at instant t [g]; MC0 is the initial water or solids amount [g]; k1 and k2 are the Peleg’s parameters; and t is the time [s]. The Mass transfer model takes into account the mass transfer between the fruit or vegetable and the liquid medium (distilled water or osmotic solution). The mass balance for the sample (fruit or vegetable) includes the mass transfer of water from the sample to the

Ultrasound in Fruit Processing

127

osmotic solution and the mass transfer of sugar and/or salt from the osmotic solution to the sample.

(

)

W dM FR W W = −K mW ⋅ A FR ⋅ C FR − C OS ⋅ VFR dt

(

)

dM SFR = −K Sm ⋅ A FR ⋅ CSFR − CSOS ⋅ VFR dt

(11)

(12)

where, AFR is the sample superficial area [m2], CFRS is the soluble solids concentration of the fruit [g/m3], CFRW is the water concentration of the fruit [g/m3], COSS is the soluble solids concentration of the osmotic solution [g/m3], COSW is the water concentration of the osmotic solution [g/m3], KmS is the mass transfer coefficient of soluble solids [1/m2.s], KmW is the mass transfer coefficient of water [1/m2.s], MFRS is the mass of soluble solid of the fruit [g], MFRW is the mass of water of the fruit [g], and VFR is the volume of the fruit [m3]. During the dehydration process the sample may shrink and this phenomenon has to be considered by the mathematical model to increase the accuracy of the mass transfer coefficients. In the model, the shrinkage effect is set to be proportional to the water mass change in the sample, according to equation 6. The sample superficial area is assumed to decrease at a proportional rate following the decrease in volume of the sample. W dVFR α dM FR = W⋅ dt dt ρ

(13)

where, α is the shrinkage factor, and ρW is the density of water [g/m3]. The mass balance for the liquid medium includes the gain of water that is removed from the sample and the loss of sugar to the sample. As the material balances are based on mass balances, the amount of water leaving the sample is equal to the amount of water entering the liquid medium. The opposite occurs with the mass balance of sugar, where the amount of solids entering the sample is equal to the amount of solids leaving the liquid medium. W W dM OS dM FR =− dt dt

(14)

dM SOS dM SFR =− dt dt

(15)

W dVOS 1 dM FR = W⋅ dt dt ρ

(16)

where, MOSS is the mass of soluble solid of the osmotic solution [g], MOSW is the mass of water of the osmotic solution [g], and VOS is the volume of the osmotic solution [m3].

128

Sueli Rodrigues and Fabiano A.N. Fernandes

To estimate the mass transfer coefficients of the ultrasound-assisted osmotic dehydration experimental data should be gathered and used with a parameter estimation procedure based on the minimization of the error sum of squares. The model is solved by numerical integration using the Runge-Kutta method or similar integration technique. The mass transfer coefficients obtained for the experiments carried out with pineapples are presented in table 7. Except for the treatment carried out with distilled water as the liquid medium, the Mass transfer model represents very well the data points as shown in figure 11 and through the regression errors. The F-values calculated for all experiments were above the listed F-values confirming the validity of this model within a 95% level of confidence. Table 7. Mass transfer coefficients for the ultrasound-assisted osmotic dehydration of pineapples Liquid medium

Distilled water Sucrose solution (35ºBrix) Sucrose solution (70ºBrix)

Mass Transfer Coefficient for Water [1/h.m2] 12.8 435.6

Mass Transfer Coefficient for Solids [1/h.m2] 87.0 84.6

R2

Ftest*

0.752 0.991

2.6 106.6

824.4

210.6

0.995

96.3

*Listed F-test: 18.5 for 95% level of confidence.

Figure 11. Normalized water and sugar content for pineapples submitted to ultrasound pre-treatment at 30ºC.

The mass transfer coefficient of water for pineapples presented an almost linear relationship with the concentration of sucrose in the liquid medium going from 12 h-1.m-2 when distilled water was used to 800 h-1.m-2 when an osmotic solution of 70ºBrix was

Ultrasound in Fruit Processing

129

employed, which was expected based on the larger osmotic pressure gradient between the fruit and the osmotic solution. The mass transfer coefficient of sucrose did not show significant change when distilled water and an osmotic solution of 35ºBrix was employed, but increased 2.5 fold when an osmotic solution of 70ºBrix was used.

Process Optimization The drying process of fruits and vegetables comprehends the ultrasonic-assisted osmotic dehydration process followed by air-drying. Total processing time can be optimized in order to reduce the drying process to a minimum, which can reduce costs and increase the overall productivity. To optimize the process, the osmotic dehydration should be used while the water loss rate of the sample is higher than the rate that would be obtained by the air-drying process. When the water loss rate in the ultrasound-assisted osmotic dehydration becomes lower than the rate that would be obtained in the air-drying process, then the sample should be transferred from the osmotic dehydration to the air-drying equipment, where the sample should stay till drying is completed (figure 12). The optimization can be carried out based on the estimated parameters for the ultrasound-assisted osmotic dehydration process and the airdrying process.

Figure 12. Illustrative scheme of the water removal rate in ultrasound-assisted osmotic dehydration and airdrying and the optimum time to change from osmotic dehydration to air-drying.

Table 8 shows the total processing time to achieve a final moisture content of 1.25 g of water/g of dry solids for pineapples submitted to ultrasound-assisted osmotic dehydration. In this case the use of ultrasound-assisted osmotic dehydration reduced the total drying time. In this case, the high water loss observed during the pre-treatment and the high mass of dry solids reduces the moisture content of the fruit (in dry solids basis) and the final moisture content is more rapidly achieved. The lowest total drying time for pineapples was observed

130

Sueli Rodrigues and Fabiano A.N. Fernandes

when the pre-treatment was carried out during 20 minutes using an osmotic solution of 70ºBrix, condition which reduced the total drying time in 21 minutes and the air-drying time in 41 minutes. The results for the ultrasound-assisted osmotic dehydration showed that the total drying time tend to increase with increasing sugar gain by the fruit. Table 8. Total processing time (pre-treatment + air-drying) for pineapples to achieve a final moisture content of 1.25 g water/g dry solids

No pre-treatment (air-drying only) 10 minutes of ultrasound (35ºBrix) 20 minutes of ultrasound (35ºBrix) 30 minutes of ultrasound (35ºBrix) 10 minutes of ultrasound (70ºBrix) 20 minutes of ultrasound (70ºBrix) 30 minutes of ultrasound (70ºBrix)

Total Time [min] 473 462 483 489 455 452 463

Pre-treatment Time [min] --10 20 30 10 20 30

Air-Drying Time [min] 473 452 463 459 445 432 433

Table 9 shows the total processing time to remove 80% of the initial water content of pineapples reducing the moisture content of the fruit by 0.67 g of water/g of fresh fruit. The use of ultrasound-assisted osmotic dehydration showed not to be a viable process to remove large quantities of water from the fruit. As presented in table 9, the ultrasound-assisted osmotic dehydration was not able to reduce the air-drying time at any of the conditions that were studied. Although this pre-treatment presents high water loss, the incorporation of sugars into the fruit increases the solid content of the fruit. High mass of dry solids reduces the moisture content of the fruit and the moisture content gradient (Fick’s law). As such, if the objective of the drying process is to reduce the initial water content in 80% (as an example) it will have to remove moisture strongly attached to the sugar incorporated into the fruit increasing the time required for drying. The results for the ultrasound-assisted osmotic dehydration showed that the total drying time increased with increasing sugar gain by the fruit. Table 9. Total processing time (pre-treatment + air-drying) for pineapples to remove 80% of the initial water content of the fruit

No pre-treatment (air-drying only) 10 minutes of ultrasound (35ºBrix) 20 minutes of ultrasound (35ºBrix) 30 minutes of ultrasound (35ºBrix) 10 minutes of ultrasound (70ºBrix) 20 minutes of ultrasound (70ºBrix) 30 minutes of ultrasound (70ºBrix)

Total Time [min] 473 577 588 685 595 655 677

Pre-treatment Time [min] --10 20 30 10 20 30

Air-Drying Time [min] 473 567 568 655 585 635 647

Ultrasound in Fruit Processing

131

CONCLUSIONS The use of ultrasound in the food industry is increasing as well as the use of ultrasound in fruit processing. Ultrasonic processes are still under development and more studies are required to fully comprehend the effects of ultrasound not only in the fruit tissue but also in some sensory characteristics such as texture, adhesiveness and other. Recent studies have shown a good potential of ultrasonic treatments. For most fruits studied the use of ultrasound pre-treatment increased the water diffusivity of the fruit leading to faster air-drying of the fruit. This phenomenon may happen due to the process of formation of micro-channels during the application of ultrasound, phenomenon that has to be further studied to understand how the micro-channels are formed and how the cell membrane and fruit cell structure change during the process. The increase in water diffusivity at the air-drying stage makes the use of ultrasound as a pre-treatment an interesting methodology complementary to classical air-drying. Fruits pre-treatment with ultrasonic waves presented in some cases significant loss of sugars and therefore the process can be applied to produce dried low sugar fruit that can be employed in foodstuffs with reduced calories.

ACKNOWLEDGEMENTS The authors thank Dr. Marisa Narciso Fernandes from Universidade Federal de São Carlos for allowing the use of the microscopy system; Dr. Maria Izabel Gallão from Universidade Federal do Ceará for the help with the preparation of the laminas.

REFERENCES AOAC (1990). Official methods of analysis. Washington: Association of Official Analytical Chemists. Ahmed, F. I. K., and Russell, C. (1975). Synergism between ultrasonic waves and hydrogen peroxide in the killing of micro-organisms. Journal of Applied Bacteriology, 39, 31-40. Allinger, H. (1975). Ultrasonic disruption. American Laboratory, 10, 75-85. Apfel, R. E. (1981). Ultrasonics. In P. D. Edmonds (Ed.). Methods of experimental physics. New York: Academic Press, Vol. 19, pp. 356–411. Azoubel, P. M., and Murr, F. E. X. (2004). Mass transfer kinetics of osmotic dehydration of cherry tomato. Journal of Food Engineering, 61, 291–295. Barbanti, D., Mastrocola, D., and Severine, C. (1994). Drying of plums. A comparison among twelve cultivars. Sciences des Aliments, 14, 61-73. Boucher, R. M., and Lechowich, R. V. (1979). Ultrasonics: A tool to improve biocidal efficacy of sterilants or disinfectants in hospital and dental practice. Canadian Journal of Pharmacological Science, 14, 1-12. Bustabad, O. M. (1999). Weight loss during freezing and the storage of frozen meat. Journal of Food Engineering, 41, 1–11.

132

Sueli Rodrigues and Fabiano A.N. Fernandes

Cao, H., Zhang, M., Mujumdar, A. S., Du, W. H., and Sun, J. C. (2006). Optimization of osmotic dehydration of kiwifruit. Drying Technology, 24, 89-94. Ciccolini, L., Taillandier, P., Wilhem, A. M., Delmas, H., and Strehaiano, P. (1997). Low frequency thermosonication of Saccharomices cerevisiae suspensions: Effects of temperature and ultrasonic power. Chemical Engineering Journal, 65, 145-149. Corzo, O., and Gómez, E. R. (2004). Optimization of osmotic dehydration of cantaloupe using desired function methodology. Journal of Food Engineering, 64, 213-219. De Gennaro, L., Cavella, S., Romano, R., and Masi, P. (1999). The use of ultrasound in food technology I: inactivation of peroxidase by thermosonication. Journal of Food Engineering, 39, 401–407. El-Aouar, A. A., Azoubel, P. M., and Murr, F. E. X. (2003). Drying kinetics of fresh and smotically pretreated papaya (Carica papaya L.). Journal of Food Engineering, 59, 85– 91. Fellows, P. (2000). Food Processing Technology––Principles and Practice. Chichester, UK: Ellis Horwood Ltd, pp. 418–440. Fernandes, F. A. N., and Rodrigues, S. (2007). Ultrasound as pre-treatment for drying of fruits: dehydration of banana. Journal of Food Engineering, xxx. Fernandes, F. A. N., Gallão, M. I., and Rodrigues, S. (2007). Effect of osmotic dehydration and ultrasound pre-treatment on cell structure. Proceedings of the 3rdCIGR (in CD-Rom). Fernandes, F. A. N., Rodrigues, S., Gaspareto, O. C. P., and Oliveira, E. L. (2006). Optimization of osmotic dehydration of papaya followed by air-drying. Food Research International, 39, 492-498. Fernandes, F. A. N., Rodrigues, S., Gaspareto, O. C. P., and Oliveira, E. L. (2006). Optimization of osmotic dehydration of bananas followed by air-drying. Journal of Food Engineering, 77, 188-193. Fito, P. (1994). Modeling of vacuum osmotic dehydration of food. Journal of Food Engineering, 22, 313–328. Fito, P. and Chiralt, A. (1997). Osmotic dehydration. An approach to the modeling of solid food-liquid operations. In Food Engineering 2000, eds. P. Fito, E. Ortega-Rodriguez and G.V. Barbosa-Canova. Chapman and Hall, New York, pp. 231-252. Fleming, I. D., and Pegler, H. F. (1963). The determination of glucose in the presence of maltose and isomaltose by a stable specific enzymic reagent. Analyst, 88, 967-968. Fuente-Blanco, S., Sarabia, E. R. F., Acosta-Aparicio, V. M., Blanco-Blanco, A., and Gallego-Juárez, J. A. (2006). Food drying process by power ultrasound. Ultrasonics Sonochemistry, 44, e523-e527. Gallego-Juárez, J. A., Rodríguez-Corral, G., Gálvez-Moraleda, J.C., and Yang, T. S. (1999) A new high intensity ultrasonic technology for food dehydration. Drying Technology, 17, 597-608. Kalichevsky, M. T., Knorr, D., and Lillford, P. J. (1995). Potential food applications of highpressure effects on ice-water transitions. Trends in Food Science and Technology, 6, 253–258. Karnovisky, M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. Journal of Cellular Biology, 27, 137-138. Kuldiloke, J. (2002). Effect of ultrasound, temperature and pressure treatments on enzyme activity and quality indicators of fruit and vegetable juices. PhD thesis, Technical University of Berlin.

Ultrasound in Fruit Processing

133

Li. B., Sun, D.W. (2002). Effect of power ultrasound on freezing rate during immersion freezing of potatoes. Journal of Food Engineering, 55, 277-282. Lima, M., and Sastry, S. K. (1990). Influence of fluid rheological properties and particle location on ultrasound-assisted heat transfer between liquid and particles. Journal of Food Science, 55, 1112–1115. López, P., and Burgos, J. (1995a). Lipoxygenase inactivation by manothermosonication: effects of sonication physical parameters, pH, KCl, sugars, glycerol, and enzyme concentration. Journal of Agricultural and Food Chemistry, 43, 620–625. López, P., and Burgos, J. (1995b). Peroxidase stability and reactivation after heat treatment and manothermosonication. Journal of Food Science, 60, 451–455. López, P., Sala, F. J., Fuente, J. L., Cardon, S., Raso, J., and Burgos, J. (1994). Inactivation of peroxidase lipoxigenase and phenol oxidase by manothermosonication. Journal of Agricultural and Food Chemistry, 42, 253-256. López, P., Vercet, A., Sánchez, A. C., and Burgos, J. (1998). Inactivation of tomato pectic enzymes by manothermosonication. Zeitschrift fur Lebensmittel-Untersuchung und Forschung, 207, 249–252. Madamba, P. S., and Lopez, R. I. (2002). Optimization of the osmotic dehydration of mango (Mangifera indica L.) slices. Drying Technology, 20, 1227-1242. Martino, M. N., Otero, L., Sanz, P. D., and Zaritzky, N. E. (1998). Size and location of ice crystals in pork frozen by high-pressure-assisted freezing as compared to classical methods. Meat Science, 50, 303–313. Mason, T. J. (1998). Power ultrasound in food processing - the way forward. In: Ultrasounds in Food Processing; Povey, M.J.W.; Mason, T.J. Eds.; Blackie Academic and Professional: Glasgow, pp. 104-124. Mason, T. J., Paniwnyk, L., and Lorimer, J. P. (1996). The use of ultrasound in food technology. Ultrasonics Sonochemistry, 3, S253-S256. Miller, G. L. (1959). Use of dinitrosalicilic acid reagent for determination of reducing sugar. Analitical Chemistry, 31, 426-428. Ngapo, T. M., Babare, I. H., Reynolds, J., and Mawson, R. F. (1999).Freezing rate and frozen storage effects on the ultrastructure of samples of pork. Meat Science, 53, 159–168. Oliveira, I. M., Fernandes, F. A. N., Rodrigues, S., Sousa, P. H. M., Maia, G. A., Figueiredo, R. W. (2006). Modeling and optimization of osmotic dehydration of banana followed by air-drying. Journal of Food Processing Engineering, 29, 400-413. Ordoñez, J. A., Sanz, B., Hernández, P. E., and López-Lorenzo, P. (1984). A note on the effect of combined ultrasonic and heat treatment on the survival of thermoduric streptococci. Journal of Applied Bacteriology, 56, 175–177. Otero, L., Martino, M., Zaritzky, N., Solas, M., and Sanz, P. D. (2000). Preservation of microstructure in peach and mango during high pressure-shift freezing. Journal of Food Science, 65, 466–470. Palou, E., Lopez-Malo, A., Argaiz, A., and Welti, J. (1993). Osmotic dehydration of papaya. Effect of syrup concentration. Revista Espanola de Ciencia y Technologia de Alimentos, 33, 621–630. Pan, Y. K., Zhao, L. J., Zhang, Y., Chen, G., and Mujumdar, A. S. (2003). Osmotic dehydration pre-treatment in drying of fruits and vegetables. Drying Technology, 21, 1104-1114.

134

Sueli Rodrigues and Fabiano A.N. Fernandes

Parjoko, Rahman, M. S., Buckle, K. A., and Perera, C.O. (1996). Osmotic dehydration kinetics of pineapple wedges using palm sugar, Lebensm.-Wiss.u.-Technol., 29, 452-459. Park, K. J., Bin, A., Brod, F. P. R., and Park, T. H. K. B. (2002). Osmotic dehydration kinetics of pear D’anjou (Pyrus communis L.). Journal of Food Engineering, 52, 293– 298. Peleg, M. (1988). An empirical model for the description of moisture sorption curves. Journal of Food Science, 53, 1216–1219. Perry, R. H., Green, D. W. (1999). Perry’s Chemical Engineer’s Handbook. McGraw Hill, New York. Prinzivalli, C., Brambilla, A., Maffi, D., Scalzo, R.L. and Torreggiani, D. (2006). Effect of osmosis time on structure, texture and pectic composition of strawberry tissue. European Food Research and Technology, 224, 119-127. Prokharkar, S. M., Prasad, S., and Das, H. (1997). A model for osmotic concentration of bananas slices. Journal Food Science and Technology, 34, 230-232. Raoult-Wack, A. L. (1994). Advances in osmotic dehydration. Trends in Food Science and Technology, 5, 255–260. Raoult, A. L., Lafont, F., Rios, G., and Guilbert, S. (1989). Osmotic dehydration: Study of mass transfer in terms of engineering properties. In: A. S. Mujumdar and M. Roques (Eds.), Drying '89, Hemisphere, 487–495. Rastogi, N. K., and Niranjan, K. (1998). Enhanced mass transfer during osmotic dehydration of high pressure treated pineapple. Journal of Food Science, 63, 508–511. Raso, J., Condon, S., and Sala Trepat, J. (1994). Mano-thermosonication: A new method of food preservation? Food Preservation by Combined Processes. Final report, Flair Concerted Action N° 7, subgroup B, EUR 15776 EN. Rastogi, N. K., and Raghavarao, K. S. M. S. (1997). Water and solute diffusión coefficients of carrot as a function of temperature and concentration. Journal of Food Engineering, 34, 429–440. Rastogi, N. K., and Raghavarao, K. S. M. S. (2004). Mass transfer during osmotic dehydration of pineapple: considering Fickian diffusion in cubical configuration, Lebensm.-Wiss.u.-Technol., 37, 43-47. Rastogi, N. K., Eshtisghi, M. N., and Knorr, D. (1999). Accelerated mass transfer during osmotic dehydration of high intensity electrical field pulse pretreated carrots. Journal of Food Science, 64, 1020–1023. Ravivan P., Zhang, Z., and Feng, H (2005). Ultrasonication for tomato pectinmethylesterase inactivation: effect of cavitation intensity and temperature on inactivation. Journal of Food Engineering, 70, 189-196. Rodrigues, S., and Fernandes, F. A. N. (2007a). Dehydration of melons in a ternary system followed by air-drying. Journal of Food Engineering, 80, 678-687. Rodrigues, S., and Fernandes, F.A.N. (2007b). Use of ultrasound as pré-treatment for dehydration of melons. Drying Technology (in press). Rooney, J. A. (1981). Ultrasonics. In P. D. Edmonds (Ed.). Methods of experimental physics. New York: Academic Press, Vol. 19, pp. 299–353. Sanz, B., Paklacios, P., Lopez, P., and Ordonez, J. A. (1985). Effect of ultrasonic waves on the heat resistance of Bacillus stearothermophilus spores. Federal European Microbiology Society, 18, 251–259.

