Focus on Food Engineering Research and Developments 1600218989, 9781600218989, 9781606925676

This new book presents new research in the growing field of food engineering which refers to the engineering aspects of

345 76 7MB

English Pages 562 Year 2007

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
FOCUS ON FOOD ENGINEERING RESEARCH AND DEVELOPMENTS......Page 3
NOTICE TO THE READER......Page 6
CONTENTS......Page 7
PREFACE......Page 9
1. INTRODUCTION......Page 17
3. COMPUTER VISION BASED MEASUREMENT OF COLOR......Page 18
3.1. Description of the Technique......Page 19
3.2. Accuracy of Color Measurement......Page 22
REFERENCES......Page 26
APPENDIX 1. MATLAB CODE FOR EXTRACTING THE COLOR INFORMATION OF A POLYGONAL MARKED REGION IN AN IMAGE......Page 27
ABSTRACT......Page 29
Tray Dryers......Page 30
Rotary Dryers......Page 31
Microwave Dryers......Page 32
Superheated Steam Dryers......Page 33
Hybrid Dryers......Page 34
Porous Materials......Page 36
Hygroscopic and Non-Hygroscopic Materials......Page 37
Dimensionless Numbers......Page 38
Specific Humidity......Page 39
Gas Pressure......Page 40
Capillary Pressure......Page 41
Chemical Potential......Page 44
Concentration......Page 45
Saturation......Page 47
Mass Capacity......Page 48
Water Activity......Page 49
Equilibrium Moisture Content......Page 50
Sorption Isotherms......Page 51
Bound and Unbound Water......Page 52
Density......Page 53
Dry Solid Density......Page 55
Specific Heat......Page 56
Thermal Conductivity......Page 58
Thermal Diffusivity......Page 61
Inverse Approaches for the Prediction of Thermophysical Properties......Page 62
Enthalpy of Vaporization......Page 63
Mass Transfer Coefficients......Page 64
Diffusibility......Page 65
Porosity......Page 66
Permeability......Page 67
Flux and Driving Force......Page 70
Velocity and Mass Average Velocity......Page 71
Anisotropic Shrinkage......Page 72
Clapeyron-Clasius Equation......Page 73
Kelvin Equation......Page 74
5. DRYING KINETICS OF FOOD MATERIALS......Page 75
Drying in the Constant-Rate Period......Page 76
Controlling Resistance......Page 77
6. TRANSPORT PHENOMENA......Page 78
GAS TRANSPORT MECHANISMS......Page 79
Knudsen or Free Molecule Diffusion......Page 80
Slip Flux......Page 81
Viscous Flux......Page 84
Molecular Diffusion......Page 86
Stefan Diffusion......Page 87
Condensation-Evaporation......Page 89
Molecular Diffusion......Page 90
Surface Diffusion......Page 92
Hydrodynamic Flow......Page 93
Theoretical Models......Page 94
Molecular Diffusion Model......Page 95
Solutions of Diffusion Equation for Some Simple Geometries......Page 97
Conventional Shapes......Page 99
Nonconventional Shapes......Page 101
DIFFUSION MODEL FOR SHRINKING MEDIA......Page 102
Relation Between Solid Displacement and Water Loss......Page 103
Rheological Behavior......Page 105
Elastic Behavior......Page 106
Mathematical Transformation......Page 107
ANOTHER DIFFUSION MODEL FOR SHRINKING MEDIA......Page 109
RECEDING FRONT MODEL......Page 110
Wet Zone (1)......Page 111
PHILIP AND DE VRIES MODEL......Page 112
Water Vapor......Page 113
Potentials for Heat and Mass Transfer......Page 114
Transfer Fluxes......Page 115
Temperature Equation......Page 118
Pressure Equation......Page 119
KRISCHER MODEL......Page 122
Balance Equations......Page 123
Liquid Phase......Page 125
Balance Equations......Page 126
WHITAKER MODEL......Page 128
Mass Conservation......Page 129
Energy Conservation......Page 130
Gas Phase......Page 131
Boundary Conditions......Page 133
Volume Averaging Method......Page 134
Energy Conservation......Page 135
Capillary Pressure Formulation......Page 138
EMPRICAL AND SEMI-EMPRICAL MODELS......Page 139
Two Term, Modified Henderson and Pabis Models......Page 140
Geometric, Thompson, Wang and Singh, Logistic, Weibull Distribution Models......Page 141
8. SHRINKAGE......Page 142
Simplified Methods......Page 146
Regression Analysis Method......Page 148
Temperature......Page 149
Shrinkage......Page 150
CONCLUSION......Page 151
NOMENCLATURE......Page 168
Greek Letters......Page 170
Subscripts......Page 171
REFERENCES......Page 172
ABSTRACT......Page 181
MICROBIOLOGICAL QUALITY OF TVAROGS AND COTTAGE CHEESE......Page 182
THE DYNAMICS OF CHANGES OF FACULTATIVE ANAEROBIC MICRO-FLORA IN LACTIC ACID CHEESE......Page 186
PACKING AS A FACTOR DETERMINING THE PRODUCT QUALITY......Page 190
ASSESSMENT OF INTERACTIONS OCCURRING AMONG TVAROG MICRO-FLORA IN MODEL CONDITIONS......Page 195
THE INFLUENCE OF HERMETIC PACKING ON STARTER MICROFLORA AT HIGH CONTAMINATION LEVEL OF TVAROGS......Page 205
INVESTIGATION OF INTERACTIONS IN TVAROGS OF HIGH MICRO-FLORA CONTAMINATION LEVEL......Page 208
CHARACTERISTICS OF INTERACTIONS OCCURRING AMONG MICRO-ORGANISMS IN HERMETICALLY-PACKED TVAROGS......Page 213
DEVELOPMENT OF MATHEMATICAL MODEL FOR EVALUATION OF QUALITY OF LACTIC ACID CHEESE......Page 215
THE INFLUENCE OF INTERACTIONS AMONG MICRO-ORGANISMS ON PHYSICO-CHEMICAL PROPERTIES OF LACTIC ACID CHEESE STORED IN REFRIGERATING CONDITIONS......Page 221
SAFETY OF LACTIC ACID CHEESE......Page 224
DEVELOPMENT OF THE PROGRAM FOR EVALUATION OF STAPHYLOCOCCI GROWTH......Page 229
SURVIVAL RATE OF STAPHYLOCOCCI ON THE SURFACE OF TVAROGS AND THE SYNTHESIS OF ENTEROTOXIN DURING STORAGE......Page 231
METHODS OF OPTIMISING QUALITY OF TVAROGS......Page 236
Optimisation with Biocides......Page 237
Optimising the Quality through Modification of the Packing System......Page 242
Optimisation with the Help of Prediction Models......Page 244
REFERENCES......Page 245
ABSTRACT......Page 255
INTRODUCTION......Page 256
1.1. Post-Harvest Loss and Physical Characteristicsof Selected Tropical Fresh Fruit......Page 257
1.2. Mechanical and Textural Properties......Page 259
1.3. Testing and Measurement......Page 264
1.4. Sound and Ultrasound Properties......Page 267
1.5. Light Properties......Page 268
1.6. Near Infrared Properties......Page 270
1.8. Recommended Future Research Directions......Page 271
2.2. Mangosteen Sizing Machine......Page 279
2.3. Semi-Automatic Young Coconut Fruit Trimming Machine......Page 281
2.4. Semi-Automatic Young Coconut Fruit Opening Machine......Page 282
2.6. General Purpose Coconut Dehusking Machine......Page 283
2.7. Unripe Durian Cutting Machine......Page 284
2.8. Baby Corn Husking and Flesh Separating Machine......Page 285
2.9. Recommended Research Trend......Page 286
3.1. Present Status......Page 295
3.2. Interior Packaging Management......Page 297
3.3. Edible Films and Coatings for Fresh Produce......Page 299
3.4. Refrigerated Storage and Packaging......Page 302
3.5. Recommended Research Trend......Page 304
CONCLUSION......Page 312
REFERENCES......Page 313
ABSTRACT......Page 323
INTEREST OF DEVELOPMENT OF GEL PRODUCTS CONTAINING FRUIT PIECES......Page 324
THE USE OF OSMOTIC DEHYDRATION FOR THE FORMULATION OF FRUIT-GEL PRODUCT’S......Page 326
OPTIMIZATION OF PRODUCT FORMULATION IN A PROCESS WITHOUT BY-PRODUCT GENERATION......Page 336
CHARACTERIZATION AND STABILITY OF THE PRODUCT......Page 341
CONCLUSIONS......Page 348
REFERENCES......Page 349
ABSTRACT......Page 355
DEHYDRATED AND REHYDRATED FRUITS, A TRADITIONAL MARKET......Page 356
FOOD DRYING PROCESS. APPLICATION OF MICROWAVE TECHNOLOGY......Page 357
PRE-DRYING TREATMENTS......Page 358
HEAT TRANSFER MECHANISM AND DRYING KINETICS......Page 360
Optical Properties......Page 362
Structural Aspects......Page 365
CONCLUSION......Page 371
REFERENCES......Page 372
ABSTRACT......Page 377
INSECTS......Page 378
Adult......Page 379
Rusty Grain Beetle......Page 380
Rice Weevil......Page 381
Yellow Mealworm......Page 382
Confused Flour Beetle......Page 384
Cooling and Heating......Page 387
Diatomaceous Earth......Page 388
Chemicals......Page 389
HIGH PRESSURE CARBON DIOXIDE PROCESS (HPCO2PR) FOR INSECT CONTROL......Page 390
Researches and Developments on the HPCO2Pr......Page 392
Commercial Applications......Page 395
Operations......Page 399
Products Applied with High Pressure Carbon Dioxide System......Page 405
Cost of HPCO2Pr......Page 406
CHEMICAL AND PHYSICAL PROPERTIES OF CO2 (CITED FROM BOC)......Page 407
REFERENCES......Page 409
ABSTRACT......Page 413
2.1. Flow Through a Tube......Page 414
2.2. Flow through a Rectangular Channel......Page 429
3.1. Tubular Geometry......Page 443
3.2. Rectangular Geometry......Page 453
NOMENCLATURE......Page 463
REFERENCES......Page 465
A.1. Effective Viscosity ( eff μ )......Page 466
A.3. Effective Viscosity for Power Law Fluid......Page 467
Flow through a Rectangular Thin Channel......Page 468
ABSTRACT......Page 469
INTRODUCTION......Page 470
Sample Preparation......Page 472
Analysis......Page 473
Sample Characteristics......Page 474
The Effect of Roll Speed......Page 475
The Effect of Roll Differential......Page 478
REFERENCES......Page 481
ABSTRACT......Page 483
INTRODUCTION......Page 484
Roasting of Cocoa Beans......Page 485
RESULTS AND DISCUSSION......Page 486
Pyrazines......Page 492
Carbonyl Derivatives......Page 493
REFERENCES......Page 495
ABSTRACT......Page 499
INTRODUCTION......Page 500
Bulgurated-Wheat......Page 501
Bulgurated Triticale......Page 502
PRINCIPLES OF BULGURATION......Page 504
BULGURATION OPERATION......Page 506
REFERENCES......Page 509
ABSTRACT......Page 513
2. MODELS......Page 514
3. TEMPERATURE DEPENDENCE OF THE PARAMETERS OF SORPTION ISOTHERMS......Page 516
4. THE EQUATIONS FOR ISOSTERIC ENTHALPY OF SORPTION FROM THE STUDIED MODELS......Page 517
5. THE DESCRIPTION OF EXPERIMENTAL DATA......Page 518
6. EXPERIMENTAL DATA AND RESULTS......Page 519
6.2. Fitting the Multitemperature Data......Page 523
7. CONCLUSION......Page 527
REFERENCES......Page 528
NOMENCLATURE......Page 530
ABSTRACT......Page 533
1. INTRODUCTION......Page 534
3. RESULTS......Page 535
3.1. Food Calorie Supply......Page 539
3.2. Food Protein Supply......Page 540
4.1. Conclusions......Page 541
4.2. Balance of Food Nutrition Supply and Demand......Page 542
REFERENCES......Page 545
INDEX......Page 547
Recommend Papers

Focus on Food Engineering Research and Developments
 1600218989, 9781600218989, 9781606925676

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

FOCUS ON FOOD ENGINEERING RESEARCH AND DEVELOPMENTS

FOCUS ON FOOD ENGINEERING RESEARCH AND DEVELOPMENTS

VIVIAN N. PLETNEY EDITOR

Nova Science Publishers, Inc. New York

Copyright © 2007 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 Focus on food engineering research and developments / Vivian N. Pletney (editor). p. cm. Includes index. ISBN-13: 978-1-60692-567-6 1. Food industry and trade--Research. I. Pletney, Vivian N. TP370.8.F63 2006 664--dc22 2007028557

Published by Nova Science Publishers, Inc.

New York

CONTENTS Preface Expert Commentary

vii Computer-Vision Based Analysis of Color as a Tool for Food Process Control Vural Gökmen and İdris Süğüt

Chapter 1

Transport Phenomena During Drying of Food Materials Kamil Kahveci and Ahmet Cihan

Chapter 2

The Influence of Interactions Occurring Between Micro-Organisms on Predicting the Safety of Lactic Acid Cheese Izabela Steinka

Chapter 3

Chapter 4

Chapter 5

Chapter 6

1

13

165

The Development of Engineering Technology to Improve the Quality of Production of Tropical Fruit in Developing Countries B. Jarimopas, P. Sirisomboon, R. Sothornwit and A. Terdwongworakul

239

Development of Gel Products Containing Fruit Pieces Using Osmotic Treatments without byProduct Generation N. Martínez-Navarrete, M.M. Camacho, E. García-Martínez and M.E. Martín-Esparza

307

Quality Aspects of Dehydrated and Rehydrated Fruit in Relation to Drying Method C. Contreras, M.E. Martín-Esparza, A. Chiralt and N. Martínez-Navarrete

339

Pest Control Using High Pressure Carbon Dioxide as an Advanced Technology Mustafa Bayram

361

vi Chapter 7

Contents Effects of Permeation on Mass Transfer Coefficient for Laminar Non-Newtonian Fluid Flow in Membrane Modules During Clarification/ Concentration of Fruit Juice Sirshendu De, Sunando DasGupta and S. Ranjith Kumar

397

Effect of Smooth Roll Grinding Conditions on Reduction of Sizings in the Wheat Flour Milling Process Aleksandar Fistes and Gavrilo Tanovic

453

An Effect of Relative Air Humidity on the Content of Volatile Compounds in Roasting Cocoa Beans Wieslawa Krysiak, Teresa Majda and Ewa Nebesny

467

Chapter 10

Bulguration: Combined Cooking and Drying Operation Mustafa Bayram

483

Chapter 11

Water Sorption on Foodstuffs - Alternative Models 497 Sylwester Furmaniak, Artur P. Terzyk, Leszek Czepirski, Ewa Komorowska-Czepirska, Joanna Szymońska and Piotr A. Gauden

Chapter 12

A Forecast Analysis on Food Nutrition Supply and Demand Worldwide Wenjun Zhang, Wengang Zhou, Xiyan Zhang, Yongkai Xia and Wei He

Chapter 8

Chapter 9

Index

517

531

PREFACE Food engineering 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 - Drying has been one of the most important techniques used in food preservation for long years. The drying process has to be performed considering energy economy and the quality standards for the product. Therefore, it is of great importance to understand the physical phenomena taking place in the drying processes. Various mass transfer mechanisms such as molecular diffusion, capillary flow and hydrodynamic flow may take place during the drying process of food materials. Drying is generally composed of a series, parallel and/or series-parallel combination of these mechanisms. In addition to the complexity because of these various transport mechanisms in the drying processes, the structures of materials are also too complex. These constitute the main reasons that make the understanding and modeling the drying process difficult. There are three basic approaches used in modeling as empirical, semi-empirical, and theoretical. Empirical and semi-empirical approaches consider only external resistance to mass transfer between product and air while the theoretical approaches consider only internal resistance to mass transfer. At the theoretical modeling two kinds of approaches are used. These are discrete approach and continuum approach. In discrete approach, transport is examined in a network structure representing the material structure and generally the purpose of use of this approach is to determine transport parameters as an alternative to the experimental measurements. On the other hand, continuum approach is commonly used for describing the transport taking place at macroscopic level. In continuum approach, the food material is considered as a fictitious continuum and the effects

viii

Vivian N. Pletney

of the physical phenomena taken into consideration are lumped into effective transport coefficients. There are many models suggested based on continuum approach. The main difficulty in using a model based on a continuum approach arises from determination of these effective transport parameters. Most of the transport parameters are strongly dependent on concentration, temperature and material structure. Various models have been suggested to clarify the effect of temperature and concentration on transport parameters. However, relatively little is known on the effect of structure on transport parameters. In conclusion, it may be stated that drying of food materials is a complex unit operation and main problem to overcome on the way to better understand and describe drying processes is to reveal the effect of structure on transport. Chapter 2 - This paper discusses numerous problems occurring in relation to microbiological quality of lactic acid cheese. Lactic acid cheese constitutes the source of various nutritive substances, what results in a possibility of allochthonous micro-flora to grow despite the presence of starter micro-flora. One of the issues discussed herein comprised the results of microbiological research depending on tvarog packing system. The influence of packing system on surface micro-flora population was assessed. Moreover, the problem of growth of enterococci and LAB (Lactic Acid Bacteria) populations depending on stage of tvarog production as well as packing system was also raised. The issue of interactions occurring among micro-organisms that re-infect tvarogs and the influence of these interactions on the growth of individual micro-organisms was also discussed. The author also presented the possibility to apply JMTPH computer program for assessment of the dynamics of changes of tvarog micro-organisms during product storage. Another chapter includes assessment of the influence of lactic acid bacteria on the behaviour of individual groups of micro-organisms occupying tvarog surface, depending on packaging hermetic properties. It was also very important to assess the safety of tvarogs in the context of a possibility of enterotoxin synthesis in conditions of various packing systems. Finally, the models of optimising lactic acid cheese quality were presented, what included application of plant additives of biostatic character, modification of used packaging as well as employing the probabilistic mathematical model helpful in evaluation of enterotoxin synthesis, depending on the level of staphylococci and yeast populations. Chapter 3 - Many developing countries are rich in agricultural and food resources but are unable to maximize the export income they earn from them because they lack value-adding technology. In other words, developing countries typically must sell their products in cheap unfinished form to nations which possess the technology that adds profitability to these goods. Accordingly, if developing countries wish to earn more revenue for the improvement of their people’s employment and education, they must develop food engineering technology alongside other food science technologies. These efforts at technological self-improvement should be supported by the developed countries as the reduction of the knowledge and income gaps between the industrialized and developing worlds will do much to further global peace and happiness. The desired trend for food engineering research is to focus on developing engineering technology that will help to improve tropical fresh produce quality. This chapter discusses three facets of this trend. The first aspect concerns the physical properties of tropical fruit and vegetables, which consist of post-harvest loss, physical characteristics, mechanical properties, firmness, friction, and non-destructive quality grading techniques relating to mangoes, mangosteen, durian, sweet tamarind, guava, tangerines, snake egg plants, white long radish

Preface

ix

and lime. The second aspect concerns innovations in machinery and devices used with mangosteen, durian, young coconut, dry over-mature coconut and baby corn. State of the art design, operating principles and key performance tests of tropical fruit machinery and inventions will be reviewed. The third aspect concerns packaging technology, particularly that which is directed towards the extension of the shelf life of the aforementioned tropical fresh produce. There are three current realities which inform this book. They are as follows: that there is a high incidence of post-harvest loss and a corresponding magnitude of shortage in research and development work on tropical fresh produce; that the global flow of information is increasing while agricultural labor is becoming scarcer and more expensive; and that tropical produce engineering technology must be thoroughly understood. Accordingly, we make two recommendations: for producer countries to instigate a dramatic increase in the research and development that they conduct into tropical fresh produce, and in the support that they provide for this research; and that the research trend should cover all economic tropical fruit and vegetable goods grown in the producer countries and all aspects of engineering technology that they use, with a particular emphasis on developing computerized nondestructive techniques for quality assurance. Chapter 4 - Fruits are products of a very important nutritional interest. Nevertheless, and mainly due to their relatively short shelf-life and modern-day eating habits, the level of consumption is below that recommended by the World Health Organization. In this sense, the development of foods with a high fresh or processed fruit content, that maintain the nutritional and sensorial properties of the fresh fruit, may contribute to stimulating the interest of the consumer, thus increasing the product consumption. Osmotic dehydration (OD) techniques have been widely applied in fruit processing, since they require little energy and allow us to obtain high quality products. However, its industrial use may be limited by the management of the osmotic solution (OS). To solve this problem, the re-use of the OS in more than one OD cycle, with or without a previous re-concentration stage, may be considered. When there is no re-concentration, the re-use will be limited by the possible microbiological contamination and by the progressive dilution that takes place after each OD cycle, which may affect the kinetics of the osmotic process. On the other hand, as some native hydrosoluble compounds, such as volatiles, acids, minerals, vitamins and phytochemicals, will be released together with water into the OS during OD, its management as an ingredient in some product formulation seems to be an interesting alternative. To this end, this work analyses the viability of formulating a fruit-gel product with the osmodehydrated fruit (strawberry, kiwi or grapefruit) and the re-used OS obtained from the dehydration step, in order to diminish the loss of flavour, aroma and functional components of the fruit and avoid the generation of by-products in the process. In this study, the number of OS re-use cycles has been optimized, on the basis of its microbial recounts, the dilution level, the solution enrichment in fruit bioactive compounds, the fruit-solution ratio used during the dehydration step and the fruit-gel ratio in the final product. The kind and concentration of gelling agents, which best favour the properties of aspect (transparency) and texture of the gels, taking the peculiar composition of the re-used OS used as gelling medium into account, have been identified. The conditions in which the fruit pieces are mixed with the gelling solution have also been studied and defined. Finally, the fruit-gel product formulation conditions have been optimized, on the basis of its sensory acceptance and its compositional stability during storage, ensuring the thermodynamic equilibrium between the fruit and the gel when mixed.

x

Vivian N. Pletney

The microbiological stability of the product was of at least 15 days in refrigerated storage. During this time, the evolution of some properties such as phytochemicals, vitamins, acids, volatile compounds, colour and texture was studied. Chapter 5 - The development of new attractive dehydrated fruit-based products, to be consumed as dried or rehydrated, with high quality and reasonable shelf-life, will increase and diversify its availability in the market. In this sense, it is necessary to optimize the dehydration operation conditions to achieve not only the maximum process efficiency and control, but also various characteristics in the final product in relation to colour, texture, water activity, nutritive value, etc. Air drying has been the most frequently selected process for industrial food dehydration, due to its efficiency, versatility and easy management. However, it is known that it provokes considerable changes in sensory and nutritional quality. Some research works refer to the advantages of applying microwaves to convective drying associated with the fast volumetric heating of the product due to its high penetration power. On the other hand, the application of certain pre-treatments before drying operation, such as vacuum impregnation or vacuum pulsed osmotic dehydration, could help to enhance the stability and quality attributes, as high temperatures are not employed and specific solutes can be incorporated into the porous structure. In this chapter the advantages of microwave application to convective drying of apple and strawberry are pointed out. These are related to the great reduction in process time and to the fact that they allow obtaining a dehydrated product with a greater resistance to deformation and fracture and a greater stability during commercialization. Nevertheless, its use is not recommendable when the product has to be used or eaten after its rehydration, as the structural damage caused by microwaves decreases the mechanical resistance and the retention capacity of the incorporated liquid phase. The colour of dehydrated or rehydrated product is more affected by microwave treatments when the fruit pigment content is relevant, as occurs with strawberry anthocyanins. Application of a previous vacuum impregnation/osmotic dehydration step with sugared solutions is always recommended. Chapter 6 - Food products are always under the risk of infestation by pests. In view of the competitive markets, there has been increasing demand for quality in foods in terms of freedom from pest and pesticide contaminants. Also, it is very important for trade purpose suffer economic and quality losses. Zero tolerance of insect pest in foods has been adopted in some of countries and there is a tendency to achieve this goal in overall the world. The governments, the food industries and exporters are dependent on fumigation as a quick and effective tool for insect pest control in food commodities. Fumigants are widely used for pest elimination in these commodities. Toxic substances have therefore been used to destroy for example pests, as well as their eggs, larvae, cocoons and adults. Currently used substances, such as methyl bromide, hydrogen phosphide, ethylene dioxide, malathion etc., are characterized by more or less serious problems. In recent years, that fumigation technology based on the chemical control of products has been facing threats/constraints because of regulatory concerns, the development of resistance, handling hazards, residues, food safety, cost, carcinogenicity, involvement in ozone depletion, resurgence, environmental pollution and other factors. Reliance upon fumigation as an overall solution to infestation problems in food products has become questionable. The chemical action of fumigants upon commodities and the environment has necessitated the withdrawal of many fumigants from the market. Also, some of them are being phase out their uses at the international level.

Preface

xi

Due to becoming the target of increasing criticism of toxic substances, such concerns have led to the development of non-chemical methods for the control of insect pests that infest food commodities. One such method is the high pressure carbon dioxide application, which mainly involves the use of CO2 at high pressure (10-40 bar) for food fumigation. It is a new effective, non-chemical, non-residual, safe, fast and environmentally friendly method for the food industry. It has been generated and developed within last 20 years. Carbon dioxide is a fumigant and being used to control pests in the food industry. After extensive testing, high pressure carbon dioxide fumigation can be accepted as the advanced pest control technology for the future. Nowadays, it is particularly indispensable for the gentle, safe, natural and organic food products. If operation time for the fumigation is constraint and nonchemical treatments are required, this technique is suitable for conventional products. Chapter 7 - Membrane based clarification and concentration of fruit juice has become a popular unit operation in modern fruit juice processing industries. The well known membrane modules used for this purpose are tubular and spiral wound modules. Therefore, design of these modules is of utmost industrial importance. The key parameter for design of membrane modules is mass transfer coefficient. Most of the fruit juices have non-Newtonian rheology, e.g., power law, ellis fluid, etc. Till today, the mass transfer coefficient for such systems used is approximated from the corresponding relations developed for Newtonian fluids. Hence, a detailed fluid flow modeling with non-Newtonian rheology is urgently warranted. In the present work, this aspect is attempted. The expressions of the mass transfer coefficients are derived from the first principles for laminar, non-Newtonian fluid flow in a porous conduit. The effects of the permeation are incorporated quantitatively in the mass transfer coefficient from a theoretical basis. The analysis is carried out for various non-Newtonian rheologies. Effects of the operating conditions, i.e., Reynolds number, permeate flux, etc. on mass transfer coefficient are also investigated. Two flow geometries are considered. Flow through a tube and that through a rectangular thin channel, which are useful for the design of the tubular and spiral wound cross flow membrane modules. The developed relations of mass transfer coefficients would be of tremendous help to the design engineers. Chapter 8 - A laboratory roll stand Variostuhl, equipped with smooth rolls (250 mm diameter, 100 mm length), was used to examine, under simulated commercial conditions, the effect of roll speed and roll differential on the reduction of sizings and coarse middlings from the primary break passages of the wheat flour milling process. The samples were obtained from the industrial mill, intercepting the sizings and coarse middlings from the 1st, 2nd and 3rd break stage that normally would have gone to the purification system, as well as intercepting the purified sizings (cleaned middlings) that normally would have gone to the reduction system of the wheat flour milling process. As roll velocity increases flour release was increased, milling energy consumption rose while flour quality (as determined by ash content) was not affected. By increasing roll velocity it is possible to increase feed rate to the rolls and, therefore, the disposable roll surface is used more efficiently. Flour release rose when differential was increased from 1.1 up to 1.25 but decreased when differential increased from 1.25 up to 5.0. Increasing roll differential led to an increase in milling energy consumption. These effects can be explained by the relative contribution of compressive and shearing forces acting on the particles passing through the grinding zone of the smooth rolls. Considering the results obtained in this study (flour release, flour quality and milling energy

xii

Vivian N. Pletney

consumption) a differential of 1.25, relative to a fast roll speed of 5 m/s could be designated as optimal. Chapter 9 - The Ivory Coast cocoa beans were convectively roasted at 135°C, at the air flow rate of 1.0 m/s and relative air humidity (RH) of 0.4%, 2.0% and 5.0%. Volatile components of raw and roasted beans were analyzed by SPME/GC/GCMS and identified by comparing their retention indices with that of standards included in a database and their mass spectra with standard spectra included in NIST computer library. Almost 100 different volatile compounds were identified in examined samples of roasted cocoa. They ranked among aldehydes, ketones, alcohols, esters, monoterpenes, pyrazines, acids, lactones, furan derivatives, and sulfur-containing compounds. It was found that a rise in the relative air humidity from 0.4% to 2.0 and 5.0% increased the contents of pyrazines, volatile acids, esters, furan derivatives, and sulfur-containing compounds in a headspace of roasted cocoa. In contrast, the contents of alcohols and aldehydes in the headspace were considerably lower when the cocoa beans were roasted at the relative air humidity of 5.0% as compared to that when less humid air was used for convective heating. Chapter 10 - Cooking and drying are two main unit operations used widely in food processing. Consecutive cooking and drying operations supplies perfect properties to gain to food products and called as bulguration. Individually, the former method is used nearly for all food products before consumption. Cooking is a well-known way to destruct microorganisms, insect, insect eggs and larvaes for food safety. Also, it increases the digestive property of food with starch gelatinization, protein gelation and textural softening. However, it is very difficult to store this product without drying due to its high moisture content after cooking. Therefore, food products should be dried. Drying is required to prolong storage time of food products. Bulguration is the gaining of the some functional characteristics on the finished product such as the resistance to mold contamination, insect attacks and radiation, inactivation of enzymes and microorganisms, encapsulation of numerous nutritional components in food products, easy preparation after bulguration due to semi and ready-to-eat form, obtaining long shelf-life having economical products with safety, decreasing undesired components e.g. phytic acid in contrast to increasing desired one e.g. folate/folic acid. As raw materials, cereals, pulses, seeds, vegetables, fruits etc. can be used. Recently, the use of bulguration in the food industry dramatically increases as an optimal method due to above situations. Bulguration is an ancient technique; however, the modern technology re-discovered it. In this chapter, the techniques of bulguration are explained with examples. Also, the results of the recent researches are given. Chapter 11 - It is well known that sorption isotherms of foodstuffs are very important for design, modeling and optimization of important processes for example drying, aeration, predicting of stability and quality during packaging and storage of food. Many literature reviews conclude that the BET (and its modifications) and the GAB sorption isotherm equations are the most popular and applicable for description of isotherms of foodstuffs. The authors showed recently the applicability of the GDW model for description of water sorption on different foodstuffs. Moreover, it was also shown that the GAB model (also widely applied in food science) is the special case of the GDW equation. In this review the authors present the current state of art and also an attempt of application of different models of water sorption, namely CMMS, DD and modified CDS for description of water sorption data on different starch samples and other foodstuffs.

Preface

xiii

Chapter 12 - This paper aimed to make a longer-term forecast analysis on global food nutrition supply and demand. The forecasts of supplies of food calories and proteins for the world and various regions over the period 2010-2030 were given, and food nutrition supply and demand balance in the forecast period was discussed. If the past pattern continues, the global total food calorie supply would grow at the annual rate of 13.43±0.71 kcal/cap/day and reach 3210.4±67.3 kcal/cap/day in 2030. Total food calorie supplies for all of the regions would grow during the forecast period and, in most regions they are forecast to be greater than 3000 kcal/cap/day from 2015-2020. Total food protein supply for all regions but not Oceania, is forecast to grow during the forecast period. The proportion of animal sourced protein in total food protein supply is in 2030 forecast to increase and reach 35.5%, 61.6%, 56.8%, and 21.7% for Asia, Europe, South America, and Africa. Food calorie supply in the world is expected to exceed the adequate energy intake after around 2015. Strong focus should be worldwide put on the over-intake of food calorie in the near future. Global food protein supply is not expected to be greater than the adequate range during the period 2010-2030. Food protein supply in Africa and Caribbean would be just a little greater than the basic demand in the forecast period. Food protein intake in these regions should be improved in the coming years.

In: Focus on Food Engineering Research and Developments ISBN: 978-1-60021-898-9 Editor: Vivian N. Pletney, pp. 1-12 © 2007 Nova Science Publishers, Inc.

Expert Commentary

COMPUTER-VISION BASED ANALYSIS OF COLOR AS A TOOL FOR FOOD PROCESS CONTROL Vural Gökmen and İdris Süğüt Department of Food Engineering, Hacettepe University, 06800 Beytepe, Ankara, Turkey

ABSTRACT The color is the first sensation that the consumer perceives and uses as a tool to accept or reject, because the color observation allows the detection of certain anomalies or defects of a product. Commercial color-measuring devices are designed to contact the materials to perform the color measurements. The main drawback of using these devices is the limitations in size and geometry of the area that subjected to color measurement, because they measure a small area with a fixed geometry making the measurement quite unrepresentative for heterogeneous materials like many processed foods. With a digital imaging system, it is possible to register the color of foods using three-color sensors. In this case, surface color of the food is detected online in a non-contact manner and monitored throughout the process. By using appropriately developed computer algorithms, highly accurate and reliable information can be obtained about the color changes in a food during processing. This kind of algorithm can be used as a process control tool for automatic visual inspection in an industrial production process and can improve the overall quality of the product. The advantage of computerized visual inspection over inspection by humans is that machines can evaluate color continuously and objectively. This chapter describes novel approaches for a non-contact computer vision based color measurement system and its potential applications in food processing.

1. INTRODUCTION The overall appearance of any object is a combination of its chromatic and geometric attributes. Both of these attributes should be accounted for when making visual or instrumental assessments of appearance. The color is the first sensation that the consumer

Vural Gökmen and İdris Süğüt

2

perceives and uses as a tool to accept or reject, because the color observation allows detecting certain anomalies or defects of a product. The human eye has receptors for short (S), middle (M), and long (L) wavelengths, also known as blue, green, and red receptors. This means, in principle, three parameters are required to describe a color sensation. A specific method for associating three parameters (or tristimulus values) is called a color space which specifies how color information is represented. A color component is also referred to as a color channel. The XYZ is the first color space that mathematically defined by the Commission Internationale d'Eclairage (CIE) in 1931. The L*a*b* color space is perceptually uniform and the most complete model defined by the CIE in 1976 to serve as a device-independent, absolute model to be used as a reference. It is based on the XYZ color space as an attempt to linearize the perceptibility of color differences, using the color difference metric described by the Macadam ellipse. The non-linear relations for L*, a* and b* are intended to mimic the logarithmic response of the human eye. Here, L* is the luminance or lightness component, which ranges from 0 to 100, and parameters a* (from green to red) and b* (from blue to yellow) are the two chromatic components, which range from –120 to 120 [1-3].

2. INSTRUMENTAL MEASUREMENT OF COLOR A lot of color-measuring instruments are available in the market for several applications. Most of them are designed to contact the materials to perform the color measurements. These instruments are successfully used for color measurement of homogenous materials. However, the main drawback of using a color-measuring instrument is the limitations in size and geometry of the area that subjected to measurement. A commercial instrument usually measures a small area with a fixed geometry. This makes the measurement quite unrepresentative for heterogeneous materials like many food items [4]. Repetitive measurements are therefore required to increase the accuracy. In most cases, increasing the repetitions is not a viable approach for irregularly shaped objects. Instead, an approach taking the overall surface into account is required to obtain meaningful information about the color. This is especially important for industrial applications in which the color homogeneity is an important feature of the material. In such a case, commercial color measuring instruments are not fit for purpose as a process control and/or product quality control tool.

3. COMPUTER VISION BASED MEASUREMENT OF COLOR A typical image captured by a digital camera consists of an array of vectors called pixels. Each pixel has red, green and blue color values:

⎡ x r ( n, m ) ⎤ ⎥ ⎢ x[n, m] = ⎢ x g (n, m)⎥ ⎢ x (n, m ) ⎥ ⎦ ⎣ b

Computer-Vision Based Analysis of Color as a Tool for Food Process Control

3

where, Xr, Xg and Xb are values of the red, green and blue components of the (m,n)th pixel, respectively. In digital images, Xr, Xg and Xb color components are represented in 8 bits, i.e., they are allowed to take integer values between 0 and 255 (=28-1) [5]. RGB values of an image captured by a digital camera can be converted into device-independent L*a*b* units. Computational approaches that convert RGB values into L*a*b* units have been previously reported using the standard equations [2,6,7]. However, usefulness of a computer vision system as a tool for color measurement depends on the accuracy of color transformation. Digital cameras have built-in white-balancing systems modifying actual color values, therefore pixel values in an image captured by a camera of a machine vision system or a consumer camera may not correspond to true colors of imaged objects. Therefore, it is better to build calibrated models by using the charts which reflect the variations in color space. The necessity of a calibration process to obtain device-independent L*a*b* color units have been previously underlined [8]. However, the information is lacking on how the accuracy of color measurement could be improved by calibration process in these reports.

3.1. Description of the Technique Following section describes a computer-vision based technique calibrated with ANN modeling to measure color in foods. The model allows a user to determine a polygonal region of interest. This feature increases the accuracy in color measurement when compared to commercial color-measuring devices. To perform the color measurement, digital images of food samples are taken using a color digital camera under well controlled conditions (illumination, lamp angle and distance). The calibration of computer vision based color measurement system was performed by using an Agfa 5x7 inch reflective color chart (figure 1) which is an internationally accepted IT8 standard (IT8.7/2-1993) with device-independent color definitions. The chart consists of 288 colored squares, and has been designed to represent the color space from full saturation to near neutrals at highlight, mid-tone and shadows. Figure 2 shows the algorithm used to convert camera RGB values to spectrophotometric CIE L*a*b* values. Based on this algorithm, the first step is the conversion of color values from RGB to L*a*b* using the standard conversion equations, and the second step is the correction of L*a*b* values through an ANN model.