Ultrasound in Fruit Processing

135

Sastry, S. K., Shen, G. Q., and Blaisdell, J. L. (1989). Effect of ultrasonic vibration on fluidto-particle convective heat transfer coefficients. Journal of Food Science, 54, 229–230. Shi, X.Q., and Fito, P. (1994). Mass transfer in vacuum osmotic dehydration of fruits: a mathematical model approach. Lebensmittel Wissenschaft und Technologie, 27, 67-72. Simal, S., Benedito, J., Sánchez, E. S., and Roselló, C. (1998). Use of ultrasound to increase mass transport rate during osmotic dehydration. Journal of Food Engineering, 36, 323– 336. Spiazzi, E. A., Raggio, Z. I., Bignone, K. A., and Mascheroni, R. H. (1998). Experiments on dehydrofreezing of fruits and vegetables: mass transfer and quality factors. Advances in the Refrigeration Systems, Food Technologies and Cold Chain, International Institute of Refrigeration Proceeding Series, Vol. 6, pp. 401–408. Suslik, K. S. (1988). Ultrasounds: its Chemical, Physical and Biological Effects. New York: VHC Publishers. Tarleton, E. S. (1992). The role of field-assisted techniques in solid/liquid separation. Filtration Separation, 3, 246-253. Tarleton, E. S., and Wakeman, R. J. (1998). Ultrasonically assisted separation process. In: M.J.W. Povey and T.J. Mason (Eds.), Ultrasounds in Food Processing, Blackie Academic and Professional, Glasgow. pp.193-218. Teles, U. M., Fernandes, F. A. N., Rodrigues, S., Lima, A. S., Maia, G. A., and Figueiredo, R. W. (2006). Optimization of osmotic dehydration of melons followed by air-drying. International Journal of Food Science and Technology, 41, 674-680. Torreggiani, D. (1993). Osmotic dehydration in fruit and vegetables. Process and Food International, 26, 59–68. Vercet, A., Lopez, P., and Burgos, J. (1997). Inactivation of heat-resistant lipase and protease from Pseudomonas fluorescens by manothermosonication. Journal of Dairy Science, 80, 29–36. Vercet, A., López, P., and Burgos, J. (1999). Inactivation of heat-resistant pectinmethylesterase from orange by manothermosonication. Journal of Agricultural and Food Chemistry, 47, 432–437. Vercet, A., Sánchez, C., Burgos, J., Montanes, L., and López-Buesa, P. (2002). The effects of manothermosonication on tomato pectic enzymes and tomato paste rheological properties. Journal of Food Engineering, 53, 273–278. Vidal, B. C. (1977). Acid glycosaminoglycans and endochondral ossification: microspectrophotometric evaluation and macromolecular orientation. Cell Molecular Biology, 22, 45-64. Wan, P.J., Muanda, M.W. and Covey, J.E. (1992). Ultrasonic vs. nonultrasonic hydrogenation in a batch reactor. Journal of American Organics Chemical Society, 69, 876-879. Zheng, L., and Sun, D.W. (2006). Innovative applications of power ultrasound during food freezing processes – a review. Food Science and Technology, 17, 16-23.

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 137-168

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 4

OPTIMISATION OF THE CONVERSION OF ERGOSTEROL IN MUSHROOMS TO VITAMIN D, AND ITS BIOAVAILABILITY Conrad O. Perera* and Viraj J. Jasinghe Department of Chemistry, Food Science Programme National University of Singapore

ABSTRACT The conversion of ergosterol in mushrooms to vitamin D2 by exposure to ultra violet (UV) light was studied under different UV lamps (UV-A, UV-B, and UV-C) and was found to be significantly different (p<0.05). Analysis of ergosterol content in different tissues of Shiitake mushrooms showed a significant difference (p < 0.01) in its distribution. Thus, the conversion of ergosterol in whole mushrooms to vitamin D2, by exposure to UV light was significantly affected (p < 0.01) by the orientation of the mushroom tissues to the UV source. The conversion of ergosterol to vitamin D2 was about four times higher when gills were exposed to UV light compared to when the outer caps were exposed to the same. The highest vitamin D2 content (184.22 ± 5.71 μg/g DM) was observed in Oyster mushrooms exposed to UV-B light at 35 oC and around 80% moisture. On the other hand, under the same conditions of UV-exposure, the lowest vitamin D2 content (22.90 ± 2.68 μg/g DM) was observed in Button mushrooms. The kinetics of conversion of ergosterol to vitamin D2 showed that Oyster mushrooms (Pleurotus ostreatus) had the highest conversion rate followed by Shiitake (Lentinula edodes) and Abalone (Pleurotus cystidus), whereas the lowest conversion rate was observed in Button mushrooms (Agaricus bisporus). Both initial moisture content and temperature of UV exposure influenced the conversion of ergosterol. The conversion of ergosterol to vitamin D2 followed zero-order kinetics, where the rate constant varied with temperature according to the Arrhenius equation (K0 = 7.32 s-1; Ea = 51.5 kJ mol-1).

*

Corresponding author’s present address: Professor, Department of Chemistry, Food Science Programme; University of Auckland; Private Bag 92019; Auckland, New Zealand; Email: [email protected]

138

Conrad O. Perera and Viraj J. Jasinghe For the bioavailability of vitamin D, thirty male Wistar rats were fed for one week with a diet deficient in vitamin D. After the first week, six rats were randomly selected and sacrificed for analysis of initial Bone Mineral Density (BMD), and serum level of 25hydroxyvitamin D [(25(OH)D]. The remaining animals were divided into two groups of 12. One group received 1 μg of vitamin D2/day from UV-exposed mushrooms for a period of four weeks until sacrificed. The other group was fed the same amount of mushrooms that was not exposed to UV light, and was use as the control. At the end of four weeks, the mean serum 25(OH)D level of the experimental group was 129.42 ± 22.00 nmol/L, whereas, it was only 6.06 ± 1.09 nmol/L in the control group. The Femur BMD and the serum calcium concentration of the experimental group of animals were significantly higher (p < 0.01) than the control group. It may be concluded from the results that vitamin D2 from UV-exposed mushrooms is well absorbed and metabolised in this model animal system.

Keywords: mushrooms, vitamin D, ergosterol, kinetics of conversion, bioavailability, bone mineral density, UV irradiation.

1. INTRODUCTION In 1919, vitamin D, sometimes referred to as the “sunshine vitamin” was discovered by Sir Edward Mellanby (Mellanby, 1919) as part of his experiments on rickets. The main role of vitamin D is it’s function as a hormone in maintaining calcium homeostasis, important in the mobilization, retention, and bone deposition of calcium and phosphorous (Webb, 1990; Morgan, 2001; Holick, 2001;). Even though the role of vitamin D in invertebrates is not clear, phytoplanktons and zooplanktons have been producing vitamin D for more than 500 million years (Holick, 2003). This suggests that there are some other hidden functions of vitamin D in the human body, which are yet to be elucidated. Vitamin D is the generic name of a closely related group of vitamins exhibiting similar biological activity to cholecalciferol (vitamin D3). Ergocalciferol (vitamin D2) is the form of vitamin D that can be formed from plant sterol, ergosterol, by UV exposure. Vitamin D2 and D3 can be further classified into vitamin D4 (22,23 dihydroergocalciferol); vitamin D5 (sitosterol or 24-ethylcholecalciferol); and vitamin D6 (stigmasterol) according to their side chain structures (Napoli et al. 1979). Vitamins D2 and D3 have very similar structures except that vitamin D2 has an additional double bond and a methyl group compared to vitamin D3. Figure 1 illustrates the chemical structures of pro-vitamin D3 (7-dehydrocholesterol), vitamin D3, pro-vitamin D2 (ergosterol), and vitamin D2. Vitamin D, along with the vitamins A, E, and K has been categorized into the group of “fat soluble” vitamins. Overdosing of vitamin D is potentially toxic in view of its hypercalcemic effect (Marriott, 1997). However there are no reported cases of vitamin D overdoses (Marriott, 1997) and on the contrary, there are concerns about the validity of current recommended dietary allowances (RDA). There is general consensus that the current recommended intake of vitamin D (5 micrograms up to age 50, 10 micrograms between the ages of 51 and 70, and 15 micrograms after age 70) is inadequate (Cheetham, 1999; Heaney, 2000; Vieth, 2000, Vieth et al. 2001). Optimal intakes are higher, though, with 25 micrograms (1000 IU) recommended for those over age 2 (Nutrition Source, 2006).

Optimisation of the Conversion of Ergosterol in Mushrooms…

139

28

CH3 21

26

22

H 3C 20

21

CH3

24 25

23

20

CH3

17

11

16

13

11

19

H 3C

15

14 9

1

2

8

2

10

8

3

5

7

3

5

7

6

4

16

14

15

9

10

HO

27

13

19

CH3 1

CH3

17

12

27

CH3 25

23

H3C18

H3C18 12

26

24

22

H 3C

6

4

HO

7-Dehydrocholesterol

Ergosterol 28

H 3C 21

20

23

CH3 12

9

11 9

26

24

23

CH3 12

27

15

14

20

18

16

13

11

25

CH3

17

22

H 3C

CH3

24

18

21

26

22

H 3C

CH3 25

CH3

17

27

13

16

14

15

8

8

7 7

6

6 19 19

CH2

5 4

10

10 1

3

1

3

HO

CH2

5 4

HO

2

2

V ita m in D 3 (ch o le ca lcife ro l)

V ita m in D 2 (e rg o c a lc ife ro l)

Figure 1. The chemical structures of ergosterol (pro-vitamin D2), 7-dehydrocholesterol (pro-vitamin D3), vitamin D2, and vitamin D3 (source: Horst and Reinhardt, 1997).

Table 1. Illustrates vitamin D intakes by age according to FAO and WHO recommendations Age group Infants 0-6 months 7-12 months 1-3 years 4-6 years 7-9 years Adolescents, 10-18 years Adults, 19-50 years Older adults, 51-65 years Elderly adults, 65+ years Pregnant women Lactating women

Recommended Dietary Allowances (μg/day) 5 5 5 5 5 5 5 10 15 5 5

Source: FAO/WHO, 1998 expert consultation on human vitamin and mineral requirements (ftp://ftp.fao.org/es/esn/nutrition/Vitrni/pdf/CHAPTER08.pdf) (last accessed 20/03/2007).

140

Conrad O. Perera and Viraj J. Jasinghe

2. VITAMIN D METABOLISM Vitamin D undergoes a series of metabolic changes, in order to form biologically active analogues as illustrated in figure 2.

H3C

21

18 CH3 12 11 19 CH 3 1

HO

9

2

10

8

3

5

6 7

4

22 20 17

13

16

14

15

DIET

26 CH3

24 23

21 H3C

22

25 18 CH

20

26 CH3

24

25

23

3

CH3 27

11

12

9

UV LIGHT

8

HEAT

4 3

(Skin)

17

CH3 27

16 15

7

6

7-dehydrocholesterol

13 14

19 CH2

5 10 2

1

HO

Vitamin D3 H3C CH3

LIVER

CH3

H3C

OH CH3

CH3

CH3

OH CH3

Low Calcium / PTH Low Phosphorus KIDNEY CH2

HO

CH2

OH

1,25-dihydroxyvitamin D3 (Calcitriol)

HO

Kidney

25-hydroxyvitamin D3 (Calcidiol)

Figure 2. Pathways of Vitamin D3 metabolism (source: Horst and Reinhardt, 1997).

Vitamin D activation is initialized by 25-hydroxilation in the liver (Horst and Reinhardt, 1997), and its metabolism is controlled by the physiological loop, which starts with calcium sensing by the calcium receptor of the parathyroid gland (Brown et al., 1998). In vitamin D deficiency, low serum calcium levels or elevated serum phosphate concentrations, stimulate parathyroid gland to release Para Thyroid Hormone (PTH) (Feldman, 1999). Increase in serum PTH concentration cause increased renal phosphate excretion which causes decreased intracellular phosphate. The combined effects of increased PTH and decreased phosphate, induce 1α-hydroxylase, which stimulates the production of 1,25-dihydroxyvitamin D (1,25(OH)2D) in the kidney (Feldman et al. 1996; Green et al. 2007). This process is auto regulated by inhibiting the production of PTH by increased serum calcium concentrations (Feldman el al. 1996) and is linked with calcium homeostasis. Apart from its unique action on mineral homeostasis, a number of additional benefits of 1,25(OH)2D have been discovered.

3. CLINICAL IMPORTANCE OF VITAMIN D Vitamin D deficiency has been implicated in a number of common diseases. Vitamin D deficiency has been associated with increased risks of deadly cancers, cardiovascular disease, multiple sclerosis, rheumatoid arthritis, and type 1 diabetes mellitus (Ponsonby et al. 2002;

Optimisation of the Conversion of Ergosterol in Mushrooms…

141

Heaney 2003; Holick, 2004; Heath and Elovic, 2006; Luong and Nguyen, 2006; Holick, 2006; Peterlik, and Cross, 2006; Cheng, and Coyne, 2007).

3.1. Cancer Vitamin D deficiency has been shown to be associated with several types of cancers such as, breast (Grant, 2002a; Berube et al. 2004; O’Kelly and Koeffler 2003), prostrate (Chen and Holick, 2003; Chen et al. 2003; Wang et al. 2003), skin (Majewski et al. 2000; Grant, 2002b; Kamradt et al. 2003), and a number of reports showing evidence of relationship between vitamin D deficiency and colon cancers are now available (Burton, 2001; Peters et al. 2001; Ogunkolade et al. 2002; Peterlik, and Cross, 2006). It is now well established that apart from having an important role in calcium homeostasis and skeleton maintenance, the active analogs of vitamin D act as growth regulators on hyperproliferative cells including cancer cells (Holick 2004; Nutrition Source, 2006).

3.2. Cardio Vascular Diseases (CVD) The relationship between vitamin D and CVD has been reported in the literature (Luong, and Nguyen, 2006). Genetic factors have been known to cause both vitamin D deficiency and CVD. Vitamin D receptor is found in the heart muscle. Vitamin D is reported to be involved in the pathogenesis of many cardiovascular problems. Vitamin D has an important role in modulating CVD (Luong, and Nguyen, 2006). Congestive Heart Failure (CHF) has been found to correlate with serum vitamin D concentrations (Zittermann et al. 2003), and therefore it has been suggested that vitamin D deficiency may be a contributing factor in the pathogenesis of CHF in adults and children (Carlton-Conway et al. 2004). It has been well elucidated how vitamin D is associated with muscle weakness (Zittermann, 2003) and the CHF associated with vitamin D deficiency may also be explained in the same way. There are a number of observations of cardiovascular diseases, which are associated with vitamin D insufficiency (Norman et al. 2002; Zittermann et al. 2003). All these findings suggest that vitamin D plays a favourable role in the prevention of heart diseases.

3.3. Diabetes In the past few decades, there has been a rapid increase in the incidence of insulin dependant diabetes mellitus (IDDM) worldwide, especially in Europe (EURODIAB ACE study group, 2000). It has been reported in some European countries, that children who later develop IDDM, have been found to be born in spring and summer (Songini and Casu, 2001; McKinney et al. 2001). In addition, fewer than expected diabetic children have been born in these countries at the end of summer in October (Samuelsson et al. 1999) when serum vitamin D concentrations are high (Vieth et al. 2001). Vitamin D deficiency in infancy or pregnancy has been found to be associated with IDDM (THE EURODIAB Substudy 2 Study Group, 1999; Hypponen et al. 2001). Furthermore, Syndrome ‘X’, is the term used to

142

Conrad O. Perera and Viraj J. Jasinghe

describe a cluster of disorders linked to insulin resistance with the potential risk of glucose intolerance (Reaven, 1995), and the risk of Syndrome ‘X’ is associated with vitamin D deficiency (Boucher, 1998). Controversial evidence suggests that vitamin D may help prevent type 1 diabetes mellitus and hypertension (Holick, 2004, and 2006)

3.4. Obesity Obesity is recognized as a major public health problem of global significance (Gill et al. 1999; WHO, 1997). The current estimates of global prevalence exceed 250 million (WHO, 1997). There is evidence reported in the literature that links vitamin D deficiency with obesity (Speer et al. 2001; Kamycheva et al. 2002). Obesity in humans and rodents is associated with high circulating leptin levels (ElHaschimi et al. 2000). Alpha-melanocyte stimulating hormone (alpha-MSH) acts on the brain to control the hormone leptin, which is produced by fat cells and insulin, which regulate food intake and body weight (Schwartz, 2001). Decrease in body weight, serum concentrations of leptin and insulin have been observed in a human study where alpha-MSH has been given to human subjects over a period of six weeks (Fehm et al. 2001). However the long-term effect of alpha-MSH on the control of body fat has yet to be fully elucidated. In addition, people who are obese are likely to be deficient in vitamin D because of decreased bioavailability of vitamin D due to its deposition in the adipose tissue (Wortsman et al. 2000). In addition to the above described major chronic diseases, vitamin D deficiency has been also been associated with arthritis (Braun and Tucker, 1997; Holick, 2004), hypertension (Pfeifer et al. 2001; Holick, 2005) psoriasis (Kira et al. 2003) etc. Moreover, Vitamin D has been suggested for therapeutic applications in the treatment of several diseases including hyperproliferative diseases, secondary hyperparathyroidism, post transplant survival, and various malignancies (Mehta and Mehta, 2002;). The evidence strongly suggests that vitamin D deficiency is not only associated with skeleton bone disease but also with a number of chronic diseases. Hence, maintenance of healthy vitamin D status could be useful in the prevention of a wide spectrum of chronic diseases throughout the general population (Holick 2004; 2005 and 2006).

4. COMMON VITAMIN D DEFICIENCY DISORDERS The most common results of severe vitamin D deficiency disorders are rickets in children and osteomalacia in adults (Feldman, 1999; Morgan, 2001). There are number of investigations that have been carried out on vitamin D deficiency disorders all over the world. Out of 824 elderly persons from 11 European countries, 36% of men and 47% of women had 25(OH)D concentrations below 30 nmol/L (van der Wielen et al. 1995), a level considered to be marginal in terms of vitamin D deficiency. Vieth et al. (2001) observed vitamin D deficiency status is common in winter in Canadian women and revealed that their vitamin D intake was not sufficient to prevent it. They have suggested that the current RDA for vitamin D is too low to prevent the insufficiency. On the other hand in Finland, vitamin D intake was low and hypovitaminosis D was common in 9 – 15 year old apparently healthy Finnish girls,

Optimisation of the Conversion of Ergosterol in Mushrooms…

143

(Lehtonen-Veromma et al. 1999), and suggested that the daily dietary vitamin D supplementation with 10 μg/day was insufficient in preventing hypovitaminosis. Furthermore, it has been reported that British pre-school children are at risk of vitamin D deficiency (Lawson et al. 1999), and it was observed that most of the children with low haemoglobin levels showed low plasma vitamin D values. Almost all the countries, which have conducted surveys in order to investigate the prevalence of vitamin D deficiency, have reported high prevalence of vitamin D deficiency among their populations (Wilkinson et al. 2000; Du et al. 2001; Chan, 2004; Holvik et al. 2005; Mona et al. 2005). Vitamin D deficiency incidences have been reported in a number of countries. Furthermore, in 2001, Vitamin D deficiency was reported as an unrecognized epidemic among the elderly population, and more than 50% of elderly persons, living in their own homes and nursing homes in the USA were found to be deficient in vitamin D (Holick, 2001).

5. VITAMIN D AND SUN EXPOSURE Naturally, humans obtain vitamin D through cutaneous synthesis in the presence of ultraviolet B (UV-B) from sunlight and as well as from the diet. UV-B (wave length 290 – 315 nm) represents approximately 1.5% of the total solar spectrum (Hollosy, 2002). The precursor of vitamin D3, 7-dehydrocholesterol found in the adipose tissues of the body can be converted to vitamin D3 in the skin, and this process is supported by sunlight (Feldman et al. 1996). Sunlight is the most important source of vitamin D for most of the people in the UK since the content of vitamin D in the largely unfortified British diet is low (Burns et al. 2003). Furthermore, sunlight is the major determinant of vitamin D stores in southern Tasmanian population (Jones et al. 1999). The cutaneous production of vitamin D under exposure to sunlight depends on number of factors such as latitude, season, exposure to direct sunlight, skin colour, and age (Need et al. 1993). Sunscreens suppress cutaneous vitamin D synthesis (Matsuoka et al. 1987). Age-related decline in skin thickness may contribute to the age-related decline in 25(OH)D (Need et al. 1993). Environmental factors such as Latitude, Season, and time of the day influence the cutaneous production of vitamin D. The prevalence of vitamin D deficiency is higher in people with darker skins than people with white skins (Shanna et al. 2002). Moreover, people with dark skins need to spend up to six times longer in the sun to obtain the same amount of vitamin D as a white person since the increased skin pigment can greatly reduce the penetration of ultraviolet radiation into the skin (Clemens et al. 1982). However, “Sun bingeing” may cause skin cancers (Wharton and Bishop, 2003). In addition, excessive exposure to ultraviolet radiation, produce undesirable inactive byproducts of previtamin D, such as tachysterol and lumisterol by photoisomerization (Havinga et al.1960). The evidence suggests that there are number of factors involved in cutaneous production of vitamin D under exposure to sunlight, and therefore the adequate exposure is not easily defined. On the other hand, still there are pro and counter arguments on the risks and benefits of sunlight among the scientific community, which keeps the question unreciprocated.

144

Conrad O. Perera and Viraj J. Jasinghe

6. DIETARY SOURCES OF VITAMIN D Vitamin D3 may be obtained in limited amounts from animal food products such as butter, margarine, milk and milk products, liver and other meats, and eggs. Oily fish (including mackerel, sardines, salmon and trout etc.) and fish liver oils provide more substantial amounts of vitamin D but are eaten only by a minority of people. Vitamin D rich dietary sources are tabulated in table 2. Table 2. Vitamin D rich food sourcesa Dietary source Cod liver oil Salmon (raw) Halibut Greenland (raw) Rainbow trout (raw) Salmon (canned) Sardine in tomato source (canned) Mackerel (raw) Egg yolk (chicken)(raw) Tuna (raw) Mackerel in tomato source (canned) Chicken Milk, dry, whole, (powder)

Vitamin D3 (μg/100g) 250 30.0 15.0 13.0 13.0 12.0 5.5 4.0 2.9 2.4 1.5 1.2

a

: Source: Danish food composition databank; http://www.foodcomp.dk/fcdb_foodcomplist.asp? CompId=0023 (last accessed 09/04/2007).

In the United States, vitamin D is added to milk and recently; in 2003, the Food and Drug Administration (FDA) released a regulation allowing the addition of vitamin D to calciumfortified juices (Linda, 2003). After subjects consumed orange juice fortified with 1000 IU vitamin D3 daily for 12 wk, serum 25(OH)D3 concentrations increased by 150%, and serum parathyroid hormone concentrations decreased by 25% compared with baseline (Tangpricha et al. 2003). In 2004, additional food fortifications as well as dietary and supplement have been recommended in the USA (Moore et al. 2004). However, milk fortified with vitamin D is not permitted in the UK and some other European countries.

7. FEASIBILITY OF USING CULTIVATED EDIBLE MUSHROOMS AS A SOURCE OF VITAMIN D The evidence gathered suggests that vitamin D deficiency among the word population is dramatically increasing. Accumulating clinical evidence suggests that the vitamin D deficiency increases the risk of a large spectrum of chronic diseases including cancers, heart diseases, diabetes, obesity, arthritis, hypertension and psoriasis. Since the feasibility of sunbathing for vitamin D is still complicated and the risk/benefit has yet to be elucidated, the best idea is to look for alternative dietary sources. Edible mushrooms are very popular among the world populations for their unique flavour and medicinal value. Furthermore, mushrooms are considered a delicacy, highly accepted by

Optimisation of the Conversion of Ergosterol in Mushrooms…

145

vegetarians as well as non-vegetarians and could be used to supplement vitamin D in the diets of those populations at risk of vitamin D deficiency. Vitamin D2 is the form of vitamin D that could be provided from mushrooms, and this form has some remarkable advantages over vitamin D3. Vitamin D2 is more effective on bone mineralization than that of the effect of vitamin D3 (Tjellesen et al. 1985), and vitamin D2 is less toxic compared to vitamin D3 at high doses (Mehta and Mehta, 2002). In addition, vitamin D2 does not have hypercalcemic effects (Mawer et al. 1995). In nature, a limited amount of vitamin D2 has been reported in some wild edible mushrooms, however, cultivated edible mushrooms have been shown to be devoid of vitamin D2 (Mattila et al. 2002; Perera et al. 2003; Jasinghe and Perera 2005). Naturally, wild mushrooms may be exposed to UV radiation, which comprise 8 – 9% of the total solar spectrum (Hollosy, 2002), and this could be the reason for the presence of a limited amount of vitamin D2 in wild mushrooms. The commercially available cultivated mushrooms may not be exposed to the sunlight, which is essential in the natural production of vitamin D2. Nevertheless, ergosterol in mushrooms can be easily converted to vitamin D2 by UV irradiation (Mau et al. 1998; Perera et al. 2003; Jasinghe and Perera 2005).

8. ERGOSTEROL AND VITAMIN D2 CONTENT IN DIFFERENT PARTS OF SHIITAKE MUSHROOMS The analysis of ergosterol content of three different parts (cap, gills, and stalk) showed that mushrooms contain remarkably high amounts of ergosterol and its distribution varied in different parts of the mushroom tissue (table 3). Table 3. Ergosterol content of the different parts of Shitake mushrooms* Part of the mushroom Gills (n = 27) Outer layer of the Cap (n = 27) Stalk (n = 27)

Mean Ergosterol (mg/g, DMa ± S.D) 10.6 ± 0.99 5.34 ± 0.64 2.97 ± 0.56

*Adapted from Perera et al. (2003); aDry matter.

The results showed that the distribution of ergosterol within different mushroom tissues was significantly different (p < 0.01). The highest concentration of ergosterol was found in the gills, while the lowest was present in the stalk of mushrooms. The concentration of ergosterol in gills was almost twice that found in the outer layer of the caps, which in turn had almost twice that found in the stalks (Perera et al. 2003). On the other hand, the analysis of non-irradiated mushrooms showed that cultivated Shiitake mushrooms were totally devoid of vitamin D2 (results not shown). Mattila et al. (2002) also reported that vitamin D2 was almost totally absent in cultivated mushrooms and the results of this study support their observation as well. This was probably due to nonexposure of cultivated mushrooms to sunlight. In contrary, a few studies carried out in Japan reported very low amounts (0.0004 μg/g DM) of vitamin D2 (Takeuchi et al. 1984), and 0.23 – 1.10 μg/g DM (Takamura et al. 1991) in cultivated Shiitake mushrooms. Mattila et al.

146

Conrad O. Perera and Viraj J. Jasinghe

(1994) also reported limited amounts of vitamin D2 in wild edible mushrooms, and this was probably due to the exposure of mushrooms to sunlight.

8.1. Effect of UV Exposure on the Conversion of Ergosterol to Vitamin D2 Since it was observed that ergosterol content in gills of mushrooms was almost double than that of outer layer of the cap, the effect of orientation of the different mushroom tissues to the source of UV would have an effect on the conversion of ergosterol to vitamin D2. The dosage of VU irradiation is an important factor in considering the conversion of ergosterol to vitamin D. The intensity of a UV lamp (Mineralight UVGL – 25, San Gabriel, U.S.A.) at a distance of 15 cm away from the source was measured by an optical radiometer [MS-100, UVP, Inc, Upland, CA, USA, equipped with UV sensor (MS-136 UV sensor, UVP, Inc, Upland CA, USA)]. The calculated irradiation dose after two-hour exposure period, at a distance of 15 cm away from the source was 25.2 kJ/m2. The effect of exposure of shiitake mushrooms to UV light is shown in figure 3.

Vitamin D2 content (μg/g DM)

22.84 ± 25

2.71a

20 15 10

5.16 ± 0.61b

5 0 Irradiated with theirIrradiated with their Gills facing UV Caps facing UV source source

a,b: Values shown are mean values of 12 replicates ± SD. Values with different superscript letters are significantly different (p < 0.01). Mushrooms were exposed to UV-A at ambient temperature (27 oC). Moisture content of mushrooms was found to be around 89 % (w.b.). Figure 3. Vitamin D2 contents of Shiitake mushrooms subject to the two different orientations of the tissues to the source of irradiation (Adapted from Jasinghe and Perera, 2005).

The results showed a much higher than expected rate of conversion of ergosterol to vitamin D2 when the mushrooms were irradiated with their gills facing the UV source. When the gills were facing the source of UV light, the yield of vitamin D2 was 22.84 ± 2.71 μg/g DM, whereas, when they were facing away from the source (caps facing the source of UV light), the yield of vitamin D2 was only 5.16 ± 0.61 μg/g DM (Jasinghe and Perera, 2005). Vitamin D2 yield values obtained from these two orientations were significantly different (p < 0.01). These values are three to four times higher than those reported by Mau et al. (1998) for the conversion of ergosterol to vitamin D2 in Shiitake mushrooms by UV-B and UV-C exposure.

Optimisation of the Conversion of Ergosterol in Mushrooms…

147

Even though the concentration of ergosterol in gills of Shiitake mushrooms was only about twice higher than that of the outer layer of the cap (table 3), figure 3 clearly shows a conversion factor of approximately 4. This high level of conversion of ergosterol to vitamin D2 may be due to the fine morphology of the gills, which allows greater exposure of the surfaces to irradiation than in the case of the caps. In addition, as discussed earlier, the dark colour of caps could also retard the penetration of UV radiation into mushroom tissues.

8.2. Ergosterol and Vitamin D2 Contents in Different Types of Edible Mushrooms A range of different types of mushrooms was investigated for their ergosterol content (Jasinghe and Perera, 2005). The overall ergosterol contents of different types of mushrooms varied. The highest ergosterol content was found in Button mushrooms (7.80 ± 0.35 mg ergosterol/g DM) while the lowest was in Enoki mushrooms (0.68 ± 0.14 mg/g DM). Oyster mushrooms contained 4.40 ± 0.08 mg/g DM and the value was more or less the same for Abalone mushrooms (4.35 ± 0.16 mg/g DM). Shiitake mushrooms contained 6.05 ± 0.07 mg/g DM of ergosterol. This is in agreement with the values found by Mattila et al. (2002) who observed a value of 6.79 mg/g DM of ergosterol in Shiitake mushrooms. Ergosterol contents of different types of mushrooms are shown in figure 4. 7 . 8 ±0 . 3 5 c

8

6 . 7 6 ±0 . 19 f

Ergosterol content (mg/g DM

7 6 . 0 5 ±0 . 0 7 a

6 5

4 . 4 ±0 . 0 8 d

4 . 3 5 ±0 . 16 d

4 3 . 0 7 ±0 . 3 1 g

3 2 0 . 6 8 ±0 . 0 5 b

0 . 6 8 ±0 . 14 b

1

0 . 3 2 ±0 . 0 9 e

0 S hiit a k e

But t on

A ba l one

Whi t e e a r

B r own be e c h

a-d: Values shown are mean values of 6 replicates ± SD. Values with different superscript letters are significantly different (p < 0.01). Figure 4. Ergosterol contents of different types of mushrooms (Adapted from Jasinghe et al. 2006).

148

Conrad O. Perera and Viraj J. Jasinghe

8.3. Conversion of Ergosterol to Vitamin D2 by UV Exposure Vitamin D2 content of different types of mushrooms subjected to 2 h of UV-B exposure with their gills facing the source of irradiation is shown in figure 5. The calculated irradiation dose after two-hour of exposure period, at a distance of 15 cm away from the source was 25.2 kJ/m2.

Vitamin DD g/g DM)DM) 2 ( (ug/g Vitamin 2

80

76.35±5.24c

70 60 50 40

35.35±2.51a

33.65±1.12a

32.24±1.51a

28.45±2.09b

28.32±2.11b

30 20

10.54±0.69f

10

4.35±0.11d

e

6.34±0.64

0 Shiitake

Enoki

Button

Oyster

Abalone

Woody

White ear

Portabello

Brown beech

a-f: Values shown are mean values of 6 replicates ± SD. Values with different superscript letters are significantly different [a and b (p < 0.05), others (p < 0.01)]. Mushrooms were irradiated with their gills facing UV-B source at 27 oC. Moisture contents of the mushrooms were found to be around 89 % (w.b.). Figure 5. Vitamin D2 contents of the different types of mushrooms subjected to irradiation for two hours; with their gills facing the UV-A source (Adapted from Jasinghe et al. 2006).

Button mushrooms showed a relatively low vitamin D2 content (32.24 ± 1.51 μg/g DM) despite having the highest levels of ergosterol in them. This may be due to the fact that the button mushrooms were closed and the gills were not exposed to UV irradiation, thus UV light was unable to penetrate into the gills area effectively. Vitamin D2 may also be subjected to a greater action of tissue mono-oxygenases in button mushrooms than in others that transform vitamin D2 to 25(OH)D2 and 25(OH)2D2, thus reducing the overall conversion of ergosterol to Vitamin D2. Yet another possible reason for this lower than expected conversion of ergosterol to Vitamin D2 may be due to the lower penetration of UV-B into the mushroom tissues. UV-A is known to penetrate to a depth of 60-90 μm, whereas UV-B and UV-C penetrates only to a depth of less than 10 μm in human skin (Freeman et al. 1962; Anderson and Parrish, 1981). UV penetration in different mushrooms could also vary depending on the presence of pigments on the mushroom tissue. Oyster mushrooms on the other hand showed the highest vitamin D2 content (76.35 ± 5.24 μg/g DM) among the different mushrooms tested. The yield of vitamin D2 obtained from Abalone mushroom, which had more or less similar ergosterol content compared with Oyster, was only 33.65 ± 1.12 μg/g DM. Once again this may be due

Optimisation of the Conversion of Ergosterol in Mushrooms…

149

to their morphological differences and / or to the presence of active mono-oxygenases, which converts vitamin D2 to the hydroxy forms. Vitamin D2 content in Shiitake mushrooms after two-hour UV-B irradiation was 35.35 ± 2.51 μg/g DM. Mau et al. (1998) have reported 6.58 μg/g DM and 12.48 μg/g DM of vitamin D2 from Shiitake mushrooms and Button mushrooms respectively, after two-hour UV-B irradiation at 12 oC. However, no indication of the orientation of the mushroom tissues to the source of UV irradiation was indicated in their study. The values obtained for Shiitake mushrooms in this study were significantly higher than those reported by others and this may be due to the specific orientation of the mushroom to the UV source.

8.4. Effect of Moisture Content of Mushrooms on the Conversion of Ergosterol to Vitamin D2

Vitamin D2 content (μg/g DM)

Jasinghe and Perera (2005) showed that the best conversion of ergosterol to vitamin D2 in mushrooms takes place at a moisture content of around 78 % on a wet weight basis (figure 6). At high moisture contents (the moisture content of fresh mushrooms), the conversion was significantly (p < 0.01) lower than that at 78.4 % moisture content. This may be due to the dilution effect of ergosterol at the high moisture levels, which is likely to bring about a lower rate of conversion. At low moisture levels, the specific surface area of the tissue is increased due to evaporation of moisture, and consequently the exposure to oxygen may be increased resulting in the oxidation of vitamin D2. Furthermore, UV exposure also contributes to an oxidative atmosphere (Vayalil et al. 2003), and photo-degradation of vitamin D2 may occur. It can be concluded from the results that irradiation of mushrooms at a moisture-content of around 70 % - 80 %, enhances the yield of vitamin D2 significantly (p < 0.01). The calculated irradiation dose after two-hour exposure period, at a distance of 15 cm away from the source was 25.2 kJ/m2. 30

27.48 ± 0.73 c

25 20

22.37 ± 1.19 b

15.29 ± 1.44 a

23.52 ± 3.02 b 21.25 ± 0.90 d

16.36 ± 0.53 a

15 10 5 0 60,28

62,16

71,24

72,65

78,41

82,65

M oisture content of the m ushroom (%)

a-d: Values shown are mean values of 6 replicates ± SD. Mushrooms were irradiated at 27 oC. Values with different superscript letters are significantly different [b and d (p < 0.05), others (p < 0.01). Figure 6. Effect of moisture content of mushrooms on the conversion of ergosterol to vitamin D2 (Adapted from Jasinghe and Perera, 2005).

150

Conrad O. Perera and Viraj J. Jasinghe

8.5. Effect of Temperature on the Conversion of Ergosterol to Vitamin D2 To study the effect of temperature on the conversion of ergosterol to vitamin D2, Shiitake mushrooms were irradiated for two hours with their gills facing UV-A source, at different temperatures. The mushroom samples were placed 15 cm away from the UV source during irradiation, and the calculated irradiation dose after two-hour irradiation period was 25.2 kJ/m2. The moisture content of mushrooms was 89 % on a wet basis. Figure 7 shows the effect of irradiation temperature on the conversion of ergosterol to vitamin D2 in Shiitake mushrooms. The yields of vitamin D2, after UV irradiation at 12, 27, 35, 45, and 65 oC were significantly different (p < 0.01). The results clearly suggest that irradiation of mushrooms at about 35 oC, enhanced the conversion leading to the highest yield of vitamin D2. The decrease in conversion rate beyond 35 oC was probably due to many concurrent events that may occur: heat stress (oxidative), cell death, formation of browning pigments, further transformation of vitamin D2 as well as photo-degradation by irradiation.

45

41.34 ± 3.94c

40

Vitamin D2 content (μg/g DM)

35 30

26.52 ± 3.68b

25 19.37 ± 2.65d 20 14.37 ± 2.42e 15

10.4 ± 4.25a

10 5 0 12

27

35

45

65

o

Temperature of irradiation ( C)

a-d

: Values shown are mean values of 6 replicates ± SD. Values with different superscript letters are significantly different (p < 0.01). Moisture content of mushrooms was found to be around 89 % on a wet basis (w.b.).

Figure 7. Effect of temperature of irradiation on the conversion of ergosterol to vitamin D2 (Adapted from Jasinghe and Perera, 2005).

Optimisation of the Conversion of Ergosterol in Mushrooms…

151

8.6. Effect of Orientations of Mushrooms to the UV Source and Duration of Exposure on the Conversion of Ergosterol to Vitamin D2 In a previous study (Jasinghe and Perera, 2005), it was found that the conversion of ergosterol was higher in mushrooms, when they were irradiated with their gills facing the UV source, than when they were irradiated with their caps facing the UV source. Threfore, the effect of orientation of mushroom tissues and duration of exposure was studied. The first lot of mushrooms was irradiated for two hours with their gills facing the UV-A source, the second lot of mushrooms was irradiated with their gills facing the UV source for one hour and then they were further irradiated for another hour with their caps facing the UV source. The third lot was irradiated with their gills facing the UV source for two hours and then they were further irradiated for another two hours with their caps facing the UV source. The calculated irradiation doses were 25.2, 25.2, and 50.4 kJ/m2 respectively. Vitamin D2 contents of Shiitake mushrooms subjected to different orientations and times of exposure to UV irradiations are shown in figure 8.