Vural Gökmen and İdris Süğüt

4

Figure 1. Agfa 5x7 inch color chart (IT8.7/2-1993) used to build a calibrated ANN model for the correction of monitor L*a*b* values. = 2.4

RG B va lue s b y d ig ita l c a me ra

RG B to XYZ

/ / O b se rve r:2o ; Illumina nt: D50 va r_R = ( R / 255 ) va r_G = ( G / 255 ) va r_B = ( B / 255 )

C o nve rsio n o f c o lo r va lue s RG B C IE L*a *b * b y using e q ua tio ns

/ / Whe re R = 0 ÷ 255 / / Whe re G = 0 ÷ 255 / / Whe re B = 0 ÷ 255

if ( va r_R > 0.04045 ) va r_R = ( ( va r_R + 0.055 ) / 1.055 ) ^ 2.4 e lse va r_R = va r_R / 12.92 if ( va r_G > 0.04045 ) va r_G = ( ( va r_G + 0.055 ) / 1.055 ) ^ 2.4 e lse va r_G = va r_G / 12.92 if ( va r_B > 0.04045 ) va r_B = ( ( va r_B + 0.055 ) / 1.055 ) ^ 2.4 e lse va r_B = va r_B / 12.92 va r_R = va r_R * 100 va r_G = va r_G * 100 va r_B = va r_B * 100

C o rre c tio n o f c o lo r va lue s C IE L*a *b * C IE L*a *b * b y using a n ANN mo d e l

C IE L*a *b * va lue s b y ANN

X = va r_R * 0.4360 + va r_G * 0.3851 + va r_B * 0.1431 Y = va r_R * 0.2225 + va r_G * 0.7169 + va r_B * 0.0606 Z = va r_R * 0.0139 + va r_G * 0.0971 + va r_B * 0.7142 XYZ to L*a *b *

va r_X = X / re f_X / / re f_X = 96.422 va r_Y = Y / re f_Y / / re f_Y = 100.000 va r_Z = Z / re f_Z / / re f_Z = 82.521 if ( va r_X > 0.008856 ) va r_X = va r_X ^ ( 1/ 3 ) e lse va r_X = ( 7.787 * va r_X ) + ( 16 / 116 ) if ( va r_Y > 0.008856 ) va r_Y = va r_Y ^ ( 1/ 3 ) e lse va r_Y = ( 7.787 * va r_Y ) + ( 16 / 116 ) if ( va r_Z > 0.008856 ) va r_Z = va r_Z ^ ( 1/ 3 ) e lse va r_Z = ( 7.787 * va r_Z ) + ( 16 / 116 ) CIE-L* = ( 116 * va r_Y ) - 16 CIE-a * = 500 * ( va r_X - va r_Y ) CIE-b * = 200 * ( va r_Y - va r_Z )

Figure 2. Algorithm used to convert camera RGB values to spectrophotometric CIE L*a*b* values.

Computer-Vision Based Analysis of Color as a Tool for Food Process Control

5

An ANN is a nonlinear mathematical model that learns from the examples through iterations. ANNs are made of a large number of nodes or artificial neurons, which are disposed in a parallel structure. Each ANN has one input layer containing one node for each independent variable, one or more hidden layers, where the data are processed, and one output layer, containing one node for each dependent variable. The data from the input layer are propagated through the hidden layer and then to all network, which are associated with a scalar weight. Neurons in the hidden and output layers calculate their inputs by performing a weighted summation of all the outputs they receive from the layer before. Their outputs, on the other hand, are calculated by transforming their inputs using a non-linear transfer function. Then, the network output is compared with the actual output provided by the user. The difference is used by the optimization technique to train the network. Thus, the training process requires a forward pass to calculate an output and a backward pass to update the weights in feed-forward back-propagation networks [9,10]. A great advantage of ANN models is that, they do not require prior knowledge of the relationship between the input and output variables, and instead of that, they figure out these relationships through training. Therefore, complex processes can be optimized to produce the desired outputs using successfully trained ANN models. A feedforward backpropagation network was used for the conversion of monitor color values to spectrophotometric color values in CIE L*a*b* units. Sigmoid function (Eq.1) was used in the hidden layers as the transfer function which gave outputs in the range [-1,1].

f ( x) =

1 1 + exp(− x)

Eq. (1)

In the output layer, the training function was purelin (Eq. 2) which gave outputs in the range [-∞,+∞].

f ( x) = x

Eq.(2)

Representative examples, obtained by experimental data sets, were presented to the network so that it could integrate this knowledge within its structure. The learning function was the gradient descent with momentum and Bayesian regularization was used as the training algorithm. Several ANN topologies were trained using the experimental data sets. The data set consisting of 288 data points were divided into two parts for training (252 data points) and testing (36 data points). Input and output layers consisted of three neurons which corresponded to monitor color values and spectrophotometric color values in CIE L*a*b* units, respectively. After each training session, train and test data sets were simulated by ANN. The error measures used for comparing the performance of various ANN configurations were estimated using Eq. (3);

SSE = ∑ (VD − VP ) 2 n

i =1

Eq. (3)

Vural Gökmen and İdris Süğüt

6

where, n is the number of data points, VD and VP are the measured and the predicted color values.

3.2. Accuracy of Color Measurement Among the different network architectures, the 3-6-6-3 network topology appeared as the best performing one which exhibited a great capability for calibrating the monitor color values to the spectrophotometric color values in CIE L*a*b* units (figure 3).

Mo nito r Co lo r

Sp e c tro p ho to me tric Co lo r va lue s

L*

L*

a*

a*

b*

b*

Figure 3. Topology of the ANN model used to correct monitor L*a*b* values.

The accuracy in color measurement is usually determined as the Euclidian distance (Eq. 4) between two colors (ΔE). ΔE of less than 1.0 is usually desired for accuracy of a color measurement system.

ΔE = ( L1 − L2 ) 2 + (a1 − a2 ) 2 + (b1 − b2 ) 2

Eq. (4)

ΔE of less than 1.0 is usually considered as an indication of high accuracy for a color measuring instrument. The ΔE values for the estimated values and the real spectrophotometric values of 288 colored squares in the IT8 chart used for training and testing is shown in figure 4. The maximum ΔE value was determined to be 0.45 in the computer vision based color measurement system described above.

Computer-Vision Based Analysis of Color as a Tool for Food Process Control

7

a.

b. Figure 4. Change of ΔE value for the patches on IT8 color chart (a) the color patches used for training, (b) the color patches used for testing.

Vural Gökmen and İdris Süğüt

8

The commercial colorimeters measure color in a small area with fixed geometry (usually circle in a few square centimeters). This is the major drawback of the commercial instruments which makes the color measurement problematic especially for heterogeneous materials. It is the usual way to make repetitive measurements over the surface in order to obtain meaningful information about the color. From the industrial point of view, it is unattractive to make several measurements and averaging the results for a single product to get useful information about the color. In practice, a system which has the capability of measuring color in the overall area of a material is advantageous and of commercial interest. In this respect, averaging the values of all pixels in a digital image of a material using a non-contact computer vision technique is absolutely better than averaging the values of repetitive measurements using a contact colorimeter in terms of accuracy and analysis time. For example, after the frying process, different kinds of image pixels appear in a typical fried potato image (figure 5a). The color of this potato crisp can be measured by marking a polygonal area over the image which represents the whole surface as shown in figure 5b. Color measurements were also performed by a color spectrophotometer (Minolta model CM3600d spectrophotometer) or a portable colorimeter (ColorSavy model CM2C colorimeter) for the same potato crisp sample making a dozen of measurements from different regions.

a.

b.

Figure 5. (a) Digital image of a potato crisp sample, (b) polygonal area marked on the image of potato crisp which subjected to color measurement by computer vision based analysis.

Figure 6 shows the results of pixel by pixel analysis of color performed by the computer vision based color measurement system and the results of 12 repetitive measurements performed by the spectrophotometer and the portable colorimeter. In order to emphasize the capability of computer vision based color measurement system, data points are represented in colors as the monitor RGB counterparts of each data points in CIE L*a*b* units determined by the appropriate technique. It is clear from the results that the color values determined by the spectrophotometer and by the portable colorimeter are perceptually not accurate. This

Computer-Vision Based Analysis of Color as a Tool for Food Process Control

9

case exemplifies that commercial instruments have certain limitations when dealing with the color of rough and heterogeneous materials like potato crisps. In addition to the advantage of averaging color values of all pixels, the measurement of color in a region of interest over a material’s surface may be of greater importance under certain circumstances. In that case, measuring the color in a user defined polygonal area may be useful to get appropriate information about the color of different regions over the surface of a single material. A commercial device is not capable of extracting color information for a specific region of a heterogeneous material. Figure 7 shows a cookie sample composed of two regions with and without cocoa. Two regions with different shapes in this image can be processed pixel by pixel to determine average L*, a* and b* values separately by defining polygonal area of interest.

Figure 6. The results of color measurements by means of computer vision technique (pixel by pixel measurements), color spectrophotometer and portable color mouse (12 repetitive measurements).

Vural Gökmen and İdris Süğüt

10

a.

b.

Figure 7. (a) Digital image of a cookie sample composed of two regions, (b) polygonal areas marked on the image of cookie sample which subjected to color measurement by computer vision based analysis.

4. CONCLUSION The computer vision based image analysis system described here offers some advantages over the commercial color-measuring instruments namely the possibility of performing noncontact color measurement without sample preparation, and extracting meaningful information in a specific region of interest over a material surface. This kind of system can be used as a tool for automatic visual inspection of colors in an industrial production process and can improve the overall quality of the product. The advantage of computerized visual inspection over inspection by humans is that machines can evaluate color continuously and objectively.

REFERENCES [1] [2]

[3] [4]

Papadakis, S.E.; Abdul-Malek, S.; Kamdem, R.E.; Yam, K.L. (2000). A versatile and inexpensive technique for measuring color of foods. Food Technology 54(12), 48-51. Segnini, S.; Dejmek, P.; Öste, R. (1999). A low cost video technique for colour measurement of potato chips. Lebensmittel-Wissenschaft and Technologie. 32(4), 216-222. Yam, K.L.; Papadakis, S. (2004). A simple digital imaging method for measuring and analyzing color of food surfaces. Journal of Food Engineering. 61, 137-142. Antonelli, A.; Cocchi,M.; Fava, P.; Foca, G.; Franchini, G.C.; Manzini, D.; Ulrici, A. (2004). Automated evaluation of food colour by means of multivariate image analysis coupled to a wavelet-based classification algorithm. Analytica Chimica Acta. 515, 313.

Computer-Vision Based Analysis of Color as a Tool for Food Process Control [5] [6] [7] [8] [9]

[10]

11

Gonzales, R. C.; Woods, R. E. (2002). Digital Image Processing, Prentice Hall, New Jersey. Paschos, G. (2001). Perceptually uniform color spaces for color texture analysis: An empirical evaluation. IEEE Transactions on Image Processing. 10(6), 932-937. Mendoza, F.; Aguilera, J.M. (2004). Application of image analysis for classification of ripening bananas. Journal of Food Science. 69, 471-477. León K.; Mery, D.; Pedreschi, F.; León, J. (2006). Color Measurement in L*a*b* units from RGB digital images. Food Research International. 39, 1084-1091. Gonçalves, E.C.; Minim, L.A.; Coimbra, J.S.R.; Minim, V.P.R. (2005). Modeling sterilization process of canned foods using artificial neural networks. Chemical Engineering Process. 44, 1269-1276. Bishop, M.C., (1994). Neural network and their applications, Review in Scientific Instruments. 65(6), 1803–1832.

APPENDIX 1. MATLAB CODE FOR EXTRACTING THE COLOR INFORMATION OF A POLYGONAL MARKED REGION IN AN IMAGE RGB=imread(‘image.extension’); RGB=im2double(RGB); Z=roipoly(RGB); [d1,d2]=size(Z); c=0; L=[ ]; for a=1:d1 for b=1:d2 if Z(a,b)==1 n=1; c=c+1; L(n,c)=a; n=2; L(n,c)=b; end end end P=[ ]; for n=1:c P(n,:)=impixel(RGB2,L(2*n),L(2*n-1)); end roired=[ ];roigreen=[ ];roiblue=[ ]; sum_red=0;sum_green=0;sum_blue=0; for n=1:c sum_red=sum_red + P(n,1);

12

Vural Gökmen and İdris Süğüt sum_green=sum_green + P(n,2); sum_blue=sum_blue + P(n,3); end roired=sum_red/c; roigreen=sum_green/c; roiblue=sum_blue/c; RGB__value=[roired roigreen roiblue]

In: Focus on Food Engineering Research and Developments ISBN: 978-1-60021-898-9 Editor: Vivian N. Pletney, pp. 13-163 © 2007 Nova Science Publishers, Inc.

Chapter 1

TRANSPORT PHENOMENA DURING DRYING OF FOOD MATERIALS Kamil Kahveci and Ahmet Cihan Mechanical Engineering Department, Trakya University, 22180 Edirne, TURKEY

ABSTRACT Drying has been one of the most important techniques used in food preservation for long years. The drying process has to be performed considering energy economy and the quality standards for the product. Therefore, it is of great importance to understand the physical phenomena taking place in the drying processes. Various mass transfer mechanisms such as molecular diffusion, capillary flow and hydrodynamic flow may take place during the drying process of food materials. Drying is generally composed of a series, parallel and/or series-parallel combination of these mechanisms. In addition to the complexity because of these various transport mechanisms in the drying processes, the structures of materials are also too complex. These constitute the main reasons that make the understanding and modeling the drying process difficult. There are three basic approaches used in modeling as empirical, semi-empirical, and theoretical. Empirical and semi-empirical approaches consider only external resistance to mass transfer between product and air while the theoretical approaches consider only internal resistance to mass transfer. At the theoretical modeling two kinds of approaches are used. These are discrete approach and continuum approach. In discrete approach, transport is examined in a network structure representing the material structure and generally the purpose of use of this approach is to determine transport parameters as an alternative to the experimental measurements. On the other hand, continuum approach is commonly used for describing the transport taking place at macroscopic level. In continuum approach, the food material is considered as a fictitious continuum and the effects of the physical phenomena taken into consideration are lumped into effective transport coefficients. There are many models suggested based on continuum approach. The main difficulty in using a model based on a continuum approach arises from determination of these effective transport parameters. Most of the transport parameters are strongly dependent on concentration, temperature and material structure. Various models have been suggested to clarify the

14

Kamil Kahveci and Ahmet Cihan effect of temperature and concentration on transport parameters. However, relatively little is known on the effect of structure on transport parameters. In conclusion, it may be stated that drying of food materials is a complex unit operation and main problem to overcome on the way to better understand and describe drying processes is to reveal the effect of structure on transport.

1. INTRODUCTION Drying of food materials is used as a preservation technique. Microorganisms that cause spoilage and decay can not grow and multiply in the absence of water [1]. In addition, enzymes that cause chemical changes can not function in an environment lacking water. It is therefore necessary to expose food materials to a proper drying process to reduce their water content. It is possible to classify drying processes in several ways. For example, drying processes can be classified as batch and continuous [1]. In batch drying process, food material is inserted into the drying equipment and drying proceeds for a given period of time. In the continuous drying process, however, food materials are continuously added to the dryer and dried material continuously removed. Drying processes can also be categorized depending on the physical conditions used to add heat and remove water vapor. The most common type of process under this categorization is the process in which heat is added by direct contact with heated air at atmospheric pressure and the resulting water vapor is removed by the air. In addition, there are other processes, in which heat is added indirectly through a metal wall or by radiation. Types of driers used in the food industry display a considerable diversity particularly depending on the type of process used in drying. Schematic views of some common driers are shown in figure 1. Brief information on various types of driers is also given below.

Sun Drying and Solar Dryers Drying of food materials by spreading them on an appropriate ground is called sun drying. In this case, part or all of the heat required for drying is supplied by direct radiation from the sun. Solar drying means the processes where solar collectors are used for heating the air.

Tray Dryers In tray dryers, food materials are usually laid on trays as a very thin layer. The required heat is supplied by the air sweeping over the trays or by conduction or radiation from heated trays.

Transport Phenomena During Drying of Food Materials

15

Tunnel Dryers In tunnel dryers, trays or trolleys containing food materials move through a tunnel, in which heat is supplied and water vapor is removed. Food material generally moves along the tunnel parallel or opposite to the direction of air flow.

Roller or Drum Dryers In this type of driers, food material is spread on the surface of a heated drum and the drum is rotated. The food remains on the drum surface for the greater part of the rotation during the time that drying takes place.

Fluidized Bed Dryers In fluidized bed dryers, food material is held suspended during the drying process with the help of the upward flow of the drying air. The air flow may be horizontal to help the food material to convey in the dryer. In this type of dryers a major part of the heat is transferred by convection.

Spray Dryers In this type of driers, liquid or fine solid material is sprayed into a heated air flow as fine droplet dispersion form. Drying speed is extremely high in this type of drying process. For this reason, this process is generally used in drying of food materials which damage due to prolonged exposure to hot air stream.

Pneumatic Dryers In pneumatic dryers, solid food particles are conveyed rapidly with an air stream. The heat necessary for drying is supplied by the air. Generally, a classifier is used in this type of driers. Dried particles are separated in this classifier. The remaining humid part is recirculated for additional drying.

Rotary Dryers In rotary dryers, food material is taken into a horizontal inclined cylinder. The necessary heat for drying is provided by an air stream moving along the cylinder or by the transfer of the heat from the walls of the cylinder by conduction. The cylinder rotates in certain types and in other types the cylinder is stationary and the material is conveyed by a paddle or screw rotating in the cylinder.

16

Kamil Kahveci and Ahmet Cihan

Trough Dryers In this type of dryers, food material is loaded on a trough-shaped conveyor belt made of mesh. Drying air is blown through the bed of material. The movement of the conveyor continually turns over the material and hence exposes wet surfaces to hot air.

Bin Dryers In bin dryers, food material is contained in a perforated bottom bin. Drying process is performed by vertically upward conveyance of hot air along the bin.

Belt Dryers In this type of dryers, food material is laid on a horizontal mesh or a solid belt and hot air is passed over the material. Mostly, the belt is mobile. In some cases, the belt is stationary and material is conveyed by scrappers.

Vacuum Dryers In vacuum drying, material is inserted in an evacuated cabinet and the drying process is performed in this cabinet. The required heat is mainly transferred by conduction or radiation. This process which allows drying at lower temperatures is generally used for food materials that are damaged by exposure to high temperatures.

Freeze Dryers In freeze dryers, food material is loaded on shelves or belts in a chamber under vacuum. Material is generally frozen before loading into the dryer. The heat is transferred to food by conduction or radiation. The resulting vapor is removed by a vacuum pump and is condensed. In certain cases, sheets of expanded metals and heated plates are inserted between food materials to enhance heat transfer and moisture removal.

Microwave Dryers Drying process in microwave dryers is performed using polarization occurring at molecular and atomic level. The heat developed in a material by an alternating electromagnetic field results from the polarization process within the product when molecules within the material rotate and move laterally millions of times per second in an attempt to align with the changing field. Microwave heating provides a uniform heat flux throughout the material.

Transport Phenomena During Drying of Food Materials

17

Radio Frequency Dryers In a radio frequency drying system, a RF generator creates an alternating electric field between electrodes. The material is conveyed between the electrodes. In this case, alternating energy causes polar molecules in the water to continuously reorient themselves to face opposite poles. This frictional movement causes the water content of food material to heat up rapidly and leave the medium by vaporization.

Infrared Dryers In the infrared dryers, the electromagnetic energy of infrared rays is used for drying. The depth of penetration of infrared is a function of its wavelength. Generally, the shorter the wavelength, the greater is its penetration power. The most important advantage of this type of drying is economy. For biological materials, however, infrared heater temperatures greater than 830°C should be avoided as this can char the product and cause surface damage [3].

Heat Pump Dryers Heat pump dryer consists of a drying chamber equipped with a circulation system and the components of an air-conditioning refrigeration system. The drying air is dehumidified by the evaporator and reheated by the condenser of the heat pump. The maximum drying temperature is determined by the condensing temperature of the refrigerant used. Heat pump drying is essentially a low-temperature process which can be controlled from -20°C to 70°C by selecting an appropriate refrigerant and regulating the compressor capacity and air flows within the system [4].

Superheated Steam Dryers In the superheated dryer, wet solids are fed into the flow of pressurized superheated transport steam by means of a pressure tight rotary valve, plug screw or similar. The transport steam is superheated indirectly via a tubular heat exchanger, by a heating media such as medium pressure steam, flue gases or thermal oil [5]. Also, electrical heating can be applied. In the subsequent drying ducts, moisture is vaporized from the product, forming excess transport steam and lowering its degree of superheat. Normally the residence time in the system is 5-60 seconds only. For some materials, a second superheater is necessary to achieve the required dryness. The dry material is separated in a high efficiency cyclone and the material is discharged from the dryer by means of another pressure tight rotary valve. From the cyclone, the transport steam is recycled by a centrifugal fan to the inlet of the first heat exchanger. The excess steam generated is continuously bled off.

18

Kamil Kahveci and Ahmet Cihan

Hybrid Dryers Particularly, dryers manufactured using latest technology consist of a combination of different types of drying processes. They are preferred mostly because the combination allows utilization of the most advantageous aspect of each process for food material and drying process. Common hybrid dryers used in drying are listed below. • • •

Microwave-Convective Dryers, Microwave-Infrared Dryers, Microwave-Vacuum Dryers, Microwave-Superheated Steam Dryers, Infrared-Convective Dryers, Infrared-Vacuum Dryer, Infrared-Heat Pump Dryers, Radio Frequency-Heat-Pump Dryers, Radio Frequency-Vacuum Dryers. centrifugal fan

heater

feed

product

steam

product

ROLLER DRYER

TRAY DRYER

vapour

vapour

feed

feed

heater product hot air FLUIDISED DRYER

Figure 1. continued on the next page.

air product PNEUMATIC DRYER

19

Transport Phenomena During Drying of Food Materials feed

feed

steam

atomiser

product hot air

condensate

ROTARY DRYER

SPRAY DRYER product

steam heated plates vapour

steam jet vacuum pump steam

food

steam jet vacuum pump

compression mechanism FREEZE DRYER

feed

infrared heater

condenser

water

INFRARED DRYER

heating media

aspirator

excess steam

infrared heater

feed product

product SUPERHEATED STEAM DRYER

vacuum pressure controller

cooler

INFRARED-VACUUM DRYER

Figure 1. Schematic view of various types of dryers used in food industry [2].

product

20

Kamil Kahveci and Ahmet Cihan

2. STRUCTURE OF FOOD MATERIALS Food materials are porous and hygroscopic materials and their structure has a strong effect on transport and transport parameters. Structure has a strong effect particularly on mass diffusivity, permeability and thermal conductivity. On the other hand, its effect on thermal diffusivity is relatively weak [6].

Porous Materials A porous medium is a multi-phase system consisting of a solid phase and one or more fluid phases that occupy the pore space. The pore space of the porous medium consists of pores (nodes) and throats (bond) as connections between pores. The pores and throats are distributed randomly inside the porous medium. They have irregular shapes and therefore the structure of porous medium is very complex (see figure 2). Porous materials are divided into two groups as porous and capillary porous materials. This distinction is based on pore size. Materials with pore diameter equal to or higher than 10-7m are called porous materials, and smaller than 10-7m are called capillary porous materials [8]. The majority of food materials are capillary-porous materials [9].

Figure 2. Examples of porous media (10 x magnified): a) beach sand, b) sandstone; c) limestone; d) rye bread; e) wood; f) human lung (adapted from[7]). with permission.

Transport Phenomena During Drying of Food Materials

21

Pores in porous materials may be classified in three groups: (a) interconnected pores, (b) isolated or closed pores and (c) dead-end or blind pores (see figure 3). The interconnected pores are usually accessible from many directions. Blind or dead end pores are accessible from one direction only. Isolated pores are inaccessible and behave as part of the solid. Isolated pores decrease the diffusivity characteristics of the porous medium. Blind Pore

Closed Pore Interconnected Pore

Figure 3. Different types of pores in a porous medium.

Hygroscopic and Non-Hygroscopic Materials In non-hygroscopic materials, pore spaces are filled with liquid if the material is fully saturated, and with air if it is completely dry (see figure 4). In non-hygroscopic materials bound water content is quite low and vapor pressure is the function of temperature only [8]. Non-hygroscopic materials do not shrink during drying process. However, hygroscopic materials contain large amounts of physically bound water and therefore these materials usually shrink during the drying process. Food materials are in the hygroscopic material class where the modeling of drying process is more complex due to shrinkage. Solid Phase

Bound Gas Hygroscopic

Figure 4. continued on next page.

22

Kamil Kahveci and Ahmet Cihan Liquid Water

Gas Phase

Solid Phase

Non-Hygroscopic Figure 4. Hygroscopic and non-hygroscopic materials (adapted from [10]).

3. TRANSPORT PARAMETERS Dimensionless Numbers Various dimensionless numbers are encountered when dealing with the transport phenomena during the drying process. Most of these numbers are given in table 1 together with their physical meanings. Table 1. Dimensionless numbers and heir physical meanings D.less Number

Equation

Physical Meaning

Biot

h L Bi m = m D

mass transfer across the boundary/mass transfer within the solid

Biot

Bi q =

heat transfer across the boundary/heat transfer within the solid

Fourier

Fo =

Grashof

Gr =

Lewis

Le =

α D

Nusselt

Nu =

hqL

Peclet

Pe =

Prandtl

Pr =

Reynolds

Re =

Schmidt

Sc =

Sherwood

Sh =

hqL λ

Dt L2

gL3β v ρ 2 ΔT ν

dimensionless time in the unsteady state regime buoyancy forces/viscous forces thermal diffusivity/mass diffusivity

λ

vL ν

convective heat transfer/conductive heat transfer convection/diffusion

ν α

momentum diffusivity/mass diffusivity

vL ν

inertial force/viscous force

ν D

h mL D

momentum diffusivity/mass diffusivity mass transfer/mass diffusivity

Transport Phenomena During Drying of Food Materials

23

Table 1. Continued D.less Number Stanton

Equation h St m = m v St q =

Stanton

Physical Meaning wall mass transfer/mass transfer by convection

h q / ρC p v

heat transferred to fluid/heat transported by fluid

where L is the characteristic dimension of a solid body (m).

Relative Humidity Relative humidity is the ratio of the mole fraction of water vapor in a given moist air sample to the mole fraction in a saturated air sample at the same temperature and pressure. By using the perfect gas law, it can be expressed as the ratio of the actual vapor pressure Pv (Pa) to the vapor pressure of the saturated air at the same temperature Pv,sat (Pa). ϕ = Pv / Pv,sat (T )

(1)

Humidity Ratio The humidity ratio of moist air is defined as the ratio of the mass of water vapor Mv (kg) to the mass of dry air contained in the moist air Ma (kg).

ω = Mv / Ma

(2)

Saturation Humidity Ratio The saturation humidity ratio ωsat is the humidity ratio of moist air saturated with respect to water at the same temperature and pressure.

Specific Humidity The specific humidity ωs is the ratio of the mass of water vapor to the total mass of air in a particular volume of air. ωsp =

Mv Mv + Ma

The specific humidity is related to the humidity ratio by the following way:

(3)

24

Kamil Kahveci and Ahmet Cihan ωsp =

ω 1+ ω

(4)

Dry-Bulb Temperature The dry bulb temperature Tdb (°C) is the temperature measured by a (dry) thermometer immersed in a vapor-gas mixture.

Wet-Bulb Temperature The psychrometric wet -bulb temperature Twb (°C) is measured by placing a thermometer having a water-moistened wick covered bulb into a fast moving stream of ambient air. If the air surrounding the wet-bulb thermometer is not saturated, evaporation of water from the wick will occur. This will cause the bulb to cool. The amount of cooling is proportional to the evaporation rate. Having waited long enough, a steady-state is reached. This final equilibrium temperature is called the wet-bulb temperature of the moist air. The psychrometric wet-bulb temperature is not the same as the adiabatic temperature which is the temperature reached by moist air and water if the air is adiabatically saturated by the evaporating water. However, the adiabatic and psychrometric wet-bulb temperatures are nearly equal for moist air.

Dew-Point Temperature The dew-point temperature Tdp (°C) is the temperature at which a given unsaturated airvapor mixture becomes saturated.

Vapor Pressure The vapor pressure Pv (Pa) is the partial pressure exerted by the water vapor molecules in moist air. When air is fully saturated with water vapor, its vapor pressure is called the saturated vapor pressure Pv,sat (Pa).

Gas Pressure The total gas pressure Pg (Pa) is the sum of partial pressures of air and vapor. Pg = Pv + Pa

(5)

If it is assumed that gas phase obeys ideal gas law: ˆ = ρ RˆT Pi M i i

(6)

25

Transport Phenomena During Drying of Food Materials

ˆ is the molecular mass (kg/mol) Rˆ is the universal gas constant (8.314 J/(mol K) where M i and T is the temperature (K). An average molar mass can also be written as: ˆ = (M ˆ +M ˆ −M ˆ )P / P M g a v a v g

(7)

ˆ and M ˆ are the molar mass of gas, air and vapor (kg/mol), respectively. ˆ , M where M a v g

Capillary Pressure Capillarity can be explained by considering the effects of two opposing forces: adhesion, the attractive force between the molecules of two dissimilar substances, and cohesion, the attractive force between the molecules of a single substance. The magnitude of attraction between gas molecules is smaller when compared to liquids due to greater distance. This results in higher attraction of liquid molecules in a liquid gas interface towards the interior liquid compared to the surrounding gas and a surface tension takes place at the interface. Surface tension values between water and air at different temperatures under atmospheric pressure are given in table 2 together with other physical properties of the water. Table 2. Surface tension between water and air and other properties of water (adapted from [11]) T(°C) 0 10 20 30 40 50 60 70 80 90 100

σx103(N/m) 75.64 74.23 72.75 71.20 69.60 67.94 66.24 64.47 62.67 60.82 58.91

ρ(g/cm3) 0.99984 0.99970 0.99821 0.99565 0.99222 0.98803 0.98320 0.97778 0.97182 0.96535 0.95840

Cp(J/gK) 4.2176 4.1921 4.1818 4.1784 4.1785 4.1806 4.1843 4.1895 4.1963 4.2050 4.2159

Pv(kPa) 0.6113 1.2281 2.3388 4.2455 7.3814 12.344 19.932 31.176 47.373 70.117 101.325

η(Pa s) 1793 1307 1002 797.7 653.2 547.0 466.5 404.0 354.4 2314.5 281.8

λx103(W/mK) 561.0 580.0 598.4 615.4 630.5 643.5 654.3 663.1 670.0 675.3 679.1

Surface tension will also occur in a similar way at the solid material/liquid interface, and at the solid material/gas interface. The angle between the edge of the meniscus and the solid material is called contact angle (see figure 5 and figure 6). The contact angle acquires a value in compliance with the balance of surface tensions at the interface between gas, liquid and solid. Thus, the contact angle is a function of the characteristics of the liquid, the gas and the solid material. The forces are balanced if σ sg = σ sl + σ lg cos θ

(8)

26

Kamil Kahveci and Ahmet Cihan

where σ is the surface tension (N/m) and the indices sg, sl and lg denote the solid-gas, solidliquid and liquid-gas interfaces, respectively. The contact angle θ (rad) can then be given as: cos θ =

σ sg − σ sl σ lg

(9)

If the contact angle is between 0dS and letting cP =

ˆ ψ SM g ρ TRˆ

(348)

s

we obtain: ρs

∂Pg ∂(u v + u a ) = ρs c p ∂t ∂t

(349)

where cp is the air capacity (kg m2/(kg N)). Substituting Eqs. (345) and (349) into Eq. (346) yields:

ρs c P

∂Pg ∂t

= δ P ∇ 2 Pg − ρs ε pc C m

∂W ∂t

(350)

With the substitution Eq.(342) into Eq. (348), we obtain:

ρs c P

∂Pg ∂t

= −ε pc ρs C m D∇ 2 W − ε pc ρs Dδ T ∇ 2 T + (1 − ε pc )δ P ∇ 2 Pg

(351)

Let us write the balance equations again: ρs C P

∂T = Δh vap ε pc ρ s C m D∇ 2 W + (λ + Δh vap ε pc δ T )∇ 2 T + Δh vap ε pc δ P ∇ 2 Pg ∂t

(352)

Transport Phenomena During Drying of Food Materials ρs C m ρs c P

∂W = ρs C m D∇ 2 W + ρs Dδ T ∇ 2 T + δ P ∇ 2 Pg ) ∂t

∂Pg ∂t

105 (353)

= −ε pc ρs C m D∇ 2 W − ε pc ρs Dδ T ∇ 2 T + (1 − ε pc )δ P ∇ 2 Pg

(354)

A general set of boundary conditions for this system can be defined as follows [16]: W = WS δm

δ δ ∂T ∂W + h m ( W − Wa ) = 0 + Jm + m T C m ∂n ∂n

T = TS λ

∂T + J q + h q (T − Ta ) + h m Δh vap (1 − ε pc )( W − Wa ) = 0 ∂n

Pg = PS

on

Γ1

(355)

on

Γ2

(356)

on

Γ3

(357)

on

Γ4

(358)

on

Γ5

(359)

where hm is the convective mass transfer coefficient (kg/(m2s)) and hq is the convective heat transfer coefficient (W/(m2K)). Subscripts a and S stand for ambient and surface respectively. The first term in Eq. (358) is the amount of heat passing into the body, the second term and the third term are the heat supplied at the surface, and the last term is the amount of heat expended in the phase change of the fluid. The first term in Eq. (356) is the moisture flux in the direction normal to the surface, while the last two terms describe the amount of moisture removed from the surface. Equations (356) and (358) can be written in a general form as [95]: k11

∂T + J *q* = 0 ∂n

k 22

∂W + J *m* = 0 ∂n

(360)

where J *q* = a q (T − Ta ) + a ε ( W − Wa ) + J *q J *m* = a δ (T − Ta ) + a m ( W − Wa ) + J *m

aε =

Δh vap h m λ

(1 − ε pc )K 11

aq =

K11h q λ

DJ q ⎤ ⎡J J *m = K 22 ⎢ m − ⎥ ⎣ h m Cmλ ⎦

δ T Δh vap K 22 δ T h q ⎡ 1 ⎤ a m = K 22 h m ⎢ − (1 − ε pc )⎥ a δ = − Cmλ Cm ⎣hm ⎦

J *q = (

K11 ) λ

(361)

(362)

(363)

106

Kamil Kahveci and Ahmet Cihan

The partial differential equations for temperature, pressure and moisture potential are not symmetric; however, by multiplying Eq. (352) by δ T / C m , Eq. (353) by Δh vap ε pc and Eq.

(354) by −Δh vap δ p / δ m .a symmetric set of equations can be obtained. c′q

∂T = k11∇ 2 T + k12 ∇ 2 W + k13∇ 2 Pg ∂t

(364)

c′m

∂m = k 21∇ 2 T + k 22 ∇ 2 W + k 23∇ 2 Pg ∂t

(365)

c′p

∂P = k 31∇ 2 T + k 32 ∇ 2 W + k 33∇ 2 Pg ∂t

(366)

where c′q = ρ s C P δ′T

k11 = (λ + Δh vapε pc δ T )δ′T k12 = k 21 = Δh vap ε pc δ m δ′T

(367)

k13 = k 31 = Δh vap ε pc δ P δ′T

(368)

c′m = ρ s Δh vap ε pc C m

c ′P =

−ρ s c P Δh vap δ P δm

k 22 = Δh vap ε pc δ m

k 33 =

− Δh vap (1 − ε pc )δ 2P δM

k 23 = k 32 = Δh vap ε pc δ P

(369)

where δ m = ρs C m D

δ′m = δ m / C m

(370)

Analytical solution of Luikov heat and mass transfer differential equation system are only obtained for simple geometrical shapes and boundary conditions. For a slab, a cylinder, and a sphere, Luikov and Mikhailov [96] used the Laplace transform technique to obtain their solutions. These same problems were also taken into account by Mikhailov and Özisik [97] using the finite integral transform technique. They obtained the same solutions as those of Luikov and Mikhailov [96]. However, it was seen subsequently that these solutions ignored the possible existence of complex eigenvalues. If complex eigenvalues exist, these solutions can be grossly in error [98]. More satisfactory solutions were also obtained for Luikov’s differential equation system using different approaches (see Ref. [98, 99]).

KRISCHER MODEL This model assumes that moisture transfer occurs with the combined effect of capillary flow of liquid and diffusion of vapor. The main difference between Luikov and this model is that, in the Luikov model, the total moisture content is assumed as the driving force for both

Transport Phenomena During Drying of Food Materials

107

water and water vapor transport, while this model considers liquid water and water vapor transport separately. In this model, the driving force for water vapor transport is assumed to be the gradient of its partial pressure in the air and the driving force for liquid water transport is taken as the gradient of the liquid moisture content. The flux and balance equations for this model are as follows.

Liquid Phase r r J ml = −ρ l K l ∇u

(371)

r where J ml is the liquid flux (kg/m2s), ρl is the density of liquid water (kg/m3), Kl is the liquid

diffusivity (m2/s) and u is the liquid content by volume (m3 liquid/m3 solid).

Vapor Phase ˆ r r D M J mv = − v,free v ∇Pv ζ RˆT

(372)

r where J mv is the flux of water vapor (kg/m2s), Pv is the partial pressure of water vapor in the

air (Pa), D v,free is the gas diffusion coefficient in open space (m2/s), ζ (>1) is the diffusion

resistance factor, which describes the decrease of the vapor flow in the porous medium in comparison with that in stagnant gas. Rˆ is the gas constant (8.314 J/(mol K)) and T is the temperature (K).

Balance Equations Let us look at, before expressing balance equations, what the approach of Krischer was while constructing flux equations. ρ l u term in liquid mass flux equation can be written as follows [58]. ρl u =

m l Vl m l = = Cl Vl V V

(373)

If this equation is substituted into the flux equation, we obtain: r r J ml = −K l∇C l

(374)

As can be seen, the equation used by Krischer for liquid transport is analogous to Fick’s law.