38.54 ± 3.18

b

36.06 ± 2.32

b

Vitamin D2 content (μg/g DM)

40,00 35,00 30,00

24.00 ± 1.72a

25,00 20,00 15,00 10,00 5,00 0,00

Mushrooms irradiated for two-hours with their Gills fac Each side of the mushrooms Irradiated for two-hours w Each side of the mushrooms Irradiated for one-hour wi Values shown are mean values of 6 replicates ± SD. Values with different superscript letters are significantly different (p < 0.01). Mushrooms were irradiated at 27 oC and the moisture content of mushrooms was found to be around 89 % (w.b.). Figure 8. Effect of orientation of mushrooms and the duration of irradiation on the conversion of ergosterol to vitamin D2 (Adapted from Jasinghe and Perera, 2006).

152

Conrad O. Perera and Viraj J. Jasinghe

The yield of vitamin D2 after the irradiation of each side of the mushrooms (cap and gills) for one hour and vitamin D2 yield after the irradiation of each side of the mushrooms for two hours were found to be significantly higher (p < 0.01) than those observed from the mushrooms irradiated with their gills facing the UV source for two hours. The vitamin D2 yield obtained from irradiation of each side of the mushrooms for two-hours was 38.54 ± 3.18 μg/g DM) whereas it was 36.06 ± 2.32 μg/g DM when each side of the mushrooms were irradiated for one-hour. However, the difference between two values was not significant (p = 0.154). In humans, UV penetrates the outermost layers of the skin (epidermis, dermis) and there are some other factors such as skin colour, thickness, and body fat that effect the penetration of UV into the skin (Need et al. 1993; MacLughlin and Holick, 1985). Similar factors may interfere with UV penetration into mushrooms as well and therefore it is not clear to what extent that UV penetrates into the mushrooms tissues. Figure 9 shows the effect of period of irradiation of each side of the mushrooms on the conversion of ergosterol to vitamin D2. The conversion of ergosterol in mushrooms to vitamin D2 is almost completed within one hour, and this could be the reason that prolonged irradiation of each side after one-hour, does not contribute further to this conversion. The calculated irradiation dose in this study was 0.21 kJ/m2/min and average moisture content of mushrooms was found to be around 89% (w.b).

Figure 9. The effect of time of UV-A irradiation of Shiitake mushrooms on the conversion of ergosterol to vitamin D2 (Adapted from Jasinghe and Perera, 2006).

Optimisation of the Conversion of Ergosterol in Mushrooms…

153

It could be seen from the graph that prolonged irradiation does not increase vitamin D2. Previtamin D intermediates also absorb UV to produce tachysterol and lumisterol by photoisomerization (Havinga et al. 1960), and prolonged irradiation produces irreversible “over-irradiation products” by dimerization, and ring cleavage of the sterols (Braun et al. 1991). These may be the reasons for the slight reduction in vitamin D2 content close to the two-hour period of irradiation, when each side of the mushrooms was subjected to UV-A irradiation. In addition, irradiation also contributes to an oxidative atmosphere (Vayalil et al. 2003), and prolonged exposure of vitamin D to UV may result in photo-degradation of vitamin D2 (Webb et al. 1989).

8.7. Conversion of Ergosterol to Vitamin D2 by Different Bands of UV (UV-A, UV-B, and UV-C) In this study, the moisture content of mushrooms was adjusted to around 80% by removing the moisture in a vacuum dryer at ambient temperature, and irradiation was performed at 35 oC, since these were found to be the optimum conditions for this conversion (Jasinghe and Perera, 2005). Figure 10 illustrates the effect of different bands of UV light on the conversion of ergosterol to vitamin D2. The yields of vitamin D2 under UV-A, UV-B, and UV-C are significantly different (p < 0.01). The calculated radiation doses of UV-A, UV-B, and UV-C after two-hour period of exposure (one-hour each side) were 25.2, 35.3, and 23.0 kJ/m2 respectively. The results clearly indicate that the conversion of ergosterol to vitamin D2 under UV-A was significantly higher (p < 0.01) than that under UV-C and the conversion under UV-B was significantly higher (p < 0.01) than those under UV-A or UV-C. Since the conversion of ergosterol to vitamin D2 is also dose dependent, it is not possible to draw conclusions, since the dosage of UV radiation received for the three bands were different for the same length of time of exposure. The highest yields of vitamin D2 were obtained under UV-B irradiation. However, under UV-B, mushrooms received almost 50% more irradiation dose than under UV-C. Therefore the vitamin D2 yields under UV-B and UV-C cannot be reconciled. Mau et al. (1998) have reported a value of 6.58 μg/g DM of vitamin D2 from Shiitake mushrooms after a two-hour period of UV-B irradiation. However, the orientation of mushrooms to the UV source and the moisture content were not reported. The temperature of irradiation, reported in their study (12oC) was much lower than that was found to be optimum (35oC) in our study (Jasinghe and Perera, 2005). Temperature of irradiation plays an important role in this conversion as shown by our earlier studies and this may be one of the reasons why they obtained low conversion rates. In addition, the irradiation dose used in their study (9.86 kJ/m2) was much lower than the irradiation dose used in this study (35.3 kJ/m2) and finally the orientation of the mushrooms to UV source is most important, as shown earlier (Jasinghe and Perera 2005).

154

Conrad O. Perera and Viraj J. Jasinghe

200,00

e

184.22 ± 5.71

180,00 f

Vitamin D2 content (μg/g DM)

160,00

146.80 ± 2.48

140,00 120,00 d

96.86 ± 5.73

100,00

h

79.51 ± 1.43

i

75.84 ± 2.53

80,00

b

56.54 ± 6.00

b

60,00

53.90 ± 3.22 c 47.35 ± 1.74 a

40,00

a

37.67 ± 3.70

34.39 ± 2.42 g

27.39 ± 2.70

j

22.90 ± 2.68

20,00 0,00 Shiitake

Irradiated with UV-A a-j

Oyster

Abalone

Irradiated with UV-B

Button

Irradiated with UV-C

: Values shown are mean values of 6 replicates ± SD. Values with different superscript letters are significantly different [b and c, h and i (p < 0.05), others (p < 0.01). The moisture content of mushrooms was around 80%, and irradiation was performed at 35 oC.

Figure 10. The conversion of ergosterol to vitamin D2 under UV-A, UV-B, and UV-C (Adapted from Jasinghe and Perera, 2006).

9. KINETICS OF CONVERSION OF ERGOSTEROL TO VITAMIN D Havinga et al., (1960) and Horst and Reinhardt, (1997) showed that photochemical cleavage of the B ring of ergosterol takes place under UV radiation, and then the intermediate (pre-vitamin D2) formed, undergoes subsequent thermal rearrangement to form vitamin D2 (ergocalciferol). However, at elevated temperatures, byproducts such as lumisterol and tachysterol could be formed, and the yield of the products is dependent on the thermal versus photochemical reactions (Havinga, 1973). Therefore in this conversion, the temperature of irradiation is very important since it plays a vital role in thermal rearrangement of pre-vitamin D2 to vitamin D2. In addition, high moisture content of mushrooms essentially has a dilution effect on the concentration of ergosterol, which is likely to bring about a lower conversion rate. We studied the combined effect of temperature of irradiation and moisture content of mushrooms on the conversion of ergosterol to vitamin D2.

Optimisation of the Conversion of Ergosterol in Mushrooms…

155

9.1. Conversion of Ergosterol in Mushrooms to Vitamin D2 Vitamin D2 contents in different types of edible mushrooms, subjected to UV-A irradiation for different time periods, are shown in figure 11. The results show that the conversion of ergosterol to vitamin D2 follows zero order reaction kinetics. The amounts of vitamin D2 produced over a period of time of irradiation were well correlated (R2 ≥ 0.98). The highest rate of conversion of ergosterol to vitamin D2 was observed in Oyster mushrooms while that of the lowest was observed in Button mushrooms. The conversion of ergosterol to vitamin D2 in different types of mushrooms under the said conditions can be predicted from the following equations: Oyster : amount of D2 = 6.78 + 0.890 * t Shiitake : amount of D2 = 11.97 + 0.643 * t Abalone : amount of D2 = 11.38 + 0.422 * t Button : amount of D2 = 2.85 + 0.278 * t where the amount of vitamin D2 converted from ergosterol is in µg/g dry matter and t is the time of irradiation in minutes.

45 y = 0,8901x + 6,7779 R2 = 0,9804

Vitamin D2 content (μg/g DM)

40 35

y = 0.6426x + 11.969 R2 = 0.9864

30 25

y = 0,4221x + 11,375 R2 = 0,9818

20 15 y = 0.2776x + 2.8532 R2 = 0.9813

10 5 5

10

15

20

25

30

35

40

45

Time of irradiation (Minutes) Button

Oyster

Shiitake

Abalone

Linear (Oyster)

Linear (Shiitake)

Linear (Abalone)

Linear (Button)

Figure 11. Effect of time of irradiation on the conversion of ergosterol to vitamin D2 in different types of edible mushrooms, irradiated at 30 oC and 80 ± 2.54% moisture content.

156

Conrad O. Perera and Viraj J. Jasinghe

9.2. Factorial Design A 2 x 2 full factorial design was used in our experiments to observe the effect of the variables on conversion of ergosterol in Shiitake mushrooms. The combined effects of temperature of irradiation, and moisture content of Shiitake mushrooms, on the conversion of ergosterol to vitamin D2, for four experimental conditions are given in table 4. The highest average vitamin D2 yield (44.8 μg/g DM) was observed from mushrooms, containing 80% moisture irradiated at a temperature of 35 oC, followed by those containing 60% moisture irradiated at 35 oC (29.7 μg/g DM), and those containing 80% moisture irradiated at 25 oC (21.3 μg/g DM). The lowest yield of vitamin D2 (8.05 μg/g DM) was observed from the mushrooms containing a moisture content of 60% irradiated at 25 oC. The average values were calculated from six experimental replicates. The tabulated data were analyzed by fitting them to 2 x 2 factorial model. The complete analysis of two-way ANOVA with repeated measures on both factors was carried out using the Vassar statistical analysis software and it is summarized in table 5. Based on p-values, it can be concluded that the main effects [factor A (temperature of irradiation, T oC) and factor B (moisture content of mushroom, M)] were statistically significant (p < 0.0001), however, interaction effect between the two factors was found to be insignificant (p > 0.05). A multiple regression analysis was performed to correlate vitamin D2 yield in Shiitake mushrooms with regards to temperature of irradiation and moisture content of the mushrooms, which resulted in following equation: D2 = -91.3 + 2.25 * T + 71 * M

Eqn (3)

The two factors considered, namely, temperature (T) and moisture content (M) were well correlated in the regression model equation (R2 = 0.98). The mean relative error and standard deviation of relative error were 5.24% and 4.90%, respectively. Table 4. Vitamin D2 content in Shiitake mushrooms irradiated at different moisture contents and temperatures Factors ToC 25 35 25 35

M (%) 60 60 80 80

Replicate analysis of vit D2 content (µg/g dry matter) 1 2 3 4 5 7.78 8.33 9.80 7.90 7.78 31.1 29.0 29.4 30.6 27.4 21.0 22.3 20.5 20.2 22.9 44.7 44.1 46.1 45.5 40.4

Averag e 8.05 29.7 21.3 44.8

6 6.81 30.5 21.0 47.6

The data was obtained from a 2 x 2 factorial design experiment.

Table 5. Analysis of variance for the experiment data obtained in 2 x 2 factorial design Source of variation T M TM Total

Sum of squares 3044.0 1206.0 4.98 4304.9

Degrees of freedom 1 1 1 3

Mean square

F

p

3044.03 1206.01 4.98

554.5 1747.8 2.69

<0.0001 <0.0001 <0.1619

Optimisation of the Conversion of Ergosterol in Mushrooms…

157

Figure 12. Modeling of the kinetic parameters for the conversion of ergosterol to vitamin D2 at 3 different temperatures (25oC, 30oC and 35oC).

9.3. Kinetic Model Parameters The results of our experimental studies showed that the conversion of ergosterol to vitamin D2 during irradiation of mushroom increased linearly with time (figure 11, table 4). This trend was consistent with all three temperatures used in this study. Hence a zero order reaction equation was assumed to model kinetics of conversion of ergosterol to vitamin D2: dC/dt = k C0

Eqn (1)

158

Conrad O. Perera and Viraj J. Jasinghe

where C is the concentration of ergosterol (g/g dry matter), t is the time of irradiation (s) and k is the reaction rate constant (s-1). One of the most common practices to model temperature dependence of reaction rates is to use Arrhenius equation (Banga and Singh, 1994). Yang et al. (1998) also used this approach to model influence of temperature on reaction rates of conversion of ergosterol into vitamin D2. Hence, in this study the temperature dependence of k is also described by an Arrhenius equation: k = k0 e(-Ea/RT)

Eqn (2)

where k0 is the reaction frequency factor, Ea is the activation energy of conversion of ergosterol to vitamin D2 (J mol-1), R is the gas constant (8.314 J K-1 mol-1) and T is the absolute temperature (K). The kinetic model described in Eqs. (1) and (2) was used as the basis of an effort to derive physical parameters for the experimental data obtained in this study. Thus, Eq. (1) was used to derive kinetic parameters at different irradiation times for each of the three experimental temperatures. The gradient of each of the straight lines in figure 12 constitutes the constant k for each temperature, which was used in Eq. (2) to obtain the reaction frequency factor (k0 = 7.32 s-1) and the activation of energy of conversion (Ea = 51.5 kJ mol-1 and R2 = 0.94. Yang et al. (1998) reported the value of activation energy of conversion of ergosterol to vitamin D2 to be in the range of 4.2–28.7 kJ mol-1 depending upon the solvent and light absorbance used. However, activation energy for the conversion of ergosterol in Shiitake mushrooms to vitamin D2 by UV-A irradiation was much higher, indicating a lower degree of conversion in the natural product compared to that in an organic solvent. These kinetic parameters can be used to estimate the amount of vitamin D2 yield for different times of irradiation within the temperature range studied.

10. BIOAVAILABILITY OF VITAMIN D2 FROM IRRADIATED SHIITAKE MUSHROOMS Bioavailability has been defined as ‘that fraction of an oral dose (parent compound or active metabolite) from a particular preparation that reaches the systemic circulation’ (Schumann et al. 1997). The serum concentration of 25(OH)D is the barometer of vitamin D status (Holick, 2001), and therefore this measurement can be used in bioavailability studies of vitamin D. Bioavailability of vitamin D2 from irradiated Shiitake mushrooms was studied in a rat model. All animals involved in this study were treated in a humane fashion in accordance with the guidelines of the National University of Singapore, and disposed of in a manner prescribed by the animal holding unit. All the subjects survived until they were sacrificed at the end of the study and neither physiological nor behavioural abnormalities were observed in any group. The ranges of measurements are given in table 6. The bodyweights at the beginning and end of the study did not differ among groups (p < 0.01). Furthermore, the lengths of femur bones did not differ among groups. No significance

Optimisation of the Conversion of Ergosterol in Mushrooms…

159

difference (p < 0.01) was shown in daily dietary intakes of Group 3 and Group 2. Group 1 was used to evaluate vitamin D deficiency status of animals before the administration of test diets. Table 6. Basic measurements of group physical parameters(1) Measurement Average body weight at commencement of the experiment (g) Body weight when sacrificed (g)

Group 12

Group 23

Group 33

54.32 ± 5.12 89.40 ± 4.63

99. 24 ± 4.64 311.87 ± 23.36

Average daily dietary intake (g) Length of right femur (mm) Length of left femur (mm)

10.56 ± 2.51 18.89 ± 1.15 18.70 ± 1.23

22.78 ± 3.42 20.68 ± 1.96 19.87 ± 3.56

93.92 ± 7.68 294.55 ± 19.06 22.35 ± 3.14 20.86 ± 2.05 20.11 ± 1.33

1

Group 1, on vitamin D deficient diet for one-week; Group 2, on vitamin D deficient diet for one-week and then irradiated mushrooms + vitamin D deficient diet for four weeks; Group 3, on vitamin D deficient diet for one-week, and then non-irradiated mushrooms + vitamin D deficient diet for four weeks. 2 Measurments after one week; n = 6, mean ± SD. 3 Measurments after five weeks; n = 12, mean ± SD.

Figure 13 shows the growth charts and daily dietary intakes of the test groups. The growth chart shows slight increment of the growth curve of Group 2 over the Group 3 after one week, when the test diets were begun to administrate. However, this difference was not significant (p = 0.603). In this study, the difference in daily dietary intake of Groups 2 and 3 was found to be not significant (p = 0.853). Assuming only vitamin D2 was formed in mushrooms under UV radiation; Group 2 received 1 μg of vitamin D2 from irradiated mushrooms while Group 3 received a similar diet but lacking the vitamin D2. Since the animals were housed under incandescent light, cutaneous synthesis of vitamin D was not expected to interfere with the results. The bone mineral density (BMD) of femur bones was measured by Lunar DPX-L Dual-Energy X-ray Bone Densitometer (DEXA); software version 1.3, (Lunar DPX-L, Lunar Corp., Madison, WI, USA). Figure 14 illustrates femur BMD of the three different groups. Femur BMDs of Group 2 were significantly higher (p < 0.01) than those of the other two groups. Difference in BMD between Group 1 and Group 3, was found to be not significant (p = 0.332). In addition, the BMD values of right and left femurs within the groups were similar and the difference between values was not significant (Group 1, p = 1.00; Group 2, p = 0.434; Group 3; p = 0.487). DEXA is a useful tool for measuring intact and excised rat leg bone mineral density (Nagy et al. 2001). In this study, it was shown that vitamin D2 from irradiated mushrooms increased femur BMD of laboratory rats. Since vitamin D is directly involved in bone mineralization (Kaastad et al. 2001; Erben et al. 2002), the results of the current study show that in laboratory rats, vitamin D2 from irradiated edible mushrooms has an important positive effect on the femur bone mineralization, especially during the period that the rats lay-down their skeleton. Serum 25(OH)D was analysed using Gamma-B 25(OH)D 125I RIA kit (DiaSorin and IDS Ltd, Boldon, UK.), serum calcium levels were measured by an automated VITROS 950

160

Conrad O. Perera and Viraj J. Jasinghe

Daily Dietary intake & Weight of the rats (g)

chemistry system (Ortho-Clinical Diagnostics, Inc, Raritan, NJ, USA), at the Singapore National University Hospital. Serum 25(OH)D, and calcium concentrations of the groups are shown in table 7.

300 250 200 150 100 50 0 0

2

4

6

8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 Days

Weight of the rats (Vitamin D+ group) Weight of the rats (Vitamin D- group) Dietary intake (Vitamin D+ group) Dietary intake (Vitamin D- group)

Figure 13. The growth charts and daily dietary intakes of the experimental group and the control group.

b

265.00 ± 20.40

300

b

250.31 ±14.17

250

228.25 ± 10.76

230.25 ± 15.56

200 (mg/cm2)

Femur Bone Mineral Density

a

a

a

a 220.50 ± 7.97 222.50 ± 14.94

150 100 50 0 Left

Right

Group 1

Left

Right

Group 2

a

Left

Right

Group 3

: Values with different superscript letters are significantly different (p < 0.01). (Adapted from Jasinghe et al. 2005).

Figure 14. Femur BMD of initial, control, and experimental group.

Optimisation of the Conversion of Ergosterol in Mushrooms…

161

Table 7. Serum 25(OH)D and serum calcium concentrations of groups1 Variable Serum 25(OH)D (nmol/L) Serum Calcium (mmol/L)

Group 12 18.06 ± 5.26a 2.61 ± 0.32a

Group 23 129.42 ± 22.0b 2.28 ± 0.11b

Group 33 6.06 ± 1.09c 1.60 ± 0.24c

1

Each sample was subjected to triplicate analysis. Group 1, on vitamin D deficient diet for one-week at the beginning; Group 2, on irradiated mushrooms + vitamin D deficient diet for four weeks; Group 3, on non-irradiated mushrooms + vitamin D deficient diet for four weeks. Means with different superscript letters are significantly different, P < 0.01.The statistical analyses were based on ANOVA and Tukey’s HSD test. 2 mean value ± SD; n = 6. 3 mean value ± SD; n = 12.