108

Kamil Kahveci and Ahmet Cihan If the state equation is used, vapor transport equation can be expressed as follows: r D 1 r RˆT ∇( J mv = − Cv ) ˆ ζ R vT M v

(375)

Let us assume that the temperature is constant too. In this case, it is possible to write: r r D J mv = − v,free ∇C v ζ

(376)

In addition, if D v = D v,free / ζ is considered as the diffusion coefficient for porous material, again a relation analogous to Fick’s law is obtained. Assuming that transport parameters are constant, balance equations may be expressed as follows. ∂u = K l∇ 2 u ∂t

(for the liquid water)

(377)

∂ρ v D v,free 2 = ∇ Cv ∂t ζ

(for the water vapor)

(378)

In order to get the same physical dimensions in both transport equations, the transport equation for liquid water can be multiplied by ρw and the equation for water vapor transport ˆ /(RˆT) . Therefore: by M v

ρl

∂u = ρl K l∇ 2 u ∂t

ˆ D ˆ ∂P M M v v = v v ,free ∇ 2 Pv RˆT ζ RˆT ∂t

(379)

(380)

Krischer introduced two additional corrections in the mass balance equation for water vapor. The first is the reduction of the accumulation term on the left-hand side by the multiplication factor (ψ-u), which expresses the fact that water vapor can appear only in that part of the porous space where water in the liquid state is not present, and its accumulation is therefore limited to only that space [58]. There is no need for reduction in the term at the right side of the equation. Because, the reduction for this term has already been made through the diffusion resistance factor ζ. The second correction contains the inclusion of Stefan diffusion [58]. Mass flux for Stefan diffusion is defined as follows (see Eq. (178)):

Transport Phenomena During Drying of Food Materials r Pg r D 1 J v = − v,free ∇Pv ζ R v T Pg − Pv

109

(381)

where Stefan diffusion is defined by the multiplication factor Pg/(Pg-Pv). Therefore, mass balance equation for vapor phase becomes: (ψ − u )

ˆ D ˆ ∂P Pg M M v v = v v,free ∇ 2 Pv ˆ ˆ RT ζ Pg − Pv RT ∂t

(382)

The total mass balance of moisture can then be written as follows: ρl

ˆ D ˆ ∂P Pg M M ∂u ∇ 2 Pv + (ψ − u ) v v = ρ l K l ∇ 2 Pv + v v ,free ˆ ˆ ∂t RT ζ Pg − Pv RT ∂t

(383)

BERGER AND PEI MODEL In this model, the total internal moisture transfer is assumed to consist of two different mechanisms. These are capillary flow of liquid due to gradient in liquid content, and the diffusion of vapor through empty pores due to a gradient in partial vapor pressure. The internal heat transfer is assumed to be governed by heat conduction and enthalpy of vaporization. In addition, it is assumed that external heat and mass transfer is proportional to the temperature and the partial vapor pressure difference between the surface of the drying solid and the external drying media. The main difference of this model from the Krischer model is that this model takes sorption isotherms into consideration, as opposed to Krischer model, by using two coupling equations between the three dependent variables u, ρv and T. These two equations are: the Clasius-Clapeyron’s equation and the equation of the sorptional isotherm of the system. Flux and balance equations for this model where Fick’s law is used to express both internal mass transfer mechanisms are as follows:

Liquid Phase r r J ml = −K l ρ l ∇u

(384)

where Kl is the liquid conductivity in the solid (m2/s), ρl is the density of the liquid (m3/kg) and u is the liquid content by volume (m3 liquid/m3 solid).

110

Kamil Kahveci and Ahmet Cihan

Vapor Phase ˆ r r M J mv = −D v (ε void − u ) v ∇Pv RˆT

(385)

where Dv is the vapor diffusivity (m2/s) and εvoid is the void fraction of solid (m3 air/m3 solid), Rˆ is the gas constant (8.314 J/(mol K)), T is the temperature (K), Pv is the vapor pressure (Pa). It is assumed that the temperature changes are small. Therefore: ˆ ˆ r r⎛ P M M v ∇Pv = ∇⎜ v v ⎜ ˆ RˆT ⎝ RT

⎞ r ⎟ = ∇C v ⎟ ⎠

(386)

r r J mv = −D v (ε void − u )∇C v

(387)

Balance Equations If the Kl and Dv are assumed to be constant, the mass balance equations for the liquid and vapor transfer can be written as follows: ρl

∂u = K l ρ l ∇ 2 u − I ev ∂t

[

(388)

]

r r ∂ [(ε void − u )C v ] = D v∇ ⋅ (ε void − u )∇C v + I vap ∂t

(389)

where Ivap is the rate of evaporation (kg/(m3s). Using the two mass balance equations above, the total moisture transfer equation can be expressed as:

(ρ l − C v ) ∂u + (ε void − u ) ∂t

[

D v (ε void

∂C v = K lρl∇ 2 u + ∂t r r − u )∇ 2 C v − (∇u ) ⋅ (∇C v )

]

(390)

The rate of enthalpy of vaporization can be defined as: q vap = Δh vap I vap , where Δh vap is the evaporation enthalpy (J/kg). In this case, the energy balance equation can be written as follows:

r r Δh vap ⎧⎪ ⎡ ⎤ ∂ 2ρ v ∂T − ∇ ⋅ ∇ = α∇ 2 T + ( u ) ( C v )⎥ ⎨D v ⎢(ε void − u ) 2 ∂t ρ s C p ⎪⎩ ⎢⎣ ∂x ⎥⎦ ∂C ∂u ⎫ − (ε void − u ) v + C v ⎬ ∂t ⎭ ∂t

(391)

Transport Phenomena During Drying of Food Materials

111

where α is the thermal diffusivity (m2/s). In the original model, proposed by Berger and Pei [100] it is considered a one dimensional system seen in figure 23 and the following boundary conditions at x=0: K lρl

∂C ∂u + D v (ε void − u ) v = h m (C v − C va ) ∂x ∂x

h q (Ta − T) = Δh vap K l ρ l

∂u ∂T −λ ∂x ∂x

(392)

(393)

where hm is the mass transfer coefficient (m/s), hq is the heat transfer coefficient (W/(m2K)) ∂u denotes the amount of and λ is the thermal conductivity (W/(m K)) the term Δh vap K l ρ l ∂x x =0 heat required to evaporate the liquid flux at the surface. It is assumed that no mass and heat is transferred across the surface of the drying material at x=L. Therefore: K lρl

∂C ∂u = D v (ε void − u ) v = 0 ∂x ∂x

∂T =0 ∂x

(394)

(395)

In the original model, moisture content with respect to volume has been used when expressing the liquid flux. Today, it is preferred to use moisture content with respect to mass. Furthermore, in the original model the temperature changes were assumed to be small and therefore vapor concentration gradient has been used as driving force instead of partial vapor pressure gradient. The flux and balance equations may be expressed as follows in terms of the moisture content with respect to mass and vapor pressure: r r J ml = −K l ρs ∇m

ˆ ⎛ εg − εl ⎞ r r r r M ⎟∇Pv J mv = −D v (ψ − u )∇ρ v J mv = −D v v ⎜⎜ RˆT ⎝ ε s ⎟⎠ K lρs

Kv

∂m ∂ 2m − I ev = ρ s 2 ∂t ∂x

ˆ ∂ ⎡⎛ ε − ε l ⎞ ⎤ ˆ r ⎡⎛ ε g − ε l ⎞ r ⎤ M M v ⎟⎟∇Pv ⎥ + I ev = v ⎢⎜⎜ g ⎟Pv ⎥ ∇ ⋅ ⎢⎜⎜ RˆT RˆT ∂t ⎣⎢⎝ ε s ⎟⎠ ⎦⎥ ⎣⎢⎝ ε s ⎠ ⎦⎥

(396)

(397)

(398)

(399)

112

Kamil Kahveci and Ahmet Cihan K l ρs ∇ 2 m + D v

( )( )

ˆ ⎡⎛ ε g − ε l ⎞ 2 r ⎤ ρ r M v ⎜⎜ ⎟⎟∇ Pv − s ∇m ⋅ ∇Pv ⎥ = ⎢ ρl RˆT ⎣⎢⎝ ε s ⎠ ⎦⎥ ˆ ⎛ ε g − ε l ⎞ ∂P ˆ P ⎞ ∂m M ⎛ 1 M v v ⎟ ⎟⎟ v + v ⎜⎜ ρs ⎜1 − ⎜ ρ RˆT ⎟ ∂t ˆ ε R T l s ⎝ ⎠ ∂t ⎠ ⎝

ˆ ⎧⎪ ⎡⎛ ε g − ε l λ ∂ 2 T Δh vap M v ∇ 2T 2 + ⎨D v ⎢⎜ ρs C p ρ s C p RˆT ⎪⎩ ⎣⎢⎜⎝ ε s ∂x ⎛ εg − εl ⎜⎜ ⎝ εs

(400)

( )( )

r ⎤ ⎞ 2 ρ r ⎟⎟∇ Pv − s ∇m ⋅ ∇Pv ⎥ − ρl ⎥⎦ ⎠

(401)

⎞ ∂Pv ρ s ∂m ⎫⎪ ∂T ⎟⎟ + Pv ⎬= ∂t ⎪⎭ ∂t ⎠ ∂t ρ l

where m is the moisture content by mass (kg moisture/kg dry solid) and εs, εl and εg are the volume fraction of solid, water and gas, respectively. External Drying Media Jm

Jq,ext

x=0 Jmv

Jml -Jq

q& vap

x

Ivap

Drying Material

x=L Figure 23. Schematic view for Berger and Pei model (adapted from [100]).

WHITAKER MODEL This model was proposed by Whitaker [71, 72] to describe the simultaneous heat, mass and momentum transfer in a porous media. The biggest advantage of this model is that the physics of the model is better understood, the assumptions are very clear and the parameters are well defined. In addition, all mechanisms for mass and heat transfer were taken into consideration: liquid flow due to capillary forces, vapor and gas flow due to convection and diffusion, internal evaporation of moisture and heat transfer by convection, diffusion and conduction.

Transport Phenomena During Drying of Food Materials

113

Microscopic Equations In this model a representative volume is considered as shown in figure 24 to express mass and heat transfer. The figure at the left represents the microscopic scale and the figure at the right the macroscopic scale. Porous structure is assumed to consist of three phases, namely, solid, liquid and gas. The solid phase will be represented here by s, the liquid phase by l and gas by g. The gas phase is assumed to consist of air (denoted by a) and vapor (denoted by v).

Figure 24. Pore and macroscale in a porous body (adapted from [93]).

At microscopic level, mass and heat transport equations can be derived using the conservation laws as follow:

Mass Conservation If it is assumed that no chemical reaction exists during the transport process, mass conservation equation for each phase can be written as follows: r ∂ρ r + ∇ ⋅ (ρv) = 0 ∂t

(402)

where ρ is the concentration of phase considered (kg/m3). The mass conservation equation for each species in the phase may be expressed as follows. r ∂ρ i r + ∇ ⋅ (ρ i v i ) = 0 for i=1,2,..n ∂t

(403)

where ρi is the concentration of species i (kg/m3) and vi is the velocity of species i (m/s). The following relations can be written for the velocities and concentrations of species:

114

Kamil Kahveci and Ahmet Cihan r ρ r v = ∑ i vi , i ρ

ρ = ∑ ρi

(404)

i

where v is the mass average velocity (m/s) and ρ is the total concentration (kg/m3).

Linear Momentum Principle For each phase, momentum conservation can be written as: r Dv r ~ ρ =∇⋅σ Dt

(405)

~ is the stress tensor (Pa). In the above equation, body forces such as gravitational where σ force are neglected. The angular momentum principle requires this tensor to be symmetric. ~=σ ~T σ

(406)

Energy Conservation Energy conservation for each phase can be written as follows: r r r ∂ (ρh ) r DP ~ r r + ∇ ⋅ (ρhv) = −∇ ⋅ J q + + τ : ∇v Dt ∂t

(407)

where h is the enthalpy per unit mass (J/kg), J q is conductive heat flux vector (W/m2), ~τ is rr the viscous stress tensor (Pa). The term ~τ : ∇v is the viscous dissipation, P is the pressure (Pa) and DP / Dt is the compression work. In the above equation, the source or sink of electromagnetic radiation was neglected. The viscous dissipation and the compression work for the liquid and gas phase can also be neglected. Therefore, the energy equation becomes as follows:

r

r r r ∂ (ρh ) r + ∇ ⋅ (ρhv) = −∇ ⋅ J q ∂t

(408)

We can also assume that the enthalpy is independent of pressure. Therefore: h = C P (T − TR )

(409)

where CP is the specific heat capacity (J/(kgK)) and TR is the reference temperature (°C). The conservation laws can then be written for each phase as follows:

115

Transport Phenomena During Drying of Food Materials

Solid Phase r The solid phase is considered to be rigid and fixed in space ( v s = 0 ). In this case, we

must deal only with the conservation of energy. ρs

r r ∂h s = −∇ ⋅ J qs ∂t

(410)

If the assumption defined in Eq. (409) is taken account and the conductive heat flux is expressed as J q = −λ s ∇Ts according to the Fourier law, the energy equation can be written as:

r

ρs C Ps

r

∂Ts = λ s ∇ 2 Ts ∂t

(411)

where λs is the thermal conductivity of the solid phase (W/(mK)).

Liquid Phase The mass and energy conservation equation for the liquid phase can be expressed asllows: r ∂ρ l r + ∇ ⋅ (ρ l v l ) = 0 ∂t

ρ l C Pl

(412)

r r ∂Tw + ρ l C Pl v ⋅ ∇Tl = λ l ∇ 2 Tl ∂t

(413)

Gas Phase The gas phase is more complicated than the other phases since it contains two components: air and vapor. The total mass conservation can be expressed as follows: ∂ρ g ∂t

r r + ∇ ⋅ (ρ g v g ) = 0

v

(414)

v

The species velocity v gi can be written in terms of the mass average velocity v g and the

r

diffusion velocity u i : r r r v gi = v g + u i

(i=a,v)

Therefore, the mass conservation of air and vapor can be given as follows:

(415)

116

Kamil Kahveci and Ahmet Cihan ∂ρ gi ∂t

r r r r + ∇ ⋅ (ρg i v g ) = −∇ ⋅ (ρ gi u i ) i =a,v

(416)

Moreover, by expressing the diffusive flux as: r r ρ gi u i = −ρ g D va ∇(ρ gi / ρ g )

(417)

where D va is the binary molecular diffusion coefficient for vapor and air (m2/s), we obtain:

∂ρ gi ∂t

[

r r r r + ∇ ⋅ (ρg i v g ) = ∇ ⋅ − ρ gi D va ∇(ρgi / ρ g

]

i=a,v

(418)

For a multicomponent phase, the energy conservation equation can be written as:

[

]

r r r r ∂ ∑ (ρ gi h gi ) + ∇ ⋅ ∑ ρ gi h gi v gi = −∇ ⋅ J q ∂t i i

(419)

The mass average enthalpy can be defined in a similar manner to the mass average velocity as: h g = ∑ (ρ gi h gi / ρ g )

(420)

i

Therefore, Eq. (419) becomes as follows: ρ g C Pg

∂Tg ∂t

r r r r r + ρ g C Pg v g ⋅ ∇Tg = λ g ∇ 2 Tg − ∇ ⋅ (ρ a h a u a + ρ v h v u v )

(421)

where C Pg = (ρ a C Pa + ρ v C Pv ) / ρ g

(422)

For the species of gas phase, ideal gas law can be assumed.

ˆ Pi = ρi RˆT / M i

i=a,v

(423)

ˆ is the molar mass of species i where Rˆ is the ideal gas constant (8.314 J/(mol K)) and M i (kg/mol). The total gas pressure can be written as: Pg = Pa + Pv

(424)

Transport Phenomena During Drying of Food Materials

117

Boundary Conditions A general set of the boundary conditions must be defined to complete the set of equations. Let us assume that Γlg represents the interface between liquid and gas phases, Γsl the interface between solid and liquid and Γsg the interface between solid and gas. The following relations are valid for these interfaces: Γlg=Γgl

Γsl=Γls

Γsg=Γgs

(425)

The boundary conditions for the solid-liquid interface Γsw can be written as:

r vl = 0

(426a)

r r r r J qs ⋅ n ls = J ql ⋅ n ls

(426b)

Ts = Tl

(426c)

r r r where n ls ( n ls = −n sl ) is the unit normal vector directed from the liquid phase toward to the

solid phase. The boundary conditions for the solid-gas interface Γsg can be given as follows: r vg = 0

(427a)

r r r r J qs ⋅ n sg = J qg ⋅ n sg

(427b)

Ts = Tg

(427c)

The boundary conditions for the liquid-gas interface Γwg can be written as follows: r r r r r r ρ v ( v v − w ) ⋅ n gw = ρ w ( v l − w ) ⋅ n gw

(428a)

r r r ρ a ( v a − w ) ⋅ n gw = 0

(428b)

r r r r r r ρ g ( v g − w ) ⋅ n gw = ρ w ( v l − w ) ⋅ n gw

(428c)

r r r r r r ρ l (h v − h l )(v l − w ) ⋅ n gl = ( J ql − J qg ) ⋅ n gl

(428d)

Tl = Tg

(428e)

r where w is the velocity of water-gas interface (m/s).

118

Kamil Kahveci and Ahmet Cihan

Volume Averaging Method The equations defined above cannot be solved on a microscopic level since the geometry of the porous medium and the distribution of the phases are not observable and are too complex to describe. Whitaker [71, 72, 101] introduced the concept of the averaging elementary volume to relate the microscopic geometrical and physical properties of the real porous medium to the macroscopic properties of the continuum model. With this method, governing equations are spatially smoothed leading to continuum equations defining the transport process. Although the equations become solvable in this case, the method has its own drawbacks, especially difficulties in determination of effective parameters that appear in the macroscopic equations. In volume averaging method, an averaging volume is associated to each point in porous material. This volume can be in any shape. In figure 24 averaging volume is represented by a circle. This representative volume provides upscaling of transport equations from pore scale to macroscopic scale. There are three types of average used in the study of drying. These are: spatial average, phase average and intrinsic average and they are defined as: 〈 y〉 =

1 ∫ ydV VV

〈 yi 〉 =

1 ∫ y i dV V Vi

〈 yi 〉 i =

1 ∫ y i dV Vi Vi

spatial average

(429)

phase average

(430)

intrinsic average

(431)

When it is interested in the average of some quantity related with a single phase, the phase average is employed. However, it should be noted that if yi is constant, the phase average is not equal to this value. In this situation, the use of intrinsic average is more appropriate. The phase average and intrinsic average are related as. 〈 yi 〉 = εi 〈 yi 〉 i

(432)

where ε i is the volume fraction of the phase i ( ε i = Vi / V ). Common expressions used in averaging are given below: r r 1 r 1 r n ws y w dΓ + 〈 ∇y w 〉 = ∇ 〈 y w 〉 + ∫ ∫ n wg y w dΓ V Γws V Γwg

(433)

r r r r 1 r r 1 r r vw ⋅ n ws dΓ + 〈 ∇ ⋅ v w 〉 = ∇〈 v w 〉 + ∫ ∫ v w ⋅ n wg dΓ V Γws V Γwg

(434)

Transport Phenomena During Drying of Food Materials 〈

r r r r ∂〈 y w 〉 1 ∂y w 1 y w w ⋅ n wg dΓ − y w w ⋅ n ws dΓ − 〉= ∫ ∫ V Γwg V Γws ∂t ∂t

119 (435)

Macroscopic Equations The following macroscopic transport equations can be obtained by averaging the pore scale transport equations over the averaging volume [71, 93]:

Mass Conservation liquid phase ρl

r r ∂ε l + ρl∇ ⋅ 〈 v l 〉 + 〈 J m 〉 = 0 ∂t

vapor in the gas phase

(

)

(

(436)

[

)

r r r r ~ ∂ ε g 〈ρ v 〉 g + ∇ ⋅ 〈ρ v 〉 g 〈 v g 〉 − 〈 J m 〉 = ∇ ⋅ 〈ρ g 〉 g D eff ⋅ ∇(〈ρ v 〉 g / 〈ρ g 〉 g ∂t

air in the gas phase

(

)

(

)

[

r r r r ∂ ~ ε g 〈ρ a 〉 g + ∇ ⋅ 〈ρ a 〉 g 〈 v g 〉 = ∇ ⋅ 〈ρ g 〉 g D eff ⋅ ∇(〈ρ a 〉 g / 〈ρ g 〉 g ∂t

Energy Conservation 〈ρ〉 C P

[

]

[

]

r r r ~ r ∂〈T〉 + ρ l C P l 〈 v l 〉 + 〈ρ g 〉 g 〈C P 〉 g 〈 v g 〉 ⋅ ∇〈T〉 + Δh vap 〈 J m 〉 = ∇ ⋅ λ eff ∇(T〉 ∂t

]

(437)

(438)

]

(439)

where 〈ρ〉 C P = ε s ρ s C Ps + εl ρl C Pl + ε g 〈ρ v 〉 g C Pv + ε g 〈ρ a 〉 g C Pa

(440)

〈ρ〉 = ε s ρs + εl ρl + ε g 〈ρ v 〉 g + ε g 〈ρ a 〉 g

(441)

with

and Δhvap is the evaporation enthalpy (J/kg). In Eq. (439), local thermal equilibrium is assumed:

120

Kamil Kahveci and Ahmet Cihan 〈T〉 = 〈Ts 〉 s = 〈Tl 〉 l = 〈Tg 〉 g

(442)

~ ~ The effective diffusivity D eff and effective thermal conductivity λ eff in Eqs. (437)-(438) are obtained from the process of upscaling from pore scale transport equations to macroscopic equations [71]. The capillary pressure is defined as the difference between gas pressure and liquid pressure: 〈 Pc 〉 = 〈 Pg 〉 g − 〈 Pl 〉 l . Vapor pressure is determined from sorption isotherm. With the analysis of the conservation of linear momentum, the following equations are obtained for the liquid and gas phase: ~ ~ K i ⋅ k rg r r .∇〈 Pg 〉 g 〈vg 〉 = − ηg

~ ~ r K ⋅k r 〈 v l 〉 = − i rl .∇〈 Pl 〉 l , ηl

(443)

~ ~ ~ where K i is the intrinsic permeability tensor, k rl and k rg are the relative permeability

tensors. To apply control volume method, Eqs. (436)-(439) must be reformulated as follows: ∂y r r −∇⋅J ∂t

(444)

r where y is a scalar quantity, and J is mass or energy flux vector. To reformulate the system, a set of main variables must be selected to define the whole drying process. One possible choice can be average temperature 〈T〉 , volume fraction of the liquid water ε l and intrinsic

phase average of air density in the gas phase 〈ρ a 〉 g . In this case, the system can be written as

follows if additionally the average notation 〈 〉 is dropped for simplicity [93]:

(

)

(

[

)

r r r r r r ∂ ρ l ε l + ε g ρ v + ∇ ⋅ ρl v l + ρ v v g = ∇ ⋅ ρ g D eff ⋅ ∇(ρ v / ρ g ∂t

(

)

(

)

[

r r r r r ∂ ε g ρ a + ∇ ⋅ ρa v g = ∇ ⋅ ρ g D eff ⋅ ∇(ρ a / ρ g ∂t (ε s ρ s

]

]

∂h s ∂h ∂h ∂h + ε lρl l + ε g ρ v v + ε g ρa a ) + ∂t ∂t ∂t ∂t r r v r r r ρ l C pl v l + (ρ v C Pv + ρ a C Pa ) v g ⋅ ∇T + Δh vap J m = ∇ ⋅ (λ eff ∇T)

[

]

(445)

(446)

(447)

In addition, the following relations can be written: ∂ ∂ ∂ (ε i ρ i h i ) = ε i ρ i ( h i ) + h i (ε i ρ i ) ∂t ∂t ∂t

(448)

Transport Phenomena During Drying of Food Materials ∂ (ε s ρ s ) = 0 ∂t

hs

121 (449)

∂h s ∂h v ∂h a ∂h l )= + εlρl + εgρv + εgρa ∂t ∂t ∂t ∂t

(ε s ρ s

∂ (ε s ρ s h s + ε l ρ l h l + ε g ρ v h v + ε g ρ a h a ) − ∂t ∂ ∂ ⎤ ⎡ ∂ ⎢h l ∂t (ε l ρ l ) + h v ∂t (ε g ρ v ) + h a ∂t (ε g ρ a )⎥ ⎦ ⎣

(450)

r r r r r r ∇ ⋅ (ρhv) = ρv ⋅ ∇h + h∇ ⋅ (ρv)

(451)

In this case, the second term on the left hand side of Eq. (447) becomes as follows:

[ρ C

]

[

]

r r r r r r v l + (ρ v C Pv + ρ a C Pa ) v g ⋅ ∇T = ∇ ⋅ ρ l h l v l + (ρ v h v + ρ a h a ) v g − r r r r r r h l∇ ⋅ (ρ l v l ) + h v ∇ ⋅ (ρ v v g ) + h a ∇ ⋅ (ρ a v g )

[

l

Pl

]

(452)

The enthalpy of vaporization can be expressed as: Δh vap = h v − h l at temperature T. Therefore: Δh vap J m ,vap = h v J m,vap − h l J m,vap

(453)

The evaporation rate Jm,vap can be computed in two different ways: one from the conservation equation for liquid water, the other from the conservation equation for water vapor. J m,vap = −ρ l J m,vap =

r r ∂ (ε l ) − ρ l ∇ ⋅ ( v l ) ∂t

(454)

[

r r r r r ∂ (ε g ρ v ) + ∇ ⋅ (ρ v v g ) − ∇ ⋅ ρ g D eff ⋅ ∇(ρ v / ρ g ) ∂t

]

(455)

With the substitution of the equations given above, the energy equation becomes as follows:

(

)

r r r ∂ ∑ ε i ρi h i + ∇ ⋅ ρl h l v l + ( ρv h v + ρa h a ) v g = ∂t i=s,w ,v,a r r r r r r r r r h a ∇ ⋅ ρ g D eff ⋅ ∇(ρ a / ρ g + h v ∇ ⋅ ρ g D eff ⋅ ∇(ρ v / ρ g + ∇ ⋅ λ eff ∇T

[

]

[

]

(

)

(456)

The energy equation can be further simplified by assuming that, within the averaging volume, the variation of enthalpy is small compared with its value.

122

(

Kamil Kahveci and Ahmet Cihan

)

r r r ∂ ∑ ε i ρi h i + ∇ ⋅ ρl h l v l + ( ρv h v + ρa h a ) v g = ∂t i=s,w ,v,a r r r r r r r r r ∇ ⋅ ρ g h a D eff ⋅ ∇(ρ a / ρ g + ∇ ⋅ ρ g h v D eff ⋅ ∇(ρ v / ρ g + ∇ ⋅ λ eff ∇T

[

]

[

]

(

)

(457)

CAPILLARY FLOW MODEL If the capillarity is the primary mode for transport, two different formulations may be expressed under isothermal conditions. One of these is capillary pressure formulation and the other capillary diffusivity formulation [8].

Capillary Pressure Formulation ⎞ ∂C l r ⎛ Kk r + ∇ ⋅ ⎜⎜ − ρ l i rl ∇(P − Pc ) ⎟⎟ = −I vap ∂t ηl ⎝ ⎠

(458)

where C is the concentration (kg/m3), ρ is the density (kg/m3), Ki is the intrinsic permeability (m2), kr is the relative permeability, η is the viscosity (Pa s), Pc is the capillary pressure (Pa) and Ivap is the rate of evaporation (kg/(m3s)). If Pc >>P (capillary pressure is large) and Ivap=0 (no significant evaporation): ∂C l r ⎛ K i k rl r ⎞ + ∇ ⋅ ⎜⎜ ρ l ∇Pc ⎟⎟ = 0 ∂t ηl ⎝ ⎠

(459)

In this form of transport equation both water concentration and capillary pressure are involved. This equation is also known as Richard’s equation. Generally, capillary head h (m) is used instead of capillary pressure [8]. In this case, Eq. (459) can be written as: ∂C l r ⎛ K i k rl ρ l g r ⎞ + ∇ ⋅ ⎜⎜ ρ l ∇h ⎟⎟ = 0 ∂t ηl ⎠ ⎝

(460)

⎛ ∂C l ⎞ ∂h r ⎛ K i k rl ρ l g r ⎞ + ∇ ⋅ ⎜⎜ ρ l ∇h ⎟⎟ = 0 ⎜ ⎟ ηl ⎝ ∂h ⎠ ∂t ⎠ ⎝

(461)

or

where g is the gravitational acceleration (m/s2), ∂C l / ∂h is the specific moisture capacity.

Transport Phenomena During Drying of Food Materials

123

Capillary Diffusive Formulation Another formulation equivalent to capillary pressure formulation for capillary flow may be expressed as follows [8]: ∂C l = D l∇ 2 Cl ∂t

(462)

This equation is similar to the commonly used diffusion equation. In Eq. (462), Dl is the capillary diffusivity of liquid water and is defined as: D l = −ρ l2 g

K i k rl ∂h ηl ∂C l

(463)

Datta [8] reported that it will be better to use capillary pressure formulation in multi-domain foods, since the capillary pressure variation with moisture content will be different in different domains of material. When a formulation is used in which capillary pressure (head) is the driving force, the most important problem arises from the determination of the relationship between capillary head and moisture content. Such a relation can be obtained only using other material properties data. One of the relations to be used for this purpose is Kelvin’s equation [102].

EMPRICAL AND SEMI-EMPRICAL MODELS Due to the complexity of transport mechanisms, empirical and semi-empirical models are often used to describe the thin-layer drying behavior of food materials. Of these models, those used frequently are given in table 14. The empirical models constitute a direct relationship between the average moisture content and the drying time. They neglect the fundamentals of the drying process and therefore their parameters have no physical meaning. The semitheoretical models are generally derived by simplifying general series solutions of Fick's second law or they are modified forms of simplified models. The empirical and semiempirical models require small time compared to theoretical models and do not need assumptions of geometry of a typical food, its mass diffusivity and conductivity, et cetera. [103]. Therefore they are useful for automatic control processes. Empirical and semiempirical models are valid within the temperature, relative humidity, air flow velocity and moisture content range for which they were developed [104]. Among these drying models, the Page models, the Henderson and Pabis model, the two-term exponential model and the Midilli et al. model are widely used models to simulate the thin layer drying behavior of food materials.

124

Kamil Kahveci and Ahmet Cihan

Lewis Model In this model, mass transfer is expressed by analogy to heat flow from a body immersed in a cool fluid, in other words, to Newton’s law of cooling [105]. The drying rate is assumed to be proportional to the difference in moisture content between the material being dried and equilibrium moisture content at the drying air condition. The most important drawback of this model is that it generally overpredicts the early stages and underpredicts late stages of the drying process.

Page and Modified Page Models Page model has been proposed by Page [106] to eliminate the drawback of the Lewis model. For this purpose, Page [106] introduced an exponent to time term in Levis model. However, the introduction of the exponent causes the model to become a purely empirical model. This parameter has an effect of moderating the time and the model in this case gives better results for the prediction of moisture loss.

Henderson and Pabis Model Various approximations and variations of the diffusion model have been used to simulate the drying behavior of food materials. The model proposed by Henderson and Pabis [107] is of this type and corresponds to the first term of a general series solution of Fick’s second law. The slope of this model, coefficient k, is related to effective diffusivity when drying process takes place only in the falling rate period and liquid diffusion controls the process [108].

Two Term Exponential, Diffusion Approach, Verma et al., Logarithmic, Midilli et al., Jena and Das Models These models are modified forms of the Henderson and Pabis model, based on simplification of a general series solution of Fick's second law. Among these models, two term exponential and Midili et al. model have been used frequently because they produce better fit to experimental data as compared to other models.

Two Term, Modified Henderson and Pabis Models It is not possible to express drying behavior of food materials having several compartments with the first term of a general series solution of Fick’s second law with the sufficient accuracy due to different drying resistance in each compartment. In such cases, exponential terms equal to the number of compartments must be used to define drying in each compartment. Two term and Modified Henderson and Pabis models have been proposed to simulate the drying process of such type of materials. Sharma et al. [109] used these two- and

Transport Phenomena During Drying of Food Materials

125

three-term models to simulate the thin-layer drying behavior of rough rice with three compartments, the hull, bran and endosperm.

Geometric, Thompson, Wang and Singh, Logistic, Weibull Distribution Models All of these models are empirical models and have been proposed for simulating the drying behavior of various types of food materials. For example, Thompson model has been suggested to simulate the drying behavior of shelled corn in the temperature range 60–150°C [110]. Wang and Singh model has been proposed to simulate the drying of medium-grain rough rice [111]. The coefficient of correlation (r) can be used to determine the suitability of the empirical and semitheoretical models in describing the experimental drying data. Correlation coefficient values close to one mean the fit is good. In addition to correlation coefficient; standard deviation (es) and mean squared deviation (χ2) can be used to determine suitability of the fit. These parameters are defined as follows [112]: r=

es =

χ2 =

n obs ∑ mrpre ,i mrexp,i − ∑ mrpre ,i ∑ mrexp,i i =1

⎞ ⎛ n obs ∑ (mrpre ,i ) 2 − ⎜ ∑ mrpre ,i ⎟ i =1 ⎠ ⎝ i=1 n obs

n obs

n obs

n obs

n obs

∑ (mrpre,i − mrexp,i ) 2

i =1

2

i =1

⎞ ⎛ n obs n obs ∑ ( mrexp,i ) 2 − ⎜ ∑ mrexp,i ⎟ i =1 ⎠ ⎝ i=1 n obs

2

(464)

n obs i =1

n obs

∑ (mrpre,i − mrexp,i ) 2

(465)

n

i =1

n obs − n const

(466)

where mrpre,i is the ith predicted moisture ratio, mrexp,i is the ith experimental moisture ratio, nobs is the number of observations and nconst is the number of constants in the drying model.

126

Kamil Kahveci and Ahmet Cihan Table 14. Thin-layer drying models

Name Lewis Page Modified Page Modified Page II Henderson&Pabis Geometric Wang&Singh Two term exponential Thompson Logarithmic Logistic Diffusion approach Verma et al. Two term Midilli et al. Jena and Das Weibull distribution Modified Henderson and Pabis

Model equation mr = exp(−kt )

N.C. 1

mr = exp(−kt ) n

2

mr = exp(−(kt ) ) n

mr = exp(−c( t / L2 ) n )

mr = a exp(−kt )

mr = at

2 2 2

−n

2

mr = 1 + at + bt 2

2

t = a ln mr + b(ln mr ) 2

2

mr = a exp(− kt ) + (1 − a ) exp(−kat ) mr = a 0 + a exp(−kt )

mr = a 0 /(1 + a exp(kt ))

mr = a exp(− kt ) + (1 − a ) exp(−kbt )

mr = a exp(−kt ) + (1 − a ) exp(−gt ) mr = a 1 exp(− k 1 t ) + a 2 exp(−k 2 t )

mr = a exp(− kt ) + bt n

mr = a exp(−kt + b t ) + c

mr = a − b exp(−(kt )) n

mr = a exp(−kt ) + b exp(−gt ) + c exp(−ht )

2

3 3 3 3 4 4 4 4 6

where mr is the moisture ratio, t is the time (s), L is the characteristic length (m) and a, b, c, n, k, g and h are the empirical constants

8. SHRINKAGE The drying process causes shrinkage in volumes and heat and mass exchange areas of food materials. This shrinkage in the drying process particularly affects the diffusion coefficient of the material significantly. Heat and mass transport and shrinkage are generally coupled. The following procedure is followed in coupling. Shrinkage is modeled with respect to moisture content, including sometimes other physical properties, and this predetermined relationship of dimensional changes is used to modify the geometry with time in the simulation [8]. Shrinkage models used in the literature are given in the comprehensive review prepared by Mayor and Sereno [46]. Of these models, those given in tables 15 and 16 are empirical models. Mayor and Sereno [46] states that linear empirical models are adequate to describe materials and process conditions leading to negligible porosity development during the drying process, or to a uniform development of porosity, corresponding to a linear decrease of volume in the whole range of humidity. If development of porosity increases sharply during the final stage of drying, linearity is lost and the behavior is best described by nonlinear models seen in table 16 such as exponential models or a quadratic model [46, 113]. These models usually produce a good fit with the experimental data, but their use is limited because of their dependence on the drying conditions and on the material and they require

Transport Phenomena During Drying of Food Materials

127

extensive experimental testing and should not be extrapolated. For modeling shrinkage, some fundamental models are also used. They are based on mass balances, density and porosity definitions and assume in most cases additivity of the volumes of the different phases in the system [46]. These fundamental methods have been classified by Mayor and Sereno [46] in three groups: models which show a linear shrinkage behavior throughout the whole drying process (table 17); models which include deviations of this linear behavior (table 18) and models which include explicitly variations of the porosity through the drying process (table 19). The fundamental models allow the prediction of moisture content and/or change in volume to be obtained without complicated mathematical calculations. Furthermore, it is not usually necessary to obtain experimental shrinkage values at every process conditions, as in the case of empirical models. Table 15. Linear empirical models for shrinkage (adapted from [46]) Model

Geometry and reduced dimension slab-thickness; cylinder-volume sphere-radius ellipsoid-(x,y,z coordinates); sphere-radius cylinder-volume

D R = a 1m + a 2

D R = a 1ε w + a 2

D R = 1 + βm

E ⎤ ⎡ D R = 1 + ⎢a 1 exp(− a )⎥ m RˆT ⎦ ⎣ T (Kelvin)

D R = a 1 + (a 2 + a 3 φ + a 4 T)Δm

D R = (a 1T + a 2 ) + (a 3 T + a 4 ) m

Material Apple Soybean Apricot Carrot, Amylose starch gel, Broccoli stem

slab-(thickness,width,length); sphere-volume; (cube,cylinder)-volume sphere-volume cylinder-(volume, radial, axial) slab-(thickness, width, length) slab-(thickness, width, length) sphere-volume

Grape Green bean Fish muscle (shark) Fish muscle (ocean perch) Cherry

Potato

slab-(thickness, width, length)

Fish muscle (ocean perch)

cube-volume slab-thickness cylinder-volume

Apple, carrot, potato gelatin gel, carrot Apple, carrot, banana, potato

sphere-volume

Grape

sphere-bed volume

wheat, canola

cylinder-volume

Potato

where ai is the empirical constants, m moisture content (kg water/kg dry solid), εw the volume fraction of water (volume of water/total volume), β shrinkage coefficient, Ea activation energy (J/mol), Rˆ universal gas constant (8.314 J/(mol K)), T temperature (°C), φ relative humidity.

128

Kamil Kahveci and Ahmet Cihan Table 16. Non-linear empirical models for shrinkage (adapted from [46])

Model D R = 0.16 + 0.816(m / m o ) + 0.022 exp(

0.018 m ) + b1 (1 − ) m + 0.025 mo

b1 = 0.209 − b 2

b2 =

0.966 m o + 0.796

A v / A vo = a 1 + a 2 m + a 3 m 2 + a 4 m 3

D R = a1 + a 2 m + a 3m 2 + a 4 m3 D R = a 1 + a 2 exp(−a 3 t ) D R = a1 + a 2

m m + exp(a 3 ) 1+ m 1+ m

D R = a 1 + a 2 m + a 3 m 3 / 2 + a 4 exp(a 5 m)

Geom.& Red. dim.

Material

cylinder-volume

Carrot, pear, potato, sweet potato

slab-volume

Garlic

cylinder-surface area to volume ratio

Apple, carrot, potato

cylinder-bed volume

Apple, carrot, potato

slab-surface area slab-thickness

Potato, squash Apple

hemispherediameter; cylinder-length

Cauliflower

slab-thickness

Garlic

D R = a 1 + a 2 ( m / m o ) + a 3 (m / m o ) 2

(cylinder, slab)-volume

D R = a 1 exp(a 2 m / m o )

(cylinder, slab)-volume

Apple, carrot, potato, squid Apple, carrot, potato, squid

where ai is the empirical constants, m moisture content (kg water/kg dry solid), Av the surface area to the volume ratio (1/m).

Table 17. Linear fundamental models for shrinkage (adapted from [46]) Model

V ⎛ m + 0.8 ⎞ A ⎛ V ⎞ ⎟ ⎟; =⎜ =⎜ Vo ⎜⎝ m o + 0.8 ⎟⎠ A o ⎜⎝ Vo ⎟⎠

Geom.& Red. dim.

V m = b1 + b 2 Vo mo

b1 =

Material

2/3

m o (ρ s / ρ w ) 1 , b2 = m o (ρ s / ρ w ) + 1 m o (ρ s / ρ w + 1

Vegetables

volume

Sugar beet root

cube-area

Carrot, potato, sweet potato, radish

cube-area

Carrot, potato, sweet potato, radish

Uniform drying model A ⎛ V =⎜ A o ⎜⎝ Vo

a) b)

⎞ ⎟ ⎟ ⎠

2/3

m + b1 1 1 V = ; b1 = m e ( − 1) + ρe ρe Vo m o + b1

ρo V = b1 m + b 2 , b 1 = , b 2 = 1 + b1 − ρ o Vo mo + 1

Core drying model

1 − b2 V = b1 m + 1 , b1 = , Vo mo − me

(m e + 1)ρ o V A b2 = , = ( )2/3 (m o + 1)ρ e A o Vo

where m is the moisture content (kg water/kg dry solid), A area (m2), V volume (m3), ρ density (kg/m3).

Transport Phenomena During Drying of Food Materials

129

Table 18. Non-linear fundamental models for shrinkage (adapted from [46]) Model Semi-core drying model V = b1 m + b 2 Vo

b1 =

Material

cube-area

Carrot, potato, sweet potato, radish

cylinder-volume

Cassava root

V A = ( )2/3 Vo Ao

1 − b3 m o − m e + b 4 (b 3 m o − m e + b 3 − 1)

b m − m e − b 4 (b 3 m o − m e + b 3 − 1) b2 = 3 o m o − m e − b 4 (b 3 m o − m e + b 3 − 1) b3 =

Geom.& Red. dim.

(m e + 1)ρ o ρ − (1 − m)ρ e , b4 = e ρo (m o + 1)ρ e

D r = b1 + b 2 (m / m o ) + 0.26b 3 (1 − m / m o ) 3 b1 =

b3 =

m o (ρ s / ρ w ) 1 , b2 = m o (ρ s / ρ w ) + 1 m o (ρ s / ρ w + 1 0.966 m o + 0.796

where m is the moisture content (kg water/kg dry solid), A area (m2), V volume (m3), ρ density (kg/m3).

Table 19 Fundamental models for shrinkage including porosity (adapted from [46]) Model

Geom.&Red.dim.

Material

cylinder-volume

Carrot, pear, potato, sweet potato

cylinder-volume

Carrot, pear, potato, sweet potato

slab-volume

Beef meat

slab-volume

Squid

cylinder-volume

Apple, potato, carrot, squid

Model A (inclusion of initial porosity) ⎛ ⎞ m + b 2 (m) ⎟⎟ b 3 D R = ⎜⎜ b1 ⎝ mo ⎠

(

)

χ sg. + ρ sn (m)b 4 b1 ⎞ ⎛ χ sg. ρ sn ,o + b1 = ⎜⎜1 + b 4 ⎟⎟ , b 2 = mo mo mo ⎠ ⎝ −1

b3 =

χ χ 1 − ψ (m o )ρ sn (m o ) , b 4 = cw + st 1 − ψ (m)ρ sn (m) ρ cw ρ st

Model B (without inclusion of initial porosity) 1 (b1 + χ sg / ρ sg + χ / ρ sn )ρ o DR = (1 − ψ ) mo + 1

χ χ b1 = cw + st ρ cw ρ st DR = DR =

1 ⎡ ρ o (m − m o ) ⎤ ⎢1 + ⎥ ρ w (1 + m o ) ⎦ (1 − ψ ) ⎣

ρo 1 + m (1 − ψ ex − ψ ) , ρ= m ) ρ 1 + mo ∑ m i /(ρ T ) i i =1

DR =

⎤ 1 ⎡ ρ o (m − m o ) − ψo ⎥ ⎢1 + ρ w (1 + m o ) (1 − ψ ) ⎣ ⎦

) where m is the moisture content (kg water/kg dry solid), m mass fraction (kg/kg total mass), χ constituent concentration 3 (kg/kg dry solid), ψ porosity, ρ density (kg/m ).

130

Kamil Kahveci and Ahmet Cihan

9. DIFFUSIVITY In a porous material, various diffusion coefficient definitions exist depending on the transport mechanism. Diffusion may in general be divided into two parts as molecular and capillary diffusion. Molecular diffusion can occur in liquid and gas phase. Molecular diffusion in the gas phase becomes more important when the water saturation decreases. Generally, liquid and gas diffusion are expressed in one single equation instead of defining individually. In this case, diffusivity becomes an effective parameter. This effective diffusivity value is different from diffusivity value which is only for liquid or for gas. Capillary diffusivity has two components one due to moisture gradient and the other due to temperature gradient. Capillary diffusivity due to the temperature gradient is mostly omitted. Capillary diffusivity data is generally unavailable. However, effective diffusivity data available in the literature is close to capillary diffusivity when the material is very wet because the molecular diffusion in that condition is insignificant [102]. For materials with low moisture content, effective diffusivity is close to molecular diffusivity. Diffusion coefficient for a porous material is generally determined experimentally. Commonly used techniques are: sorption kinetics method, permeation method, concentrationdistance curve method and drying method. Among these techniques, drying method is the one widely used. Although there are alternative approaches to determine diffusivity by drying method, all approaches are based on diffusion equation. Different approaches based on the drying method are given below.

Simplified Methods In these methods, solution of diffusion equation is used to determine diffusion coefficient. Series solutions of the diffusion equation for some simple geometries are as follows. mr =

mr =

8 π2

⎡ π 2 D eff t 1 π 2 D eff t π 2 D eff t 1 − + − + − exp( ) exp( 9 ) exp( 25 )+ ⎢ 9 25 4L2 4L2 4L2 ⎣ ⎤ π 2 D eff t 1 exp(−49 ) + ..⎥ 2 49 4L ⎦

π 2 D eff t π 2 D eff t π 2 D eff t 1 1 6 ⎡ )+ exp( 9 − + − + − ) exp( 4 ) exp( ⎢ 9 4 L2 π2 ⎣ L2 L2 ⎤ π 2 D eff t 1 exp(−16 ) + ..⎥ 2 16 L ⎦ D t Deff t ) + 0.131 exp(−30.5 eff2 ) + L L2 D t D t 0.0534 exp(−74.9 eff2 ) + 0.029 exp(−139.1 eff2 ) + .. L L

(467)

(468)

m r = 0.692 exp(−5.78

(469)

Transport Phenomena During Drying of Food Materials

131

where L is the characteristic length (m). Since the values of the exponential terms of Eq. (467) (for infinite slab), Eq. (468) (for sphere) and Eq. (469) (for infinite cylinder) except the first terms contribute little to moisture ratio when the value of D eff t / L2 is greater than 0.2, the first term can be taken into account for finding the effective moisture diffusivity. If, also, natural logarithms are taken of these equations, they become as: ln(m r ) = ln ln(m r ) = ln

8 π 2 D eff − t π2 4L2

(470a)

6 π 2 D eff − t π2 L2

(470b)

π 2 D eff (470c) t L2 From Eqs. (470), plots of ln(mr) versus drying time t give straight lines with the slopes of ln(m r ) = ln 0.692 − 5.78

Slope =

π 2 D eff 4L2

Slope =

π 2 D eff L2

Slope = 5.78

π 2 D eff L2

(471)

These slopes are used as a measure for the diffusivity (see figure 25). x

ln(mr)

x x x

Slope=f(Deff;L)

x x

x x

x x

t Figure 25. Experimental drying curve.

This method can not be used when the diffusion coefficient is strongly dependent on concentration. In this case, the following procedure is followed in determining the diffusivity. The theoretical moisture ratio is evaluated numerically for a range of the Fourier numbers. Then, the same ratio is evaluated using experimental data. Subsequently, both theoretical and experimental moisture ratio curves are plotted versus time and the Fourier number on a semilogarithmic diagram as shown in figure 26. Moisture diffusivity is determined with the help of the following equation by comparing the slopes of both curves [114]:

132

Kamil Kahveci and Ahmet Cihan ⎛ ∂m r ⎞ ⎟ ⎜ ⎝ ∂t ⎠ exp 2 L D= ⎛ ∂m r ⎞ ⎟ ⎜ ⎝ ∂t ⎠ the 1.0

(472)

x x x

mr

x x

experimental x

0.1

x

theoretical

x x x

x

0.0 0

2

4

6

0.0

0.2

0.4

0.6

8 t (h) 10 0.8 Fo

1.0

Figure 26. Theoretical and experimental drying curves (adapted from [114]).

Regular Regime Method The regular regime method with which concentration dependent diffusivity can be calculated is based on the experimental measurement of the regular regime curve. Regular regime curve is the drying curve when it becomes independent of the initial concentration profile. The method assumes that material is homogeneous and nonporous, however it can also be applied to cellular tissue foods such as apples and potatoes [115]. In this method, the diffusion equation is solved numerically for the regular regime period of an isothermal diffusion process to obtain diffusivity. Regular regime method is rather complicated and needs successive interpolations and differentiations of the experimental drying data. However, a short-cut method has also been proposed to avoid this rigorous procedure [116].

Regression Analysis Method In this method, first partial differential equations governing drying process are created. Geometry and particular boundary conditions are taken into consideration. Diffusivity is mostly expressed as a parametric model of local moisture content, temperature, or any other property. Following assignment of initial guess for parameters contained in the diffusion coefficient, the partial differential equation system is solved numerically. The values obtained are compared with experimental values with non-linear regression analysis. If the criterion of the least sum of squares is not satisfied, a new guess of the model parameters is fed back to the numerical calculation. Procedure is continued until a final convergence is obtained.

Transport Phenomena During Drying of Food Materials

133

Diffusion coefficient of food materials are mostly obtained by using one of the drying techniques given above. For various food materials, diffusion coefficients or correlations obtained using the diffusivity data are given in table 20. Diffusion coefficient is affected by many physical properties. Brief explanations on these factors are given below.

Temperature The dependence of the diffusivity on temperature is generally described by the Arrhenius equation as follows: D eff = D o exp(−

Ea ) RˆT

(473)

Potential Energy

Here Do is the diffusion coefficient factor, Rˆ is the universal gas constant (8.314 J/(mol K)), T is the air temperature (K) and Ea is the activation energy (J/mol). Activation energy values for various food materials are given in table 21. Activation energy is used to describe temperature dependence of diffusion coefficient. According to the energy levels involved in a reaction and to the collision theory of reactive molecules, enough energy must be generated to provide the necessary activation energy to be able to develop the reaction (figure 27) [114, 117]. Activation energy will not itself provide any idea of the reactivity of a given system, only information on temperature dependence of the reaction. Activation energy is also related to moisture content. The activation energy for diffusion increases at lower moisture contents since generally the interaction forces between moisture and solid are higher at lower moisture contents.

energy level in an activated state

activation energy for reverse reaction average energy of reactants

activation energy

average energy of reactants heat of reaction

Figure 27. Potential energy levels during a given endothermic reaction [114, 117].

134

Kamil Kahveci and Ahmet Cihan

Pressure Under certain conditions, there may be a significant increase in the gas pressure inside the material during drying. As it can be seen from Eqs. (161)-(163), gas pressure is inversely proportional to diffusion coefficient and any increase in the gas pressure causes gas diffusion coefficient to decrease.

Composition Moisture diffusivity of most food materials shows a strong dependence on composition. Diffusivity is generally small for small values of moisture content and shows an increase with increasing moisture content and approaches to a constant value for values of moisture content higher than a certain value. Concentration dependence of diffusion coefficient is generally expressed by relations in linear, polynomial and exponential forms. Also, some studies show that an increase in the salt content of food materials leads to a decrease in moisture diffusivity particularly for high temperature and low moisture content. A decrease is also observed in the diffusion coefficient of some food materials with an increase in fat content. Some other studies show that with an increase in protein content, moisture diffusion coefficient can take higher values.

Shrinkage Diffusion coefficient is highly affected by shrinkage. This effect is due to the decreasing diffusion path in shrinking media [118]. Some researchers incorporated the volume change into the diffusion coefficient in order to take account the shrinkage effect on transport properties. They suggested multiplying the diffusion coefficient by a power of the volume changing factor, which is the ratio between the actual volume and a reference volume that is either the basic initial volume or the volume of the totally dried samples [119]: ⎛V⎞ D eff = ⎜⎜ ⎟⎟ D eff ,R ⎝ Vo ⎠

n

(474)

The power exponent used by Crank [74] was 2 and by Fish [120], 2/3. Another approach used in the literature is based on calculating first a reference diffusion coefficient by using the initial thickness of the product. The diffusion coefficient obtained can thereafter be corrected for the shrinkage by applying: ⎛ L ⎞ D eff = ⎜⎜ ⎟⎟ D eff ,R ⎝ L o ⎠

n

(475)

The shrinkage during drying is sometimes neither ideally three-dimensional, nor onedimensional. For these cases, the following equation is suggested for the diffusion

Transport Phenomena During Drying of Food Materials

135

coefficient [121, 122]: ⎛V⎞ D eff = ⎜⎜ ⎟⎟ D eff ,R ⎝ Vo ⎠

2/ n

(476)

where the exponent n equals to 1 for one-dimensional shrinkage and equals to 3 for isotropic three-dimensional shrinkage. This parameter n may be viewed as a measure of the degree of isotropicity of the deformation and is related to volume shrinkage by [119]: Sb = Sdn

(477)

where Sb and Sd are defined as Sb =

V Vo

Sd =

L d or d o Lo

(478)

Other Factors In food materials enzymatic and microbiological changes may also have an effect on diffusion coefficient. Enzymatic reactions may change the internal properties of food material. Furthermore, the micro-organisms may also influence the surface of the food material, changing the equilibrium moisture content of food at the surface. Another factor acting on diffusion coefficient is pretreatment. Food materials may be exposed various types of pretreatment before drying. Pretreatment generally causes a change in the porosity of food materials and this leads to a change in diffusion coefficient.

CONCLUSION A large number of models have been proposed to simulate drying process of food materials. The basic reason of such an abundance of model propositions is the variety of transport mechanisms involved in drying process and complexity of material structures. Models based on continuum approach are preferred generally to simulate drying at macroscopic level. Among proposed models based on continuum approach, Whitaker model is one step ahead of other models in that physical basis of this model is stronger. However, the empirical character of transport coefficients as in other models based on the continuum approach constitutes the most important drawback of this model. The majority of transport coefficients show a stronger dependence on concentration, temperature and material structure. The effect of concentration and temperature on transport parameters is relatively well known and various models have been proposed to express these effects. However, there is relatively limited knowledge on the effect of structure on transport. It may be said that determination the effects of structure on transport will hereafter be one of the basic topics in studies related drying behavior of food materials.

Table 20. Moisture diffusion coefficients of various food materials Material

Meth&Geom

Apple (Red Delicious)

Slope Cylinder

Apple (organic)

Slope Slab

Apple pomace

Slope Slab

Apricot

Slope Slab

Banana

Regression Cylinder

Banana

Slope Slab

Barley

-

Basil leave

Slope Slab

Bean

Regression Sl.-Sph.

Drying Conditions Hot air m=0.06-8.5, T= 40-70°C v=3 m/s ,d=0.7cm , L=10d PT: water bath Hot air mw=0.11-0.82, T=40-60°C v=0.8m/s, L=5-9 mm PT: 5% lemon solution Microwave mw=0.25-0.40 N=150-600W Hot air mw=0.16-078, T=50-80°C v=0.2-1.5m/s Hot air m=0.2-3.8, T=60-80°C v=1.3 m/s Hot air (Tunnel) m=0.24-4, T=40-60°C v=0.3-0.7m/s, L=4mm m=0.10-0.27 T=30-70°C Open sun mw=0.04-0.87 Te=30-36.5°C ϕe=0.24-0.28 qs=195-796 W/m2 T=25-40°C

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

D eff = 1.65x10 −6 exp(−

123

19.34 ) RˆT ab

Deff= 2.27x10-10 –Deff=4.97x10-10

124

Deff=1.0465x10-8 for N=150W , Deff=2.2130x10-8 for N=30W Deff=2.7157x10-8 for N=450W , Deff=3.6854x10-8 for N=600W

125

D eff = 1.14x10 −7 + 2.25x10 −9 T + 1.47 x10 −7 v

126

D eff = 1.36 x10 −7 exp(−

127

13.4 ) RˆTab

Deff =6.61x10-11 for T=40°C, v=0.3 m/s Deff =1.95x10-10 for T=60°C, v=0.3 m/s Deff =1.61x10-10 for T=50°C, v=0.5 m/s Deff =9.80x10-11 for T=40°C, v=0.8 m/s Deff =2.41x10-10 for T=60°C, v=0.8 m/s

, , , ,

Deff =1.57x10-10 for T=50°C, v=0.3 m/s Deff =7.33x10-11 for T=40°C, v=0.5 m/s Deff =2.23x10-10 for T=60°C, v=0.5 m/s Deff =1.74x10-10 for T=50°C, v=0.8 m/s

128

Deff=1.31x10-11-6.52x10-10

129

Deff =6.44x10-12

130

Deff=4.35x10-11-3.79x10-9

131

Table 20. (Continued). Material

Meth&Geom

Beef Meat

Slope Slab

Biscuit

Black grape

Black tea particle

Broccoli floret

Carrot

-

Slope Sphere

Slope Sphere

Drying Conditions Hot air m=0.1-2.8, T=6.6-40.4°C d=38mm h=10mm m=0.10-0.60 T=20-100°C Hot air mw=0.25-0.79, T=60°C v=1.1 m/s PT: Potassium carbonate solution: 5% K2CO3 + 0.5% olive oil PT: Ethyl oleate plus potassium carbonate solution PT: Ethyl oleate plus potassium hydroxide solution PT: Ethyl oleate plus sodium carbonate solution Hot air T=80-120°C v=0.25-0.65m/s

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

D eff = 5.09x10 −6 exp(

132

24.643 ) RˆT ab

Deff=8.6x10-10-9.4x10-10

133

Deff =3.82x10-10

Deff =1.05x10-9 134

Deff =1.28x10-9

Deff =8.64x10-10 Deff =7.97x10-10 D eff = 1.68x10 −7 exp(−

406.02 ) RˆT

135

ab

Regression

Hot air T=50-75°C v=1.2-2.25m/s l=10mm

Deff=3.5780x10-6 for T=50°C, v=1.2m/s , Deff=5.9747x10-6 for T=60°C, v=1.2m/s Deff=10.6760x10-6 for T=75°C, v=1.2m/s , Deff=6.0112x10-6 for T=50°C, v=1.75m/s Deff=8.2654x10-6 for T=60°C, v=1.75m/s , Deff=12.8230x10-6 for T=75°C, v=1.75m/s Deff=4.8886x10-6 for T=50°C, v=2.25m/s , Deff=7.8897x10-6 for T=60°C, v=2.25m/s Deff=16.6770x10-6 for T=75°C, v=2.25m/s

136

Slope Sphere

Hot air mw=0.06-088, T=50-70°C v=0.5-1.0m/s Dims: 1x1x1cm 2x2x2cm

Deff=0.776x10-9 - 9.335x10-9

137

Table 20. (Continued). Material

Carrot

Meth&Geom

Slope Slab

Drying Conditions Hot air m=0.2-14, T=55-75°C v=1.6 m/s, Dims:1x1x1cm

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5 D eff = 3.0x10 −6 exp(−

Ref.

22.1426 ) RˆT ab

PT: NaCl

D eff = 4.0 x10 −8 exp(−

PT: Sucrose

D eff = 2.0 x10

−7

PT: Sucrose+NaCl

D eff = 3.0x10

−7

10.0017 ) RˆT ab

14.8729 ) exp(− RˆT

138

ab

16.2130 ) exp(− RˆT ab

Carrot

Slope Slab

Carrot core

Slope Cylinder

Carrot cortex

Slope Cylinder

Celery

Inverse M.

Chelwa

Slope Cylinder

Cherry (sweet)

Regression Sphere

Infrared mw=0.0-0.9, T=50-80°C L=1-2mm Hot air m=1.17-6.39, T=40-70°C v=1.5,3m/s, d=0.7cm PT: water bath Hot air m=1.05-5.31, T=40-70°C v=1.5-3m/s, d=0.7cm PT: water bath m=0.09-9.82 T=49.1°C ρc=461-1428kg/m3 L=3-10mm Open sun m=0.1-2.5,Te=32.5-42.5°C ϕe=0.15-0.32 qs=460-820W/m2 Hot air m=0.1-2.6, T= 50-80°C φ=0.05-0.5, v=1-5 m/s

Deff =7.295x10-11 for T=50°C , Deff =9.309x10-11 for T=60°C Deff =1.140x10-10 for T=70°C , Deff =1.501x10-10 for T=80°C

139

Deff =6.42x10-10 -14.7x10-10

140

Deff =6.68x10-10 -13.6x10-10

140

Deff = 7.98x10−4 exp(0.130m −

3217.3 1323.7 + ) ρc Tab

D eff = 17.57 x10 −11 exp(−1.591m m )

Deff =6.814x10-11 for T=50°C , Deff =12.516x10-11 for T=60°C Deff =25.026x10-11 for T=70°C , Deff =34.739x10-11 for T=80°C

141

142

143

Table 20. (Continued). Material

Meth&Geom

Cherry (sour)

Slope Sphere

Chestnut

Slope Sphere

Drying Conditions Hot air mw=0.2-0.82, T=55-65°C v=1 m/s

Hot air m=0.04-0.50, T=70-90°C

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

Deff =4.75x10-10 for T=55°C, Deff =5.57x10-10 for T=65°C

144

D eff = 1.2152 x10 −5 exp(− D eff = 9.8684 x10 −6 exp(−

22.578 ) RˆT

for Longal variety

21.926 ) RˆT

for Martainha variety

20.458 ) RˆT

for Viana variety

ab

145

ab

Chestnut Chickpea

Coconut presscake

Coffe cherry

Corn

D eff = 6.6979 x10 −6 exp(−

Slope Sphere

Hot air m=0.04-0.40, T=70-90°C

Regression Sphere

T=15-40°C

9.71x10-11-5.98x10-10

Vacuum m=0.02-1.038 P=65 mmHg T=65-75°C, L=2-4 mm Hot air mw=0.11-0.66T=45,60°C Hot air mw=0.14-0.3, T=55-75°C

D eff = 0.3753x10 −8 exp(549.8L −

Slope Slab Regression Sphere Slope Sphere

PT: alkali solution

145

ab

129 171.9 ) T

146

Deff =0.1x10-10-1x10-10 for T=45°C , Deff =0.3x10-10-3x10-10 for T=60°C

147

Deff =9.488x10-11 for T=55°C,Deff =1.153x10-10 for T=65°C,Deff =1.768x10-10 for T=75°C Deff =1.424x10-10 for T=55°C,Deff =1.733x10-10 for T=65°C,Deff =2.716x10-10 or T=75°C

148

Cracker

-

m=0.03-0.14 T=40-90°C

Deff=1.41x10-11-1.81x10-9

129

Date (red soft )

Slope Sphere

Hot air m=0.4-1.6, T=50-80°C v=1.5 m/s

Deff =5.89x10-10 for T=50°C , Deff =7.48x10-10 for T=60°C Deff =1.45x10-9 for T=70°C , Deff =1.78x10-9 for T=80°C

149

Date (tempo 2)

Slope Sphere

Hot air m=0.4-1.5, T=50-80°C v=1.5 m/s

Deff =3.22x10-9 for T=50°C , Deff =4.52x10-9 for T=60°C Deff =7.32x10-9 for T=70°C , Deff =8.16x10-9 for T=80°C

149

Table 20. (Continued). Material

Meth&Geom

Drying Conditions

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

Date (tempo 3)

Slope Sphere

Hot air m=0.5-1.5, T=50-80°C v=1.5 m/s

Deff =7.53x10-9 for T=50°C , Deff =1.11x10-8 for T=60°C Deff =2.30x10-8 for T=70°C , Deff =2.98x10-8 for T=80°C

149

Dough

-

M=0.20-0.50 T=15-203°C

Deff=1.30x10-10-1.00x10-9

129

Dill leave

Slope Slab

Hot air mw=0.05-0.82, T=50-70°C v=1.1 m/s

Deff =6.69x10-10 for T=50°C , Deff =9.21x10-10 for T=60°C , Deff =1.43x10-9 for T=70°C

150

Eggplant

Slope Slab

Fig

Slope Sphere

Fig

Slope Sphere

Vacuum m=0.0-0.16, T=30-50°C P=2.5,5,10kPa Dims: 45x25x20mm Open sun mw=0.25-0.74,Te=3547°C Hot air (conditioned) m=0.14-1.9, T=55-85°C ϕ=0.10 , v=0.5-3 m/s

D eff = 2.012x10 −4 exp(−

29.52 ) Tab

151

Deff =2.47x10-10 D eff = 4.83x10 −4 exp(− D eff = 2.01x10

152 37.27 30.81 ) , D eff = 8.77 x10 −5 exp(− ) for v=0.5 and 1 m/s RˆT RˆT ab

−2

45.81 exp(− ) , RˆT ab

D = 5.99 x10

ab

−2

48.47 exp(− ) eff RˆT

for v=2.0 and 3

153

ab

m/s Garlic

Garlic

Slope Slab Regression Slab

Hot air m=0.08-1.56, T=40-60°C v=0.8m/s, L=3-5mm Hot air m=0.27-1.35, T=50-80°C v=2-4m/s

D eff = 1.61x10 −6 exp(−

23.48 ) RˆT

154

ab

D eff = 7.490x10 −6 exp(−

27.84 ) RˆT ab

155

Table 20. (Continued). Material

Meth&Geom

Grape

Regression Sphere

Grape (Sultana)

Slope Slab

Green bean

Slope Slab

Drying Conditions Hot air m=0.2-2.4, T=40-70°C v=1-2.3m/s PT: alkali solution (1% of sodium hydroxide) Hot air m=0.18-3 , T=65°C v=0.45m/s Hot air mw=0.14-0.9, T=50-70°C v=1 m/s , L=4cm (length) Hot air mwo=0.94, T= 25-45°C v=4.1 m/s , l=10mm Hot air mw=0.16-0.86, T= 3060°C v= 1 m/s L=10-50mm (layer thick) Hot air m=0.15-4.65, T=30-90°C L=6mm

Green pepper

Slope Slab

Kale

Slope Slab

Kiwi

Regression Slab

Lentil

Regression Slab

Mango

Slope Slab

Minced meat

-

Milk (skimmed)

-

m=0.3-0.8, T=30-50°C

-

Hot air m=0.3-1.8, T= 30-120°C

Minced meat

T=15-40°C Hot air m=0.08–8, T=55-65°C L=2.8 mm Hot air m=0.3-1.8, T= 30-120°C

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5 D eff = 0.522 exp(−

54 ) exp[− (0.0075Tab + 1.829)m ] RˆTab

D eff = 1.6x10 −3 exp(−

49 ) exp[− (0.0012Tab + 0.309)m ] RˆT

Ref.

156

ab

Deff=6.44 x10-10

157

D eff = 5.53x10 −4 exp(−

35.43 ) RˆT

158

ab

D eff = 1.320x10 −2 exp(−

51.4 ) RˆTab

159

Deff =14.8894x10-10 – 55.9451x10-10

160

D eff = 1.476 x10 −5 exp(−

161

27 ) RˆTab

Deff=3.53x10-10-1.33x10-9

131

Deff =2.62x10-10 for T=55°C , Deff =2.95x10-10 for T=60°C , Deff =3.19x10-10 for T=65°C

162

Deff =5.0x10-11-53.0x10-11

114

Deff= 0.24x10-10-2.1x10-10

163

-11

Deff =5.0x10 -53.0x10

-11

114

Table 20. (Continued). Material

Meth&Geom

Mint leave

Slope Slab

Mint leave

Slope Slab

Mulberry

Regression Slab

Mullet roe

Slope Slab

Murici

Regression Slab

Olive cake

Slope Slab

Olive extraction

oil

Olive extraction

oil

Slope Slab Regression Slab

Okra

Slope Sphere

Onion

-

Drying Conditions Open sun mw=0.1-0.86, Te=30-36.5°C φe=0.24-0.28 qs=195-796W/m2 Hot air mw=0.1-0.85, T=35-60°C v=4.1m/s Hot air m=0.1-4.6, T=60-80°C v=1.2 m/s Hot air m=0.3-1, T= 20-40°C v=1.0 m/s Dims.20x10x3cm Hot air mwo=0.88, T= 50-70°C v=1.5 m/s Hot air mw=0.05-0.45 T=50-110°C, v=1.2m/s Hot air m=0.1-2 , T=20-80°C v=1.0m/s Hot air mw=0.05-0.45 T=80-110°C, v=1.2m/s Hot air mw=0.15-0.9 , T=50-70°C v=1m/s m=0.10-10 T=40-80°C

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

Deff =7.04x10-12

130

Deff =3.067x10-9 for T=35°C , Deff =5.837x10-9 for T=45°C, Deff =1.237x10-8 for T=55°C , Deff =1.941x10-8 for T=60°C

164

Deff =2.32x10-10, Deff =2.84x10-10, Deff =3.58x10-10 for T=60,70,80°C (CRP) Deff =5.03x10-10, Deff =6.68x10-10, Deff =8.43x10-10 for T=60,70,80°C (FFRP) Deff =2.63x10-9 , Deff =2.30x10-9 , Deff =2.76x10-9 for T=60,70,80°C (SFRP)

165

D eff = 19.8x10 −4 exp(−

37.2 ) RˆTab

166 D

Deff =1.275x10-9 for T=50°C , Deff =1.975x10-9 for T=60°C, Deff =2.906x10-9 for T=70°C D eff = 3.128x10 −6 exp(−

167

17.97 ) RˆT

168

15.77 ) RˆT

169

ab

D eff = 1.615x10 −7 exp(−

ab

Deff=4.89x10-8 - 9.98x10-8

170

Deff =4.27x10-10 for T=50°C , Deff =7.76x10-10 for T=50°C , Deff =1.30x10-9 for T=50°C

171

Deff=1.38x10-11-6.60x10-9

129

Table 20. (Continued). Material

Meth&Geom

Orange skin

Regression Slab Regression Slab

Drying Conditions Hot air m=0.25-3.66, T=30-90°C v=2.5 m/s, L=0.45cm Hot air mw=0.3-088, T=40,60°C v=1.25-3.25m/s

Papaya PT: osmotical

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

D = 3.957 x10 −4 exp(−

172

36.36 −0.0496 m ) RˆT ab

Deff =1.72x10-9 v=1.25m/s Deff =2.21x10-9 v=3.25m/s Deff =1.03x10-9 v=1.25m/s Deff =1.11x10-9 v=3.25m/s

for T=40°C and v=1.25m/s , Deff =2.71x10-9

for T=60°C and

for T=40°C and v=3.25m/s , Deff =4.78x10-9

for T=60°C and

for T=40°C and v=1.25m/s , Deff =1.48x10-9

for T=60°C and

for T=40°C and v=3.25m/s , Deff =1.78x10-9

for T=60°C and

173

Parsley leave

Slope Slab

Open sun mw=0.06-0.84 Te=30-36.5°C ϕe=0.24-0.28 qs=195-796 W/m2

Parsley leave

Slope Slab

Hot air mw=0.05-0.82, T=50-70°C v=1.1 m/s

Deff =9.00x10-10 for T=50°C , Deff =1.36x10-9 for T=60°C , Deff =2.34x10-9 for T=70°C

150

Pasta

-

mw=0.05-0.27, T=40-74°C

Deff=0.8x10-11-9.3x10-11

163

Pasta (Semolina)

Regression Slab

Hot air m=0.14-0.50, T=40-100°C Dims:100.0x20.0x1.3mm

Deff = 1.2x10−11 exp(−3036.95(

Pea

-

T=30-65°C

Deff=3.1x10-10-6.6x10-10

Slope

Hot air m=0.16-7.52, T= 55-65°C v=0.8 m/s, L=3.5mm

Deff =3.04x10-6 for T=55°C , Deff =3.62x10-6 for T=65°C

PT: 1% KMS

Deff =3.44x10-6 for T=55°C , Deff =4.41x10-6 for T=65°C

PT: 1% ascorbic acid

Deff =3.51x10-6 for T=55°C , Deff =4.04x10-6 for T=65°C

Peach

Deff =4.53x10-12

130

1 1 − ) exp(6.46m) T 293

174 163

175

Table 20. (Continued). Material

Meth&Geom

Pear

Regression Sphere

Pineapple

-

Pistachio nuts

Slope Sphere

Drying Conditions Hot air (conditioned) mw=0.2-0.8, T=30-50°C v=0.5-1.5m/s, ϕ=0.4-0.6 L=4mm

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5 D eff = 1.1276 x10 −5

1 + 1.2987 x10 −2 m 24.3 ) exp(− 1 + 4.5111x10 −1 m s RˆTab

Ref. 176

m=3.80-5, T=30-50°C

Deff=5.38×10-12-2.64×10-9

129

Hot air mw=0.37-0.05 T=25-70°C ϕ=0.05- 0.2 v=0.5-1.5m/s

Deff=1.37x10-10 Deff =2.26x10-10 Deff =5.24x10-10 Deff =7.01x10-10 Deff =5.42x10-11 Deff =2.29x10-10 Deff =4.91x10-10 Deff =8.82x10-10 Deff =1.70x10-10 Deff =3.26x10-10 Deff =5.25x10-10 Deff =8.07x10-10 Deff =1.36x10-11 Deff =2.98x10-10 Deff =5.82x10-10 Deff =8.40x10-10 Deff =1.80x10-10 Deff =3.48x10-10 Deff =5.75x10-10 Deff =9.29x10-10 Deff =1.81x10-11 Deff =3.25x10-10 Deff =5.90x10-10 Deff =8.90x10-10

177

for for for for for for for for for for for for for for for for for for for for for for for for

T=25°C, φ=0.05, v=0.5 m/s T=40°C, φ=0.05, v=0.5 m/s T=55°C, φ=0.05, v=0.5 m/s T=70°C, φ=0.05, v=0.5 m/s T=25°C, φ=0.20, v=0.5 m/s T=40°C, φ=0.20, v=0.5 m/s T=55°C, φ=0.20, v=0.5 m/s T=70°C, φ=0.20, v=0.5 m/s T=25°C, φ=0.05, v=1.0 m/s T=40°C, φ=0.05, v=1.0 m/s T=55°C, φ=0.05, v=1.0 m/s T=70°C, φ=0.05, v=1.0 m/s T=25°C, φ=0.20, v=1.0 m/s T=40°C, φ=0.20, v=1.0 m/s T=55°C, φ=0.20, v=1.0 m/s T=70°C, φ=0.20, v=1.0 m/s T=25°C, φ=0.05, v=1.5 m/s T=40°C, φ=0.05, v=1.5 m/s T=55°C, φ=0.05, v=1.5 m/s T=70°C, φ=0.05, v=1.5 m/s T=25°C, φ=0.20, v=1.5 m/s T=40°C, φ=0.20, v=1.5 m/s T=55°C, φ=0.20, v=1.5 m/s T=70°C, φ=0.20, v=1.5 m/s

Table 20. (Continued). Material

Meth&Geom

Drying Conditions

Pistachio nuts

Slope Sphere

Hot air (conditioned) mw=0.37-0.05 T=25-70°C ϕ=0.05- 0.2 v=0.5-1.5m/s

Plantain (Musa AAB)

-

Plum

Slope Slab

Hot air (conditioned) m=0.12-1.2, T=37-62°C φ=0.01, v=3.6m/s L=7.2-20mm Hot air m=0.09-9.86, T=55-65°C L=3.5 mm

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5 Deff=1.37x10-10 for T=25°C, φ=0.05, v=0.5 m/s Deff =2.26x10-10 for T=40°C, φ=0.05, v=0.5 m/s Deff =5.24x10-10 for T=55°C, φ=0.05, v=0.5 m/s Deff =7.01x10-10 for T=70°C, φ=0.05, v=0.5 m/s Deff =5.42x10-11 for T=25°C, φ=0.20, v=0.5 m/s Deff =2.29x10-10 for T=40°C, φ=0.20, v=0.5 m/s Deff =4.91x10-10 for T=55°C, φ=0.20, v=0.5 m/s Deff =8.82x10-10 for T=70°C, φ=0.20, v=0.5 m/s Deff =1.70x10-10 for T=25°C, φ=0.05, v=1.0 m/s Deff =3.26x10-10 for T=40°C, φ=0.05, v=1.0 m/s Deff =5.25x10-10 for T=55°C, φ=0.05, v=1.0 m/s Deff =8.07x10-10 for T=70°C, φ=0.05, v=1.0 m/s Deff =1.36x10-11 for T=25°C, φ=0.20, v=1.0 m/s Deff =2.98x10-10 for T=40°C, φ=0.20, v=1.0 m/s Deff =5.82x10-10 for T=55°C, φ=0.20, v=1.0 m/s Deff =8.40x10-10 for T=70°C, φ=0.20, v=1.0 m/s Deff =1.80x10-10 for T=25°C, φ=0.05, v=1.5 m/s Deff =3.48x10-10 for T=40°C, φ=0.05, v=1.5 m/s Deff =5.75x10-10 for T=55°C, φ=0.05, v=1.5 m/s Deff =9.29x10-10 for T=70°C, φ=0.05, v=1.5 m/s Deff =1.81x10-11 for T=25°C, φ=0.20, v=1.5 m/s Deff =3.25x10-10 for T=40°C, φ=0.20, v=1.5 m/s Deff =5.90x10-10 for T=55°C, φ=0.20, v=1.5 m/s Deff =8.90x10-10 for T=70°C, φ=0.20, v=1.5 m/s

Ref.