The results show that serum 25(OH)D concentration of Group 2 clearly differs from that of Group1 and 3. Serum 25(OH)D concentration of Group 2, which received 1 μg of vitamin D2 daily from mushrooms for four weeks, was 129.42 ± 22.00 nmol/L whereas it was only 6.06 ± 1.09 nmol/L in Group 3, which received no vitamin D2. A decrease in 25(OH)D concentration was observed in Group 3 compared with Group 1 but on the other hand, a remarkable increase was observed in Group 2. The serum calcium levels among groups were also significantly different. In contrast to what might have been expected, the serum calcium level of Group 2 was significantly lower compared with Group 1. This could be due to a higher rate of bone meneralization in Group 2 (which received vitamin D2 from mushrooms) compared with Group 1. This is supported by the observation that there was a significantly higher BMD and lengths of femur bones in Group 2. In addition, lowered serum levels of PTH, raised serum ionised calcium levels, and an age related decline in duodenal calcium absorption have also been reported and could be contributing factors to this difference (Agnusdei et al. 1998; Schulz and Morris, 1999). The current results clearly indicate that vitamin D2 from irradiated mushrooms was well absorbed in the laboratory rats since the serum concentration of 25(OH)D of the experimental group was remarkably higher than that of the control group. Vieth and Milojevic (1995) reported a value of 58 ± 8 nmol/L of 25(OH)D in a similar rat study using vitamin D3 as a supplement. In this study, the quantities of vitamin D administered were considerably higher than the amount of vitamin D given by Vieth and Milojevic (1995), and this may be the reason for the observation of high values of serum 25(OH)D in this study. Since vitamin D influences several steps in the active calcium transport system, (Gueguen and Pointillart, 2000), measurement of serum calcium concentration is a useful tool to predict vitamin D deficiency. Serum calcium concentration of Group 2 was significantly higher than that of the value for Group 3. Thus, it was clearly indicated that Group 3, fed only on vitamin D deficient diet, was indeed deficient in vitamin D. The dose of vitamin D2 for rats, which was used in this study, was around 3 μg/Kg body weight. If this dose is converted to an average body weight of a human (70 Kg), it is around 200 μg/day. This is around 20 times higher compared with current RDA of vitamin D for adults (10 μg/day), which some workers believe to be inadequate (Cheetham, 1999; Heaney, 2000; Vieth, 2000), and even up to100 μg vitamin D3/day is a safe intake (Vieth et al. 2001).

162

Conrad O. Perera and Viraj J. Jasinghe

Irradiated edible mushroom powder could be used in fortification of human food supplements or form of fresh irradiated mushrooms could be used for human consumption.

REFERENCES Anderson, R. R., and Parrish, J.A. (1981). The optics of human skin. J. Invest. Dermatol, 77, 13-19. Agnusdei, D., Civitelli, R., Camporeale, A., Parisi, G., Gennari, L., Nardi, P., and Gennari, C. (1998). Age-related decline of bone mass and intestinal calcium absorption in normal males. Calcif. Tissue Int, 63, 197-201. Banga JR, and Singh RP (1994) Optimization of air drying of foods, J Food Eng 23, 189-211. Berube, S., Diorio, C., Verhoek-Oftedahl, W., and Brisson, J. (2004). Vitamin D, calcium, and mammographic breast densities. Cancer Epidemiol. Biomarkers Prev, 13, 1466-72. Boucher, B.J. (1998). Inadequate vitamin D status: does it contribute to the disorders comprising syndrome ‘X’?. Br. J. Nutr, 79, 315-327. Braun, M., Fub, W., and Kompa, K.L. (1991). Improved photosynthesis of previtamin D by wavelengths of 280-300 nm. J. Photochem. Photobiol. A: Chem, 61, 15-26. Braun, M.M., and Tucker, M.A. (1997). A role for photoproducts of vitamin D in the etiology of cutaneous melanoma? Med. Hypotheses, 48, 351-4. Carlton-Conway, D., Tulloh, R., Wood, L. and Kanabar, D. (2004). Vitamin D deficiency and cardiac failure in infancy. J. R. Soc. Med, 97, 238-239. Chan, T. Y. K. 2004. Vitamin D Deficiency and Susceptibility to Tuberculosis. Calcified Tissue International, 66(6), 476-478 Cheetham, C. H. (1999). Time for a tablet containing high doses of vitamin D alone. Br. Medic J., 318, 1284. Chen, T.C., and Holick, M.F. (2003). Vitamin D and prostate cancer prevention treatment. Trends Endocrinol. Metab, 14, 423-30. Chen, T.C., Holick, M.F., Lokeshwar, B.L., Burnstein, K.L., and Schwartz, G.G. (2003). Evaluation of vitamin D analogs as therapeutic agents for prostate cancer. Recent Results Cancer Res, 164, 273-88. Cheng, S., and Coyne, D. (2007). Vitamin D and outcomes in chronic kidney disease. Curr. Opin. Nephrol. Hypertens, 16(2), 77-82 Clemens, T.L., Henderson, S.L., Adams, J.S., and Holick, M.F. (1982). Increase skin pigment reduces the capacity of skin to synthesize vitamin D3. Lancet, 319, 74-76. Danish food composition data bank (2004), Danish Institute for Food and Veterinary Research. http://www.foodcomp.dk/fcdb_foodcomplist.asp?CompId=0023 (Last accessed 09/04/2007) Du, X., Greenfield,H., Fraser,D.R., Ge,K., Trube, A. and Wang, Y. 2001. Vitamin D deficiency and associated factors in adolescent girls in Beijing. Am J Clin Nutr, 74, 494– 500. El-Haschimi, K., Pierroz, D.D., Hileman, S.M., Bjorbaek, C., and Flier, J.S. (2000). Two defects contribute to hypothalamic leptin resistance in mice with diet-induced obesity. J. Clin. Invest, 105, 1827-32.

Optimisation of the Conversion of Ergosterol in Mushrooms…

163

Erben, R.G., Mosekilde, L., Thomsen, J.S., Weber, K., Leyshon, A., Smith, S.Y., and Phipps, R. (2002). Prevention of bone loss in ovariectomized rats by combined treatment with risedronate and 1α,25-dihydroxyvitamin D3. J. Bone Miner Res, 17, 1498-1511. EURODIAB ACE Study Group (2000). Variation and trends in incidence of childhood diabetes in Europe. Lancet, 355, 873-76. FAO/WHO, 1998 expert consultation on human vitamin and mineral requirements (ftp://ftp.fao.org/es/esn/nutrition/Vitrni/pdf/CHAPTER08.pdf) (last accessed 20/03/2007). Fehm, H., Smolinik, R., Kern, W., McGregor, G.P., Bickel, U., and Born, J. (2001). The melanocortin melanocyte-stimulating hormone/adrenocorticotropin4-10 decrease body fat in humans. J. Clin. Endocrinol. Metab, 86, 1144-48. Feldman, D., Malloy, P.J., and Gross, C. (1996). Vitamin D: metabolism and action. In: R. Marcus, D. Feldman, J. Kelsey, (Eds). Osteoporosis. San Diego, CA. Academic Press, pp. 205-235. Feldman, D. (1999). Vitamin D. parathyroid hormone, and calcium: A complex regulatory network. Am. J. Med, 107, 637-39. Freeman RG, Cockerell EG, Armstrong J, and Knox JM (1962) Sunlight as a factor influencing the thickness of epidermis. J. Invest. Dermatol 39, 295-98. Gill, T.P., Antipatis, V.J., and James, W.P.T. (1999). The global epidemic of obesity. Asia Pacif. J. Clin. Nutr, 8, 75. Grant, W.B. (2002a). An ecologic study of dietary and solar ultraviolet-B links to breast carcinoma mortality rates. Cancer. 94, 272-81. Grant, W.B. (2002b). An estimate of premature cancer mortality in the U.S. due to inadequate doses of solar ultraviolet-B radiation. Cancer, 94, 1867-75. Green, T.J., Skeaff, C.M., Rockell, E., Venn, B.J., Lambert, A., Todd, J., Khor, G.L., Loh, S.P., Muslimatun, S., Agustina, R., and Whiting, S.J. (2007). Vitamin D status and its association with parathyroid hormone concentrations in women of child-bearing age living in Jakarta and Kuala Lumpur. Eur. J. Clinical Nutr. 2007 Mar 7; [Epub ahead of print] Gueguen, L., and Pointillart, A. (2000). The bioavailability of dietary calcium. J. Am. Coll. Nutr, 19, 119S-136S. Havinga, E., Kock, R.J.D., and Rappoldt, M.P. (1960). The photochemical interconversions of provitamin D, lumisterol, previtamin D and tachysterol. Tetrahedron, 11, 276-284. Havinga, E. (1973). Vitamin D, example and challenge. Experientia, 29,1181–1193. Heaney, R.P. (2000). Vitamin D: how much do we need, and how much is too much? Osteoporos Int, 11, 553-55. Heaney, R.P. (2003). Long-latency deficiency disease: insights from calcium and vitamin D. Am. J. Clin. Nutr, 78, 912-19. Heath, K.M., and Elovic, E.P. (2006). Vitamin D deficiency: implications in the rehabilitation setting. Am. J. Phys. Med. Rehabil. 85(11): 916-23. Holick, M.F. (2001). Meeting the vitamin D needs of the elderly. Nutrition and the M.D, 27, 1-4. Holick, M.F. (2003). Evaluation of function of vitamin D. Recent Results Cancer Res, 164, 328. Holick, M.F. (2004). Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and cardiovascular disease. Am. J. Clin. Nutr. 80(6) Suppl, 1678S-88S.

164

Conrad O. Perera and Viraj J. Jasinghe

Holick, M.F. (2005). The vitamin D epidemic and its health consequences. J. Nutr. 135(11), 2739S-48S Holick, M.F. (2006). High prevalence of vitamin D inadequacy and implications for health. Mayo Clinic Proc. 81(3), 353-73 Hollosy, F. (2002). Effect of ultraviolet radiation on plant cells. Micron, 33, 179-97. Holvik, K., Meyer, H.E., Hung, E., Brunvand, L. 2005. Prevalence and predictors of vitamin D deficiency in five immigrant groups living in Oslo, Norway: The Oslo immigrant health study. Eur. J. Clin. Nutr, 59(1), 57-63. Horst, R.L., and Reinhardt, T.A. (1997). Vitamin D metabolism. In Vitamin D. Feldman D., Glorieux F. H., Pike J. W., (Eds.), pp. 13 – 31, Academic Press, New York. Hypponen, E., Laara, E., and Reunanen, A. (2001). Intake of vitamin D and risk of type 1 diabetes: a birth-cohort study. Lancet, 358(9292), 1500-503. Jasinghe, V. J., Perera, C.O. (2005). Distribution of ergosterol in different tissues of mushrooms and its effect on the conversion of ergosterol to vitamin D2 by UV irradiation. Food Chem, 92(3), 541-546 Jasinghe, V. J., Perera, C. O., Barlow, P.J. (2005). Bioavailability of vitmin D from irradiated mushrooms; an in-vivo study. British J. Nutrn, 93(6), 951-955 Jasinghe, V. J., Perera, C. O. (2006). Ultraviolet irradiation: the generator of Vitamin D2 from edible mushrooms. Food Chemistry, 95(4), 638-643. Jasinghe, V. J., Perera, C.O., and Barlow, P.J. (2006). Vitamin D2 from irradiated mushrooms significantly increase femur bone mineral density in laboratory rats; an in-vivo study. J. Toxicology and Environmental Health Part A, 69, 1979-1985. Jasinghe, V.J., Perera, C.O., Sablani, S. (2007). Kinetics of conversion of ergosterol to vitamin D in shiitake mushroom. J. Food Engineering, 79, 864-869 Jones, G., Blizzard, C.L., Riley, M.D., Parameswaran, V., Greenaway, T.M., and Dwyer, T. (1999). Vitamin D levels in prepubertal children in Southern Thasmania: prevalence and determinants. Eur. J. Clin. Nutr, 52, 824-29. Kaastad, T.S., Reikeras, O., Halvorsen, V., Falch, J.A., Obrant, K.J., and Nordsletten, L. (2001). Vitamin D deficiency and ovariectomy reduced the strength of the femoral neck in rats. Calcif. Tissue Int, 69, 102-108. Kamradt, J., Rafi, L., Mitschele, T., Meineke, V., Gartner, B.C., Wolfgan, T., Holick, M.F., and Reichrath, J. (2003). Analysis of the vitamin D system in cutaneous malignancies. Recent Results Cancer Res, 164, 259-69. Kamycheva, E., Joakimsem, R.M., and Jorde, R. (2002). Intakes of calcium and vitamin D predict Body Mass Index in the population of northern Norway. J. Nutr. 132, 102-106. Kira, M., Kobayashi, T., and Yoshikawa, K. (2003). Vitamin D and skin. J. Dermatol, 30, 429-437. Lawson, M., Thomas, M., and Hardiman, A. (1999). Dietary and lifestyle factors affecting plasma vitamin D levels in Asian children living in England. Eur. J. Clin. Nutr, 53, 268272. Lehtonen-Veromaa, M., Mottonen, T., Irjala, K., Karkkainen, M., Lamberg-Allardt, C., Hakola, P., and Viikari, J. (1999). Vitamin D intake is law and hypovitaminosis D common in healthy 9 to 15-year-old Finish girls. Eur. J. Clin. Nutr, 53, 746-51. Linda, M.O. (2003). Getting more from vitamins and minerals. Food Tech, 57, 87-88. Luong, K.V., and Nguyen, L.T. (2006). Vitamin D and cardiovascular disease. Curr. Med. Chem, 13(20), 2443-7.

Optimisation of the Conversion of Ergosterol in Mushrooms…

165

MacLaughlin, J., and Holick, M.F. (1985). Aging decreases the capacity of human skin to produce vitamin D3. J. Clin. Invest, 76, 1536-38. Majewski, S., Kutner, A., and Jablonska, S. (2000), Vitamin D analogs in cutaneous malignancies. Curr. Pharm. Des, 6, 829-38. Marriott, B,M, (1997) Vitamin D supplementation: a word of caution. Ann. Intern. Med, 127, 231-33. Matsuoka, L.Y., Ide, L., Wortsman, J., Maclaughlin, J.A., and Holick, M.F. (1987). Sunscreens suppress cutaneous vitamin D3 synthesis. J. Clin. Endocrinol. Metab, 64, 1165-68. Mattila, P.H., Lampi, A.M., Ronkainen, R,. Toivo, J., and Piironen, V. (2002). Sterol and vitamin D2 contents in some wild and cultivated mushrooms. Food Chem, 76, 293 -98. Mattila, P.H., Piironen, V.I., Uusi-Rauva, E.J., and Koivistoinen, P.E. (1994). Vitamin D contents in edible mushrooms. J. Agric. Food Chem, 42, 2449–53. Mau, J.L., Chen, P.R., and Yang, J.H. (1998). Ultraviolet irradiation increased vitamin D2 content in edible mushrooms. J. Agric. Food Chem, 46, 5269-72. Mawer, E.B., Davies, M., Still, P.E., Jones, G., Knutson, J.C., and Bishop, C.W. (1995). 1, 24-Dihydroxyvitamin D2, a biologically active analog of vitamin D, is a naturally occurring metabolite in humans. Bone, 17, 321. McKinney, P.A. (2001) on behalf of the EURODIAB Seasonality of Birth Group*. Seasonality of birth in patients with childhood type 1 diabetes in 19 European regions. Diabetologia, 44(S3), B67-B74. Mehta, R.G., and Mehta, R.R. (2002). Vitamin D and cancer. J. Nutr. Biochem, 13, 252-264. Mellanby. E. (1919). An experimental investigation on rickets. Lancet, 196, 407-412. Mona, S. Calvo, M.S., Whiting, S.J., Barton, C.N. 2005. Vitamin D Intake: A Global Perspective of Current Status. J. Nutr. 135, 310-316 Moore, C., Murphy, M.M., Keast, D.R., and Holick, M.F. (2004). Vitamin D intake in the United States. J Am Diet Ass, 104, 980-83. Morgan, S.L. (2001). Calcium and vitamin D in osteoporosis. Rheumatic Diseases Clinics of North America, 27, 101-130. Nagy, T.R., Charles, W., and Jing, L.I. (2001). Validation of Peripheral Dual-Energy X-Ray Absorptiometry for the Measurement of Bone Mineral in Intact and Excised Long Bones of Rats. J. Bone Miner Res, 16, 1682-87. Napoli, J.L., Fivizzani, M.A., Schnoes, H.K., and Deluca, H.F. (1979). Synthesis of vitamin D5: its biological activity relative to D3 and D2. Arch. Biochem. Biophys, 197(1), 119125. Need, A.G., Morris, H.A., Horowitz, M., and Nordin, C. (1993). Effect of skin thickness, age, body fat, and sunlight on serum 25-hydroxyvitamin D. Am. J. Clin. Nutr, 58, 882-85. Norman, P., Moss, I., Minder, S., Gosling, M., and Powell, J. (2002). Maternal and postnatal vitamin D ingestion influences rat aortic structure, function and elastin content. Cardiovascular Res, 55, 369-74. Nutrition Source, (2006). Vitamins. Harvard School of Public Health. http://www.hsph.harvard.edu/nutritionsource/vitamins.html (accessed 09/03/2007) Ogunkolade, B.W., Boucher, B.J., Fairclough, P.D., Hitman, G.A., Dorudi, S., Jenkins, P.J., and Bustin, S.A. (2002). Expression of 25-hydroxyvitamin D-1-α-hydroxylase mRNA in individuals with colorectal cancer. Lancet, 359, 1831-32.