Deff =2.32x10-10 -18.01x10-10

178

Deff =3.04x10-10 for T=55°C , Deff =3.44x10-10 for T=60°C , Deff =3.69x10-10 for T=65°C

179

177

Table 20. (Continued). Material Plum (Stanley)

Pollen

Meth&Geom Slope Sphere Sphere

Pork lean

-

Potato

Regression Slab

Potato

Slope Cylinder

Prawn

Slope Cylinder

Prune

Regression Sphere

Pumpkin

Slope Slab

Drying Conditions Hot air T=60-80°C, v=1-3 m/s PT: 2% NaOH solution Fluidized bed m=0.03-0.1, T=40-45°C Hot air m=0.4-1.0, T= 10-30°C Hot air m=0.25- 2, T=40-85°C v=0.5, 1 m/s, Dims:45x 20x10 mm Hot air m=0.24-3.55, T=40-70°C v=1.5,3m/s, d=0.7-1.4cm PT: water bath Open sun m=0.1-3, Te=32.5-42.5°C ϕe=0.15-0.32 qs=460-820W/m2 Hot air (Tunnel) m=0.02-0.2, T=70-80°C v=5m/s Hot air mw=0.1-0.92, T=50-60°C v=1 m/s , L=0.7 cm

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5 -9

Deff =1.179x10 -6.671x10

Ref.

-9

180

Deff =1.197x10-7-4.551x10-7 D eff = 1.0285x10 −6 exp(−

29.69 ) RˆT

181

ab

Deff =1.1x10-11-2.0x10-11 D eff = 7.18x10 −5 exp(−

114 31580 ) exp[(−0.0025Tab + 1.22)m ] RˆT

182

ab

Deff =4.55x10-10 -10.2x10-10

140

D eff = 18.09 x10 −11 exp(−1.279m m )

142

Deff =4.32x10-10 for T=70°C , Deff =5.48x10-10 for T=70°C , Deff =7.64x10-10 for T=70°C

183

Deff=3.88x10-10 for T=50°C, Deff=6.58x10-10 for T=55°C, Deff=9.38x10-10 for T=60°C

184

Table 20. (Continued). Material

Meth&Geom

Drying Conditions

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Hot air m=0.04 -0.67, T=40-60°C v=0.8 m/s

D eff = 1.95x10 −5 exp(−

Ref.

33.15 ) RˆT ab

Open Sun m=0.05 -0.67, Te=21.6-39.7°C ϕe=12.1-51.5 qs=205.1-796.2 W/m2

Deff=1.66 x10-11

Solar Tunnel m=0.05 -0.67

Deff=1.94 x10-11

Quince

Regression Slab

Hot air (conditioned) m=0.3-4.7, T=35-55°C v=0.2-0.6m/s, ϕ=0.4-0.7 L=4mm

Deff =0.65x10-10-6.92x10-10

Quinoa seed

Regression Sphere

Hot air m=0.09-0.26, T=30-90°C

Raisin

-

M=0.15-2.40, T=60°C

Deff=5.0x10-11-2.5x10-10

133

Red bell pepper (Lamuyo)

Slope Slab

Hot air m=0.02-10.11, T=50-80°C v=2.5 m/s, Dims:1x1x1cm

Deff=3.2x10-9 for T=50°C , Deff=6.9x10-9 for T=60°C Deff=10.2x10-9 for T=70°C , Deff=11.2x10-9 for T=80°C

188

Red pepper

Slope Slab

Red pepper

Slope Cylinder

Pumpkin seed (hull-less)

Red pepper

Slope Slab

Slope Cylinder

Hot air mwo=0.91, T= 25-45°C v=4.1 m/s , l=10mm Hot air m=3.25-0.11, T=50-65°C v=0.4 m/s Rotary dryer m=1.05-3.3, T=50-65°C v=0.8 m/s

185

186

(

)

D eff = − 2.11x10 −7 m o + 7.95x10 −7 exp(−

D eff = 0.043x10 −2 exp(− D eff = 4.92x10 −3 exp(−

37.98 ) RˆT

187

ab

42.8 ) RˆTab

159

37.76 ) RˆT

189

24.76 ) RˆT

190

ab

D eff = 1.95x10 −5 exp(−

ab

Table 20. (Continued). Ref.

D eff = 2.5x10 −5 exp(−

191

Meth&Geom

Rice (cooked)

Slope Cylinder

Rice (parboiled)

-

M=0.50-0.53 T=40-100°C

Deff=2.38x10-11-4.84x10-10

129

Rose hip

Regression Sphere

Hot air m=0.07-0.49, T=50-80°C v=15 m/s

Deff =7.501x10-11 for T=50°C , Deff =15.622x10-11 for T=60°C Deff =22.639x10-11 for T=70°C , Deff =33.674x10-11 for T=80°C

192

Rough rice (long)

-

Hot air mo=0.13, T=12-50°C

Rough rice (medium)

-

Rough rice (short)

-

Drying Conditions Hot air m=0.06-1.22 T= 110–180°C v=0.5-3 m/s

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Material

Hot air mo=0.22, T=40-70°C Hot air mo=0.32,T=35-54°C

Regression Cylinder

Hot air m=0.24-1.22 T= 40–60°C v=1.5-3 m/s

Rough rice

Regression Ellipsoid

Hot air m=0.24-1.22 T= 40–60°C v=1.5 m/s

Salami

-

Soya bean (Nidera A6381)

Regression Sphere

Rough rice

Hot air m=0.4-0.45, T= 10-20°C Hot air m=0.14-0.27, T=19-75°C v=0.23m/s

36.44 ) RˆT ab

D eff = 8.24 x10 −4 exp(−

49.19 ) RˆT

193

43.56 ) RˆT

194

53.37 ) RˆT

195

4706 ) v 0.076 Tab

84

ab

D eff = 5.08x10 −4 exp(−

ab

D eff = 9.33x10 −3 exp(−

ab

D eff = 1.64v −0.85 exp(− Deff= 4.176x10-8 Deff= 4.644x10-8 Deff= 5.508x10-8 Deff= 5.976x10-8 Deff= 7.524x10-8

for T=40°C for T=45°C for T=50°C for T=55°C for T=60°C

87

Deff =0.03x10-11-0.37x10-11

114

Deff =1.78x10-11-7.28x10-11

196

Table 20. (Continued). Material

Meth&Geom

Soya bean (Nidera A5409)

Regression Sphere

Soya bean

SlopeSlab

Soya bean (Brazilian Doko)

Regression Sphere

Squid mantle

Regression Hollow cylinder

Sugar beet

-

Tomato

Slope Slab

Drying Conditions Hot air m=0.10- 0.30, T=25-70°C v=0.23m/s

Hot air m=0.10-0. 3, T=30,50°C v=0.5-1.5m/s

Hot air mw=0.11-0.82 T= 31.5–58.5 C v= 0.33–3.17 m/s Hot air m=0.148-2.74, T= 34.3°C v=1.05 m/s

(

)

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

D eff = 4.51x10 −7 + 1.9x10 −8 m o exp(−

196

27 ) RˆTab

Deff =2.38x10-10 for T=30°C and v=0.5m/s v=0.5m/s Deff =2.62x10-10 for T=50°C and v=0.5m/s v=1.0m/s Deff =2.59x10-10 for T=40°C and v=1.0m/s v=1.0m/s Deff =2.65x10-10 for T=30°C and v=1.5m/s v=1.5m/s Deff =2.92x10-10 for T=50°C and v=1.5m/s 4848.5 Deff ) exp(3.8m) = 8.64 exp(− Tab R2

, Deff =2.50x10-10

for T=40°C and

, Deff =2.43x10-10

for T=30°C and

, Deff =2.84x10-10

for T=50°C and

, Deff =2.83x10-10

for T=40°C and

197

198

D eff = 8.23x10 −11 exp(0.0646mr)

199

T=40-84°C

Deff=0.4x10-10-1.3x10-10

163

Hot Air mw=0.11-0.94, T=55-70°C v=1.5 m/s

D eff = 6.987 x10 −5 exp(−

PT: alkaline ethyl oleate solution

Tomato (organic)

Slope Slab

Solar tunnel mw=0.12 -0.93 T=22.4-35.6°C ϕ=14.5-50.9 qs=202.3-767.4W/m2

Turnip

-

m=0.31-7, T=20-100°C

32.94 ) RˆT ab

D eff = 3.284x10

−7

17.40 exp(− ) RˆT

200

ab

Deff =1.31x10-9

201

Deff=7.61×10-12-3.62×10-9

129

Table 20. (Continued). Material

Meth&Geom

Wheat (broom)

Regression Ellipsoid

Wheat (parboiled)

Slope Sphere

White mulberry

Slope Sphere

Yam (Dioscorea alata)

Yam (Dioscorea rotundata)

Yoghurt

Regression Slab

Regression Slab

Slope Slab

Drying Conditions Hot air m=0.14-0.29, T=64-75°C v=0.64m/s Hot air mw=0.1-0.45, T=40-50°C v=3.7 m/min Hot air mw=0.17-0.82, T=50°C v=1 m/s Hot air mwo=0.69, T= 50-80°C v=1.5 m/s, Dims.=50x20x10mm PT: water bath PT: sodium metabisulphite solution Hot air mwo=0.71, T= 50-80°C v=1.5 m/s , Dims.=50x20x10mm PT: water bath PT: sodium metabisulphite solution Hot air m=0.09-3.81 ,T=40-50°C v=2 m/s

[

]

Effective Diffusion Coefficient (m2/s) Dmin=4.13x10-12 Dmax=1.041x10-5

Ref.

D eff = 3.3x10 −6 + 36 x10 −6 (m o − 0.222) exp(−

202

D eff = 2 x10 −4 exp( −

32 ) RˆTab

37 ) RˆTab

203

Deff =2.231x10-10-6.909x10-10

204

Deff =1.11x10-7 for T=50°C , Deff =1.92x10-7 for T=60°C Deff =1.93x10-7 for T=70°C , Deff =4.00x10-7 for T=80°C Deff =9.21x10-8 for T=50°C Deff =1.41x10-7 for T=70°C Deff =1.53x10-8 for T=50°C Deff =2.54x10-7 for T=70°C

, , , ,

Deff =1.23x10-7 for T=60°C Deff =2.84x10-7 for T=80°C Deff =2.37x10-7 for T=60°C Deff =4.01x10-7 for T=80°C

205

Deff =1.120x10-6 for T=50°C , Deff =2.867x10-6 for T=60°C Deff =4.902x10-6 for T=70°C , Deff =8.029x10-6 for T=80°C Deff =1.386x10-6 for T=50°C Deff =5.121x10-6 for T=70°C Deff =3.392x10-6 for T=50°C Deff =9.008x10-6 for T=70°C D eff = 2.11x10 −5 exp(

26.07 ) RˆT

, , , ,

Deff =2.548x10-6 for T=60°C Deff =1.041x10-5 for T=80°C Deff =5.449x10-6 for T=60°C Deff =1.298x10-5 for T=80°C

205

206

ab

where: Deff is the effective diffusion coefficient (m2/s), m moisture content in dry basis (kg moisture/kg dry solid), mw moisture content in wet basis (kg water/kg wet substance), mm hourly mean moisture content (kg water/kg dry matter), ms sugar concentration (kg sugar/kg dry solids), T temperature (°C), Tab

ˆ absolute temperature (K), v velocity (m/s), L thickness (m), l length (m) h height (m), d diameter(m), ϕ relative humidity, P pressure ( mmHg or kPa), R 2 universal gas constant (8.314 J/(molK)), N power (W), qs solar radiation (W/m ), PT pretreatment, CRP constant rate period, FFRP first falling rate period, SFRP second falling rate period, (Subscripts: e ambient air, o initial)

151

Transport Phenomena During Drying of Food Materials Table 21. Activation energies of various food materials for mass transfer Ea (kJ/mol)

Ref.

Materials

Apple (red delicious)

19.96-22.62

123

Mullet roe

37.2

166

Avocado Banana Banana Beef meat Black tea particle Broccoli Carrot Carrot Carrot core Carrot cortex Cauliflower leave Cherry (sweet) Chestnut (Longal) Chestnut (Martainha) Chestnut (Viena) Corn Corn Date (Red Soft) Date (Tempo 2) Date (Tempo 3) Dill leave Eggplant Fig Garlic Garlic Grape (Sultanin) Grape (Chasselas) Green bean Green bean Green pea Green pepper Kale Kiwi Lettuce leave Mango Mint leave Mint leave Mulberry

39.8 13.4 15.5-25.3 -24.64 406.02 26.2 12.7-28.7 22.43 24.78 16.53 19.82 53 22.578 21.926 20.458 29.56-30.56 27.61 35.17 29.50 44.02 35.05 29.52 30.81-48.47 23.48 27.84 54 49 35.43 39.47 24.7-28.40 51.4 36.12 27.00 19.82 22.95-27.68 82.93 62.96 21.2

207 127 208 132 135 209 207 139 140 140 210 143 145 145 145 148 211 149 149 149 150 151 153 154 155 156 156 158 212 213 159 160 161 210 214 215 164 165

Mushroom (Ag. Bisp.) Mushroom (Pl. florida) Olive cake Olive oil extraction Olive oil extraction Okra Orange skin Parsley leave Pear Pistachio nuts Plantain Pollen Potato Potato Prune Pumpkin Pumpkin seed (hulles) Quince Quinoa seed Red bell pepper Red pepper Red pepper Red pepper Rice (cooked) Rose hip Rough rice (long) Rough rice (medium) Rough rice (short) Soya bean (Nid. A6381) Soya bean (Nid. A5409) Sugar beet Tomato Wheat (broom) Wheat (parboiled) Yam (Dios. alata) Yam (Dios. rotundata) Yoghurt

19.79 23.59 17.97 15.77 26.71 51.26 36.36 43.92 24.3 30.79 38.81 29.69 16.3-108 31.5 57.00 78.93 33.15 33.83-41.52 32.50-40.99 39.70 42.80 37.76 24.76 36.44 46.00 49.19 43.56 53.37 28.80 27.00 28.8 32.94 32.00 37.13 41.71-69.45 25.26-46.46 26.07

210 210 168 169 170 171 172 150 176 177 178 181 207 216 183 184 185 186 187 188 159 189 190 191 192 193 194 195 197 197 217 201 203 204 206 206 207

Materials

Ea (kJ/mol)

Ref.

152

Kamil Kahveci and Ahmet Cihan

NOMENCLATURE a aw A A b bK Bk Bim Biq c cp C ˆ C

constant water activity Helmholtz free energy [J] area (m2) constant Klinkenberg parameter (Pa) viscous flow parameter [m2] Biot number for mass transfer Biot number for heat transfer constant air capacity [kgm2/(kgN)] concentration [kg/m3] molar concentration [mol/m3]

Cm CP d d D DR d e es Ey Ea Ec f f 〈f 〉

mass capacity [kg/(kg°M)] specific heat [J/(kgK)] diameter [m] constant Diffusion coefficient [m2/s] shrinkage dimension [volume, area, thickness] constant constant standard deviation Young modulus [Pa] activation energy [J/mol] characteristic energy [J] constant function spatial average of function f

〈f s 〉

〈f s 〉

F Fmi

phase average of a function fs, which represents a property of the s phase s

intrinsic phase average of a function fs, which represents a property of the s phase force [N] shape factor [1/m2]

Fo g gˆ

Fourier number gravitational acceleration [m2/s] molar Gibbs function [J/mol]

G G Gr Gr Gu h h hm hn hP hq H I J Jm Jm,vol

Gibbs free energy [J] slip modulus [m] relaxation model [Pa]. Grashof number Gukhman number capillary pressure head [m] specific enthalpy [J/kg] mass transfer coefficient [kg/(m2s) or m/s] mass transfer coefficient mass transfer coefficient [kg/(m2sPa)] heat transfer coefficient [W/(m2K)] enthalpy [J] the volumetric capacity [kg/(m3s)] transfer flux [[…]/(m2s)] moisture flux density [kg/(m2s)] volumetric flux density [m3/(m2s)]

Transport Phenomena During Drying of Food Materials Jˆ m

molecular flux density [mol/(m2s)]

Jn(x) Jq k kr K Ki Ko KH Kl Kn KV l L L Le M ˆ M

Bessel function of the first kind of order n heat flux density [J/(m2s)] constant relative permeability permeability [m2] intrinsic permeability [m2] Knudsen flow parameter [m] hydraulic conductivity [m/s] liquid conductivity [kg/(ms)] Knudsen number volume model [Pa] half length [m]

m mw ) m mr r n ) n N Nu P Pn(x) Pe Pr qvol q&

moisture content in dry basis [kg moisture/kg dry solid] moisture content in wet basis [kg water/kg wet substance] mass fraction

the number of moles Nusselt number pressure [Pa] Legendre polynomial of the first kind of order n Peclet number Prandtl number heat of desorption or absorption [J/m3] volumetric heating [W/m3]

Q Qdif Qvol r r r Rˆ

heat energy [J] diffusibility volumetric flow rate [m3/s] radius [m] radial coordinate [m] correlation coefficient in Eq. (463) universal gas constant [8.314 J/(molK)]

Re s sˆ

Reynolds number Laplace transformation parameter molar entropy [J/(molK)]

S S S Sb Sc Sh Stm Stq t T u u

entropy [J/K] saturation degree Surface area [m2] relative volumetric shrinkage [V/Vo] Schmidt number Sherwood number Stanton number for mass transfer Stanton number for heat transfer time [s] temperature [°C or K] velocity [m/s] Moisture content by volume [m3 liquid/m3 solid]

characteristic length (m) phenomenological transport coefficient [[…]2/s] Lewis number mass [kg] molecular mass [kg/mol]

moisture ratio unit normal vector molar fraction

153

154

Kamil Kahveci and Ahmet Cihan U Uq v V w wp W x y y z ~ 1

internal energy [J] overall heat transfer coefficient [W/(m2K)] velocity [m/s] volume [m3] velocity [m/s] dimensionless probability factor mass-transfer potential or degrees of moistness [°M] coordinate [m] coordinate [m] scalar quantity in Eq. (443) coordinate [m] unit tensor

Greek Letters α α αT β

βn

βv βT δ δm χ χ2 δP δT Δhvap Δhˆ ε ε εav εpc φ φ γ

Γ η ϕ Φ λ λ λ′ p

Λ μ μˆ

μ′

ν νp

vap

thermal diffusivity [m2/s] root of Bessel function in Eqs. (212)-(218) thermal expansion coefficient [1/K] shrinkage coefficient root of cosine function in Eqs. (217) and (218) thermal expansion coefficient [1/K] volumic expansion coefficient [1/K] Kronecker delta in Eqs. (264)-(266) moisture conductivity coefficient [kg/mh°M] constituent concentration (kg/kg dry matter) mean square deviation moisture filtration coefficient [kgm/(sN)] thermo-gradient coefficient [1/°C] evaporation enthalpy [J/kg] molar evaporation enthalpy [J/mol] volume fraction strain ratio of air and vapor diffusion coefficient phase change criterion dissipation function [J/(m3s)] zenith angle [rad] in Eq. (201) shear rate [1/s] in Eq. (16) activity coefficient in Eq. (187) pore shape factor in Eqs. (102) and (103) Boundary surface [m2] dynamic viscosity [Pa s] relative humidity thermodynamic force [J/(kg[…])] thermal conductivity [W/(mK)] mean free path length [m] Lame coefficient [Pa] eigenvalue chemical potential [J/kg] chemical potential [J/mol] Lame coefficient [Pa] kinematic viscosity [m2/s] Poisson coefficient

Transport Phenomena During Drying of Food Materials θ θ ρ σ σ τ τ

σb

contact angle [rad] azimuth angle [rad] in Eq. (200) and (201) density or concentration [kg/m3] surface tension [N/m] stress [Pa] Boltzmann constant [1.3805x10-23J/K]

υˆ

tortuosity shear stress [Pa] in Eq. (16) molar volume [m3/mol]

ω ΩD Ω ξ ξ ψ Ψg ΨH ζ

humidity ratio [kg moisture/kg dry air] diffusion collision integral eigenfunction in Eqs. (229)-(221) association factor solid based coordinate [m] porosity gravity potential [m2/s2] capillary potential [m] diffusion resistance factor

Subscripts a a ab app av c c cp cw c db dp ds e e eff ex exp f fw F h irr g g k

air ash absolute apparent average capillary cardioid closed pore cell wall material carbohydrate dry bulb dew point dry solid equilibrium epitrochoide effective excess experimental fat free water moving evaporation front haxagon irreducible gas corrugated Knudsen

l lg LM m max ml o

liquid liquid-gas log-mean moist maximum monolayer standard

155

156

Kamil Kahveci and Ahmet Cihan Subscripts (Continued) o op p P p pre R s s sat s.a. sg sg sl sn sp sp sorb st S the v vap w w wb *

initial open pore protein particle pore predicted reference solid substance saturated symmetry axis solid-gas sugar solid-liquid solution specific single pore adsorbed water starch surface theoretical vapor evaporation water wet basis wet bulb dimensionless

Superscripts * o T

dimensionless standard transpose

Overlines ∼ ^

average tensor molar fraction

REFERENCES [1] [2] [3]

Geankoplis, C. J. Transport Processes and Unit Operations; 3rd Edition; Prentice –Hall International Inc.: New Jersey, 1993. Earle, R. L.; Earle M. D. Unit Operations in Food Processing; 2nd edition; Pergamon Press: Oxford, 1983. Sheridan, P.; Shilton, N. J. Food Eng. 1999, 41, 203-208.

Transport Phenomena During Drying of Food Materials [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]

[24] [25] [26] [27] [28] [29] [30]

157

Wijesinghe, B. (1997). Low temperature drying of food materials using energy-efficient heat pump dryers [Newsletter]. www.nwfpa.org . Barr-Rosin Ltd. (2007). Superheated steam drying – environmental and energy saving drying technology. www.barr-rosin.com. Saravacos, G. D. In Engineering and Food for the 21st Century; Jorge, W.-C.; Gustavo V. B.-C.; José, M. A.; CRC Press: New York, 2002. Dullien, F. A. L. Porous media: Fluid Transport and Pore Structure; Academic Press Inc.: New York, 1992. Datta, A. K. J. Food Eng. 2007, 80, 80-95. Bruin, S.; Luyben, K. Ch. A. M. In Advances in Drying; Mujumdar, A. S.; Hemisphere Publishing Corporation: New York, 1980; Vol. 1, pp. 155-216. Yamsaengsung, R.; Moreira, R. G. J. Food Eng. 2002, 53, 1–10. Weast, R. C.; Lide, D. R.; Astle, M. J.; Beyer, W. H. Handbook of Chemistry and Physics, 70th Edition; CRC Press Inc.: Boca Raton, FL, 1989. Atkins, P. W. Physical Chemistry; 5th Edition; Oxford University Press, 1994. Janz, M. (1997) Methods of Measuring the Moisture Diffusivity at High Moisture Levels. Lund Institute of Technology, Licentiate Thesis, Sweden. Baierlein, R. Thermal Physics; Cambridge University Press, 2003. Keey, R.B. Drying Principles and Practise; Oxford:Pergamon Press, 1972. Luikov, A. V. Heat and Mass Transfer in Capillary-Porous Bodies; Chapter 6, Pergamon Pres: Oxford, 1966. Karel, M. F.; Lund D. B. Physical Preservation of fruits and vegetables by drying. UNIDO, Paper No: 29, pp. 1-9. Marinos-Kouris, D.; Maroulis, Z. B. In Handbook of Industrial Drying; Mujumdar, A.; CRC: New York, 2006; Vol. 1, pp. 82-114. Rahman, S. Food Properties Handbook; CRC: 1995. Iglesias, H.A.; Chirife, J. J. Food Sci. 1976, 41, 984-992. Fortes, M.; Okos, M. R. In Advances in Drying; Mujumdar, A. S.; Taylor&Francis: 1980; Vol.1, pp 119-154. Boguet, R; Chirife, J.; Iglesias, H. A. J. Food. Tech. 1978, 13, 319-327. Abdula, M. (1993) Calculation of water sorption isotherms for multicomponent mixtures of protein-containing products. Ph.D Thesis. Technical University of Lodz. Lodz. Poland. Urbicain, M. J.; Lozano, J. E. In Handbook of Food Engineering Practice; Valentas, K. J.; Rotstein, E.; Singh, R. P.; CRC: New York, 1997, Chapter 11. Boukouvalas, Ch. J.; Krokida, M. K.; Maroulis, Z. B.; Marinos-Kouris, D. International J. Food Prop. 2006, 9, 715–746. Sweat, V. E. In Engineering Properties of Food; Rao, M. A.; Rizvi, S. S. H.; 2nd ed.; Marcel Dekker: New York, 1995, pp. 99-138. Choi, Y; Okos, M.R. In Food Engineering and Process Applications; LaMaguer, M; Jelen, P.; Elsevier: New York, 1986; Vol. 1. USDA Handbook 8, http://www.ars.usda.gov/2007. Leninger, H. A.; Baverloo, W. A. Food Process Engineering; 1st ed.; Springer: 1975. Charm, S. E. The Fundamentals of Food Engineering; 3rd ed. AVI Publishing: Westport, CT, 1978.

158

Kamil Kahveci and Ahmet Cihan

[31] Heldman, D. R.; Singh, R. P. Food Process Engineering; 2nd ed.; AVI: New York; 1981. [32] Sweat, V. E. J. Food Sci. 1974, 39, 1080-1083. [33] Sweat, V. E. ASAE Trans. 1975, 18(3), 564-568. [34] Rahman, M. S. J. Food Eng. 1991. 15(4), 261-268. [35] Riedel, L. Chem. Ing.-Tech. 1949, 21, 340-341. [36] Hopmans, J. W.; Simunek, J. I.; Romano, N.; Durner, W. In Methods of Soil Analysis, Dane J. H.; Topp, G. C.; Soil Science Society of America Book Series No. 5, 2002, 963-1008. [37] Luikov, A. V. In Advances in Heat Transfer; Irvine, T. F.;Hartnett, J. P.; Academic Press: New York, 1964, Vol. 1, pp. 123-184. [38] Rahman, M. S. Drying Technology. 2001, 19(1), 1–13. [39] Alnajım, A. (2004) Modelisation Et Simulation Du Comportement Du Beton Sous Hautes Temperatures Par Une Approche Thermo-Hygro-Mecanique Couplee. Application A Des Situations Accidentelles. Ph.D Thesis. Universite De Marne La Vallee, Annee, France. [40] Datta, A. K. Biological and Bioenvironmental Heat and Mass Transfer; Marcel Dekker: New York, 2002. [41] Kroes, B. (1999) The Influence of Material Properties on Drying Kinetics, Ph.D dissertation;Technische Universiteit Eindhoven, Holland, 1999. [42] Whitaker, S.; Chou, W. T. H. Drying Tech. 1983, 1(1), 3-33. [43] Lowell, S.; Shields,J. E.; Thomas, M. A.; Thommes, M. Characterization of Porous Solids and Powders: Surface Area, Pore Size and Density; Springer, 2006. [44] Bassingthwaighte J. B. (2005). Osmosis and filtration: passive solute and water transport, http://bioeng.washington.edu/Courses/BIOEN589/06/handout/06osm.pdf. [45] J. Welti-Chanes, H. Mújica-Paz, A. Valdez-Fragoso, and R. León-Cruz. In Transport Phenomena in Food Processing; Edited by Jorge Welti-Chanes, Jorge F. Vélez-Ruiz, Gustavo V. Barbosa-Cánovas 2003 by CRC Press. [46] Mayor, L.; Sereno A. M. J. Food Eng. 2004, 61, 373–386. [47] Baggio P.; Bonacina C.; Schrefler B. A. Transp. Porous Med.. 1997, 28, 233-251. [48] Cunningham, R. E.; Williams, R. J. J. Diffusion in Gases and Porous Media; Plenum Press: New York, 1980. [49] Carman P. C. Flow of gases through porous media, Butterworths Scientific Publications 1956 [50] Reinecke, S. A. (2000) The effect of Water Content on Knudsen Diffusion in Unconsolidated Porous Media; MSc Dissertation, Toronto University, Toronto, Canada. [51] Kerkhof, P. J. A. M.; Geboers, M. A. M. Chem. Eng. Sci. 2005, 60, 3129-3167. [52] Gilliland, E. R. Indust. and Eng. Chem. 1934, 26(1), 681-685. [53] Fuller, E. N.; Schettler, P. D.; Giddings, J. C. Indust. And Eng. Chem. 1966, 58(19). [54] Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Properties of Gases and Liquids; Fifth Edition; The McGraw-Hill Companies: 2004; All Chapter 11. [55] Hirschfelder, J. O.; Bird, R. B.; Spotz, E. L. Chemical Reviews. 1949, 44(205). [56] Gosse, J. In Les Techniques de l'Ingénieur, Constantes Physico-Chimiques, K. 2005, 425. [57] Sercombe, J.; Vidal, R.; Gallé, C.; Adenot F. Cement and Conc. Res. 2007, 37(4), 579588.

Transport Phenomena During Drying of Food Materials

159

[58] Rovnanikova, P.; Cerny, R. Transport Processes in Concrete;Taylor&Francis: 2002. [59] Krischer, O.; Kast, W. Die Wissenschaftlichen Grundlagen der Trocknungstechnik; Erster Band, dritte Auflage; Springer: Berlin, 1992. [60] Henry, P. S. H. Proc. R. Soc. 1939, 171A, 215-241. [61] Hines, A. L.; Maddox, R. N. Mass Transfer; Upper Saddle River: Prentice-Hall, 1985. [62] Jacobson, A. (2006) Diffusion of Chemicals into Green Wood. Ph.D Thesis. School of Chemical and Biomolecular Engineering Georgia Institute of Technology. [63] Perry, R. H.; Green, D.W. Perry's Chemical Engineers' Handbook; 7th ed.; McGrawHill: 1997; 2581. [64] Vignes, A. Indust. and Eng. Chem. Funda. 1966, 5(2), 189-199. [65] Leffler, J.; Cullinan, H. T. Indust. and Eng. Chem. Funda. 1970, 9(1), 84-88. [66] Buckinghan, E. A. Studies on the Movement of Soil Moisture. U. S. Dept. Agr. Bull. 1907, 38. [67] Hill, T. L. J. Chem. Phys. 1956, 25:730. [68] Imakoma, H.; Okazaki, M.; Toei, R. In Advances in Drying; Mujumdar, A. S.; Hemisphere Publishing Corporation: New York, 1992; Vol. 5, pp. 77-108. [69] Philip, J. R.; de Vries, D. A. Trans. Am. Geophys. Union. 1957, 38(2), 222-232. [70] Bear, J. Dynamics of Fluids in Porous Media; American Elsevier Pub. Co.: 1972. [71] Whitaker, S. Advances in Heat Transfer. 1977, 13, 119-203. [72] Whitaker, S. In Heat and Mass Transfer in Granular Porous Media: Advances in Drying 1; Mujumdar, A. S.; Hemisphere Publ. Corp.: Washington DC, 1980; pp 23-61. [73] Prat, M. Chem. Eng. J. 2002, 86(1-2), 153-164. [74] Crank, J. The Mathematics of Diffusion; Clarendon Pres: Oxford, 1975; Vol. 49, pp. 4749, 90-91. [75] Sherwood T. K. Trans. AIChE. 1936, 32, 150-168. [76] Muthukumarappan, K.; Gunasekaran, S. Tr. ASAE. 1990, 33(5), 1637-1641. [77] Singh, B.; Gupta, A. K. J. Food Eng. 2007, 79, 459–470. [78] Fan, C. C.; Liaw, S. P.; Fu, W. R.; Pan, B. S. J. Food Sc. 2003, 68(3), 886-891. [79] Reyes, A.; Alvarez, P. I.; Marquardt, F. H. Drying Technology. 2002, 20(7), 1463– 1483. [80] El-Beltagy, A.; Gamea, G. R.; Amer Essa, A. H. J. Food Eng. 2007, 78, 456–464. [81] Bon, J.; Simal, S.; Rosselld, H.C.; Mulet, A. J. Food Eng. 1997, 34, 109-122. [82] Mulet, A.; Sanjuán, N.; Bon, J.; Simal, S. Eur Food Res Tech. 1999, 210, 80-83. [83] Jain, D.; Pathare, P. B. J. Food Eng. 2007, 78, 1315–1319. [84] Ece, M. C.; Cihan, A. Tr. ASAE. 1993, 36(3), 837-840. [85] Kahveci, K.; Cihan, A.; Ece, M. C. Intern. J. Energy Res. 2003, 27, 1133-1145. [86] Lima, A. G. B.; Farias, S. N.; Lima, D. R. Brazilian Journal of Chemical Engineering. 2004, 21(4), 667-680. [87] Hacıhafızoğlu, O.; Cihan, A.; Kahveci, K.; De Lima, A. G. B. Eur Food Res Tech. 2007, (in Press). [88] Rotstein, E.; Laura, P. A.; de Cemborain, E. J. Food Science. 1974, 39, 627-631. [89] Chemkhi, S.; Zagrouba, F.; Bellagi, A. Drying Tech. 2004, 22(5), 1023–1039. [90] Mihoubi, D.; Zagrouba, F.; Vaxelaire, J.; Bellagi, A.; Roques, M. Drying Tech. 2004, 22(1-2), 91–109. [91] Ketelaars, A. A. J. (1992) Drying deformable media; kinetics, shrinkage and stresses. Ph.D Thesis. Eindhoven University of Technology, The Netherlands.

160

Kamil Kahveci and Ahmet Cihan

[92] Chen, P.; Schmidt, P. S. Drying ’89. 1990, 162-168. [93] Vu, T. H. (2006) Influence of Pore Size Distribution on Drying Behaviour of Porous Media by a Continuous Model. Ph.D Thesis. Otto-von-Guericke-Universität Magdeburg. [94] Irudayaraj, J.; Haghighi, K.; Stroshine, R. L. J. Agricultural Eng. Res. 1992, 534, 209229. [95] Lewis, R. W.; Ferguson, W. J. Int. J. Num. Meth. in Eng. 1990, 29, 357-369. [96] Luikov, A.V.; Mikhailov, Yu. A. Theory of Energy and Mass Transfer; Chaps.1-2; Pergamon Press: Oxford, 1965. [97] Mikhailov, M. D.; Ozisik, M. N. Unified Analysis and Solutions of Heat and Mass Diffusion; Chap.10; John Wiley: New York, 1984. [98] Liu, J. Y.; Cheng, S. Int. J. Heat Mass Transfer. 1991, 34(7), 1747 1754. [99] Chang, W. J.; Weng, C. I. J. Heat Mass Transfer. 2000, 43, 3621-3632. [100] Berger, D.; Pei, D. C. T. J. Heat Mass Transfer. 1973, 16(2), 293-302. [101] Whitaker, S. The Method of Volume Averaging; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1999. [102] Data, A. K. J. Food Eng. 2007, 80, 96–110. [103] Parry, J. L. J. Agricultural Eng. Res. 1985, 32, 1–29. [104] Fortes, M.; Okos, M. R. Tr. ASAE. 1981, 22, 761-769. [105] Lewis, W. K. J. Industrial and Eng. Chem. 1921, 13(5), 427–432. [106] Page, C. (1949) Factors influencing the maximum rates of air drying of shelled corn in thin layers. MSc Thesis(unpublished), Purdue University, West Lafayette, IN 47907 USA. [107] Henderson, S. M.; Pabis, S. J. f Agricultural Eng. Res. 1969, 6(3), 169–174. [108] Madamba, P. S.; Driscoll, R. H.; Buckle, K. A. J. Food Eng. 1996, 29, 75–97. [109] Sharma, A. D.; Kunze, O. R.; Tolley, H. D. Tr. ASAE. 1982, 25, 221–224. [110] Thompson, T. L.; Peartt, R. M.; Foster, G. H. Tr. ASAE. 1968, 11(4), 582–589. [111] Wang, C. Y.; Singh, R. P. ASAE. 1978, Paper No. 78-6505, St. Joseph, MI, USA. [112] Chapra S. C.; Canale R. P. Numerical Methods for Engineers; McGraw-Hill Book Company: New York, 1989; p 337. [113] Mulet, A.; Tarrazo, J.; Garcia-Reverter, J.; Berna, A. 1997. Shrinkage of cauliflower florets and stems during drying. In Engineering and Food at ICEF 7; Ed. R. Jowitt; Sheffield Academic Press: Sheffield, UK, 1997; pp 97–100. [114] Beringues, J. C. (1999) Sorption isotherms and water diffusivity in Muscles of pork ham at different NaCl contents. Ph.D Thesis. UPC, Barcelona. [115] Shoeber, W. J. A. H. (1976) Regular regimes in sorption processes. Ph.D Thesis. Eindhoven, The Netherlands (Quoted by Luyben et al., 1980). [116] Schoeber, W. J. A short cut method for the calculation of drying rates in case of a concentration-dependent diffusion coefficient. Proceeding of First International Symposium on Drying. Montreal, 1978. [117] Villota, R.; Hawkes, J. G. In Handbook of Food Engineering; Eds. Heldman, D. R.; Lund, D. B.; Marcel Dekker, Inc.: New York, USA, 1992; pp 39-144. [118] Katekawa, M. E.; Silva, M. A. Drying Tech. 2006, 24, 5–20. [119] Giri, S. K.; Prasad, S. International Journal of Food Properties. 2006, 9(3), 409–419. [120] Fish, B. P. In Fundamental Aspects of Dehydration of Foodstuffs; Society of Chemical Industry: London, Macmillan Co, New York, 1958; pp 143–157.