166

Conrad O. Perera and Viraj J. Jasinghe

O’Kelly, J., and Koeffler, H.P. (2003). Vitamin D analogs and breast cancer. Recent Results Cancer Res, 164, 333-48. Peters, U., McGlynn, K.A., Chatterjee, C., Gunter, E., Garcia-Closas ,M., Rothman, N., and Sinha, R. (2001). Vitamin D, calcium, and vitamin D receptor polymorphism in colorectal adenomas. Cancer Epidemiol Biomarkers Prev, 10, 1267-74. Perera, C.O., Jasinghe, V.J., Ng, F.L., Mujumdar, A.S. 2003. The effect of moisture content on the conversion of ergosterol to vitamin D in shiitake mushrooms. Drying Technology, 21(6),1093-1101 Peterlik, M., and Cross, H.S. (2006). Dysfunction of the vitamin D endocrine system as common cause for multiple malignant and other chronic diseases. Anticancer Res, 26(4A), 2581-8 Pfeifer, M., Begerow, B., Minne, H.W., Nachtigall, D., and Hansen, C. (2001). Effects of short-term vitamin D3 and calcium supplementation on blood pressure and parathyroid hormone levels in elderly women. J. Clin. Endocrinol. Metab, 86, 1633-37. Ponsonby, A.L., McMichael, A., and van der Mei, I. (2002). Ultraviolet radiation and autoimmune disease: insights from epidemiological research. Toxicology, 181, 71-78. Reaven, G.M. (1995). The fourthmusketeer – from Alexandre Dumas to Claude Bernard. Diabetologia, 38, 3-13. Samuelsson, U., Johansson, C., and Ludvigsson, J. (1999), Month of birth and risk of developing insulin dependent diabetes in southeast Sweden. Archives of Disease in Childhood, 81, 143-46. Schulz, S.R., and Morris, H.A. (1999.) Ionized calcium and bone turnover in the estrogendeficient rat. Calcif. Tissue Int, 65, 78-82. Schumann, K., Classen, H.G., Hages, M., Langenhol, R.P., Pietrzik, K., and Biesalski, H.K. (1997). Bioavailability of oral vitamins, minerals and trace elements in perspective. Arzneimittelforschung, 47, 369-80. Schwartz, M.W. (2001). Progress in the search for neuronal mechanisms coupling type 2 diabetes to obesity. J. Clin. Invest, 108, 963-64. Shanna, N.O.D., Scanlon, K.S., Cogswell, M.E., Gillespie, C., Hollis, B.W., Looker, A.C., Allen, C., Doughertly, C., Gunter, E.W., and Bowman , B.A .(2002). Hypovitaminosis D prevalence and determinants among African American and white women of reproductive age: third National Health and Nutrition Examination Survey, 1988-1994. Am. J. Clin. Nutr. 76, 187-92. Songini, M., and Casu, A. (2001). The Sardinian Collaborative Group for Epidemiology of IDDM., Ashkenazi I., and Laron Z. Seasonality of Birth in children (0-14 years) and young adults (0-29 years) with type 1 diabetes mellitus in Sardinia differe from that in general population. J. Pediatric Endocrinol. Metab, 14, 781-83. Speer, G., Cseh, K., and Winkler, G. (2001). Vitamin D and estrogen receptor gene polymorphisms in type 2 diabetes mellitus and in android type obesity. Eur. J. Endocrinol, 144, 385-89. Takamura, K., Hoshino, H., Tatsuyuki, S. and Hisao, A. (1991). Determination of vitamin D, in shiitake mushroom (Lentinus edodes) by high-performance liquid chromatography. Journal of Chromatography, 545, 201-204. Takeuchi, A., Okano, T., Teraoka, S., Murakami, Y., and Koba-yashi, T. (1984). High performance liquid chromatographic determination of vitamin D in foods, feeds and

Optimisation of the Conversion of Ergosterol in Mushrooms…

167

pharmaceuticals by successive use of reversed-phase and strait-phase columns. J. Nutr. Sci. Vitaminol, 30, 11-25. Tangpricha, V., Koutkia, P., Rieke, S.M., Chen, T.C., Perez, A.A., and Holick, M.F. (2003). Fortification of orange juice with vitamin D: a novel approach for enhancing vitamin D nutritional health. Am. J. Clin. Nutr, 77, 1478-83 The EURODIAB Substudy 2 Study Group. (1999). Vitamin D suppliment in early childhood and risk for type 1 (insulin-dependant) diabetes mellitus. Diabetologia, 42, 51-54. Tjellesen L, Gotfredsen A, and Christiansen C (1985) Different actions of vitamin D2 and D3 on bone metabolism in patients treated with phenobarbitone / phenytoin. Calcif. Tissue Int. 37, 218-22. Van der Wielen, R.P.J., Lowik, M.R.H., Van den Berg, H., deGroot, L.C.P.G.M., Haller, J., Moreiras, O., and van Staveren, W.A. (1995). Serum vitamin D concentrations among elderly people in Europe. Lancet, 346, 207-10. Vayalil, P.K., Elmets, C.A., and Katiyar, S.K. (2003). Treatment of green tea polyphenols in hydrophilic cream prevents UVB-induced oxidation of lipids and proteins, depletion of antioxidant enzymes and phosphorylation of MAPK proteins in SKH-1 hairless mouse skin. Carcinogenesis, 24, 927-36 Vieth, R. (2000). Problems with direct 25-hydroxyvitamin D assays, and the target amount of vitamin D nutrition desirable for patients with osteoporosis. Osteoporos Int, 11, 635-36. Vieth, R., and Milojevic, S. (1995). Moderate vitamin D3 supplementation lowers serum 1, 25-dihydroxy-vitamin D3 in rats. Nutr. Res, 15, 725-31. Vieth, R., Cole, D.E., Hawker, G.A., Trang, H.M., and Rubin, L.A. (2001). Wintertime vitamin D insufficiency is common in young Canadian women, and their vitamin D intakes do not prevent it. Eur. J. Clin. Nutr, 55, 1091-97. Wang, L., Whitlatch, L.W., Flanagan, J.N., Holick, M.F., and Chen, T.C. (2003). Vitamin D autocrine system and prostate cancer. Recent Results Cancer Res, 164, 223-37. Webb, A.R., Decosta, B.R., and Holick, M.F. (1989). Sunlight regulates the cutaneous production of vitamin D3 by causing its photodegradation. J. Clin. Endocrinol. Metab, 68, 882-87. Webb AR, Pilbeam C, Hanafin N, and Holick MF (1990) An evaluation of the relative contributions of exposure to sunlight and of diet to the circulating concentrations of 25hydroxyvitamin D in an elderly nursing home population in Boston. Am. J. Clin. Nutr. 51, 1075-81. Wharton, B., and Bishop, N. (2003). Rickets. Lancet, 362, 1389-400. Wilkinson, R.J., Llewelyn, M., Toossi, Z., Patel, P., Pasvol, G., Lalvani, A., Wright, D., Latif, M., and Davidson, R. N. 2000. Influence of vitamin D deficiency and vitamin D receptor polymorphisms on tuberculosis among Gujarati Asians in west London: a case-control study. The Lancet, 355(9204), 618-621. WHO. (1997). Obesity: preventing and managing the global epidemic. 1997: Geneva. Wortsman, J., Matsuoka, L.Y., Chen, T.C., Lu, Z., and Holick, M.F. (2000). Decreased bioavailability of vitamin D in obesity. Am. J. Clin. Nutr, 72, 690-93. Yang L, Tan T, and Qi Y (1998) Study on optimum conditions and kinetics for the reaction of ergosterol under ultraviolet ray. Chemical Reaction Engineering and Technology 14, 117124.

168

Conrad O. Perera and Viraj J. Jasinghe

Zittermann A, Schleithoff SS, Tenderich G, Berthold HK, Korfer R, and Stehle P (2003) Low vitamin D status: A contributing factor in the pathogenesis of congestive heart failure? J. Am. Coll. Cardiology. 41, 105-12. Zittermann, A. (2003). Vitamin D in preventive medicine: are we ignoring the evidence? Br. J. Nutr, 89(5), 552-72.

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 169-194

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 5

PROTEIN HYDROLYSIS WITH ENZYME RECYCLE BY MEMBRANE ULTRAFILTRATION Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto Department of Chemical Engineering, University of Granada, Spain

ABSTRACT Enzymatic hydrolysis allows to improve the functional, nutritional and immunological properties of proteins. For instance, protein hydrolysates are used in the food industry as ingredients in hypoallergenic formulae and clinical nutrition. Conventional batch hydrolysis of proteins has been traditionally used to obtain protein hydrolysates due its simple operation. However, there are several disadvantages associated to this method, the high enzyme consumption being the main one. Among the solutions assayed to enhance the yield of the process, enzyme immobilisation onto highly activated supports allows to work continuously and reuse the enzyme. Since there are loss of enzyme activity and constrains for the diffusion into the support, the feasibility of this technique is limited. Continuous reaction and simultaneous separation of products from the reaction mixture can be achieved in a continuous membrane recycle reactor. Here, the low molecular weight peptides generated permeate through an ultrafiltration module with the appropriate molecular weight cut-off, while the enzyme (which acts in soluble form) is continuously recycled to the reaction tank. As important drawbacks, permeate flux decline due to membrane fouling and frequent purges are required to eliminate nonreacting substrate which involves severe difficulties in the control process. In order to profit from the advantaged of both batch and continuous membrane recycle reactor, the objective of this research work was to design and optimise a reactor for the production of low antigenicity protein hydrolysates. The operation proposed comprises 3 consecutive steps: a) hydrolysis in a stirred tank reactor; b) ultrafiltration of the reaction mixture through a membrane with full retention of the enzyme; c) enzyme recycling, in which the retentate is returned to the tank reactor for a new hydrolysis. The materials employed were a whey protein concentrate as substrate, a subtilisin from Bacillus licheniformis as protease and an 8 kDa flat polyethersulfone membrane. The pH-stat method was used to monitor the hydrolysis reaction. The molecular weight profile and the antigenicity of the hydrolysates were determined by HPLC and ELISA, respectively. Regarding the process kinetics, zero-order for the substrate and second order for the enzyme deactivation were identified. Process optimisation involved the calculation of the

170

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto optimum number of enzyme uses that minimised the enzyme consumption, subject to a required productivity. The performance of the system was compared at several temperatures to that of the conventional stirred tank reactor. Significant enzyme savings were achieved, which demonstrate the viability of this approach.

INTRODUCTION Enzymatic hydrolysis of food proteins is a well-known method to modify and improve their functional and nutritional properties. Protein hydrolysates are usually classified into three categories depending on the degree of hydrolysis achieved: a) low degree of hydrolysis products (typically lower than 10%) are used for the improvement of functional properties such as solubilisation, emulsifying, gelling or foaming power and water or oil adsorption; b) high degree of hydrolysis products (i.e. higher than 10%) are excellent sources of nitrogen in hypoallergenic formulae and enteral diets for children and sick adults. For this purpose, many sources of proteins are commonly employed: from vegetal proteins such as soy bean to animal proteins such as egg, meat or fish proteins. Nevertheless, hydrolysates are often obtained from milk proteins (both casein and whey proteins) due to their high nutritional value, low bitterness and low antigenicity; c) some particular enzyme-substrate systems yield peptides modulating biological processes. Since bioactive peptides may be ingredients in new functional foods they are receiving more and more attention by the pharmaceutical and food industry.

General Aspects on the Enzymatic Hydrolysis of Food Proteins The primary role of food proteins is to supply the adequate quantity and proportion of nitrogen and amino acids required for the synthesis of protein-based tissues of animal organisms. It is well-known that all proteins do not pose equal capacity to promote tissue creation, being the amino acid content and profile directly related to the nutritional quality of proteins. Among other parameters, quality can be quantified by means of the net protein utilization (NPU). It means the fraction of assimilated protein out of the total protein ingested. Milk proteins show the highest NPU values among the main edible sources of proteins. For instance, whey proteins show a NPU around 0.8 while soy bean or wheat gluten proteins just reach 0.6 and 0.35, respectively. The source of proteins for the manufacturing of hydrolysates is clearly linked to the final use of the product: for example meat protein hydrolysates are frequently used as flavour enhancers in soups, sauces and prepared meals. On the other hand, milk proteins are preferred for clinical applications or special food formulations due to their superior nutritional properties. With respect to the catalyst of the hydrolysis reaction, an enzyme is a protein with catalytic properties due to its power of specific activation (Dixon and Webb, 1979). Enzymes typically catalyse metabolic reactions inside cells. As a result, its catalytic behaviour is characterized by mild pH and temperature conditions, specificity towards one substrate and high reaction rates. Proteases are classified according to their source (animal, plant, fungal or microbial), their catalytic action (endopeptidases or exopeptidases) and the chemical nature of

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

171

the catalytic site. Endopeptidases are mainly used in food protein hydrolysis although exopeptidases are often combined with them in order to achieve an extensive degradation (Adler-Nissen, 1986). Regarding their catalytic nature there are four main classes of endopeptidases used in food technology applications: serine, cysteine, aspartic and metalloproteins. The catalytic properties of serine proteases are connected with the hydroxyl group into the active site and exhibit maximal activity at alkaline pH. The activity of cysteine is closely related to that of serine protease, having a thiol group into the active site instead of the –OH. As a consequence, their maximum activity is found at nearly neutral pH. The same phenomena occur for metalloproteases, which contains a metal atom (usually Zn). Eventually, aspartic proteases have a carboxyl group from aspartic acid in the active site so that the maximum activity is obtained at acid pH. Be that as it may, a high number of commercial food-grade enzyme preparations are usually mixtures of individual enzymes and stabilizing chemicals. The hydrolysis of proteins consists of the cleavage of the peptide bond that links two amino acids inside the protein chain. During this reaction, a water molecule is added per every broken peptide bond. Proteases are enzymes that specifically catalyse this reaction by means of the formation of an enzyme-substrate intermediate complex. The net reaction is as follows:

P1 − CO − NH − P 2 + H 2 O → P1 − COOH + NH 2 − P 2

(1)

In aqueous solutions the equilibrium lies so far to the right that the degradation and not the synthesis of larger macropeptides is thermodynamically favoured. Provided that the reaction products are macropeptides bonded by amide bonds as well, they can undergo further hydrolysis to yield new shorter peptides through consecutive reactions in series. Regarding the mechanism of most protease-catalyzed hydrolysis reactions, three consecutive steps are believed to occur: a) formation of a Michaelis complex between substrate (i.e. the peptide bond) and the active site of the enzyme; b) cleavage of the peptide bond to release one of the two resulting peptides; c) nucleophilic attack on the remaining complex to yield another peptide and release the free active enzyme. That is to say, the Michaelis-Menten mechanism has been traditionally used to explain the proteolysis reaction (Svedsen, 1976; Adler-Nissen, 1986). However, the hydrolysis mechanism has been explained by other routes. A “one-by-one” mechanism indicates that one protease molecule attacks one substrate molecule at a time so that no intermediate products are formed. On the contrary, the “zipper” reaction involves a rapid degradation of native proteins to yield intermediate products, which will be slowly converted into the end products (Nielsen and Olsen, 2002). With regard to the monitoring of the proteolysis process, the degree of hydrolysis (DH) is a numerical parameter aimed at the measurement of the extent of the reaction. It is expressed as the ratio between cleaved peptide bonds and the total number of peptide bonds into the protein chain. According to equation 1, both a terminal carboxyl and an amino residue are released during the hydrolysis reaction. As a result, the net balance of carboxylic acids and tertiary amine bases depends on the pH of the reaction medium. Thus, the reaction can be controlled by the addition of neutralising agent in order to maintain the pH at a constant number (Adler-Nissen, 1986; Camacho et al., 2001). Finally, the consumption of neutralising

172

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

agent is correlated with the number of cleaved peptide bonds so that the continuous monitoring of the DH is achieved. This method is called the pH-stat technique and will be discussed later on.

Applications of Protein Hydrolysates The enzymatic hydrolysis of proteins provides desirable characteristics from the food technology point of view. In fact, limited hydrolysis of proteins allows to obtain products with favourably customised functional properties such as solubility, foaming and emulsifying properties. Consequently, protein hydrolysates are typically employed as food ingredients to stabilize interfaces, bind water, lipids and aroma, solubilise other ingredients or improve organoleptic features of the product such as colour, odour and flavour. It must be pointed out that the length of the peptide chain has strong effects on the solubility, the emulsifying and foaming properties of protein hydrolysates. The solubility of corn proteins is increased by 30-50 % if just DH 2 % is achieved (Casella and Whitaker, 1990). Soy protein and casein hydrolysates show higher emulsifying capacity than the native protein if a limited 1-2 % DH is reached. However, if higher DH is achieved, the emulsifying power significantly decreases with respect to the whole protein. The same fact has been described for whey proteins: a 1 % DH hydrolysate improves and stabilises emulsions so it can be employed as emulsifying ingredient in sauces (Christiansen et al., 2004). The influence of the peptide-chain length on the foaming power of wheat gluten protein hydrolysates was studied by Popineau et al. (2002). They employed different ultrafiltration membranes to control the average molecular weight of the final hydrolysate and concluded that an optimal molecular weight associated to a certain DH gave enhanced and more stable foams than native wheat proteins or extensive hydrolysates. Regarding the functional properties of protein hydrolysates, other important usages are the production of meat extracts and flavour enhancers for ready meals, the processing of fish by-products, the formation of gel-like structures for gelly desserts and sausages. Furthermore, the objective of extensive hydrolysis is to produce reduced allergenicity products for special diets, clinical preparations and human breast milk substitutes. Due to its high nutritional value, whey proteins are ideal substrates for the latter applications. Allergic reactions to food proteins are caused by specific sequences of amino acids called epitopes. Several epitopes have been identified in β-lactoglobulin which is the main responsible for adverse reactions to milk proteins (Otani, 1987). To avoid such reactions, enzymatic hydrolysis has demonstrated to be an excellent method. As a matter of fact, significant reduction of the antigenicity of milk proteins has been reported for enzymatic hydrolysis of whey proteins (Nakamura et al., 1993). As well, Otani and Osono (1989) did not find allergic response to milk protein hydrolysates if the average molecular weight was below 1000 Da. Equally, Ena et al. (1995) showed that peptides (from whey protein hydrolysates) with molecular masses lower than 3400 Da did not provoke allergic reactions. Likewise, Svenning et al. (2000) found a maximal reduction of allergenicity in whey protein hydrolysates with molecular weight around 1000-5000 Da. In conclusion, it is worth noting that the extension of the hydrolysis reaction determines the allergic response and that short-chain peptide fractions are responsible for the reduction of allergic reactions.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

173

Finally, the hydrolysis of high nutritional quality food proteins employing selective enzymes is intended for the production of bioactive peptides. For instance, whey proteins can release biologically active peptides during the hydrolysis reaction. Such bioactive peptides are considered as potential modulators of various processes in human metabolism (PihlantoLeppala, 2001).

Protein Hydrolysates Production Routes At an industrial level, most of the hydrolysis reactions are carried out in a classical batch reactor with controlled temperature and pH. At the end of the reaction the enzyme is inactivated before recovering the final products. This conventional batch-type protein hydrolysis is simple and easy to control. However, there are several disadvantages with respect to batch hydrolysis mainly related to high enzyme and labour costs (Giorno and Drioli, 2000; Rios et al., 2004). Several solutions based on a continuous operation mode have been tried. Enzyme immobilization onto highly activated supports (Lamas et al., 2001; Sousa et al., 2004; Tardioli et al., 2005) allows working continuously and re-using the catalyst so that the global yield is improved if compared to batch procedure. From a theoretical point of view, the physical or chemical binding of an enzyme onto a solid surface should not influence on the catalytic properties of the enzyme. However, severe loss of enzyme activity has been reported due to constrains for diffusion into the support, mainly if colloidal or macromolecular substrates are concerned. The current state-of-the-art technique is the enzymatic membrane reactor which consists on the coupling of a membrane separation device to a tank reactor. Here, the membrane creates a selective barrier so that permeable products can be separated from the reaction mixture whilst the enzyme is retained into the reaction tank (Rios et al, 2004). The most widespread configuration of membrane reactors is that of direct contact membrane reactors (i.e. soluble enzymes can act directly on the substrate). The direct contact membrane reactors must be broken down into three different classes: dialysis, dead-end and recycle reactors. In dialysis configuration, both enzyme and substrate flow into the same membrane stream and the separation takes place as a consequence of a concentration gradient of reaction products through the membrane. That means the dialysis reactors is based on diffusion and no convective flow is achieved since process streams in each side of the membrane are circulated at approximately the same flow rate. Dead-end membrane reactors (also known as ultrafiltration cells) includes separation and reaction in the same compartment being the reaction media pressurised against the membrane and forced to flow through the membrane pores. Bearing in mind that the separation of solutes is carried out by means of a conventional filtration through a flat membrane, low ratio of membrane surface area to reactor volume are obtained and low permeation flux due to polarization concentration (Prazeres and Cabral, 1994). Therefore, dead-end reactors are practical for small scale purposes and laboratory research whereas they are not applied for pilot plant or industrial scale. Typical membrane recycle reactors are composed of a stirred vessel coupled to an ultrafiltration module in a semi-closed loop configuration. The membrane reactor avoids the main difficulties related to the enzymatic immobilization: soluble enzymes can be used, high substrate conversion is achieved due to the elimination of inhibitors and the molecular size of the product is controlled by an ultrafiltration module with the appropriate molecular weight

174

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

cut-off. Membrane reactors can be operated continuously or discontinuously. Continuous reaction and simultaneous separation of products from the reaction mixture can be achieved with a continuous membrane recycle reactor (Deeslie and Cheryan, 1981; Deeslie and Cheryan, 1982; Mannheim and Cheryan, 1990; Perea and Ugalde, 1996; Martin-Orue et al., 1999; Prata-Vidal et al., 2001; Guadix et al., 2005, Cheison et al, 2006). According to Rios et al. (2004), the main advantages associated to the continuous membrane recycle reactor are: continuous operation and reuse of enzyme, water-soluble enzymes are employed; inhibition phenomena are reduced, enzyme-free final hydrolysate and control of the properties of the end product. In addition, some drawbacks should be pointed out (Giorno and Drioli, 2000): loss of enzyme activity due to leakage, enzyme thermal deactivation and shear stress deactivation, irreversible fouling on the membrane surface and presence of unconverted substrate. Owing to these disadvantages, the economical profit of the continuous membrane reactor critically relies on the need of purging of the reactor to remove the remaining intact protein (Nielsen and Olsen, 2002). The aim of the research presented in this work is to design and optimise a hydrolysis process in which the enzyme recycle is performed by means of a ultrafiltration membrane. As a case study, the obtention of a hydrolysate with a reduction of antigenicity of 1000 times will be required. First, the degree of hydrolysis to achieve this reduction will be determined. Second, the membrane cut-off will be selected according to the molecular weight distribution of the hydrolysate and the enzyme. Third, a mathematical description of the system will be developed and experiments will be performed to estimate the model parameters. Eventually, the optimal operation will be determined and its performance will be compared to that of the traditional batch hydrolysis.