Transport Phenomena During Drying of Food Materials

161

[121] Gekas, V. C.; Motarjemi, Y.; Lamberg, I.; Hallström, B. In Preconcentration and Drying of Food Materials; Ed. Bruin, S.; Elsevier Science Publishers BV: Amsterdam, The Netherlands, 1988; pp 317-319. [122] Gekas, V. C.; Lamberg, I. J. Food Eng. 1991, 14, 317–326. [123] Teixeira, M. B. F.; Tobinaga, S. J. Food Eng. 1998, 36, 169-181. [124] Sacilik, K.; Elicin, A. K. J. Food Eng. 2006, 73, 281–289. [125] Wang, Z.; Sun, J.; Chen, F.; Liao, X.; Hu, X. J. Food Eng. 2007, 80, 536–544. [126] Togrul, I. T.; Pehlivan, D. J. Food Eng. 2003, 58, 23–32. [127] Baini, R.; Langrish, T. A. G. J. Food Eng. 2007, 79, 330–343. [128] Karim, Md. A.; Hawlader, M. N. A. J. Food Eng. 2005, 70, 35–45. [129] Saravacos, G. D.; Maroulis, Z. B. Transport Properties of Foods; 1st edition; CRC: New York, 2001. [130] Akpinar, E. K. J. Food Eng. 2006, 77, 864–870. [131] Gurtas, F. S.; Ak, M. M.; Evranuz, E. Ö. Turk J. Agric. For. 2001, 25, 297-304. [132] Trujillo, F. J.; Wiangkaew, C.; Pham, Q. T. J. Food Eng. 2007, 78, 74–85. [133] Mujumdar, A. S.; Devahastin, S. Practical Guide to Industrial Drying; Ed. S. Devahastin, (Exergex, Montreal, 2000). www.geocities.com/drying_guru/chapter1fundamentals.pdf [134] Doymaz, I. J. Food Eng. 2006, 76, 212–217. [135] Panchariya, P. C.; Popovic, D.; Sharma, A. L. J. Food Eng. 2006, 76, 212–217. [136] Mrkic´, V.; Ukrainczyk, M.; Tripalo, B. J. Food Eng. 2007, 79, 640–646. [137] Doymaz, I. J. Food Eng. 2004, 61, 359–364. [138] Singh, B.; Gupta, A. K. J. Food Eng. 2007, 79, 459–470. [139] Togrul, H. J. Food Eng. 2006, 77, 610–619. [140] Srikiatden, J.; Roberts, J. S. J. Food Eng. 2006, 74, 143–152. [141] Bialobrzewski, I.; Markowski, M. Drying Tech. 2004, 22(7), 1777–1789. [142] Jain, D.; Pathare, P. B. J. Food Eng. 2007, 78, 1315–1319. [143] Alejandra Mabellini; Mabel B. Vullioud; Carlos A. Márquez; Antonio De Michelis. (2005). http://www.enpromer2005.eq.ufrj.br/nukleo/pdfs/0287_drying_kinetics_of_sweet_cherries. pdf [144] Doymaz, I. J. Food Eng. 2007, 78, 591–596. [145] Guine´, R. P. F.; Fernandes, R. M. C. J. Food Eng. 2006, 76, 460–467. [146] Jena, S.; Das, H. J. Food Eng. 2007, 79, 92–99. [147] Sfredo, M. A.; Finzer, J. R. D.; Limaverde, J. R. J. Food Eng. 2005, 70, 15–25. [148] Doymaz, I.; Pala, M. J. Food Eng. 2003, 60, 125–130. [149] Falade, K. O.; Abbo, E. S. J. Food Eng. 2007, 79, 724–730. [150] Doymaz, I.; Tugrul, N.; Pala, M. J. Food Eng. 2006, 77, 559–565. [151] Wu, L.; Orikasa, T.; Ogawa, Y.; Tagawa, A. J. Food Eng. 2007 (in press). [152] Doymaz, I. J. Food Eng. 2005, 71, 403–407. [153] Babalis, S. J.; Belessiotis, V. G. J. Food Eng. 2004, 65, 449–458. [154] Sacilik, K.; Unal, G. Biosystems Eng. 2005, 92(2), 207–215. [155] Pezzutti, A.; Crapiste, G. H. J. Food Eng. 1997, 31, 113-123. [156] Azzouz, S.; Guizani, A.; Jomaa, W.; Belghith, A. J. Food Eng. 2002, 55, 323–330. [157] Margaris, D. P.; Ghiaus, A. G. J. Food Eng. 2007, 79, 1115–1121. [158] Doymaz, I. J. Food Eng. 2005, 69, 161–165.

162

Kamil Kahveci and Ahmet Cihan

[159] Ertekin, F. K. J. Food Sc.. 2002, 67(1), 168-175. [160] Mwithiga, G.; Olwal, J. O. J. Food Eng. 2005, 71, 373–378. [161] Simal, S.; Femenia, A.; Garau, M. C.; Rossell, C. J. Food Eng. 2005, 66, 323–328. [162] Goyal, R. K.; Kingsly, A. R. P.; Manikantan, M. R.; Ilyas, S. M. Biosystems Eng. 2006, 95(1), 43–49. [163] Doulia, D.; Tzia, K.; Gekas V. (1999). A knowledge Base for the Apparent Mass Diffusion Coefficient (Deff ) of Foods. Available: www.nelfood.com [164] Doymaz, I. J. Food Eng. 2006, 74, 370–375. [165] Maskan, M.; Gogus, F.; J. Food Eng. 1998, 37, 437-449. [166] Fan, C. C.; Liaw, S. P.; Fu, W. R.; Pan, B. S. J. Food Sc. 2003, 68(3), 886-891. [167] Giraldo-Zuniga1, A. D.; Arévalo-Pinedo, A.; Rezende, J. R.; Silva, C. S.; Monteiro, J. A.; J. Food Proces. Preserv. 2006, 30(6), 699–705. [168] Akgun, N. A.; Doymaz, I. J. Food Eng. 2005, 68, 455–461. [169] Celma, A. R.; Rojas, S.; Lo´pez, F.; Montero, I.; Miranda, T. J. Food Eng. 2007, 80, 1261–1271. [170] Doymaz, I.; Gorel, O.; Akgun, N. A. Biosystems Eng. 2004, 88(2), 213–219. [171] Doymaz, I. J. Food Eng. 2005, 69, 275–279. [172] Garau, M. C.; Simal, S.; Femenia, A.; Rossello, C. J. Food Eng. 2006, 75, 288–295. [173] El-Aouar, A. A.; Azoubel, P. M.; Murr, F. E. X. J. Food Eng. 2003, 59, 85–91. [174] De Temmerman, J.; Verboven, P.; Nicolai, B.; Ramon, H. J. Food Eng. 2007, 80, 892– 903. [175] Kingsly, R. P.; Goyal, R. K.; Manikantan, M. R.; Ilyas, S. M. Int. J. Food Sc. Tech. 2007, 42, 65–69. [176] Guine´, R. P. F.; Rodrigues, A. E.; Figueiredo, M. M. Ap. Math. Comp. 2007 (in press). [177] Kashaninejad, M.; Mortazavi, A.; Safekordi, A.; Tabil, L. G. J. Food Eng. 2007, 78, 98–108. [178] Johnson, P. N. T.; Brennan, J. G.; Addo-Yobo, F. Y. J. Food Eng. 1998, 37(2), 233242. [179] Goyal, R. K.; Kingsly, A. R. P.; Manikantan, M. R.; Ilyas, S. M. J. Food Eng. 2007, 79, 176–180. [180] Menges, H. O.; Ertekin, C. Energy Conv. and Manag. 2006, 47, 2337–2348. [181] Vizcarra-Mendoza, M. G.; Martínez-Vera, C.; Ruiz-Martínez, R. S. 15th International Drying Symposium (IDS 2006). 2006, pp 504-507, Budapest, Hungary. [182] Hassini, L.; Azzouz, S.; Peczalski, R.; Belghith, A. J. Food Eng. 2007, 79, 47–56. [183] Sabarez, H. T.; Price, W. E. J. Food Eng. 1999, 42, 167–172. [184] Doymaz, I. J. Food Eng. 2007, 79, 243–248. [185] Sacilik, K. J. Food Eng. 2007, 79, 23–30. [186] Kaya, A.; Aydin, O.; Demirtas, C.; Akgün, M. Desalination. 2007, 212, 328–343. [187] Gely, M. C.; Santalla, E. M. J. Food Eng. 2007, 78, 1029–1033. [188] Vega, A.; Fito, P.; Andre´s, A.; Lemus, R. J. Food Eng. 2007, 79, 1460–1466. [189] Kaleemullah, S.; Kailappan, R. J. Food Eng. 2006, 76, 531–537. [190] Kaleemullah, S.; Kailappan, R. Biosystems Eng. 2005, 92(1), 15–23. [191] Ramesh, M. N. Lebensm-Wiss u-Technol. 2003, 36, 245–255. [192] Marquez, C. A.; De Michelis, A.; Giner, S. A. J. Food Eng. 2006, 77, 566–574. [193] Lu, R.; Siebenmorgen, T. J. Tr. ASAE. 1992, 35(6), 1955–1961. [194] Aguerre, R.; Suarez, C.; Viollaz, P. E. J. Food Tech. 1982, 17(6), 679–686.

Transport Phenomena During Drying of Food Materials

163

[195] Steffe, J. F.; Singh, R. P. J. Agricultural Eng. Res. 1982, 27, 489–493. [196] Gely, M. C.; Giner, S. A. Biosystems Eng. 2007, 96(2), 213–222. [197] Akosman, C.; Kalender, M. F. Ü. Fen ve Müh. Bil. D. 2004, 16(2), 243-251. [198] Barrozoa, M.A.S.; Henriquea, H. M.; Sartorib, D. J. M.; Freireb, J. T. J. Stored Products Res. 2006, 42, 348–356. [199] Teixeira, M. B. F.; Tobinaga, S. J. Food Eng. 1998, 36, 169-181. [200] Doymaz, I. J. Food Eng. 2007, 78, 1291–1297. [201] Sacilik, K.; Keskin, R.; Elicin, A. K. J. Food Eng. 2006, 73, 231–238. [202] Gaston, A. L.; Abalone, R. M.; Giner, S. A.; Bruce, D. M. Biosystems Eng. 2004, 88(22), 175–185. [203] Mohapatra, D.; Rao, P. S. J. Food Eng. 2005, 66, 513–518. [204] Doymaz, I. J. Food Eng. 2004, 61, 341–346. [205] Falade, K. O.; Olurin, T. O.; Ike, E. A.; Aworh, O. C. J. Food Eng. 2007, 80, 1002– 1010. [206] Hayaloglu, A. A.; Karabulut, I.; Alpaslan, M.; Kelbaliyev, G. J. Food Eng. 2007, 78, 109–117. [207] Zogzas, N. P; Maroulis, Z. B.; Marinos-Kouris, D. Drying Tech. 1996, 14(10), 2225– 2253. [208] Nogueira, R.; Park, K. J. Drying Parameters to Obtain Bananapassa. In Proceedings of the 8th International Drying Symposium, Montreal, Canada, Part A, 1992, 874–883. [209] Simal, S.; Rossello, C.; Berna, A.; Mulet, A. J. Food Eng.1998, 37, 423–435. [210] Lopez, A.; Iguaz, A.; Esnoz, A.; Virseda, P. Drying Tech. 2000, 18(4–5), 995–1006. [211] Tolaba, M.; Suarez, C. Lebensmittel-Wissenschaft und-Technologie. 1988, 21, 83–86. [212] Senadeera, W.; Bhandari, B. R.; Young, G.; Wijesinghe, B. J. Food Eng. 2003, 58, 277–283. [213] Simal, S.; Mulet, A.; Tarrazo, J.; Rosello, C. Food Chem. 1996, 55, 121–128. [214] Ruiz-Lopez, I. I.; Garcia-Alvarado, M. A. J. Food Eng. 2007, 79, 208–216. [215] Park, K. J.; Vohnikova, Z.; Brod, F. P. R. J. Food Eng. 2002, 51, 193–199. [216] Simal, S.; Rossello , C.; Berna, A.; Mulet, A. Chem. Eng. Sc. 1994, 49(22), 3739–3744. [217] Vaccarezza, L.M.; Lombardi, J. L.; Chirife, J. J. Food Technol. 1974, 9, 317-327.

In: Focus on Food Engineering Research and Developments ISBN: 978-1-60021-898-9 Editor: Vivian N. Pletney, pp. 165-237 © 2007 Nova Science Publishers, Inc.

Chapter 2

THE INFLUENCE OF INTERACTIONS OCCURRING BETWEEN MICRO-ORGANISMS ON PREDICTING THE SAFETY OF LACTIC ACID CHEESE Izabela Steinka* Gdynia Maritime University Department of Commodity and Cargo Sciences Poland, 81-225 Gdynia, Morska 83

ABSTRACT This paper discusses numerous problems occurring in relation to microbiological quality of lactic acid cheese. Lactic acid cheese constitutes the source of various nutritive substances, what results in a possibility of allochthonous micro-flora to grow despite the presence of starter micro-flora. One of the issues discussed herein comprised the results of microbiological research depending on tvarog packing system. The influence of packing system on surface micro-flora population was assessed. Moreover, the problem of growth of enterococci and LAB (Lactic Acid Bacteria) populations depending on stage of tvarog production as well as packing system was also raised. The issue of interactions occurring among micro-organisms that re-infect tvarogs and the influence of these interactions on the growth of individual micro-organisms was also discussed. Author presented also the possibility to apply JMTPH computer program for assessment of the dynamics of changes of tvarog micro-organisms during product storage. Another chapter includes assessment of the influence of lactic acid bacteria on the behaviour of individual groups of micro-organisms occupying tvarog surface, depending on packaging hermetic properties. It was also very important to assess the safety of tvarogs in the context of a possibility of enterotoxin synthesis in conditions of various packing systems. Finally, the models of optimising lactic acid cheese quality were presented, what included application of plant additives of biostatic character, modification of used packaging as well as employing the probabilistic mathematical model helpful in evaluation of enterotoxin synthesis, depending on the level of staphylococci and yeast populations. *

Izabela Steinka: [email protected]

166

Izabela Steinka

INTRODUCTION Tvarog is a product formed as a result of dehydration of lactic acid curds after the processes of reheating and pressing, produced in Europe. Tvarog is manufactured from milk soured with a starter consisting of the following bacteria cultures: Lactococcus lactis spp. lactis v. diacetylactis, Lactococcus lactis spp. cremoris, Lactococcus lactis spp. lactis and 4 % Leuconostoc lactis. Hence, the non-regional name – lactic acid cheese – renders the product characteristics. This product, together with other similar ones, belongs to the group of unripened cheese. The rennin is not applied during its production. The commonly used names vary depending on the region where it is produced: in Russia it occurs as domasznij syr, while in Poland as tvarog. Slightly similar products comprise: Italian mozzarella, German quark, cottage cheese produced in North America or its form coming from Latin America - Queso blanco. Apart from these, there are also other products known in the market, however they are less similar to typical tvarogs. These include: impastata, bakers cheese, cream cheese, petit susie, gervais, fromage frias ala crème, ymer, lactofil. The above-mentioned cheese differs significantly from typical unripened lactic acid cheese with regard to production technology (acid and rennin), the presence of dressing or cream, the fat content and the consistency. Unripened lactic acid cheese such as cottage cheese, quark or tvarog constitute products of significant protein content and they are especially susceptible to the influence of microorganisms. In literature, there are numerous data available concerning in particular the microbiological quality of cottage cheese, however there are hardly any information on lactic acid cheese.

MICROBIOLOGICAL QUALITY OF TVAROGS AND COTTAGE CHEESE From the literature, it results that the improperly produced cottage cheese can constitute the source of mould comprising the following strains: Candida famata, Candida spherica Candida robusta, Pichia membranofaciens, Saccharomyces exiquus (Westal 1998). The research of Rosenthal shows that the undesired changes of taste and odour of hermetically packed cottage cheese result from the growth of mould belonging to species: Penicillium, Geotrichum, Mucor and Alternaria (Rosenthal et al. 1996). Our research conducted on cottage cheese showed the various level of mould in these products (Steinka 2000). In 37 cheese samples, yeast at the level from 0 up to 7.45 log10cfu/g and mould at the level from 0 up to 3.92 log10cfu/g were observed. Carem cheese was characterised by much better quality: in no sample yeast count exceeded the level of 3.72 log10cfu/g, and for mould - the count of 3.20 log10cfu/g. The microbiological quality of cottage cheese products is discussed in publications of Sims 1989, Ashenafi 1990, Maniar 1994. From data obtained by Asfhenafi, it results that the quality of majority of these products is influenced not only by the presence of yeast and enterococci, but also faecal coliforms. The presented results concerning cheese coming from the market arouse hygienic reservations. Bacillus cereus and Staphylococcus aureus are observed at the level of 102-103, while 55% of the samples contain faecal coliforms. It appears that cottage cheese stored in different temperature conditions and devoid of sorbic acid supports the growth of Pseudomonas fluorescens. Different types of Salmonella

Influence of Interactions Occurring Between Micro-Organisms…

167

show the ability to survive in these products (Sims et al. 1989). The authors observed 100fold increase of Salmonella typhimurium count in tvarogs with pepper and garlic, especially in those products which were not enriched with sorbic acid. During further laboratory research conducted by the authors as regards addition of proteins or vegetable ingredients, the decrease in count of staphylococci was observed at the temperature of 10oC and 20oC. Whereas in the same conditions, Bacillus cereus and S. typhimurium counts increased in cottage cheese. The count of Yersinia enterocolitica cells increased only at 10oC, while the decrease of these bacteria number was observed at the temperature of 20oC. Addition of sorbic acid to these products resulted in reduction or stabilisation of bacteria count at the constant level. Pathogenic micro-flora present in cottage cheese some in cases can include: Listeria monocytogenes and Clostridium sporogenes (Chen et al. 1993, Stañczak et al. 2000). There exist many controversies concerning the unequivocal assessment of adverse influence of certain environmental conditions on the survival rate of Listeria monocytogenes in cottage cheese. Majority of authors (Benkeroum and Sandine 1988, George and Lund 1988, Ryster and Marth 1985) prove the growth of populations or at least the presence of this pathogen at constant level in the product. Hicks et al. (1991) and Piccinnin (1995) agree that storage in refrigerating conditions results in reduction of Listeria count in cottage cheese. The condition determining survival rate in tested cottage cheese was neither concentration of H+ ions nor hermetic packaging. Listeria is able to grow in environment of acidity lower than those occurring in lactic acid cheese (George et al. 1988). From data presented by Chen et al. 1993, it results that inoculation of cottage cheese with these species gives the possibility to observe their behaviour during storage in conditions of atmosphere modified with carbon dioxide. In this research, during two months of storage at three temperatures, the reduction in count of Clostridium sporogenes was observed. The decrease in number of Listeria monocytogenes cells was observed in cottage cheese at the temperature of 4oC, while as soon as at 7o it was possible to determine the growth of these micro-organisms by one logarithmic cycle. In traditionally packed cottage cheese, the literature data showed the increase of these bacteria count by 1000 times. The significant number of conducted research concerned also the problem of survival rate of pathogenic bacteria in cottage cheese, indicating the possibility of their occurrence in these products (Piccinin and Shelef 1995, Hicks and Lund 1991 and Farrag et al. 1992). These data suggest that such products as unripened lactic acid cheese can support the growth of pathogenic micro-flora, since they do not include special biostatic additives. This is especially notable, because the composition of products includes significant number of lactic acid bacteria, what should be sufficient for the absence of pathogenic microflora resulting from the inter-microorganism antagonism. However, the prolonged refrigerating storage of products can be the reason for the presence of psychrotrophic micro-flora in lactic acid cheese. This concerns mostly food of animal origin. While listing the psychrotrophs isolated from dairy products, Champagne et al. (1994) mentions above all such bacteria as: Listeria monocytogenes, Yersinia enterocolitica or Bacillus cereus. As far as psychrotrophic micro-flora is concerned, the strains of Pseudomonas fluorescens and Enterobacter aglomerans were identified in cottage cheese (Lund et al. 1988). The behaviour of psychrotrophs depending on the level of added carbon dioxide was investigated by Maniar 1994, Fedio 1994, Moir et al. 1993.

168

Izabela Steinka

Tvarogs have no dressing or cream, what makes their micro-flora differ both qualitatively and quantitatively from the one being predominant in cottage cheese. Micro-organisms occurring in cottage cheese comprise such species as: Escherichia coli, Enterobacter sp., Bacillus subtilis, Enterococcus faecalis, Enterococcus faecium, Staphylococcus aureus, Micrococcus sp., Candida quilierimonde, Candida famata, Candida lusitaniae, Geotrichum candidum, Aspergillus flavus (Steinka et al. 1999, 2001, 2002). Due to the product instability, tvarogs are stored in refrigerating conditions. There exist many data on changes of physico-chemical and sensory properties of these products. However, there are only few micro-biological tests connected with quality of these products (Cais et al. 1998, Steinka et al. 1998, 1999a,b, 2000, 2001c, 2006b, Steinka 1999c, 2001a,b,d, 2003a, 2004, 2005a,b, 2006b,c, Ziółkowski et al. 2003). Data obtained by Steinka et al. (2002a) showed the presence of Bacillus sp. bacteria in lactic acid cheese stored in refrigerating conditions. During refrigerating storage, the intensive growth of Staphylococcus epidermidis was also observed in hermetically-packed tvarogs. The number of samples, in which the presence of staphylococci was detected increased by 77.7% in relation to amount observed on the day of buying the tvarog by the customers (Steinka et al. 2002a). Table 1 presents the micro-organisms that could be isolated from stored lactic acid cheese prior to implementation of HACCP system to their production. Table 1. Type and count of micro-organisms isolated from tvarogs during refrigerating storage Type of micro-flora

Enterococcus Haemolysing Streptococcus á Micrococcus Bacillus Staphylococcus epidermidis Candida Mould

Storage time (days) 0 7 Micro-organism count (cfu/g) 1.82 ·106 9.93·105 6 2.2 ·10 5.6·104 0 4·104 1.22 ·106 9·104 0 9·104 4 9 ·10 4·104 7.45 ·105 3.71·104

14 5.19·105 0 5·101 3.13·105 5.98·105 0 3.2·104

Steinka et al. 2002a.

From our previous research, it results that the psychrotrophic micro-flora present in lactic acid cheese could also include Micrococcus sp. and Staphylococcus sp. Prior to implementation of HACCP system into the production of lactic acid cheese, the count of psychrotrophic micro-organisms in tvarogs coming from the market was equal at the maximum to 1.4·103 cfu/g. In tvarogs packed into PA/ PE and stored for the period of 7 days after purchase, the count of populations of these micro-organisms could reach the level of 3.0•105 cfu/g. In more than 52% of tvarog samples, and after 14 days of refrigerating storage from the date of purchase, the count of psychrotrophic micro-flora increased tenfold in relation to the level observed after one week (Steinka et al. 1999b). Unripened lactic acid cheese is also illustrated by only a few mathematical descriptions (Steinka 2003, 2005c).

Influence of Interactions Occurring Between Micro-Organisms…

169

With the help of mathematical models, it is possible to predict most of all the technological details in tvarogs such as: the influence of salt level and the initial level of a starter on pathogen survival rate (Bozukart 2001). Another subject of prediction can be evaluation of time of cutting cottage cheese, taking the growth of starter cultures into account (Crofcheck et al. 1999). In ripening cheese (white cheese), the prediction concerned e.g. diffusion during long-term brining (Turhan et al. 1992). Many years of research on lactic acid cheese showed that microbiological quality of tvarogs was various, depending not only on applied technology, but also on the production plant as well as layer, from which the sample is taken for tests. Observations presented in Table 2. confirmed the differences in count of microorganisms populations on the surface and inside the cheese mass (Steinka et al. 2006b). Table 2. Level of micro-organisms in tvarogs, depending on place of taking a sample Type of microflora Escherichia coli Mould Yeast

Place of taking a sample from tvarogs Surface Inside Surface Inside Surface Inside

Average cfu/g

Standard deviation

2.48 2.10 0.31 0.13 2.52 2.43

0.2 0.2 0.09 0.07 0.1 0.2

Steinka et al. 2006 b.

Count of mould on the surface was higher by 58.1% than the number observed inside the cheese mass. In the case of yeast and Escherichia coli these values were equal to 15.0 and 3.6% respectively. Statistic analysis showed that there existed a weak correlation between count of yeast inside the a/m product mass and on its surface ( r 0.136) ( Steinka et al. 2006b). Table 3. illustrates the model of changes of micro-organisms count present in several batches of lactic acid cheese packed into PA/PE, taking surface and inside layers of tvarog cubes into account. Table 3. Models illustrating total count of micro-organisms in tvarogs Type of microflora Escherichia coli Mould Yeast

Model of changes Z=0.091-0.291x-0.33y-0.006x2 –-0.2xy-0.225y2 Z=3.844-0.897x-0.175y-0.082x2 +0.549xy-0.205y2 Z=-0.151-3.811x+4.993y+0.345x2 +0.125xy2 0.479y

Steinka et al 2006b; x-tvarog surface, y-inside of tvarog mass, z-number of micro-organisms in tvarog.

During his research, Parisi compared the behaviour of chosen bacteria and mould in dependence on the distance of product surface from the packaging laminates (Steinka and Parisi 2006). Research was conducted in order to provide knowledge on the differences

170

Izabela Steinka

between cheese produced using citric acid and using lactic acid. Cheese manufactured using citric acid was packed into cryovac, while lactic acid cheese was packed into PA/PE laminates. Determination of micro-organisms was performed from the four layers: upper sample 0-5 mm from the packaging surface, middle sample 5-15 mm and lower sample 15-30 mm from the packaging surface. The averaged sample contained micro-organisms from layers of 0-30 mm. In cheese packed into cryovac, the Escherichia coli bacteria grew better in lower and middle layers. In the surface cheese layer, the reduction in bacteria count was observed. The growth of fungi (yeast and mould) was observed by Parisi in samples taken from all cheese layers, however the dynamics of growth were the greatest in a lower layer. The behaviour of fungi in sample taken from the layers 0-30 mm from the packaging surface was similar to the one observed in cheese packed into PA/PE laminates (Steinka et al. 2006b). Test results obtained by other authors show that the intensity of micro-organism occupation of a certain layer of food product is dependant on type of bacteria and type of food. Research conducted by Ingram in meat products showed the growth of staphylococci count by 0.7 cfu/g inside the ham, whereas on its surface the population size varied by only 0.1 cfu/g (Ingram 1996). The results of both research show a different expansion of many species of micro-organisms in external and internal lots of food product.

THE DYNAMICS OF CHANGES OF FACULTATIVE ANAEROBIC MICRO-FLORA IN LACTIC ACID CHEESE The tables present average values of certain types of bacteria and fungi occurring in lactic acid cheese prior to implementation of HACCP system. Population sizes were different, depending on the packaging type. The microbiological requirements for tvarogs valid during that time (1991-2003) assumed the absence of coagulase-positive staphylococci in 0.1 g, what was later verified for the presence of these bacteria in number not greater that 10 cfu in one gram of a product. Salmonella and Listeria monocytogenes could not occur in 1 g of a product. The limits for 4

yeast assumed the maximum presence equal to 10 cfu/g, whereas mould was not permissible 2

at the level higher than 5•10 cfu/g. The presence of coliforms was accepted at the level not greater than 0.001 g (the Standard and the Decree of Polish Minister). Results of testes conducted in individual years, and presented in Table 4., show that the level of many micro-organisms present in these products was inconsistent with the a/m requirements. Table 5. below presents the levels of fungi detected after implementation of HACCP system, depending on type of packaging. From these data it results that PA/PE packaging is the best type of packaging, at high level of initial contamination. However, results concerning the behaviour of populations of other micro-organism species in tvarogs packed with different systems do not entirely confirm these observations.

Influence of Interactions Occurring Between Micro-Organisms…

171

Table 4. Levels of contamination of tvarogs with allochthonous micro-flora Type of microorganisms

Type of packaging

Staphylococci E. coli

PA/PE

Staphylococcus aureus Yeast E.coli

PA/PE/ Cryovac

Count after 14 days of storage cfu/g / log10cfu/g

Initial contamination cfu/g/ log10cfu/g 1.1 •10-1 1.2•101

2.3.103 0

2.1•103

1.1•103

1•104 1.6•103/1.6•102

1.4•106 1.7•101/0

Staphylococcus aureus Yeast

5.4•102 / 9.1•101

4.6•102/2.1•102

9.1•104 /6.6•103

4.4•106/8.2•105

Enterococci

6.1•103/1.5•105

4.8•105/1•104

2.97

1.6

2.85

1.39

E. coli

PA/PE

Staphylococcus aureus Yeast Enterococci

4.17 Parchment paper

Staphylococcus aureus Yeast

3.3•10

Author Steinka 1998 Steinka Stankiewicz Steinka, Morawska 1999

Steinka Zukowski, Hildebrand 2000

Steinka Kukulowicz 2002

5.25 3

3.7•105

6.4•101

4•101

2.9•105

6.3•105

Steinka Mieczkowska 2003

Enterococci

Steinka – collective study conducted in Microbiological Laboratory in years 1998-2003.

Table 5. Level of fungi (yeast and mould) in tvarogs prior to HACCP implementation Type of fungi

Type of packaging

Market quality log10cfu/g

Count of fungi after storage log10cfu/g

Yeast Mould Yeast Mould Yeast

Traditionally- parchment paper

4.46 4.73 6.79 4.0 3.81

7 days 4.95 5.73 6.11 3.86 4.25

14 days 3.79 6.27 6.11 4.11 5.91

1.71

2.2

2.36

Mould

Own Study.

Vacuum system - PA/PE Atmosphere modified with nitrogen addition

172

Izabela Steinka

It should be noticed that the kinetics of populations changes is different, depending on way of conducting the production. These changes are determined together by the state of raw material, packing system as well as the presence of background micro-flora and re-infecting micro-flora. Model of changes of populations occurring in lactic acid cheese produced without taking GHP rules into account was dveloped by Steinka 2002a, Steinka 2003a, 2005a. However, in order to compare the dynamics of population changes depending on the production system, the kinetics of growth should be determined.

2 1,8 1,6 y= -1,075x2 + 4,895x- 3,82 R2 = 1

1,4 1,2 1

y= 0,595x- 0,2367 R2 = 0,4789

Log cfu/h 0,8 0,6 0,4 0,2 0 0

7

14

-0,2 Time of population growth (days)

Figure 1. Kinetics of changes of Enterococcus sp. population in tvarogs produced without GHP.

Attempts were made to compare the growth of Enterococcus sp., Staphylococcus aureus bacteria and yeast during refrigerating storage of tvarogs produced in HACCP system as well as manufactured by dairy plants without implemented quality system. Significant differences were noticed in the dynamics of growth of e.g. enterococci count in tvarogs. Changes in enterococci count during 7 days of storage were equal to 1.67 log10cfu/g up to the day 7, and 1.19 log10cfu/g between day 7 and 14 (figure 1). The growth of staphylococci was not as dynamic – the observed change was equal on the average to 0.26 log10cfu/g in the first week of storage, and yeast was equal to 0.48 log10cfu/g during that period. The presented figures (figure 1 and figure 2), reflecting the behaviour of enterococci observed during 14 days of product storage, present the different behaviour of these bacteria in both production systems.

Influence of Interactions Occurring Between Micro-Organisms…

173

Evaluation of kinetics of changes of tested streptococci populations show that in the case when GHP rules are not applied, the rate of population growth is ca. 3.7-fold higher (3.68) than in the case of tvarog production conducted in HACCP system. From obtained data it results that variability of population count is significantly diversified, unless critical checkpoints are introduced.

0,0045 0,004 0,0035 0,003 y= 0,0021x- 0,0024

0,0025

R2 = 0,9196

0,002 log cfu/h

y= 0,0011x2 - 0,0022x+ 0,0011 R2 = 1

0,0015 0,001 0,0005 0 0

7

14

-0,0005 -0,001 Time of population growth (days)

Figure 2. . Kinetic changes of Enterococcus population in tvarog produced in HACCP system.

For packing lactic acid cheese, several types of packaging are applied: PA/PE laminates, foamed polystyrene trays wrapped with a thin PE film – so called: frischaltenfolie, aluminium foils with parchment paper, the parchment paper and cryovac. The hermetic level of packaging as well as its barrier properties are responsible for the behaviour of surface microflora behaviour. The rate of micro-flora growth can be inhibited or stimulated by conditions occurring inside the packaging. A specific ecological niche is created between surfaces of packaging and the product characterised by conditions dependant on the packaging hermetic level. A very significant influence of packaging material of tvarog quality is also observed. Research conducted on the influence of packaging on cottage cheese and cream cheese quality showed that both the packaging and the storage of these products in polystyrene containers also did not meet microbiological expectations. Whereas applying the packaging made of waxed cardboard significantly favoured maintaining good quality (Steinka 1999c). Cottage cheese packed into containers with sealed covers provided better protection against

174

Izabela Steinka

contamination than packaging with separate moveable covers. Sealable covers made of aluminium film allowed additional sterilisation and limiting the presence of surface microflora. During further research on cottage cheese and cream cheese it was additionally observed that contamination with yeasts was significantly dependant on technology of cheese production, while the packaging hermetic properties had no influence on mould level in products (Steinka et al. 2000). Majority of research available in literature is devoted to the influence of packing system on organoleptic properties of the products. Only a few studies raise the issue of interactions occurring among product micro-flora, depending on the packing system. It has a first-grade significance for the quality and safety of food. The packaging hermetic properties accompanied by application of vacuum packing system constitutes a special hygienic problem.

PACKING AS A FACTOR DETERMINING THE PRODUCT QUALITY Apart form packaging type and packing system, the time and conditions of product storage constitute very significant factors responsible for quality of stored cheese. In figures 3 and 4, the sizes of populations of facultative anaerobic bacteria in tvarogs coming from the market are presented. Tvarogs were packed into different types of packaging: parchment paper, PA/PE laminate and cryovac. 6 5 4 Log cfu/g 3 2 1 0 Paper

Cryovac

PA/PE

Packaging type E.coli

Enterococcus sp.

Staphylococcus aureus

Figure 3. The Influence of packaging type on growth of facultative anaerobic bacteria in tvarogs.

Influence of Interactions Occurring Between Micro-Organisms…

175

7

6

5

4 Log cfu/g

Yeast 3

Mould

2

1

0 Paper

Cryovac

PA/PE

Packaging type

Figure 4. The influence of packaging type on growth of fungi populations in tvarogs.

From conducted research it results that the level of contamination with bacteria and fungi was significantly dependent on the packaging type. The greatest count of micro-organisms was detected in cheese packed into parchment paper. Application of HACCP system and the packaging made of plastics such as Cryovac showed better protective properties in comparison to PA/PE laminate. It is probable that in the case of Cryovac application, blowing nitrogen through the surface of cheese contributed to modification of atmosphere after closing the packaging. From literature analysis it results that food products hermetically packed into plastic packaging can pose a hazard connected with production of bacterial toxins (Post et al. 1988, Adams et al. 2000, Steinka et al. 2001c, 2004, Duffranse 2000). The safety of hermetically packed products is not always improved by applying modified gaseous atmosphere into their packing. From numerous reports, it results that selection of appropriate composition of atmosphere can be adapted to inhibition of certain types of microflora, since every species reacts differently to the presence of CO2. Therefore, for example: the growth of Staphylococcus aureus is intensively inhibited at 50-100% CO2, however the 0

obligatory condition is temperature of 10 C. From research of Kimura, it appears that staphylococci are more sensitive to the influence of certain concentrations of carbon dioxide in comparison to Escherichia coli rods, what stands in contradiction to the previous opinions on reactions of these bacteria in relation to application of modified atmosphere (Kimura 1999).

176

Izabela Steinka

From literature it results that addition of carbon dioxide causes changes of generation time in laboratory conditions, depending on its concentration and type of micro-flora (table 6). Table 6. The influence of carbon dioxide on micro-organisms populations Type of micro-organisms Pseudomonas, Flavobacterium Achromobacter Moraxella-Acinetobacter

Carbon dioxide concentration

Reaction of micro-organisms

10-20 %

Inhibiting

20-70 %

Time prolongation

,

Lactic acid bacteria

Sustaining of growth

Yeast

50%

50-100-fold increase in count

Clostridium botulinum Clostridium perfringens Listeria monocytogenes

< 50%

Minimal growth inhibiting

< 50%

Staphylococcus aureus

50-100%

Depending on the temperature – inhibiting or sustaining of growth

Own study on the basis of Nguyen and Carlin 1994, FDA 2001, Zagory 1995, Philips 1996, Adams 2000.

The presence of bacterial toxins in hermetically packed food products is not always preceded by obvious organoleptic symptoms. From data presented in literature, it results that at room temperature, 60% of CO2 addition, 25% of oxygen presence and 15% of nitrogen concentration in packed fish fillets, the organoleptic changes occur at the same time as the presence of toxins in products. Whereas at the same composition of gaseous atmosphere in packaging and the temperature of 4.40C, the presence of toxins precedes the indicators of spoiled fish fillet by 7 days (Tyszkiewicz 1992). Control of parameters responsible for physico-chemical changes of products is possible in the case of meat and its products packed with application of modified atmosphere or using active packaging. In the case of lactic acid cheese, application of the above-mentioned technologies is much more difficult due to the delicate product structure. Attempts to pack cottage cheese in modified atmosphere were made by Dixon, 1988, Honer, 1988, Brody, 1989, Fedio, 1994, Maniar, 1994. The stability of quality features of mozzarella cheese was described by Alves et al. (1996) and Elliot et al. Tvarogs packed with MAP system were investigated by PanfilKuncewicz et al. (1997). However, application of modified atmosphere in technology of tvarog packing revealed that there existed a relationship between concentration of carbon dioxide and the adverse sensory changes (Fedio 1994). In Poland, there are several systems of hermetical packing of cheese into plastic packaging, among others: vacuum-packing into

Influence of Interactions Occurring Between Micro-Organisms…

177

PA/PE films and shrink-wrapping with Cryovac laminate. The non-hermetic packing methods include: packing into aluminium films with parchment paper and into the parchment paper. Evaluation of physico-chemical changes of tvarogs packed in a traditional way was conducted by Pieczonka 1993, Kujawski et al. 1994, Chojnowski et al. 1999, Molska et al. 1992, Œmietana et al. 1997. There are not many data on assessment of quality of lactic acid cheese hermetically packed with vacuum and non-vacuum systems (Panfil-Kuncewicz 1997a,b, Steinka et al. 1998, 1999c, 2001b). Technological and technical aspects of packing tvarogs and soft cheese into various types of packing films were presented by Œmietana et al. 1997 and Marcotte et al. 2001. Significant information on the influence of modified atmosphere on unripened cheese quality is presented by Gonzales-Fandos et al. 1999. However, the subject of his study is fresh cheese Cameron, which acidity differs from acidity of tvarogs and is equal to pH ~6.35. This products is characterised by the fat content of 54-56% and 7 days of stability period. Research on the effect of 100% CO2 addition during packing Cameron cheese, as well as prolonged time of storage result in decrease in number of psychrotrophic bacteria only in the early period. Whereas, their increase can be detected between day 7 and day 30 of product storage. In the case of mesophilic micro-flora, this growth is observed between day 12 and day 30 of cheese storage. From the research of Fandos, it also results that the growth of Enterobacteriaceae rods is noticeable between day 7 and 15. The research results of Fandos are similar to data obtained by Moir (1993), who observed the growth of psychrotropic bacteria in tested products despite the presence of CO2. In tvarogs vacuum-packed into PA/PE laminates, where atmosphere was not modified, Steinka et al. (1999b) observed the increase of psychrotrophs count during product storage. However, vacuum-packing did not favour the growth of Listeria monocytogenes rods in tvarogs. The number of samples indicating the presence of this rod after 14 days of storage could evidence the change of atmosphere during product breathing or the metabolism of other micro-organisms. Number of tvarog samples, where Listeria monocytogenes was present decreased by 24% after 14 days of storage (Steinka et al. 1999b). The significant factor determining the growth of micro-organisms in tvarogs is pH value characterising the products. From data obtained by Moir, it results that when CO2 additive does not exceed the value of 40%, then the acidity value does not change while packing the cheese using MAP technology. In tvarogs packed into both types of laminates, the acidity varied from 4.26 to 5.03 pH. During 14 days of storage, no significant changes were observed in acidity of vacuum-packed cheese. In tvarogs packed into Cryovac laminates, the decrease in content of organic acids was observed up to the day 7 of storage, what could result from the growth of alkalising micro-flora of the mould (Steinka 2003a). In tvarogs of low count of contaminating micro-organisms, the acidity variations during storage are insignificant (table 7).