MATERIALS AND METHODS Materials The substrate was a commercial whey protein concentrate (WPC) Lactalbumin 75L from Milei (Sttutgart, Germany) with a protein content of 76% determined by the Kjeldhal method. The protease was subtilisin Protex 6L from Genencor International Inc. (Rochester, USA) obtained by controlled fermentation of a selected strain of Bacillus licheniformis. The experimental device includes a hydrolysis tank and an ultrafiltration module. The reaction was carried out in a 200 mL stirred jacketed tank reactor. The pH was controlled by an automatic titrator Titrino 718 from Metrohm (Herisau, Switzerland). The degree of hydrolysis (DH) was monitored by the pH-stat technique, which allows to correlate the DH with the amount of alkali added to maintain constant the pH. The ultrafiltration module was a Labscale TFF (Millipore Corp., Bedford, USA) which is equipped with organic polyethersulphone flat membranes with an effective filtration area of 50 cm2.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

175

Analytical Techniques Determination of the Degree of Hydrolysis: pH-Stat Technique As it was fixed above, the DH is expressed as the ratio between cleaved peptide bonds and the total number of peptide bonds into the protein chain:

DH =

h h tot

(2)

Being h the number of equivalent of peptide bonds per gram of protein cleaved during the hydrolysis reaction and htot the total number of peptide bonds per gram of protein. The value htot is typical for each protein and is calculated taking into account the amino acids composition of the protein. If an average molecular weight of 125 g/mol for amino acids is considered, the value for htot is around 8 eq per kilogram of proteins. In fact, htot is equal to 8.8 eq/g for whey proteins. Bearing in mind that release of H+ is achieved during the hydrolysis reaction with serine proteases (see the introductory section of this chapter), NaOH is added to maintain the pH in the reaction at 8.5. The consumption of alkali can be correlated with the DH by means of the following equation:

DH =

VB ·N B M p ·α·h tot

(3)

where VB is the volume of base added during the hydrolysis (mL), NB is the concentration of the NaOH solution (N), MP is the total mass of protein in every batch (g) and htot is the total number of peptide bonds (eq/g). α is the dissociation degree of the amine groups released during the hydrolysis. It depends on the average pK of the amino groups and the medium pH as well. α is calculated as follows:

α=

10 pH −pK 1 + 10 pH −pK

(4)

For pH equal to 8.5 and whey proteins as substrate, the degree of dissociation is almost total so that it reaches a figure of 0.99-1. Furthermore, the mean length of the peptide chain can be calculated from the DH determination. As a consequence, an approximation to the peptide chain length (PCL) value is defined as follows:

DH =

1 PCL

(5)

Correlation between DH and Antigenicity Reduction A whey protein solution was hydrolysed until a degree of hydrolysis of 0.175 was achieved. During the hydrolysis, different samples were taken and their antigenicity was measured. The antigenicity of substrate and hydrolysates was measured by an enzyme-linked immunosorbent assay (ELISA) following the procedure by Knights (1985). Then, a

176

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

correlation between degree of hydrolysis and antigenicity reduction was established based on the ELISA data for different DH materials.

Selection of the Molecular Weight Cut-off of the Membrane In order to assure that the membrane molecular weight cut-off is large enough to allow the hydrolysed product to pass through the membrane module, the molecular weight distribution of a hydrolysate was determined. The molecular weight distribution was analysed by gel exclusion chromatography according to the method described by Richter et al. (1983). Two TSK-gel G2000SW (Tosoh Bioscience, Stuttgart, Germany) in series were employed with 6 M guanidine as the mobile phase. The separation was carried out at 1 mL/min for 25 min with UV detection at 280 nm. The retention times of the peaks obtained were compared to a standard mixture of polypeptides of known molecular weight obtained from Sigma (St Louis, MO, USA): ovoalbumin (44000), chimotrypsin (25000), ribonuclease A (13700), insuline (6000), insuline A (2531), sleep inducing peptide (849), Phe-Gly-Gly (279). The membrane molecular weight cut-off must be greater than the molecular weight corresponding to the minimum retention time in which protein is detected. On the other hand, the molecular weight cut-off must be small enough to ensure that no enzyme leaks through the membrane module. The enzymatic activity of the membrane permeate was determined to detect any enzyme leakage during ultrafiltration as follows: 200 mL WPC hydrolysate (5 g/L, enzyme to substrate ratio 0.01) were obtained at 60 ºC and pH 8.5 in 4 hours. The hydrolysate was ultrafiltered at 20 ºC until 150 mL of permeate were obtained. 50 mL of this permeate were added to 150 mL of WPC solution adjusted to 60 ºC and pH 8.5. The pH variation of the mixture was monitored for 2 h.

Ultrafiltration Membrane Reactor with Enzyme Recycling The operation of the discontinuous recycle membrane reactor comprises 3 steps, namely hydrolysis reaction, ultrafiltration of the mixture and enzyme recycle. The hydrolysis experiments were performed as follows: 200 mL Milli-Q water were heated up to 35 ºC and then WPC was dissolved up to 5 g/L. The protein solution was vacuum filtered. Afterwards, it is heated into the reaction tank until reaching the required operation temperature. Then, 1 N sodium hydroxide was added in order to raise the pH up to 8.5. When the appropriate temperature and pH have been reached, the enzyme is added to the medium and the reaction begins. The pH is kept constant by addition of 1 N NaOH. Being P0 the number of peptide bonds at the beginning of this step, this number at the end will be P0·(1-DH). After the reaction is completed (i.e. the required DH is reached), the reaction mixture is cooled down to 20 ºC, since the protease does not show enzymatic activity at this temperature and in order to avoid any enzyme deactivation. The cold reaction mixture permeates through the ultrafiltration membrane and the filtration continues until the permeate volume collected is 150 mL. After the filtration, the membrane module is cleaned with 0.5 N sodium hydroxide and 0.1% (w/w) sodium dodecyl sulphate. By assuming full transmission of peptides through the membrane, the final concentration of peptide bonds in the retentate will be P0·(1-DH)·(1R), where R is defined as the ratio between the final and initial retentate volumes.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

177

Finally, the retentate volume (50 mL) containing the enzyme is heated again up to the operating temperature, while a 150 mL substrate solution for the new hydrolysis is prepared. The retentate is recycled to the reaction tank to perform a new hydrolysis. The concentration of peptide bonds at the end of this step must be equal to that of the first hydrolysis. To this end, the concentration of the 150 mL solution must be (P0·(1-(1-DH)·R)/(1-R).

MATHEMATICAL MODELLING This section focuses on the mathematical description of the operation of the reactionseparation system and the formulation of an optimisation problem. Firstly, a model comprising both hydrolysis and enzyme deactivation kinetics is developed. Different approaches have been formulated in order to describe kinetics of enzyme-catalysed reactions: from a classical Michaelis-Menten mechanism (Svedsen, 1976) to first-order kinetics (Vorob’ev et al., 1996), competitive inhibition mechanisms (González-Tello et al, 1994) or empirical kinetic equations that are not based on a mechanistic explanation (Margot and Fleschel, 1997). A comprehensive kinetic study of the enzymatic proteolysis is rather complicated due to the various types of peptide bonds involved and their different reactivity during the enzyme-catalysed reactions. Finally, an objective function based on the kinetics model aims at the optimisation of the operation conditions. Optimisation has become a major enabling area in process engineering. It has evolved from a methodology of academic interest into a technology that has and continues to make significant impact in industry (Biegler and Grossman, 2004).

Hydrolysis and Enzyme Deactivation Kinetics The overall rate of hydrolysis during every batch-hydrolysis stage, r, for a tank reactor at constant pH and temperature is:

r=

dS = k h ·Sm ·e dt

(6)

being S the substrate concentration (g/L), t the reaction time (h), kh a hydrolysis kinetic constant (units depend on the reaction order) , m the order of the hydrolysis reaction (dimensionless) and e the enzyme concentration (g/L). Previously (González-Tello et al., 1994), zero-order kinetic has been identified for the hydrolysis of whey proteins with subtilisin. If the substrate concentration is expressed on a conversion basis, after separating variables, we obtain: x

t

S0 ·∫ dx = k h ·∫ e·dt 0

0

(7)

178

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

where S0 is the initial amount of substrate (g/L) and x is the conversion of the reaction (i.e. the degree of hydrolysis). If the enzyme is recovered after every batch hydrolysis and reused repeatedly achieving n successive reactions and assuming negligible enzymatic deactivation during the ultrafiltration step, we get: t

n·S0 ·x n = ∫ e·dt kh 0

(8)

As equation 8 shows, a mathematical description of the enzyme deactivation is needed to develop a model for the process. As a consequence, different expressions may arise depending on the deactivation kinetics considered. For example, if a second-order deactivation is considered:

−

de = k d ·e 2 dt

(9)

An expression for the concentration of active enzyme as a function of the reaction time is obtained by the integration of the latter equation:

e=

1 1 + k d ·t n e0

(10)

where kd is a kinetic constant for the thermal deactivation of enzyme. So that, by substituting into equation 8, we obtain:

n·S0 ·x 1 = ·ln (1 + k d ·e 0 ·t n ) kh kd

(11)

Optimisation Problem As it has been described above, the operation consists of three stages: hydrolysis, enzyme recovery by ultrafiltration and enzyme recycling. Then, only one complete batch (the last one) and n-1 incomplete batches are obtained, which allows to define the productivity in the reactor as:

P=

(n − 1)·R + 1 tT

(12)

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

179

That means P is defined as the ratio number of hydrolysate batches produced per total time of operation. The total time of operation includes the reaction time corresponding to n hydrolysis and the filtration time for n-1 ultrafiltrations: t T = t n + (n − 1)·t F (13) where tF is the filtration time, tn is the total hydrolysis time including all the reactions, tT is the total operation time and n is the number of enzyme uses. Taking into consideration that the key disadvantage for the traditional batch hydrolysis is the high cost of enzyme involved, alternative procedures based on enzyme recycling have been developed to reduce the associated cost. In our case, the main objective is to minimise the amount of enzyme employed to yield a fixed productivity of the reactor. Therefore, the objective function is:

ET =

e0 tT

(14)

That is to say, the consumption of enzyme expressed as mass of enzyme per time and volume units.

RESULTS AND DISCUSSION Correlation between Degree of Hydrolysis and Antigenicity Reduction The operative DH of the hydrolysis reactor must be preset to obtain a protein hydrolysate reducing the antigenicity 1000 times in comparison to the native protein. The antigenicity of the original WPC and the hydrolysates at different DH was measured by the ELISA technique (Figure 1). The inhibition curves show a linear relationship between the percentages of inhibition during the ELISA and the DH of different hydrolysates (from 2.5 % to 17.5 %). Several substrate concentrations have been also assayed and as far as the substrate concentration rises, the inhibition is higher as well. If the same substrate concentrations are compared, the inhibition percentage is reduced insofar as the DH goes up. That is, extensive hydrolysis makes the antigenicity to decrease significantly in comparison to the native protein or low DH. Then, the percentage of inhibition was calculated as:

I = 1−

AS A0

(15)

being AS the absorbance measured in the ELISA test for the samples (WPC and hydrolysate at different DH) and A0 the absorbance measured when neither WPC nor hydrolysate sample is added. The antigenic reduction (AR) is calculated as the logarithm of the ratio between the concentration of hydrolysate and WPC which give an inhibition of 50%. Figure 2 shows that

180

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

a linear relationship exists between DH and the antigenic reduction. The minimal-square regression line is as follows:

AR = 11.1·DH + 1.33

(16)

with a correlation coefficient r2 = 0.990. According to the aim of this research work, if the required AR is equal to 3, a DH of 0.15 is needed.

Selection of the Molecular Weight Cut-off of and Characterization of the Membrane With respect to the molecular weight cut-off, the ultrafiltration module must fulfil two conditions: first, the membrane must allow the peptides to pass through it and, second, the enzyme must be completely rejected. Thus, the molecular weight distribution of the resulting hydrolysate must be determined and accordingly, the MWCO is selected. Then, it must be confirmed that there is no enzyme leakage through the membrane. The molecular weight distribution is determined by size exclusion liquid chromatography coupled to a UV-visible detector at 280 nm. Before the chromatographic determination of the molecular weight distribution, a calibration curve has been obtained. The standard peptides employed for calibration purposes and their corresponding molecular weight and retention time are shown in Table 1. 100

80

I (%)

60

40

20

0 -5

-4

-3

-2

-1

0

1

2

3

log C (g/mL)

Figure 1. Inhibition curves of the ELISA for WPC ({) and DH=2.5 (z), 7.5 (…), 12.5 („) and 17.5 % (‘) hydrolysates.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

181

3.5

AR

3.0

2.5

2.0

1.5 0.00

0.05

0.10

0.15

0.20

DH

Figure 2. Correlation between degree of hydrolysis and antigenicity reduction.

Table 1. Standard peptides for HPLC calibration Standard peptides Ovalbumin Chymotripsinogen Ribonuclease Insuline Insuline A Sleep inducing peptide Phe-Gly-Gly

MW (g/mol) 44000 25000 13700 6000 2531 849 279

tR (min) 10.45 12.17 13.93 16.52 17.45 19.30 20.36

There are two different linear regions between the retention time (tR) and the logarithm of molecular weight separated at 7000 Da and their equations are listed below:

t R = −2.99·log(MW ) + 27.8

(17)

t R = −7.00·log(MW ) + 42.9

(18)

Equation 17 is valid for MW < 7000 Da, being r2 = 0.9876 while equation 18 is suitable from 7000 Da on and r2 is 0.9998. The hydrolysate (DH = 0.15) was analysed to assure that low molecular weight peptides are released and compared to the chromatogram of the native substrate.

182

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

(b)

23.30

14.12

21.73

10.08

20.82 22.37

11.19

10.41

18.44

14.52

20.49

12.73

(a)

Figure 3. Molecular weight distribution of (a) WPC and (b) hydrolysate DH=0.15.

With respect to the chromatogram of the native WPC (Figure 3a), there are two peaks at retention times of 12.73 min and 14.52 min for the major whey proteins, β-Lactoglobulin and α-Lactalbumin, respectively. The chromatogram of the hydrolysate (Figure 3b) shows that the amount of peptides with a molecular weight greater than 8000 Da (i.e. retention time minor than 15.62 min) is negligible if compared to the low molecular weight peptide fractions (mainly the peaks for retention times of 18.44 min and 20.49 min). There is also a fraction mostly related to small di- and tripeptides (21.73 min). It must be also pointed out that a small fraction of non-hydrolysed protein is detected at 10.08 min. This peak can be identified as bovine seroalbumin (BSA). The average peptide chain length (PCL) is equal to 6.6, that is, the hydrolysate basically consists of peptides containing 6 – 7 amino acids. If a mean molecular weight around 120 Da per amino acid is considered, then the mean molecular weight of the hydrolysate is about 800 Da. This calculation is in accordance with the chromatographic distribution of the hydrolysate. Still, it must be taken into account that the hydrolysate includes from small di- and tripeptides to non-reacting protein. As a result, considering that a DH of 0.15 is required to achieve an AR of 3 and no significant fraction of peptides is larger than 8000 Da, thus, a MWCO of 8000 Da is appropriate to be incorporated in the DMR. In addition, the enzyme activity of the permeate flow collected through the 8000 Da membrane was determined for a protein hydrolysate. The pH drop was registered and just a slight decay in pH from 8.5 to 8.3 was found. Consequently, it can be concluded that there is no significant leakage of catalyst through the membrane. Hence, the 8000 Da membrane module selected allows the hydrolysate to pass through it and simultaneously retains the enzyme.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

183

Ultrafiltration is a typical pressure-driven membrane process: a mixture of components of different sizes is brought to the surface of a semi-permeable membrane. Under the driving force of a pressure gradient, some components permeate the membrane, whereas others are retained. A feed solution is separated into a filtrate that is depleted of particles or molecules and a retentate in which these components are concentrated. A mathematical description of ultrafiltration membrane process is based on a general equation relating the convective mass flux through the membrane to the driving force (i.e. applied pressure for ultrafiltration). The relationship between mass flow and transmembrane pressure for the selected membrane (8 kDa polyethersulphone) is shown in Figure 4 for both water and protein hydrolysate. As it can be seen in the hydrolysate line, there are two different flux-pressure regions: a linear relationship exists up to 1.4 bar and an asymptotic curve is obtained for higher values of pressure. Being a protein solution the feed flow, the transmembrane flux depends on the applied pressure only at low pressures (pressure controls the filtration operation) and reaches a constant, pressure-independent value at higher pressures (mass diffusion controls the operation). In microfiltration and ultrafiltration processes the upper borderline pressure for which a linear relationship exists is associated to a critical flux (Wu et al, 1999). This critical flux relates to the minimal permeate flux provoking the fouling phenomenon (Bacchin et al, 2005) and is the maximal pressure for the pressure-control region of the mass transfer through the membrane. Thus, the applied pressure must be below this value. Besides, there is a “plateau” flux in UF called limiting flux. Above the limiting flux, an increase in the applied pressure does not make flux to rise at the same degree due to a combination of a gel layer formation and fouling of the membrane. 40

JF (mL/min)

30

20

10

0 0.0

0.5

1.0

1.5

2.0

PTM (bar)

Figure 4. Relationship between permeate flux and transmembrane pressure for water ({) and hydrolysate (z).

184

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

Both phenomena (gel layer formation and fouling) are related to deposits of substrate onto the membrane surface. Gel layer formation involves a reversible accumulation of retained material and depletion of the permeating components in the boundary layer adjacent to the membrane surface. Hence, a decrease in the transmembrane flux with a constant hydrostatic pressure driving force appears. The gel layer formation (also called as concentration-polarisation) can be controlled by adjustment of the operating pressure of the membrane. However, fouling is linked to an irreversible precipitation of macromolecules and only a chemical cleaning of the membrane can recuperate the original surface properties. The optimal operation of a membrane module aims at the mitigation of the fouling phenomena and this requirement is only achieved if the operation conditions are selected in the pressure-controlling region. Secondly, the collection of a high permeate flux is also desired. As a consequence of this trade-off, a transmembrane pressure of 0.85 bar is selected for the 8 kDa membrane, giving a permeate flux around 7.5 mL/min.

Influence of Temperature and Enzyme Concentration in the Hydrolysis Reactor In order to validate the model for the DMR experiments at different enzyme concentrations were assayed: 0.05, 0.10, 0.15, 0.20 and 0.25 g/L. The substrate concentration and pH were maintained at 5 g/L and 8.5, respectively. Three operation temperatures were assayed: 50 ºC, 60 ºC and 70 ºC. Considering five different concentrations of enzyme, the hydrolysis time for each batch reaction (ti) and the number of enzymes reuses are shown in Figure 5 for 50 ºC, 60 ºC and 70 ºC. With respect to the influence of temperature on the operating parameters, the higher temperatures increase the initial reaction rate. A high temperature activates the enzyme and contributes to the unfolding of peptide chains (i.e. denaturising the protein). Then, peptide bond is more accessible for the enzyme active site and, consequently, an extensive degradation of the protein is achieved (Van der Placken et al., 2004). Simultaneously, a high temperature increases enzyme deactivation so that the hydrolysis reaction is not efficient. As far as the long-term reaction time is concerned, it must be taken into account that thermal enzyme deactivation makes the reaction to slow down. Because of this fact, a balance between initial reaction rate and long-term operation of the reactor must be considered. For instance, shorter hydrolysis are obtained for n = 1 and n = 2 with 60 ºC when compared to 50 ºC. Nonetheless, 70 ºC is such a high temperature that the enzyme deactivation phenomenon is much faster than the hydrolysis reaction. Hence, only if high enzyme concentration is added, the reuse of the catalyst is possible. Regarding the effect of enzyme concentration, it must be deduced for a given temperature that insofar as the initial enzyme concentration rises, the total hydrolysis time is reduced due to the progressive increase of the reaction rate. In addition, the higher initial concentration of enzyme yields the higher number of enzyme uses for every temperature. An increase of the reaction rate is achieved for rising enzyme concentrations due to the fact that the extent of the hydrolysis reaction is closely related to the formation of an enzyme-substrate intermediate which depends on the active enzyme in the reaction (Deqinq et al., 2005). Nevertheless, the amount of enzyme is actually constrained by the auto-digestion and the cost of the enzyme. In fact, auto-digestion is only important at high DH because substrate protects enzyme

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

185

e0 (g/L)

molecules against enzyme cleavage. Additionally, so long as the active enzyme concentration decreases the hydrolysis time increases for a fixed operation temperature so that the hydrolysis time to obtain the last batch is always the longest among the series. With regard to the number of enzyme reuses, up to n = 9 were achieved for the operation at 60 ºC. In that case, more than 10 hours of consecutive hydrolysis were achieved for just one charge of enzyme of 0.25 g/L.