178

Izabela Steinka Table 7. Acidity of tvarogs of low level of contamination with allochthonous micro-flora Storage time of tvarogs in PA/PE packaging (days) Average 0 2 4 acidity values pH 4.73±0.06 4.59±0.05 4.63±0.04

7

14

4.56±0.03

4.63±0.08

Own study.

From the research of Steinka (2003a), it results that during the period up to day 7, no reduction is observed in number of all tested groups of micro-organisms isolated from tvarogs. It often happens that the aesthetics and convenience play a decisive role in purchase of a product by a consumer. The results of study of consumer preferences concerning tvarogs packaging revealed that among the selection factors connected with packaging, very significant were such indicators as: tightness (71.8% of respondents), aesthetic appearance (59.4% of respondents) and information on the packaging that is important for 64% of those polled (Steinka 2003a). High water content in lactic acid cheese is the reason for their low stability and the difficulties connected with selection of packaging as well as storage problems. Research conducted in recent years concerning the possibility to apply hermetic packaging for tvarogs still has not solved many problems and has not answered numerous arousing doubts (Panfil–Kuncewicz et al. 1997 a,b, Panfil–Kuncewicz et al. 1997b, Steinka 1999c, Steinka 2001b, Steinka et al. 2001c). The research of Panfil–Kuncewicz et al. proved the improvement of microbiological quality of tvarogs stored in modified atmosphere, however the composition of applied atmosphere that allowed obtaining that effect is considered to worsen the sensory quality of the products. Microbiological tests concerning hermetic packing of tvarogs (Steinka et al. 2001c, Steinka 2004) showed a possibility of staphylococcal enterotoxin to occur in stored products. In subject literature, there are no such models which would take all parameters into account at the same time: packaging type, packing system, biological and biogenic factors dependent on variable conditions of product storage in packaging. Biological factors should be understood as live organisms belonging to Eucaryota and Procaryota, constituting autochthonic micro-flora and secondary food contamination. The notion ‘biogenic factors’ refers to products of basic and secondary metabolism of these organisms formed as a consequence of transformations resulting from their physiology. The biogenic factors include fermentation products (organic acids, alcohols), aldehydes, gases, toxins, bacteriocins, antibiotics as well as products of proteolysis, liopolysis and decomposition of specific organic substances. The table below presents few mathematical models that are valid in predictive microbiology and which concern cheese quality issues. The type of tests conducted in such cheese and its products are presented in table 8. There exists only one predictive model concerning tvarog micro-flora, which presence results from certain contamination and re-infection level (Steinka 2003a). Appropriate selection of packaging type and packing system determines quality of the product and safety

179

Influence of Interactions Occurring Between Micro-Organisms…

of a consumer. This results from the fact that in hermetically-packed tvarogs, there occur interactions between product and micro-flora as well as among product, micro-flora and the packaging. The influence of the above-mentioned biological factors and biogenic factors on the type of interactions is dependent on packing system and decides about the quality of products stored in refrigerating conditions. Multidirectional character of interactions influences also safety of these products. This arouses difficulties in adapting mathematical description to changes of interaction types having the influence on safety of hermeticallypacked lactic acid cheese. Table 8. Predictive models of the micro-organisms behaviour in cheese Authors of a model Brouillaud –Delante 1997 Bolton 1999 Bozukart 2001 Steinka 2003 Tamagnini 2005 Poirazi 2006

Type of described changes

Type model Linear

Evaluation of probability of growth of Listeria monocytogenes Behaviour of Yersinia enterocolitica

Type of dairy product Micro-organisms in dairy products Mexican cheese Feta cheese

Surface micro-flora

Lactic acid cheese

Polynomial

Crottin Sheep’s milk cheese Cheese

Vitalistic, Churchil, Gompertz Neural network

Listeria monocytogenes

Yersinia enterocolitica, typhimurium Streptococcus macedonicus

Salmonella

of

Logistic Gompertz

Own study.

Nevertheless, in recent years, actions aiming at improvement of product safety has been more and more often observed. These tendencies include taking packaging into account in predictive models, as well as application of a new generation of packaging or using bacteriostatic additives. However, taking these factors into account and excluding them require new research to be conducted as regards evaluation of interactions occurring among micro-flora both in raw material and with possible re-infections.

ASSESSMENT OF INTERACTIONS OCCURRING AMONG TVAROG MICRO-FLORA IN MODEL CONDITIONS The significant element taken into account in predicting the quality and safety of products should be interactions occurring between micro-organisms. Among interactions occurring between micro-organisms based on the presence of numerous nutritive substances, there are such which result from the severe competition for food. The developed defence mechanisms work in the form of such interactions which ensure the possibility of using nutrition substances to all types of micro-organisms.

180

Izabela Steinka

Among micro-organisms, the complete dependence of species on reciprocal metabolism is commonly known. The interactions of syntrophic character cause that in order to inhibit the growth of one micro-organism, it is enough to limit the growth of another one. In some fermented dairy products, the syntrophic phenomenon is responsible for proper organoleptic properties. A very special case of syntrophic character is the dependence between cells of Streptococcus thermophilus and Lactobacillus delbrueckii bulgaricus in yoghurt. However, most often the interactions occurring between cells of Procaryota are based on the phenomena of competition for food, and the interactions are most often of antagonistic character. Knowledge on food requirements of certain type of micro-flora is necessary for proper prediction of direction of micro-organism metabolism in the presence of other species. If only two micro-organisms species are present in the environment, then their growth is different than in the situation when there are much more of them. The interactions are also different in laboratory conditions and in food. In tvarogs, the growth of micro-organisms in a monoculture also occurs differently than in the presence of many cultures. The behaviour of micro-organisms in tvarogs is significantly dependent also on other factors, which include: • • • • •

type and count of technical micro-flora (starters) in tvarog, type and count of other micro-organisms, type of packaging, packing system, production stage, during which penetration of micro-organisms occurs: raw material, curd, unpacked finished products, packed finished product.

The example of micro-organism behaviour in tvarogs, depending on the moment of infection, contamination size or the presence of other micro-organisms, can be the behaviour of Staphylococcus aureus and Enterococcus faecalis, if they are added to milk or to the produced lactic acid curd. Tests were conducted in laboratory conditions. Enterococci and staphylococci were added to milk, which was then soured using a typical starter for production of lactic acid cheese. The starter composition contained of bacteria of Lactococcus sp. type in amount of 96% and Leuconostoc lactis in amount of 4%. The table below presents changes in count of micro-flora infecting raw material and the curd, in the case when inoculum of this bacteria is added to the curd or the milk, then a curd is produced and such a semi-product is stored in refrigerating conditions (table 9). In the case of traditional packing of tvarogs into parchment paper, the growth of faecal streptococci could be observed in the final period of refrigerating storage of products (figure. 5). The behaviour of population can evidence that re-infection with enterococci occurred not during milk souring, but at further stages of tvarog production.

Influence of Interactions Occurring Between Micro-Organisms…

181

Table 9. Changes of staphylococci and enterococci in model conditions depending on production stage Type of microorganism

Time of curd storage (days) Production /inoculum log10cfu/g

Staphylococcus aureus Enterococcus faecalis

stage

Milk 2.30 cfu/mL Curd 2.75cfu/g Milk 2.74 cfu/mL Curd 2.0 cfu/g

2

4

7

14

2.30

0

0

0

2.30 3.33

1 2.90

1 2.25

0 3.74

2.47

3.14

2

3.46

Own study.

4,5 4 3,5 y= 0,0833x3 - 0,7236x2 + 1,9331x+ 2,144

3

R2 = 0,9974

2,5 log cfu/g 2 1,5 1 0,5 0 0

2

4

7

14

Storage time (days)

Figure 5. The influence of refrigerating storage on enterococci population in tvarogs packed into parchment paper.

Enterococci and yeast constitute predominant micro-flora in tvarogs (Steinka et al. 2001c). The modelling research was conducted, in order to evaluate interactions between micro-flora most often present in lactic acid cheese. The behaviour of starter micro-flora, yeast and staphylococci was tested in the presence of certain inoculum of enterococci present in milk or in semi-products (lactic acid curd). The influence of enterococci of technical micro-flora (Lactococcus spp. and Leuconostoc lactis) and their interactions with other micro-organisms occurring depending on inoculum is presented in the figures.

182

Izabela Steinka

As a result of souring milk with starter consisting of pure dairy cultures, lactic acid curds were obtained, in which count of lactic acid bacteria after 14 days of souring was equal from 8.22 to 8.51 log10cfu/g. Number of Lactococcus bacteria decreased after 14 days of storage to 7.25-7.9 log10cfu/g on the average. Inoculation of milk with Entrococcus faecalis cultures resulted in changes of count of these bacteria after curd formation by 0.67 log10cfu/g. In the moment of curd formation, the count of faecal enterococci was equal to 3.33 log10cfu/g. During curd storage, the insignificant changes of enterococci population count occurred up to day 4 of its storage (figure. 6). However, after 14 days the level of enterococii was different from the initial value only by 0.11 log10cfu/g. These changes were described by a polynomial equation of the third order. As much as 99.4% of variability of enterococci population could be explained with the help of this model (table 10.).

5 4.5 4 3.5 3 Log cfu/g 2.5 2 1.5 1 0.5 0

Curd C Curd B Milk 0

2

4

7

14

Tim e (days)

Figure 6. Changes of enterococci count depending on curd production stage and applied inoculum.

Addition of faecal streptococci of initial inoculum 2 log log10cfu/g to the formed curd resulted in detecting inhibition of growth of these bacteria after 14 days of storage (figure 6).

Influence of Interactions Occurring Between Micro-Organisms…

183

Table 10. Tendency of changes of enterococci depending on tvarog production stage and the presence of allochthonous micro-flora Experimental variant Soured milk with Enterococcus faecalis ATTC 29241 inoculum 4log cfu/g. A-produced curd

Equation form y = -0.0142x3 + 0.1839x2 - 0.6919x + 3.85

r2 0.994

B and C-curds with Enterococcus faecalis ATTC 29241 inoculum 2 (B) and 4 log cfu/g ( C)

y = 0.0883x3 - 0.8729x2 + 2.4788x + 0.84 (B)

0.844

y = 0.0467x3 - 0.2593x2 + 0.114x + 4.772 (C)

0.954

4

3

2

D-curd with Enterococcus faecalis ATTC 29241 and Staphylococcus aureus ATTC 25923 after milk fermentation

y = 0.1483x - 1.8483x + 8.0267x - 14.057x +10.42

0.999

E-curd with Enterococcus faecalis ATTC 29241 and Candida sp. after milk fermentation

y = -0.035x3 + 0.3864x2 - 1.1386x + 4.8

0.951

F–industrially produced tvarogs

y = -0.005x4 + 0.1383x3 - 1.0775x2 + 2.8392x + 0.84

0.555

y- Enterococcus faecalis count, x-storage time.

The tendency of observed population changes was characterised by polynomial equation of the third order (table 10). The coefficient of determination of this model showed that only 84.4% of observed variability of enterococci populations could be explained by the influence of time of exposure to low temperature. The addition of 4 log cfu/g of enterococci to the curd showed different dynamics of population changes (figure 6). Equation describing changes of Enterococcus faecalis during storage of a curd with additive in amount of 4 log cfu/g reflected to a great extent the real variability of streptococci populations in stored curd (table 10). Only 5.6% of observed streptococci variability did not result from the influence of curd storage time of the behaviour of enterococci. The presence of Staphylococcus aureus population in stored curd caused that count of enterococci in a curd insignificantly decreased after 14 days of storage. However, the reduction of size of streptococci population did not exceed the value of 0.28 log10cfu/g (figure 7).

184

Izabela Steinka

10 8 6 Log cfu/g 4 2 0

Lactococcus sp. Enterococcus sp. 0

2

4

7

14

Time (days) Figure 7. Changes in count of enterococci in the presence of populations: Lactococcus sp. and Staphylococcus aureus (variant D).

The addition of yeast of Candida kefyr species to the curd (inoculum 3 log cfu/g) and the presence of Enterococcus faecalis of inoculum equal to 4 log cfu/g resulted in detecting a small decrease in count of Candida and a small increase in count of faecal streptococci after 14 days of curd storage. However, the magnitude of changes of Enterococcus population did not exceed the value of 0.38 log10cfu/g (figure 8). Changes in count of Enterococcus population in the presence of lactic acid bacteria and Candida could be expressed with the help of a polynomial equation of the third order characterised by a high value of coefficient of determination (table 10).

9 8 7 6 5 Lo g cf u/ g 4 3 2 1 Lactococcus sp. 0

Candida kef yr 0

2 4

T i me ( d ays)

Enterococcus sp. 7 14

Figure 8. Changes of enterococci count in the presence of populations: Lactococcus sp. and Candida kefyr (variant E).

Influence of Interactions Occurring Between Micro-Organisms…

185

3.5 3 2.5 2 log cfu/g

1.5 1 0.5 0 0

2

4

Time (days)

7

14

Figure 9. Changes of enterococci count in vacuum-packed tvarogs.

Observation of enterococci population in vacuum-packed tvarogs showed that during storage of the products, the increase in count of these bacteria occurred by 0.82 log log10cfu/g on the average up to the second day of storage. Between day 2 and day 10 of storage, the reduction in faecal streptococci count by ca. 1 log10cfu/g was detected (figure 9). Enterococci population in tvarogs showed dynamics of changes similar to the enterococci population added to the curds before their storage at low temperature (Variant B). Equation describing changes of bacteria count in hermetically-packed tvarogs is a polynomial equation of the fourth order characterised by a low coefficient of determination value. The value of coefficient r2 showed the presence of other, apart from storage time, factors influencing the population behaviour (table 10). The polynomial equation determined for describing changes did not describe 44.5% of variability of faecal streptococci population as time-dependant function. The behaviour of enterococci population observed in variants B, C, D and E was similar, no matter if additional type of re-infecting micro-organisms was present in the curd or enterococci were present there as a monoculture. Another dynamics of changes were observed in the case when curd subjected to final technological treatment i.e. lactic acid cheese was closed in hermetic packaging (Variant F). In this experiment, no significant influence of interactions occurring between staphylococci and streptococci as well as between yeast and enterococci was observed on the dynamics of changes of bacteria present in curds in model conditions. No apparent antagonism was observed between lactic acid bacteria and enterococci. At ratio 2:1 this phenomenon was expressed as lack of evident reduction in number of faecal streptococci in a curd.

186

Izabela Steinka

In physiologically optimum conditions of the monoculture growth, the antagonistic influence of lactic acid bacteria in relation to significant number of pathogenic bacteria is known (Usajewicz 2000). The author investigated the influence of lactic acid rods on pathogenic strains of staphylococci and Salmonella. Research on interactions of enterococci with other bacteria concerned the microflora responsible for defects in ripening rennet cheese (Usajewicz 1995). The quoted research indicates the significant influence of enterococci on proteolytic activity of some bacteria of Lactococcus type that can have the influence on determining the sensory and physico-chemical properties of obtained curds (Usajewicz 1995). There exist only a few data concerning reciprocal interactions between faecal streptococci and facultative pathogenic micro-organisms (yeast and staphylococci) in conditions of lactic acid curd production (Steinka 2003a). From other research conducted e.g. by Kornacki et al. (2002), it results that the effect of inhibiting micro-organisms of faecal origin by technical micro-flora is obtained in tvarogs only in certain technological conditions. However, the observed effect concerned faecal rods of coli group, and not streptococci of Enterococcus type. Lack of significant changes of streptococci population count and their apparent synergistic interactions with yeast and staphylococci, at certain inoculum, causes that enterococci remain at the constant level for a long time of curd storage. Obtained results indicate also the low sensitivity of enterococci to the presence of Lactococcus sp., staphylococci or yeast. In model conditions, regardless of the level of infection with enterococci within the inoculum range 2-4 log cfu/g, changes of these bacteria count were not significant after 14 days of curd storage. It was also observed that at low initial contamination of tvarogs with enterococci, the reduction of these bacteria count after 14 days of storage was insignificant and did not exceed the value of 1 log10cfu/g. A small change in environmental conditions is able to change the intensity of observed phenomena or the total effect of interactions detected during production or manufacture of tvarogs. For interactions occurring among bacteria cells, the size of population occurring together in a given environment is of a great significance. The example is interaction between Staphylococcus aureus and Lactococcus rods presented in figure 10.

Influence of Interactions Occurring Between Micro-Organisms…

187

10 9 8 7 6 Log cfu/g 5 4 3 2 1 0 0

2

4

7

14

Time (days) inoculum9

inoculum7

inoculum5

inoculum3

Figure 10. The behaviour of Staphylococcus aureus in lactic acid curd during storage.

At constant number of Lactococcus sp. cells in lactic acid curds produced in model conditions, the dying-out time of staphylococci is dependent on the initial inoculum of Staphylococcus aureus. Analyses of predictions of staphylococci changes in model conditions, where only one population occurs (apart from staphylococci), show a completely different behaviour of these bacteria than the one observed in environment containing several populations (figure 11). The reduction time for monoculture in laboratory model conditions is different than for dicultures in the same conditions (table 11). Table 11. Inhibition of growth of staphylococci tvarog curd depending on the inoculum level Inoculum Log jtk/g

10 9 8 6 3

Size of reduction of Staphylococcus areus population count log10cfu number 4 5 6 Reduction time (days) 7 10 11 3 9 12 4 5 6 5 6 7 3 cycles 2 days -

Own research.

The behaviour of staphylococci investigated in lactic acid curds in the presence of three other strains of micro-organisms is presented in the figure below.

188

Izabela Steinka

8 7 6 5 Log cfu/g 4 3 2 1 0 0

2

4

7

14

Time (days) Yeast

Enterococcus sp.

Staphylococcus aureus

Figure 11. The influence of interactions among staphylococci, enterococci, yeast and lactic acid bacteria in a curd on population behaviour during storage.

Changes in count of yeast, enterococci and staphylococci populations show rather synergistic interactions among these cells as well as small inhibiting effect on staphylococci caused by present populations of Lactococcus, Enterococcus and yeast. Changes observed in the so-called ‘co-cultures’ could be determined as logit and described with the help of polynomial equations. 4

3

2

2

Y1 = 0.3771x - 4.7058x + 20.293x - 34.744x + 23.85 r =0.99 3

2

4

3

2

Y2= -0.0783 x + 0.5279 x - 0.4838x + 3.654-yeast r = 0.98 2

2

Y3= -0.1733 x + 1.9933 x - 7.2567 x + 7.6767x + 4.5 r = 0.99 where: Y1- enterococci count, Y2 - yeast count, Y3 - staphylococci count, x – time of refrigerating storage 2

r – equation coefficient of determination

(1) (2) (3)

Influence of Interactions Occurring Between Micro-Organisms…

189

THE INFLUENCE OF HERMETIC PACKING ON STARTER MICROFLORA AT HIGH CONTAMINATION LEVEL OF TVAROGS The influence of packing on micro-organism populations in tvarogs has to be considered in the aspect of the behaviour of both technological and re-infecting micro-flora. Irrespective of the number of populations infecting raw material or a curd constituting semi-product for the manufacture of tvarogs, the technical micro-flora occurs in the conditions of hermetic packing. The conducted research on the influence of vacuum packing on starter bacteria populations showed a varied behaviour of these micro-organisms, depending on whether they occur as the only cultures present in soured raw material or a curd, or whether they are present in ready products (tvarogs). Tests were performed during 14 days of curd and tvarog o storage at the temperature of 6±2 C. Evaluation of dynamics of changes of lactic acid bacteria populations was conducted during storage of lactic acid curds and tvarogs manufactured in model conditions, in dependence on vacuum packing. It was observed that vacuum packing did not influence changes of allochthonous microflora count up to day 7 of storage of curds manufactured in model conditions. The dynamics of changes of lactic acid bacteria were varied during storage of model lactic acid curds and tvarogs manufactured in industrial conditions.

9 8 4

3

2

y = 0,0363x - 0,5158x + 2,4938x - 4,7542x + 11,15 R2 = 1

7 6 4

3

2

y = -0.0742x + 0.7433x - 2.4308x + 3.0217x + 4.54

5 log cfu/g 4

2

R =1

3 2 1 0 0

2 Curd B

4

7

Time of storage (days) Curd F polynom.

Figure 12. Changes of Lactic Acid Bacteria in curd B and tvarog (F) during storage.

14 polynom.

190

Izabela Steinka

As a result of souring the milk with pure lactic acid bacteria cultures, the lactic acid curds were obtained, in which count of lactic acid bacteria after 14 hours of souring was equal from 8.22 to 8.41 log10cfu/g. In tvarogs produced industrially, number of lactic acid bacteria ranged from 5.44 log10cfu/g to 6.32 log10cfu/g (figure 12). Statistically significant differences were observed in sizes of Lactococcus sp. population in tvarogs (C) and curd manufactured in model conditions (B) for á=0.05. The average differences between count of lactic acid bacteria for C and B were significant and were equal to t = - 1.9280 and t = –3.9967. Number of lactic acid bacteria in vacuum–packed tvarog LT could be expressed with the help of mathematical equation presented below: 2 L T = 31.295-3.05•LSP r -0.706 , r 0.498 , p 0.16

á0.05

(4)

LT – count of lactic acid bacteria SP – vacuum-packed curd T – vacuum-packed tvarog Correlation coefficient between variables indicated high correlation between counts of lactic acid bacteria in both tested products, however the value of coefficient of determination 2 R showed that only 33% of variability of Lactococcus sp. count in tvarogs (C) could be explained with changes observed in stored curds produced in model conditions (B). During storage of curds A and B (Curd A was a control sample and remained without packaging, while curd B was vacuum-packed and manufactured in model conditions), no statistically significant differences in the dynamics of changes of Lactococcus sp. count were observed. However, from obtained data it results that vacuum packing constitutes factor influencing the dynamics of growth of technological micro-flora population in lactic acid curd during final period of storage in refrigerating conditions (figure 13). Regardless of the presence of packaging, the significant differences were detected in the dynamics of changes of Lactococcus sp. population between day 7 and 14 of curd storage (figure 13). In control curds remaining without packaging (A) the level of lactic acid bacteria after 14 days was the same as before storage (A), while in vacuum-packed curd (B) the decrease in population count by 1 logharithmic cycle was observed (figure 13).

191

Influence of Interactions Occurring Between Micro-Organisms…

Log cfu/g 8,5 y= -0,0133x4 + 0,1083x3 - 0,2167x2 + 0,0117x+ 8,33 R2 = 1

8,4 8,3 8,2 8,1 8 7,9

y= 0,0363x4 - 0,5158x3 + 2,4938x2 - 4,7542x+ 11,15 R2 = 1

7,8 7,7 7,6 0

2

4

7

14

Time of storage (days) Curd A

Curd B

polynom.

polynom.

Figure 13. Changes of Lactic Acid Bacteria count in curds A and B during storage.

Polynomial equation of the fourth order describing variations of Lactococcus sp. population count in curds A and B showed a different direction of observed changes (figure 13). LSP=0.25951+0.96344LSN

2

r 0.509, r 0.259 , p =0.381

á 0.05

(5)

L- number of lactic acid bacteria, SP - vacuum-packed curd SN - non-vacuum packed curd However, in the case of curds produced in similar technological conditions, the differences in dynamics of changes of lactic acid bacteria for vacuum-packed curds and curds remaining without packaging were not as significant as differences between micro-flora of curds and lactic acid cheese (figures 12 and 13). In vacuum-packed products of animal origin, changes in lactic acid bacteria population count and their metabolism are varied and dependant on type of the product, packing system, time and way of storage as well as secondary micro-flora present in a product (Holly et al. 1992, Jackson et al. 1992, Nicolai et al. 1993). Confirmation of the importance of vacuum-packing in stabilising the growth of lactic acid bacteria can also be the model developed for tvarog produced in industrial conditions.

192

Izabela Steinka

From the presented model, it results that in lactic acid cheese manufactured in industrial conditions, the behaviour of applied autochthonous micro-flora is influenced by other factors than in the case of non-packed semi-products (curds). Probably, it is dependent not only on technologies applied in production plants, but also on the presence of micro-flora re-infecting milk. Determined equation of linear regression for vacuum-packed and unpacked curds show that only 25.8% of changes of lactic acid bacteria count can be explained by application of vacuum-packing. This is consistent with observations of Maniar et al., who detected significant discrepancy between lactic acid bacteria count in cottage cheese, depending on level of atmosphere modification of the above-mentioned packaging wherein the product was stored (Maniar et al. 1994). Whereas, the predictive model of Nicolai describing the dynamics of changes in vacuumpacked meat indicates the significant influence of the environmental pH value on changes of lactic acid bacteria count during storage of packed products. The observed behaviour of lactic acid bacteria during refrigerating storage of the curds without re-infection is not significantly dependent on the presence of packaging up to day 7 of curd storage. Changes in fermentation bacteria population count during storage of vacuumpacked curds produced in model conditions and industrially manufactured curds show significant differences in relation to the semi-product.

INVESTIGATION OF INTERACTIONS IN TVAROGS OF HIGH MICRO-FLORA CONTAMINATION LEVEL From numerous research on lactic acid cheese (Steinka 2003a), it results that time of tvarog storage is not the only one and the most significant factor determining the population size in products. In minor part of available literature, the changes of microbiological quality of cottage cheese and tvarogs packed using vacuum or in modified atmosphere are described Maniar 1994, Fedio 1994, Panfil –Kuncewicz 1997a,b, Severini 1998, Steinka 1999c, Steinka 2001a,b). From data it results that packing system has a significant influence on the behaviour of micro-flora present in a product during its storage (table 12). Table 12. Equations of linear correlation for the influence of tvarog storage time on allochthonous micro-flora count Type of organisms

micro-

Enterococcus sp. Escherichia coli Staphylococcus aureus CP Mould Yeast Psychrotrophs

Lactic acid cheese packed into PA/PE laminate Equation r Y=38473+43157t 0.299 Y=1725.3-141.6t -0.348 Y=861.67-6.600t -0.021

Lactic acid cheese packed into Cryovac laminate Equation r Y=1575e2-126et -0.239 Y=170.42-14.29t -0.344 Y=129.17+10.714t 0.157

Y=7170.4+307.86t Y=6845e2+53759t Y=8445e2-102E2t

y=72.917+15.893t Y=-148e3+66775t Y=-295.6+274.02t

Steinka 2003a, t - storage time.

0.000 0.481 -0.090

0.170 0.631 0.261

193

Influence of Interactions Occurring Between Micro-Organisms…

Application of modified atmosphere in packing of lactic acid cheese does not guarantee inhibition of growth of all types of micro-organisms present in a product before storage (Kuncewicz 1997a,b). The research of Kuncewicz showed that application of 80% of CO2 additive for packing tvarogs limited the growth of mould and coliforms during refrigerating storage, however it did not inhibit the growth of yeast. Vacuum packing of tvarogs in conditions without Good Manufacture Practice also does not guarantee the reduction of micro-flora during storage of products in refrigerating conditions (Steinka 1999c). An attempt to create model predicting changes of quality and safety of hermetically packed lactic acid cheese had to take interactions occurring micro-flora present in tvarogs into account. Changes of micro-flora population count present in tvarogs could result from the change of conditions occurring in the space between the packaging and the products during storage. In lactic acid cheese packed with vacuum system and into Cryovac, the different tendency in behaviour as well as varied dynamics of change of micro-organism population count is observed. Table 13. Tendency of changes of tvarog allochthonous micro-flora Type of micro-flora

Enterococci Escherichia coli Moulds Yeasts Psychrotrophic bacteria Coagulase positive Staphylococci Coagulase negative Staphylococci

Type of packaging PA/PE Equation describing tendency of changes

R2

Cryovac Equation describing tendency of changes

r2

y = -0.82x2 + 4.28x + 0.32 y = 0.07x2 - 1.27x + 4.5 y = 1.21x2 - 4.77x + 7.6 y = -0.27x2 + 1.61x + 3.71 y = -0.15x2 + 0.59x + 5.31

0.999 0.997 0.999 0.999 0.999

y = 0.74x2 - 3.55x + 8.08 y = -1.15x + 3.4167 y = -1.675x2 + 5.515x+0.97 y = 0.41x2 - 0.44x + 3.6 y = 0.49x2 - 1.41x + 3.43

0.999 0.997 0.999 0.999 0.999

y = 0.205x2 - 0.355x + 2.97

0.999

y = 0.21x2 - 1.02x + 3.58

0.999

y = 0.255x2 - 0.525x + 3.71

0.999

y = -0.43x2 + 1.54x + 1.8

0.999

x-storage time.

Analyses of data also indicate the existence of significant variation in the course of function describing changes in count of all other tested groups of facultative anaerobic bacteria such as enterococci, staphylococci, E. coli after 14 days. Linear equations cannot be used for describing changes of tvarog micro-flora (table 12). The courses of these functions and consistency of predicted data with empirical data for both tvarog packing systems are not identical. Coefficients of linear correlation r indicate weak or lack of linear correlation between storage time and the behaviour of surface micro-flora population. Differences in tendencies of changes of micro-organisms present in tvarogs depending on packing system can be observed. It is confirmed by the derived equations describing tendencies of changes presented in table 13. Due to the possibility of occurrence of multi-species population occupying the surface of tvarogs as well as lack of predictive models concerning tvarogs, it became important and

194

Izabela Steinka

necessary to create such a predictive model for evaluation of quality of stored products, which will take interaction occurring among surface micro-flora into account. The research of Steinka et al. proved that the level of micro-flora present in tvarogs showed diversification dependent on packing method (Steinka et al. 1998, 2001c Steinka 2001a). In tvarogs packed into cryovac, after 14 days of storage, the lower number of Enterococcus sp. and Escherichia coli by 1 log cfu/g on the average was detected in comparison with tvarogs packed into PA/PE. The count of coagulase-negative Staphylococcus aureus, mould and psychrotrophs was higher by 2 logarithmic cycles on the average in tvarogs packed into PA/PE. The level of yeast and coagulase-positive staphylococci in tvarogs packed into both types of packaging was similar after storage (Steinka 2003a). Equation of linear correlation determined for evaluation of influence of storage time on the dynamics of population changes showed that only 1.53-19.7 % of variability could be explained with the help of linear models. Models of changes of micro-flora populations present in tvarogs packed into Cryovac laminates were characterised by low coefficients of determination R2 at the level of 0.0260.3716. Microbiological data were subjected to multiple regression analysis. Determined coefficients of semi-partial correlation are presented in tables 14 and 15. Values of partial correlations allowed evaluation of the influence of interactions among micro-organisms on changes in count of individual groups of micro-organisms in lactic acid cheese (table 14). The obtained test results showed that enterococci, mould and psychrotrophs had the greatest influence on determining changes of surface micro-flora of tvarogs packed into PA/PE had. In lactic acid cheese packed into cryovac, the significant reciprocal influence of facultative anaerobic bacteria on changes of other micro-organism populations was detected (table 15). Table 14. Coefficients of multifactor regression correlation in equations of micro-organism interactions in tvarogs packed into PA/PE Type of interactions between micro-organisms in tvarogs packed into PA/PE

r

2 r

Enterococcus sp. E. coli

Yeast Psychrotrophs

-0.202 -0.153

0.235 0.066

Staphylococcus aureus CP

Yeast Mould Mould Psychrotrophs Staphylococcus aureus Enterococcus sp. Staphylococcus aureus Psychrotrophs

0.257 -0.155 -0.117 0.190 -0.155 -0.175 0.215 0.214

0.013 0.024 0.036 0.060 0.046 0.060 0.046 0.046

Yeast Mould

0.249 0.184

0.062 0.033

Staphylococcus aureus CN Mould Yeast

Psychrotrophs

Steinka 2003a, CP- coagulase-positive Staphylococcus aureus, CN- coagulase-negative Staphylococcus aureus.

Influence of Interactions Occurring Between Micro-Organisms…

195

Table 15. Coefficients of multifactor regression correlation in equations of micro-organism interactions in tvarogs packed into Cryovac Type of interactions between micro-organisms in tvarogs packed into Cryovac

r

2 r

Enterococcus sp.

Staphylococcus aureus CP E.coli Enterococcus sp. Staphylococcus aureus CN Yeast Mould

0.146 0.753 0.767 0.409 0.411 0.020

0.021 0567 0.589 0.167 0.169 0.000

Staphylococcus aureus CN

Enterococcus sp. Yeast Escherichia coli

0.294 0.252 -0.120

0.086 0.063 0.014

Mould Yeast

Psychrotrophs Staphylococcus aureus CP Staphylococcus aureus CN Mould

0.678 0.329 -0.206 0.668

0.460 0.108 0.042 0.447

E. coli Staphylococcus aureus CP

Psychrotrophs

Steinka 2003a, CP- coagulase-positive Staphylococcus aureus, CN- coagulase-negative Staphylococcus aureus.

Equations presented in this paper describing variability of surface micro-flora of tvarogs, depending on storage conditions suggest the existence of factors other than time that influence changeability of micro-organism populations. Coefficients of linear correlation determined for equations describing population of staphylococci in both packing systems show that the behaviour of these micro-organisms is probably dependant not only on atmosphere occurring inside the packaging in both packing system, but also on other factors requiring further research. The above-mentioned observations are the reasons for which it is difficult to evaluate unequivocally the behaviour of staphylococci and to compare it with predictions obtained with the help of a predictive computer program called Pathogen Modelling Program. From predictions obtained as a result of simulations of Pathogen Modelling Program for model conditions at pH 4.7, it results that time necessary for achieving reduction in staphylococci count by only 1 logarithmic cycle at temperatures 4, 6 and 8°C is equal to 11.3, 13 and 14 days respectively. Data obtained through simulation of Pathogen Modelling Program show a low rate of variability in these micro-organism count in model conditions, what is also consistent with tendencies observed in tvarogs (Steinka 2003a). According to Kimur, the behaviour of staphylococci and Escherichia coli is dependant on the level of carbon dioxide and oxygen in the environment (Kimura et al. 1998). The tendencies of staphylococcus growth observed in tested cheese coming from Cryovac and PA/PE laminates show a significant meaning of atmosphere composition within the space under packaging surface for the growth of this group of micro-organisms. Temperature is not a factor that limits absolutely the growth of staphylococci. Their slow growth below the temperature of 10°C is observed in different types of food (Lee Wong et al. 2002). In tested tvarogs, the significant reduction in number of Escherichia coli rods during storage is observed. From literature, it results that this pehnomena is closely connected with sensitivity of rods to action of organic acids (Hsiao et al. 1999).

196

Izabela Steinka

Adams et al. (1995) proves that E. coli bacteria are sensitive to low pH values of the environment. However, from model research it results that during production and storage of a product at room temperature conditions, the reduction of Escherichia coli cells occurs not before the acidity reaches the value of pH 3.4. Although Buchanan et al. (1992) did not observe the growth of E. coli at the temperature of 4°C, Jordan et al. (1999) detected the ability of non-toxicogenic strains of E. coli O157:H7 to survive in the environment of pH 3.0, at the temperature of 37°C. This research showed that even after 3 days, it was possible to identify the presence of a significant number of cells that survived in such conditions. This allowed assuming that the growth of Escherichia coli in a product was not influenced by the packing system, but by a low temperature and reciprocal antagonistic interactions occurring among the micro-organisms. The size of enterococci population detected in tvarogs can be related to the literature information determining the level of contamination of unripened cheese (Baumgartner et al. 2001, Centeno et al. 1999). In the literature, there are no models available describing the growth of fungi in vacuumpacked unripened cheese such as tvarog. However, the issue of determining the number of these micro-organisms during storage of unripened cheese in conditions of modified atmosphere was undertaken in several papers (Fedio et al. 1994, Maniar et al. 1994, Westal et al. 1998). Yeast constituted quantitatively predominant micro-flora in tvarogs packed with both systems. Changes in populations showed high linear correlation with time of storage in refrigerating conditions. The dynamics of growth of yeast in vacuum-packed products confirmed tendencies observed by Westal (Westal et al. 1998). From data obtained by Westal, it resulted that as much as 25% of tested cottage cheese with additives packed in modified 6 atmosphere showed the significant increase in yeast count in a product ranging from 10 up 8 to 10 cfu/g after 2-5 days. The qualitative and quantitative composition of yeast in cottage cheese, as well as production of spores remained in a strong dependence on percentage composition of modified atmosphere. The tendencies of changes of psychrotropic micro-organisms observed in tvarogs are hard to be related to the models accessible in literature. The survival rate of Pseudomonas sp. during storage of cottage cheese was presented by Brocklehurst and Lund (Brocklehurst and Lund 1988). There are only a few publications mentioning changes of psychrotroph populations during tvarog storage. Tests conducted on cottage cheese aiming at determination of this micro-flora were performed by Fedio et al. 1994, Maniar et al. 1994, Neugebauer et al. 2005. Evaluation of changes of psychrotrophs populations in tvarogs packed with a hermetic and vacuum system was conducted by Steinka et al. 1999b. Packing tvarogs into Cryovac laminates is often preceded by blowing nitrogen through the product surface, what causes that the initial concentration of carbon dioxide at tvarog nonvacuum packing is different in comparison with its content in vacuum packaging. Complete modification of gaseous atmosphere involving addition of 100% N2 during packing of unripened cheese is commonly applied in many countries (Adams et al. 2000). Interpretation of change in micro-organism count in hermetically-packed tvarogs gives rise to many difficulties. In tested tvarogs packed with vacuum system, the existence of interactions were detected among micro-organisms that are typical in model conditions and difficult to precise unequivocally.