0.25

1

0.20

1

2

3

2

6

4

2

3

1

0.10

5

3

1

0.15

4

2

1

0.05

2

0

2

4

6

8

10

ti (h)

a. 0.25 1 2 3

4

e0 (g/L)

0.20 1 2

3

0.15 1 2

1

0.10

6

4

3

0

7

5

2

8

6

7

4

3

4

4

6 ti (h)

b. Figure 5 - Continued on next page

9

7

5

3

2

8

6

4

2

1

0.05

5

8

10

12

186

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

0.25

1

e0 (g/L)

0.20

2

1

0.15

1

0.10

1

0.05

0

1

2

3

4

5

6

ti (h)

c. Figure 5. Consecutive hydrolysis times obtained at (a) 50 ºC, (b) 60 ºC and (c) 70 ºC. 3.0

2.5

e0·tn (g/L·h)

2.0

1.5

1.0

0.5

0.0 0

1

2

3

4

5

6

7

8

9

10

n

Figure 6. Fit of experimental data to a model assuming zero-order kinetics and second order enzyme deactivation for 50 ºC (…), 60 ºC ({)and 70 ºC (‘).

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

187

Estimation of Kinetic Parameters The hypotheses of zero-order hydrolysis kinetics and second-order enzyme deactivation are verified since the experimental data fit to the equation proposed for this model. According to this, the kinetics parameters kd and kh stand for the thermal enzyme deactivation and the hydrolysis constants, respectively. They are calculated by non-linear regression and the leastsquares fit curves are shown in Figure 6. The parameters obtained are listed in Table 2. Since excellent correlation has been found, the mathematical model is validated. Table 2. Parameters estimated by nonlinear regression T (ºC) 50 60 70

kd (L/(g·h)) 1.58 4.51 50.2

kh (g/L·h) 4.567 9.97 18.1

r2 0.993 0.992 0.992

Process Optimisation If the expressions for e0, P and tT are substituted into the objective function, then, the resulting amount of enzyme employed in the hydrolysis is:

⎛ k ·n·S0 ·x ⎞ ⎟⎟ − 1 exp⎜⎜ d k h ⎝ ⎠ ET = ⎛ (n − 1)·R + 1 ⎞ k d ·⎜ − t F ·(n − 1) ⎟ P ⎝ ⎠

(19)

Thus, the consumption of enzyme just depends on the productivity P, temperature (through the kinetic parameters kd and kh) and the number of enzyme uses, n. Two optimisation problems may arise: a) fiven the productivity required for the reactor (P), determine the number of uses of enzyme (n) so as to minimise ET. And, b) given a fixed consumption of enzyme, ET, determine the number of uses of enzyme (n) so as to maximise P. If we consider that the main advantage of the system should be a more efficient enzyme consumption, the first approach can be formulated and solved by employing an enumeration technique. According to the latter calculation scheme and taking into account the estimated kinetic constants kd and kh, the value of the objective function versus the number of enzyme uses is shown in Figure 7, for each temperature assayed. As the kinetic constants only depend on the reaction temperature, the objective function for a given temperature is only function of P and n. Obviously, it can be seen that the higher the productivity, the higher the enzyme consumption. That is, a higher amount of enzyme is required to produce the same amount of hydrolysate in shorter operating periods. In addition, the number of enzyme uses is higher insofar as P decreases. In fact, its optimal value reaches a maximal value for low or moderate values of P. For example, nopt = 6 is achieved for 50 ºC and P = 0.05 h-1. Higher values for P

188

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

yield lower values for nopt until obtaining just nopt = 1 for P = 1.50 h-1. Interestingly, the maximum nopt decreases as long as temperature increases. As a matter of fact, nopt = 4 and 1 for 60 ºC and 70 ºC, respectively. It can be seen that the objective function when the temperature reaches 70 ºC rises sharply if the enzyme is recycled (n > 1). In fact, the deactivation is so fast that no long-lasting reactions can be obtained. 10

ET (g/(L·h))

1

0.1

0.05 1.50 0.01

0.001

0.0001 0

2

4

6 n

8

10

12

a. 10

ET (g/(L·h))

1

0.1

0.05 1.50 0.01

0.001

0.0001 0

2

4

6 n

b. Figure 7. Continued on next page.

8

10

12

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

189

10

ET (g/(L·h))

1

0.1

0.05 1.50 0.01

0.001

0.0001 0

2

4

6 n

8

10

12

c. Figure 7. Optimal number of enzyme uses for different productivities.at (a) 50 ºC, (b) 60 ºC and (c) 70 ºC. 7

6

5 70 ºC 60 ºC 50 ºC

nopt

4

3

2

1

0 0.0

0.5

1.0 P

Figure 8. Optimal number of enzyme uses.

1.5

2.0

190

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

The relationship between the optimal number of enzyme uses and the productivity of the hydrolysis reactor is shown in Figure 8. The lower the temperature, the higher optimum number of enzyme uses for a given P. In addition, the borderline value for which nopt = 1 (i.e. enzyme recycling is not advisable) is higher for low operation temperatures, well above P = 1 h-1 at 50 ºC. Transition between optimal values occurs at follows. At 50 ºC, nopt moves from 6 to 5, 5 to 4, 4 to 3 and 3 to 1 at P=0.07, 0.99, 1.32 and 1.36, respectively. It is worth noting that nopt ≠ 2 for any P value at this temperature. With respect to 60 ºC, less changes were observed, since nopt moved from 4 to 3, 3 to 2, and 2 to 1 at P=0.67, 1.11 and 1.13, respectively. Dramatically, nopt = 1 for the operation at 70 ºC. In terms of enzyme consumption, the performance of the hydrolysis system with enzyme recycle can be compared to that of the single batch procedure. To this purpose, an appropriate parameter is the saving of enzyme (SE), defined as:

SE = 1 −

(E T ) min (E T ) n =1

(20)

where (ET)min is the total amount of enzyme for the optimum operation mode and (ET)n=1 is the amount of enzyme for the single batch oper ation mode (i.e. n = 1). The values for ET are calculated employing the optimal operation mode (i.e. nopt) for every figure of productivity. The Figure 9 shows the saving of enzyme as a function of the productivity for each temperature. The maximal saving of enzyme was obtained for each temperature when the productivity tends to zero, ranging from 0.44 for 50 ºC to 0.32 for 60 ºC. There is no saving of enzyme for 70 ºC since nopt = 1, regardless the productivity. 0.5

70 ºC 60 ºC 50 ºC

0.4

SE

0.3

0.2

0.1

0.0 0.0

0.5

1.0 P

Figure 9. Savings of enzyme in the hydrolysis system.

1.5

2.0

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

191

The saving of enzyme allows us to quantify the advantage obtained from the enzymerecycling operation in comparison to the standard batch procedure. Nonetheless, the comparison among several operation temperatures is significant if the enzyme consumptions are considered. Therefore, even though the optimal number of enzyme uses and the saving of enzyme are maximal for 50 ºC, the best performance among different temperatures must be compared considering the values for the objective function. The values for the objective function for 50 ºC, 60 ºC and 70 ºC are shown in Figure 10. It must be pointed out that the enzyme consumption as a function of the productivity of the reactor is minimal for 60 ºC at the entire range of productivity, while 70 ºC compares favorably to 50 ºC at productivities higher than 0.72. 0.15

50 ºC 60 ºC 70 ºC

ET (g/(L·h))

0.10

0.05

0.00 0.00

0.25

0.50

0.75

1.00

P

Figure 10. Enzyme consumption for different reaction temperatures.

CONCLUSION The hydrolysis reactor incorporating a 8000 Da polyethersulphone membrane, was a successful device to enhance the enzyme use in the production of a whey protein hydrolysate in which the antigenicity was reduced 1000 times compared to the original whey protein. The experimental data fit to the theoretical model developed and zero-order kinetics for the hydrolysis and second-order deactivation for the enzyme were identified. Given the production requirements, process optimisation was performed and the number of batch reactions that minimise the total amount of enzyme used was determined. As an important result, the operation allows to save up to 44% and 32% of enzyme compared to the single batch operation mode at 50 ºC and 60ºC, respectively. In any case, the optimal operation temperature is 60 ºC since it yields lower enzyme consumption for all productivities of the reactor.

192

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

REFERENCES Adler-Nissen J. (1986) Enzymic hydrolysis of food proteins. London: Elsevier. Bacchin P., Espinasse B. and Aimar P. (2005). Distributions of critical flux: modelling, experimental analysis and consequences for cross-flow membrane filtration. Journal of Membrane Science, 250, 223–234. Biegler L. T. and Grossmann I. E. (2004). Retrospective on optimization. Computers and Chemical Engineering, 28, 1169-1192. Camacho F., González-Tello P., Páez-Dueñas M. P., Guadix E. M. and Guadix A. (2001). Correlation of base consumption with the degree of hydrolysis in enzymic protein hydrolysis. The Journal of Dairy Research, 68, 251-265. Casella M. A. and Whitaker J. R. (1990). Enzymatically and chemically modified zein for improvement of functional properties. Journal of Food Biochemistry, 14, 453-464. Cheison S. C., Wang Z. and Xu S. (2006). Hydrolysis of whey protein isolate in a tangential flow filter membrane reactor I. Characterisation of permeate flux and product recovery by multivariate data analysis. Journal of Membrane Science, 283, 45–56. Christiansen K. F., Vegarud G., Langsrud T., Ellekjaer M. R. and Egelandsdal B. (2004). Hydrolyzed whey proteins as emulsifiers and stabilizers in high-pressure processed dressings. Food Hydrocolloids, 18, 757-767. Deeslie W. D. and Cheryan M. (1981). A CSTR-hollow fiber system for continuous hydrolysis of proteins: performance and kinetics. Biotechnology and Bioengineering, 23, 2257-2271. Deeslie W. D. and Cheryan M. (1982). A CSTR-hollow-fiber system for continuous hydrolysis of proteins: factors affecting long-term stability of the reactor. Biotechnology and Bioengineering, 24, 69-82. Deqing S., Zhimin H. and Wei Q. (2005). Lumping kinetic study on the process of tryptic hydrolysis of bovine serum albumin. Process Biochemistry, 40, 1943-1949. Dixon M. and Webb E. C. (1979). Enzymes. London: Longman. Ena, J.M., Van Beresteijn, E.C.H., Robben, A.J.P.M. and Schmidt, D.G. (1995). Whey protein antigenicity by fungal proteinases and a pepsin/ pancreatin combination. Journal of Food Science, 60, 104–116. Giorno, L., and Drioli, E. (2000). Biocatalytic membrane reactors: applications and perspectives. Trends in Biotechnology, 18, 339-349. González-Tello P., Camacho F., Jurado E., Paez M. P. and Guadix E. M. (1994). Enzymic hydrolysis of whey proteins: I Kinetic models. Biotechnology and Bioengineering, 44, 523-528. Guadix A., Camacho F and Guadix E. M. (2006). Production of whey protein hydrolysates with reduced allergenicity in a stable membrane reactor. Journal of Food Engineering, 72, 398-405. Knights R. J. (1985). Processing and evaluation of the antigenicity of protein hydrolysates. In F. Lisshlpz (Ed.) Protein hydrolysates. (105-115), New York: Marcel Dekker. Lamas, E.M., Barros, R.M., Balcao, V.M., and Malcata, F.X. (2001). Hydrolysis of whey proteins by proteases extracted from Cynara cardunculus and immobilized onto highly activated supports. Enzyme and Microbial Technology, 28, 642-652.

Protein Hydrolysis with Enzyme Recycle by Membrane Ultrafiltration

193

Mannheim, A., and Cheryan, M. (1990). Continuous hydrolysis of milk protein in a membrane reactor. Journal of Food Science, 55, 381-385. Margot A., Flaschel E. and Renken A. (1997). Empirical kinetic models for tryptic wheyprotein hydrolysis. Process Biochemistry, 32, 217-223. Martin-Orue, C., Henry, G., and Bouhallab, S. (1999). Tryptic hydrolysis of κcaseinomacropeptide: Control of the enzymic reaction in a continuous membrane reactor. Enzyme and Microbial Technology, 24, 173–180. Nakamura T., Sado H., Syukunobe Y. and Hirota T. (1993). Antigenicity of whey protein hydrolysates prepared by combination of two proteases. Milchwissenschaft, 48, 667-70. Nielsen P. M. and Olsen H. S. (2002). Enzymic Modification of food protein. In R. Whitehurst, B. A. Law (Eds.) Enzymes in food technology. (109-143). Sheffield: Sheffield Academic Press. Otani H. (1986). Antigenically reactive regions of bovine milk proteins. Agricultural Research Quaterly, 21, 135-142. Otani H. and Hosono A. (1989). Immunological properties of pepsin, trypsin and/or chymotrypsin digests of bovine αs1-casein. Nippon Chikusan Gakkaiho, 60, 1143-1150. Perea A. and Ugalde U. (1996). Continuous hydrolysis of whey proteins in a membrane recycle reactor. Enzyme and Microbial Technology, 18, 29-34. Pihlanto-Leppala, A. (2001). Bioactive peptides derived from bovine whey proteins:opioid and ACE-inhibitory peptides. Trends in Food Science and Technology, 11, 347–356. Popineau, Y., Huchet, B., Larré, C., Bérot, S. (2002). Foaming and emulsifying properties of fractions of gluten peptides obtained by limited enzymatic hydrolysis and ultrafiltration. Journal of Cereal Science, 35, 327-335. Prata-Vidal M., Bouhallab S., Henry G. and Aimar P. (2001). An experimental study of caseinomacropeptide hydrolysis by trypsin in a continuous membrane reactor. Biochemical Engineering Journal, 8, 195-202. Prazeres D. M. F. and Cabral J. M. S. (1994). Enzymatic membrane bioreactors and their applications. Enzyme and Microbial Technology, 16, 738-750. Richter, W. V., Jacob, B. and Schwandt, P. (1983). Molecular weight determination of peptides by high performance gel chromatography. Analytical Biochemistry, 133, 288– 291. Rios G. M., Belleville M. P., Paolucci D. and Sanchez J. (2004). Progress in enzymatic membrane reactors - a review. Journal of Membrane Science, 242, 189-196. Sousa R., Lopes G. P., Tardioli P. W., Giordano R. L. C., Almeida P. I. F. and Giordano R. C. (2004). Kinetic model for whey protein hydrolysis by alcalase multipoint-immobilized on agarose gel particles. Brazilian Journal of Chemical Engineering, 21, 147-153. Svedsen I. (1976). Chemical modifications of the subtilisins with special reference to the binding of large substrates: a review. Carlsberg Research Communications, 41, 237-291. Svenning C., Brynhildsvold J., Molland T., Langsrud T. and Elisabeth Vegarud G. (2000). Antigenic response of whey proteins and genetic variants of β-lactoglobulin - the effect of proteolysis and processing. International Dairy Journal, 10, 699-711. Tardioli, P. W., Sousa, R., Giordano, R. C., and Giordano, R. L. C. (2005). Kinetic model of the hydrolysis of polypeptides catalyzed by Alcalase immobilized on 10% glyoxylagarose. Enzyme and Microbial Technology, 36, 555–564.

194

Antonio Guadix, Emilia M. Guadix and Carlos A. Prieto

Van der Plancken I., Delattre M., Indrawati, Van. Loey A. and Hendrickx M. E. G. (2004). Kinetic study on the changes in the susceptibility of egg white proteins to enzymatic hydrolysis induced by heat and high hydrostatic pressure pre-treatment. Journal of Agricultural and Food Chemistry, 52, 5621-5626. Vorob’ev M. M., Levicheva I. Y. and Belikov V. M. (1996). Kinetics of the initial stage of milk protein hydrolysis by chymotrypsin. Applied Biochemistry and Microbiology, 32, 219-222. Wu D., Howell J. A. and Field R. W. (1999). Critical Flux measurement for model colloids. Journal of Membrane Science, 152, 89-98.

In: New Food Engineering Research Trends Editor: Alan P. Urwaye, pp. 195-223

ISBN: 978-1-60021-897-2 © 2008 Nova Science Publishers, Inc.

Chapter 6

THE DEVELOPMENT OF THE PROCESSING OF YUBA (PROTEIN-LIPID FILM) Li Zaigui1, Shen Qun1 and Lin Qing2 1

College of Food Science and Nutritional Engineering, China Agricultural University 2 Shanghai Institute of Technology

1. INTRODUCTION 1.1. The History and Properties of Yuba Yuba is a kind of protein-lipid film formed from soymilk under continuous heating, so it is also called to be “Tofupi” or “Tofuyi”, which means “tofu sheets or tofu shirts”. Yuba is one of the most famous traditional foods which has a history of over 2000 years and is very popular now in China. The annual yield of yuba is over 200,000 ton in China. The film can be consumed directly as an ingredient of soups or be used as a sheet for wrapping and shaping meats or vegetables into various forms with different tastes. Now freezing yuba is also used as salad for sashimi in Japan and Korea. Yuba has first been found from the supernatant film of heated soymilk for Tofu making and was usually dried to 8% moisture content for storage. It can be stored for more than three months to six months. It contains about 55% protein and 25% lipid, which are important nutrients for monks to whom meat was prohibited. As its fine taste and unique texture, yuba spread fast in all of the country through the rede of people lived in temples. Compared with the other soybean food or other foods, yuba has superior nutrition (table 1) and digestibility (table 2). Shih(Shih, 1998) said compared with the foods from animal sources, the plant sourced foods are beneficial to human health. Soybean protein has the benefits of not being closely associated with saturated fat and cholesterol and less likely to cause problems such as heart diseases.

196

Li Zaigui, Shen Qun and Lin Qing Table 1. The nutrition content of yuba and other foods (%)

food

Dried yuba Tofu Freezing tofu Fried tofu Soybean flour Chicken Beef

Moisture

protein

lipid

carbohydrate saccharide

fibre

ash

8.0

50.0~55.0

25.0

12.0~19.0

0

3.0

Caloric (Cal/ 100g) 474.0

88.0~92.0 10.3

6.0 53.5

3.5 26.4

1.9 7.0

0 0.2

0.6 2.6

63.0 480.0

43.0 7.8

20.6 21.9

30.4 19.2

4.5 41.9

0.2 4.5

1.3 4.7

374.0 428.0

75.0 72.9

12.7 20.1

11.2 5.7

0 0.3

1.1 1.0

152.0 133

Table 2. The digestibility of foods (%) Food Yuba Cooked soybean Soybean flour Tofu Freezing tofu Fried tofu

digestibility 93 60 83 95 83 64

Yuba has excellent cooking properties besides its fine nutrition, taste and texture. First, Yuba is not only suitable to storage but readily to rehydrate. It is not necessary to say freezing soybean which is spread in recent years, even dried yuba with 8% moisture content, the rehydration time is just about 2 minutes in 50℃ water or 8 minutes in tap water. Rehydrated yuba has also good taste and texture though it may be affected comparing with the fresh one. Secondly, the surface of yuba is so huge that it is good at absorbing of flavoring. It is one kind of food that is indispensable for popular Chinese chafing dish. Sauce, vinegar or peppers are common flavoring for yuba dishes. Yuba is consistent to other foods such as meat, fish, vegetables or even cheese, so it can be consumed extensively in Chinese, Japanese or European dishes. Yuba has so excellent coating property that is also called edible film. The material, types or size can be altered according to need. The material includes minced meat, shrimp or fish, ham, vegetables, mushroom, vermicelli, bean sprout or even cooked rice. This kind of food can be varied and kept in refrigerator to prevent the taste, color or rigidity changes a lot. Figure 1 is dried yuba and figure 2 is some freezing yuba dishes. The color of yuba of high quality is flaxen and translucent. Mixture of yuba and other materials will rich the color of foods especially mixed with vegetables. The last but important property of yuba is the suitability for fr