Influence of Interactions Occurring Between Micro-Organisms…

197

CHARACTERISTICS OF INTERACTIONS OCCURRING AMONG MICRO-ORGANISMS IN HERMETICALLY-PACKED TVAROGS The results of tests conducted on tvarogs (Steinka 2003a) allow saying that in vacuumpacking conditions, a major part of interactions occurring among micro-organisms had antagonistic character. It was indicated by negative values of correlation coefficients. In these tvarogs, the antagonistic interactions between yeast and enterococci, yeast and staphylococci, Escherichia coli rods and psychrotrophs as well as fungi and staphylococci were detected. The interactions between bacteria and fungi are more often observed in mould cheese. The antagonistic influence of Penicillium roqueforti on such pathogens as Escherichia coli or Staphylococcus aureus was described by Larsen (1997). The antagonism of fungi in relation to pathogens was conditioned by proteolytic and lipolytic action of fungi. High activity of enzymes decomposing proteins and fats results in high activity in inhibiting bacteria cells (Larssen 1997). It turns out that metabolites of fungi such as acetaldehyde and benzaldehyde play a significant role in antagonism occurring between fungi and pathogenic bacteria. Fungi of Penicillium type show the ability to inhibit Staphylococcus aureus and Listeria monocytogenes at the temperature of 150C as soon as after 6 days. Mould occurring in tvarogs include most of all Penicillium expansum and Aspergillus flavus. Their metabolic activity could contribute to inhibition of growth of staphylococci cells, however this inhibition was not effective. In tested lactic acid cheese, the growth stimulation occurred in the case of yeast and staphylococci, as well as yeast, mould and psychrotrophs. The direction of changes indicated reciprocal actions stimulating the growth of one population by the other. In laboratory conditions, the antagonistic action of Escherichia coli described in literature concerned staphylococci. In tested tvarogs, this type of interactions was observed in the case of Cryovac packaging. The synergistic interactions between coagulase-negative staphylococci population and Escherichia coli rods was detected in tvarogs packed into cryovac. The results of multifactor analysis did not confirm such a strong antagonistic interaction between staphylococci and psychrotrophs as were observed in the previous testing (Steinka 2001b) of vacuum-packed tvarogs. In model conditions of growth, the interactions between Escherichia coli and Enterococcus faecalis both of synergistic and antagonistic character are observed (Usajewicz 1995). In tested tvarogs, the interactions of synergistic character occurring between these bacteria reflect the tendency occurring in model conditions of mixed culture described by Usajewicz (1995). The synergism between these bacteria was detected in tvarogs packed into cryovac. From conducted model research concerning growth of facultative anaerobic bacteria in lactic acid curd, it results that the behaviour of some micro-organisms determined in a tvarog differs from the one detected in curds (Steinka 2003a). Low temperature and the atmosphere occurring inside the vacuum packaging lower the dynamics of growth of populations and the metabolic activity of lactic acid bacteria. The observed changes in count of lactic acid streptococci in curds produced in model conditions and in stored tvarogs could be described with the help of polynomial equations of

198

Izabela Steinka

the third and fourth order. Significant differences in the course of curves illustrating those changes were indicated in the case of non-hermetic and vacuum packaging (Steinka 2003a). Antagonistic influence of lactic acid streptococci on micro-organisms present in tvarogs could be significantly lowered, due to conditions occurring inside the packaging. Despite the acknowledged antagonistic properties of lactic acid bacteria against pathogenic bacteria, lactic acid cheese can also constitute the medium wherein the growth of such micro-flora as staphylococci occurs (Ingham 1996, Belickova et al. 2001, Masa-Calpe 1996, Steinka 2001c, Steinka 2004). From the research of Kornacki (2002), it results than even at correctly conducted process of souring the milk with application of active lactic acid cultures, the reduction in count of micro-flora originating from postproduction contamination is not always obtained. It concerns e.g. coli rods, which are sensitive to action of lactic acid bacteria, and which number was not reduced during industrial production of tvarog, using butter cultures. Changes of staphylococci population observed in tvarogs during product storage probably do not result from the antagonism of Escherichia coli, because the count of faecal bacteria was too low to inhibit Staphylococcus aureus. In order to inhibit the growth of staphylococci effectively, more than 100-fold number of rods in relation to number of staphylococci, as well as temperature higher than 150C are necessary (Steinka 2001 b). The insignificant inhibition of growth of staphylococci population observed in tested tvarogs was probably caused by the growth of mould. It was stated that reduction in number of cells belonging to Staphylococcus aureus species could not result from antagonistic action of yeast in relation to these staphylococci. The correlation coefficients showed the stimulating influence of yeast on staphylococci in both packing system. It was confirmed by model tests conducted on lactic acid curds. In tvarogs, no inhibiting influence of yeast on majority of micro-organisms groups was detected, despite the fact that yeast constituted predominant micro-flora in hermetically-packed tvarogs. Slow reduction of Staphylococcus aureus in products did not occur similarly to the experiments performed in laboratory conditions. However, in tvarogs packed into cryovac, a strong synergistic relations among populations of yeast, mould and psychrotrophs was observed. The character of interactions occurring among micro-organisms in tvarogs could result from variability of gaseous atmosphere inside the packaging. Research of Westal proved a small inhibiting influence of modified atmosphere on the growth of yeast in products (Westal 1998). Test results obtained by Westal were controversial in relation to the theory of Fedio and Alves, who confirmed the increase in tvarog stability depending on the level of atmosphere modification in MAP system. In conditions of tvarog hermetic packing, it is hard to indicate unequivocally the synergism, antagonism or antibiosis between two micro-organisms, because tvarogs presented in mentioned cases constitute the bases for growth of multi-component population composed of many species of fungi and bacteria. It is hard to separate the influence of one group on the others, if it is considered that the growth of certain micro-organism is influenced by metabolites of 7 groups of microorganisms at the same time. It is also difficult to separate their common influence on chemical and physical characteristics of a products as well as basic packaging properties.

Influence of Interactions Occurring Between Micro-Organisms…

199

Application of response surface models as well as equations describing them in order to evaluate the microbiological quality allowed defining changes occurring in tvarogs, taking at least three groups of micro-organisms into account (Steinka 2003a). The advantage of applied method was the possibility to predict magnitude of population changes, depending on variability of two other species during storage. The exact knowledge of these interactions is necessary for conducting observations of changes in microbiological quality and safety, in order to compare obtained values with valid microbiological standards for tvarogs. The obtained surface response models allowed defining changes in time, depending on the initial size of tested populations. In this regard, mathematical equations describing this variability allowed predicting strictly defined interactions occurring in a products. However, surface response models have some certain limitations, which concern among others a small number of relationships that they can illustrate. Despite the above-mentioned limitations, the surface response models constitute an assumption for interpretation of observed phenomena. In order to illustrate the interactions occurring in both types of hermetically-packed tvarogs, the dynamic model should be developed, which will consider all interactions occurring in a given time both among the micro-organisms and between the products and packaging.

DEVELOPMENT OF MATHEMATICAL MODEL FOR EVALUATION OF QUALITY OF LACTIC ACID CHEESE Development of a model was based on an assumption that interactions occurring among micro-flora present on the tvarog surface and the interactions between them and the packaging constituted a base for determining a mathematical model of dynamic interaction among micro-organisms. Calculations were started from searching for the relationship between variables

X i (t ), i = 1,2,..., m, of the form:

X i (t ) = a i 0 (t ) + a i1 (t ) X 1 (t ) + a i 2 (t ) X 2 (t ) + ... + a ii −1 (t ) X i −1 (t ) + a ii +1 (t ) X i +1 (t ) + ...

+ a im (t ) X m (t )

(6)

during time t , t ∈< 0, T >, for i = 1,2,..., m.

Unknown coefficients a ik (t ), at defined i, i = 1,2,..., m, for k = 0,1, ,..., m, k ≠ i, are

determined using the least squares method, minimising sums of squares of deviations between empirical values given in table 16 and calculated according to the following formula (6)

Izabela Steinka

200

Δi = ∑ [ xij (t ) − a i 0 (t ) − a i1 (t ) x1 j (t ) − a i 2 (t ) x 2 j (t ) − ... − a ii −1 (t ) x1−1 j (t ) n

j =1

− a ii +1 (t ) x1+1 j (t ) − ... − a im (t ) x mj (t )] 2 , i = 1,2,..., m,

(7)

Table 16. Empirical values used for model development Variable

X 1 (t )

X 2 (t )

X 3 (t )

X m −1 (t )

X m (t )

Realisation number j

Variable realisation

1

x11 (t )

x 21 (t )

x31 (t )

x m −11 (t )

x m1 (t )

2

x12 (t )

x 22 (t )

x 32 (t )

x m −12 (t )

x m 2 (t )

3

x13 (t )

x 23 (t )

x33 (t )

x m −13 (t )

x m 3 (t )

n −1

x1n −1 (t ) x 2 n −1 (t ) x 3n −1 (t ) x m −1n −1 (t )

n

x1n (t )

x 2 n (t )

x3n (t )

x mn −1 (t )

x m −1n (t )

x mn (t )

Steinka 2003a.

From necessary condition of existence of function extreme Δi , i = 1,2,..., m,

∂Δi = 0, k = 0,1, ,..., m, k ≠ i, ∂a ik For each defined i, i = 1,2,..., m, the following system of equations was obtained

na i 0 (t ) + ∑ x1 j (t ) a i1 (t ) + ∑ x 2 j (t ) a i 2 (t ) + ... + ∑ xi −1 j (t ) a ii−1 (t ) + ∑ xi +1 j (t ) a ii+1 (t ) n

j =1

n

n

n

j =1

j =1

j =1

+ ... + ∑ x mj (t ) a im (t ) = ∑ xij (t ) n

n

j =1

j =1

∑ x1 j (t ) a i 0 (t ) + ∑ x1 j (t )x1 j (t ) a i1 (t ) + ∑ x 2 j (t )x1 j (t ) a i 2 (t ) + ... + ∑ xi −1 j (t )x1 j (t ) n

n

n

n

j =1

j =1

j =1

j =1

aii−1 (t )

+ ∑ xi +1 j (t )x1 j (t ) a ii+1 (t ) + ... + ∑ x mj (t )x1 j (t ) a im (t ) = ∑ xij (t )x1 j (t ) n

n

n

j =1

j =1

j =1

...

Influence of Interactions Occurring Between Micro-Organisms…

201

∑ xi −1 j (t ) a i 0 (t ) + ∑ x1 j (t )xi −1 j (t ) a i1 (t ) + ∑ x 2 j (t )x i −1 j (t ) a i 2 (t ) + ... n

j =1

n

n

j =1

j =1

+ ∑ xi −1 j (t )xi −1 j (t ) a ii−1 (t ) n

j =1

+ ∑ xi +1 j (t )xi −1 j (t ) a ii+1 (t ) + ... + ∑ x mj (t )x i −1 j (t ) a im (t ) = ∑ xij (t )xi −1 j (t ) n

n

n

j =1

j =1

j =1

...

∑ xi +1 j (t ) a i 0 (t ) + ∑ x1 j (t )xi +1 j (t ) a i1 (t ) + ∑ x 2 j (t )xi +1 j (t ) a i 2 (t ) + ... n

j =1

n

n

j =1

j =1

+ ∑ xi −1 j (t )xi +1 j (t ) a ii−1 (t ) n

j =1

+ ∑ xi +1 j (t )xi +1 j (t ) a ii+1 (t ) + ... + ∑ x mj (t )xi +1 j (t ) a im (t ) = ∑ xij (t )xi +1 j (t ) n

n

n

j =1

j =1

j =1

...

∑ x mj (t ) a i 0 (t ) + ∑ x1 j (t )x mj (t ) a i1 (t ) + ∑ x 2 j (t )x mj (t ) a i 2 (t ) + ... + ∑ x i −1 j (t )x mj (t ) n

n

n

n

j =1

j =1

j =1

j =1

a ii−1 (t )

+ ∑ xi +1 j (t )x mj (t ) a ii+1 (t ) + ... + ∑ x mj (t )x mj (t ) a im (t ) = ∑ xij (t )x mj (t ) n

n

n

j =1

j =1

j =1

i = 1,2,..., m, from which the unknown equation coefficients can be determined assuming that for each defined i, i = 1,2,..., m, :

⎤ ⎡bi 00 (t ), bi 01 (t ),..., bi 0i −1 (t ), bi 0i +1 (t ),..., bi 0 m (t ) ⎡a i 0 (t ) ⎤ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢bi10 (t ), bi11 (t ),..., bi1i −1 (t ), bi1i +1 (t ),..., bi1m (t ) ⎢a i1 (t ) ⎥ ⎥ ⎢ . ⎥ ⎢. ⎥ ⎢ ⎥ ⎢ ⎥ ⎢. ⎥ ⎢. ⎥ ⎢ ⎥ ⎢. . ⎥ ⎢ ⎥ ⎢ B t b t b t b t b t b t ( ), ( ),..., ( ), ( ),..., ( ) = ( ) ⎥ ⎢ Ai (t ) = ⎢a ii −1 (t ) ⎥ i ii −10 ii −11 ii −1i −1 ii −1i +1 ii −1m ⎥ ⎢ ⎢a (t )⎥ ⎢bii +10 (t ), bii +11( t ) (t ),..., bii +1i −1 (t ), bii +1i +1 (t ),..., bii +1m (t )⎥ ⎢ ii +1 ⎥ , ⎥, ⎢. ⎥ ⎢. ⎥ ⎢ ⎥ ⎢ ⎥ ⎢. ⎥ ⎢. ⎥ ⎢ ⎥ ⎢. . ⎥ ⎢ ⎥ ⎢ ⎥⎦ ⎢ b t b t b t b t b t ( ), ( ),..., ( ), ( ),..., ( ) a t ( ) im1 imi −1 imi +1 imm ⎣ im 0 ⎦⎥ ⎣⎢ im

(8)

202

Izabela Steinka ⎡c i 0 (t ) ⎤ ⎥ ⎢ ⎢c i1 (t ) ⎥ ⎥ ⎢. ⎥ ⎢ ⎥ ⎢. ⎥ ⎢. ⎥ ⎢ C i (t ) = ⎢c ii −1 (t ) ⎥ ⎢c (t )⎥ ⎢ ii +1 ⎥ , ⎥ ⎢. ⎥ ⎢ ⎥ ⎢. ⎥ ⎢. ⎥ ⎢ ⎣⎢c im (t ) ⎦⎥

(9)

where:

bi 00 = n, bi 0l = ∑ x lj (t ), l = 1,2,..., m, l ≠ i, n

j =1

(9a)

bik 0 = ∑ x kj (t ), k = 1,2,..., m, k ≠ i, n

j =1

bikl = ∑ x lj (t )x kj (t ), k = 1,2,..., m, k ≠ i, l = 1,2,..., m, l ≠ i, n

j =1

c i 0 = ∑ x ij (t ), n

j =1

c ik = ∑ x ij (t )x kj (t ), k = 1,2,..., m, n

j =1

systems of equations (8) can be expressed in a matrix form:

Bi (t ) Ai (t ) =C i (t ) for i = 1,2,..., m.

(10)

Hence, if matrix determinants Bi (t ), i = 1,2,..., m, are different from zero, then for each

defined i, i = 1,2,..., m, we determine unknown coefficients

a i 0 (t ), a i1 (t ), a i 2 (t ), a ii −1 (t ), a ii+1 (t ), a im (t ) of models from matrix equations

Ai (t ) = [ Bi (t )] −1 C i (t ) during time t , t ∈< 0, T >, where [ Bi (t )] −1 are inverse matrixes of a matrix Bi (t ).

(11)

Influence of Interactions Occurring Between Micro-Organisms…

203

By this, we identify approximate relationships between variables X i (t ), i = 1,2,..., m, in

defined time t , t ∈< 0, T > . Figure 14 presents the algorithm of determining the parameters of mathematical model of dynamic interactions between tvarog microbiological parameters and the packaging. The obtained polynomial equation served for developing a computer program called TWAROGI JMTPH, which takes the dynamic changes of micro-flora in time into account. In developed model, the predictions concerning changes in properties of applied packaging were also considered. The predictions applied to micro-biological parameters of products packed with a vacuum system and using the technology of shrink-wrapping of product with cryovac. The obtained predictions differed from each other as regards the absolute terms and values of constant coefficients. Below, the exemplary prediction of yeast population for day 11 of tvarog storage is presented: YPA/PE=126032.78+5.82x1+436.02x2-97.19x3+12.12x4-306.39x5+4.19x7

(12)

Yc=118770003.92-2.69x1+215.21x2+45.98.19x3+14.04x4-281.14x5+3.39x7

(13)

where x1-enterococci, x2-E. coli, x3-coagulase positive Staphylococcus aureus, x4-coagulase negative Staphylococcus aureus, x5 – mould, x7 –psychrotrophs Developed computer program allows quick prediction of magnitude and type of changes occurring in hermetically-packed tvarogs depending on packing technology of products. Although multi-parameter models used for predicting microbiological quality of products are criticised, they reflect conditions occurring in food to the much greater extent than other models (Baranyi 1996, Steinka 2003a). In predicting quality and safety of products, the mathematical models of a small number of parameters are most commonly acknowledged and applied. However, the disadvantage of the latter is that it omits many environmental factors responsible for the dynamics of observed changes. Up till now, one of a few available linear multi-parameter models was the model of Reinchart and Mohasci-Farkas (Ko³o¿yn – Krajewska 2003). The above-mentioned model took the influence of four environmental factors on inactivation of pathogenic bacteria into consideration. The presented evaluation of changes in quality of products packed with different technologies proves that the element of reciprocal interactions among micro-organisms, often omitted in predicting, can significantly influence the predicting quality. It should be also taken into consideration that in literature there are no models available connected with predicting the quality of tvarogs. Therefore, the presented program can constitute complementation of prediction models connected with evaluation of survival rate of micro-organisms in products of animal origin and can be useful in optimising the quality of hermetically packed tvarogs.

204

Izabela Steinka

START

v=1 i=1 Read in data Determine matrixes Bi(t), Ci(t) Determine matrix Ai(t)

i=m?

YES v=w?

NO

NO

i=i+1

v=v+1

YES Print matrix elements Ai(t), i = 1,…,m i=1 Determine matrixes E, Fik Determine matrix Dik i=m?

NO

i=i+1

YES Print matrix elements Dik, i = 1,…,m

Is it end ?

NO

YES STOP

Figure 14. Algorithm of determining parameters of mathematical model of dynamic interactions occurring among microbiological parameters of tvarogs.

Influence of Interactions Occurring Between Micro-Organisms…

205

THE INFLUENCE OF INTERACTIONS AMONG MICRO-ORGANISMS ON PHYSICO-CHEMICAL PROPERTIES OF LACTIC ACID CHEESE STORED IN REFRIGERATING CONDITIONS Multidirectional interactions cause that tvarogs packed with different systems (vacuum, MAP, atmosphere) differ from each other in physico-chemical (water content) and chemical (concentration of hydrogen ions) properties. Regardless of the applied packing system and the packaging hermetic properties, the loss in water content was observed in tvarogs relating to the packing system and product storage time. As soon as after two days of tvarog storage, the decrease in water content from 0.6 g up to 0.8 g could be observed in products packed into cryovac, depending on the packaging tightness (figures 15 and 16). In tvarogs hermetically packed with vacuum system, the drop in water content was equal to 0.9 g. Tvarogs coming from depressurised packaging indicated 1.2 g less water in a product, in comparison with control sample. The significant loss in weight by drying of surface of tvarog packed into Cryovac coming from both hermetic and non-tight packaging was detected after 7 days of product storage. Between day 7 and day 14 of storing the product packed into cryovac, the insignificant decrease in water content was observed in non-tight packaging and the insignificant increase in hermetic packaging.

75 74,5 74 y= 0,1164x2 - 0,9956x+ 75,246

73,5

R2 = 0,9857

%

73 72,5 72 71,5

y= 72,78x-0,0075 R2 = 0,2858

71 70,5 70 0

2

4

7

14

Time (days) Vacuum

Non-vacuum

polynom.

Figure 15. Changes in water content in tvarog packed hermetically with different system

power

206

Izabela Steinka

After 4 days, loss of water content in vacuum-packed products was equal to 1.3 g and 1.2g for tight and non-tight packaging respectively, in comparison with control samples. In vacuum packaging, the increase of water content in product was observed after 7 days of storage resulting probably from falling of water present on the internal packaging surface. After 14 days, water loss equal to 1.1g was detected in tvarogs coming from Cryovac laminate, and it was higher by 0.5 g than the change observed on day 2 of product storage. Changes in water phase of a product after 14 days of storage were equal to 1.3g for tvarogs packed into PA/PE laminates and 1.1g for tvarogs packed into Cryovac. The dynamics of changes of water content during storage of tvarogs were different in PA/PE and Cryovac packaging. However, in the final period of storage, the magnitude of changes in water phase of products packed hermetically with both systems were comparable, despite the fact that during initial period of storage the higher loss in water content was observed in vacuum packed tvarogs. Changes in water content during storage of tvarogs packed into Cryovac could be 2

described by the equation y = - ax + bx + c, whereas the tendency of changes in water phase n

of tvarogs packed into hermetic PA/PE laminates was illustrated by a power function y=ax . However, the above-mentioned equation described only 28.6% of water phase variability in a function of time. Unfortunately, coefficients of determination of other functions showed even lower values. This fact evidences the existence of complex packaging-product interactions observed in this packing system of tvarogs.

75 74,5 74 73,5

y= 74,225x-0,0108 R2 = 0,9541

73 Log cfu/g 72,5 72 y= 0,2257x2 - 1,4023x+ 74,066 R2 = 0,7407

71,5 71 70,5 0

2

4

7

14

Time (days) Vacuum

Non-vacuum

power

polynom.

Figure 16. Changes of water content in tvarogs packed with different systems coming from depressurised packaging.

Influence of Interactions Occurring Between Micro-Organisms…

207

In this research, the average water content on the day of purchase of tvarogs packed into Cryovac was equal to 74.3%. Water content in vacuum-packed tvarogs was at the level of 73.0% during that time, what is consistent with values contained within the Standard for this type of cheese. However, as soon as after two days of storage, the differences connected with packaging hermetic properties were observed. Changes in water phase of tvarogs manufactured in experimental conditions presented by other authors showed significant fluctuations during refrigerating storage (Œmietana et al. 1997). From tests performed on lactic acid cheese coming from trade and subjected to storage (Steinka 2004), it resulted that after 7 and 14 days of storage, tvarogs in vacuum packaging showed a similar tendency as the one quoted by Œmietana. Whereas the tendency of changes in water content in tvarogs packed into Cryovac was different from the one presented by Œmietana. Conducted numerous research showed that at significant level of contamination of tvarogs with micro-flora, the biological factors should be taken into account for mathematical description of the water fluctuation process. Water fluctuation between product surface and the packaging is determined not only by breathing of a product and its barrier properties, but also by micro-organism metabolism (Steinka 2005b). Equations of trend lines observed during changes in water content within time periods Ä2, Ä4, Ä7, Ä14 during tvarog storage are presented below. Water fluctuations were tested in vacuum-packed tvarogs containing level of micro-flora taken into account while developing the computer program (Steinka 2003a). The trend of changes was described by quadratic equation of the following form Y = 3.6x2 + 8.9333x – 6.9 r2 0.999

(14)

The obtained results proved the existence of the average linear correlation between initial water content in tvarogs and its level after 14 days, what is showed by the equation: Y14 = 67.651 + 0.0707• x0 r 0.4320 where: y- water content (15) At significant contamination with micro-flora, during storage of tvarogs, the insignificant influence of micro-organisms on fluctuation of water phase between packaging and product surface was observed. Table 17 below presents equations of linear correlation reflecting the influence of microorganism count on changes of water content during storage of tvarogs. Table 17. The influence of water on populations occupying tvarog surface Type of micro-organism

Equation

Enterococcus sp. Yeast Psychrotrophs

Y=11003E-148e3x Y=8593E3-104e3x Y=1157E4-150e3x

Steinka 2003a, x-water.

2 r -0,268 -0,244 -0,185

208

Izabela Steinka

As a result of analysis of multifactor regression of data obtained during tests on tvarogs showing a high level of contamination with micro-flora, it was observed that the growth of chosen types of micro-flora could contribute to the water fluctuation. The analysis of coefficient values of semi-partial and partial correlations showed that in vacuum-packed tvarogs, the populations of enterococci, yeast and psychrotrophs inter-reacted with other micro-organisms and influenced the water fluctuation in tvarogs in 12.9%, 9% and 4.6% respectively (Steinka 2003a). From data analysis, it resulted that water phase fluctuation in tvarogs after 14 days of storage was mostly influenced by interactions of yeast and psychrotrophs with other groups of micro-organisms.

SAFETY OF LACTIC ACID CHEESE Many years of microbiological research on lactic acid cheese (Steinka 1998-2006) showed the presence of coagulase-positive and coagulase-negative staphylococci both in tvarogs packed in a traditional way (parchment paper) and into hermetic packaging. From literature it results that the growth of staphylococci in food is determined by numerous factors (Neumeyer et al. 1990, Taub et al. 2003, Castillo-Rodriquez et al. 2002, Pereira et al. 1991, Feehery et al. 2004, Inhram et al. 2004, Schaffener et al. 2001, VernozyRozynand et al. 1998, Post et al. 1988, Zurea-Cosano et al. 2004, Lindquist et al. 2002, Sameshima et al. 1998, McCann et al. 2003, Pepe et al. 2006, Schaffener et al. 2001). Application of different types of additives in food should limit staphylococci growth in products. Among other, Zurera–Cosano et al. (2004) tested the combined influence of temperature o

7-19 C, acidity equal to 4.5-8.5 pH, content of NaCl equal to 0-8% and 0-200 ppm of nitrates on the growth and lag phase of staphylococci in aerobic and anaerobic conditions. Authors estimated parameters, comparing suitability of a surface response model and Davey model. Two kinetic parameters were evaluated, and it turned out that using surface response model allowed achieving better prediction as regards the bacteria growth in comparison with the second model. It was evidenced by values of bias and accuracy factors equal to bf1=1.06 and bf2=1.31 and af1= 1.17 and af2=1.37 respectively for aerobic and anaerobic conditions. In lag phase, Davey model indicated higher values of coefficients for these conditions. Stewart et al. (2002) investigated the influence of humidity, pH 4.5-7.0 and sorbate concentration on the growth of staphylococci. Tests results showed that saccharose and fructose constituted factors inhibiting the growth of staphylococci populations at neutral pH value. Sodium chloride turned out to be a significant inhibitor of bacteria only at lower pH values. Authors also presented that the addition of potassium sorbate contributed to inhibition of growth, especially when pH value was lower than 6.0.

Influence of Interactions Occurring Between Micro-Organisms…

209

Table 18. The influence of bacteria on the behaviour of facultative anaerobic bacteria populations in tvarogs depending on gaseous conditions Type of micro-organism Staphylococcus aureus

Escherichia coli

Yeast

Vacuum conditions

Variable atmosphere, micro-aerophilic conditions

N=5.941+1.5789NL

N=22.202-3.269NL

r 0.9045

r -0.8053

N=-1.706 +0.6677NL

N=25.407-4.137NL

r 0.2026

r -0.6378

N=7.6124-0.6321NL

N=-10.39 + 2.5435NL

r -0.3622

r 0.2382

Steinka et al. 2003b, N-population count, Nbfm – LAB count.

Tests conducted on tvarogs indicate the significant meaning of the environment on type and intensity of interactions occurring among micro-organisms. Inside the tvarog packaging and above the product surface, various conditions are formed depending on packing system. These could be micro-aerophilic or aerobic conditions, or modified atmosphere (MAP). This is important not only for technical micro-flora and micro-flora re-infecting tvarogs. Table 18 presents interactions occurring among micro-organisms present on the surface of lactic acid cheese in environment of different proportions of oxygen and carbon dioxide. From presented data it results that the growth of facultative anaerobic micro-organisms is dependent on the level of lactic acid bacteria, however the type and magnitude of this influence is dependent on environment wherein these interactions are observed (Steinka 2003a). In the case of staphylococci, values of linear correlation coefficients were equal to r 0.904 for vacuum-packed tvarogs and r -0.805 for tvarogs packed into parchment paper. An important issue is different direction of changes of staphylococci population under the influence of lactic acid bacteria in both types of packaging. The size of staphylococci population was not large in tested products. It was at the level of 2 log jtk/g. No statistically significant differences were observed between sizes of staphylococci populations in vacuum-packed tvarogs (hermetically) and tvarogs packed in a traditional way (non-hermetically). In traditionally packed tvarogs, the growth of lactic acid bacteria populations was corresponding to the decrease in number of staphylococci, whereas in vacuum-packed tvarogs – the growth of streptococci population count was correlated with the growth of staphylococci. The obtained results indicate high importance of packaging hermetic properties in determining the dynamics of changes of population re-infecting lactic acid cheese. Due to their physiology, staphylococci constitute the greatest problem among facultative anaerobic micro-flora. They have a strong ability to repair sub-lethal damages resulting from the process of product manufacture, what can be confirmed by research of many authors (Steinka 1999a, 2003a). Moreover, from conducted research it results that their cluster distribution in products is also the reason for improper evaluation of food.

210

Izabela Steinka Table 19. Presence of staphylococci in tvarog samples and on the packaging during 7 days of storage Tvarogs before storage

Tvarogs after 7 days of storage (40C)

Surface layer of a product

Packaging 2 [cm ]

Surface layer of a product

Packaging 2 [cm ]

Presence of staphylococcus

23.1%

30.1%

61.5%

61.5%

Absence of staphylococcus

66.9%

69.9%

38.5%

38.5%

Percentage contribution of tvarogs with presence and absence of staphylococcus

Steinka et al. 1999a.

From data presented by many authors (Castillo-Rodriqueza et al. 2002, Walls et al. 1996, Schaffener et al. 2001), it results that staphylococci give rise to difficulties in predicting food safety and the predictive microbiology is moderately useful for determining safety of products, in which these bacteria can occur. From conducted observations it results that products of metabolism of these bacteria can appear not before refrigerating storage of the products. Staphylococci can appear in tvarogs after 7 days or later, although during postproduction period their presence was not detectable (table 19). Nevertheless, tvarog composition can have the significant influence on the behaviour of staphylococci populations. From conducted research it results that after dropping 3

2

staphylococci in amount of 1 cm on 1cm of tvarog surface, and then vacuum-packing of the products into PA/PE packaging, the behaviour of these micro-organisms differed depending on the fat content. Lactic acid cheese manufactured by the same producer and originating from one lot contained 15 and 30% of fat. Products after addition of Staphylococcus aureus of 0

inoculum 3 log cfu/mL were stored after repacking at the temperature of 6±2 C for the period of 14 days. Analyses were carried out after 2, 4, 7 and 14 days. The tendency of changes of staphylococci population count in fat tvarogs could be expressed with a quadratic equation. In low-fat tvarogs, the decrease in lactic acid bacteria (LAB) count was accompanied by the growth of staphylococci population. Variability in number of these micro-organisms could be also described with a quadratic equation (figure 16). In the case of fat tvarogs, the growth of lactic acid bacteria was accompanied by the decrease of staphylococci count. From research it resulted that variability of lactic bacteria count in both types of tvarogs was not significantly dependant on storage time and temperature in 18.7% and 40% respectively (table 20). In tested fat tvarogs a weak linear correlation (r 0.235) between storage time and staphylococci count was observed. This can evidence the influence of lactic acid bacteria, other micro-flora not defined in this research and the fat content in products. Similar relationships were observed in the case of low-fat tvarogs (r 0.283).

Influence of Interactions Occurring Between Micro-Organisms…

6 y = -0,1733x3 + 1,4386x2 - 3,2981x + 6,38

5

R2 = 0,8138

4

y = -0.2658x4 + 3.1433x3 - 12.834x2 + 20.957x - 7.7 R2 = 1

Log cfu/g 3

2

1

0 0

2

4

7

14

Tim e (days) LAB

Staphylococcus aureus

polynom.

polynom.

Figure 17. The behaviour of staphylococci and lactic acid bacteria in fat tvarogs.

6 y = -0,277x + 5,075

5

R2 = 0,8113

4

y = -0,255x4 + 3,0267x3 - 12,17x2 + 19,148x - 5,98

Log 3 cfu/g

R2 = 1

2

1

0 0 LAB

2

4

Tim e (days) Staphylococcus aureus

7 linear

Figure 18. The behaviour of staphylococci and lactic acid bacteria in low-fat tvarogs.

14 polynom.

211

212

Izabela Steinka

Table 20 below presents equations of linear correlation, as well as equations describing the tendency of changes in both types of tvarogs. Table 20. Trend equations of changes of LAB and staphylococci in stored tvarog of different fat content Type tvarog

of

Type of micro-organisms Staphylococcus aureus

Fat

Low-fat

4 3 2 Y = -0.1808x + 2.2083x - 9.3492x + 15.772x – 5.15 2 r 0.999 Y = -0.098x + 3.586; 2 r 0.235 3 2 Y = 0.0383x - 0.385x + 1.0467x + 2.662 2 r 0.3416 Y = 3.5854e-0.0297x 2 r 0.2419 Y = 3.4786x-0.0614 2 r 0.1673 Y = -0.255x4 + 3.0267x3 - 12.17x2 + 19.148x 5.98; 2 r 1 Y = 0.195x + 3.281 2 r 0.2839 Y = 0.3612Ln(x) + 3.5201 2 r 0.1574 3 2 Y = -0.0333x + 0.4707x - 1.616x + 5.036 2 r 0.6005 Y = 3.3189e0.0477x 2 r 0.2307

Starter cultures (Lactococcus spp + Leuconostoc lactis) 3 2 y = -0.1733x + 1.4386x - 3.2981x + 6.38 2 r 0.8138 y = 0.2092Ln(x) + 4.3097 2 r 0.0846 y = 4.3287e0.0124x 2 r 0.0414 y = 0.064x + 4.318 2 r 0.049

y = -0.277x + 5.075 2 r 0.8113 y = -0.7409Ln(x) + 4.9534 2 r 0.9376 y = 4.962x-0.1685 2 r 0.938

Own study.

Table 21. Size of Staphylococcus aureus population simulated with Pathogen Modelling Program Concentration of lactic acid

0 0.3 0.5 0.7 0.8 1

Magnitude of reduction of Staphylococcus areus count Log10cfu 4 Reduction time (days) 92.87 61.65 52.21 48.17 47.78 50.12

Developed on the basis of PMP v 5.1.

5

6

116.09 77.06 65.26 60.21 59.72 62.65

139.3 92.47 78.31 72.25 71.67 75.18

Influence of Interactions Occurring Between Micro-Organisms…

213

Computer simulation developed with Pathogen Modelling Program shows that in conditions corresponding to different magnitudes of changes connected with action of alkalising micro-flora, the level of staphylococci is varied (table 21). From data obtained by Pathogen Modelling Program, it results that time of reduction of 0 staphylococci count by 1 logarithmic cycle at temp. 4 C and pH 4.7 is equal to 11.3 days. In tested tvarog, no reduction in count of these bacteria was observed after 14 days. In low-fat tvarog the increase in bacteria number by 1 logarithmic cycle during that time was detected. It evidences that in case of infection of tvarogs, starter cultures present therein will not inhibit the growth of staphylococci in products of low fat content. From previous research it also results that reciprocal interaction between lactic acid bacteria and staphylococci present at this level in fat tvarogs will be dependent on the stage of hermetic properties and type of applied packaging (Steinka et al. 2003c). In literature, there are a few predictive models available, which allow evaluation of the growth of these bacteria in food (table 22). Table 22. Predicting changes in staphylococci count in model conditions and in food Type of product Model conditions Bread

Sterile food

Model conditions Model conditions Tvarogs

Type of described changes Davey and surface response model of the growth at several environmental conditions Kinetics of dying out, death in variable environmental conditions, quasi-chemical model was compared with Gompertz model, and probabilistic model integrated with quasichemical was applied Defined environmental conditions. Gompertz model and comparison of obtained data with PMP and FMM predictions Environmental conditions. Gompertz model – growth of staphylococci. Kinetic models, environmental factor influence on „growth/no growth”

Source Zurera-Cosano 2004

Polynomial survival rate

Steinka 2003

Taub 2003

Walls 1996

McCann 2003 Stewart 2002

Own study.

An attempt of developing such a model was also made by Steinka (2003a), Steinka et al. (2005a).

DEVELOPMENT OF THE PROGRAM FOR EVALUATION OF STAPHYLOCOCCI GROWTH The program was developed in Borland Delhi 2.0. language and serves for evaluation of microbiological quality of lactic acid cheese vacuum-packed into PA/PE as well as shrinkwrapped with Cryovac packaging. Values of differences in predictions for tvarogs in both types of packaging showed the higher dynamics of changes in bacteria count in tvarogs

214

Izabela Steinka

packed into Cryovac. It was observed that JMTPH computer program was useful for predicting changes of staphylococci count in stored tvarogs. Type of packaging and the packing system had the influence on evaluation of tvarog safety. In the computer program used for predicting changes of these bacteria populations in tvarogs packed into PA/PE laminates, 4 microbiological parameters were applied up to 7 days of storage and 5 parameters during further storage period. In the case of prediction regarding tvarogs packed into Cryovac, 6 microbiological parameters were applied. In conditions of shrink-wrapping the product with Cryovac, no influence of yeast on the level of determined staphylococci was observed in the second week of storage (table 23). From predictions obtained for vacuum-packed tvarogs, it resulted that at maximum product contamination equal to 3.32 log cfu/g, the level of staphylococci equal to 4 log10 cfu in one gram of tvarogs remained between day 3 and day 21 of storage. The prediction of change of coagulase-positive staphylococci population in tvarog during storage is presented in table 23. Table 23. Predictions of staphylococci growth in tvarog depending on packing system Day of storage

Parameters of mathematical model

Prediction result Log cfu/g

TVAROGS PACKED INTO PA/PE 3

Y=-3353.77+0.04 x1 +5.84 x2-0.11 x4+1.07 x5

4.11

7

Y=-1768.74+0.02x1+4.78x2-0.1x4+0.94x5

4.09

13

Y=608.79-0.01x1-3.17x2 +0.09x4+0.75x5+0.01x7

4.08

15

Y=1040.13-0.02x1+2.64x2-0.09x4+0.62x5+0.01x7

4.31

21

Y=3778.84-0.04 x1+1.04x2-0.08x4+0.49x5+0.02x7

4.06

TVAROGS PACKED INTO CRYOVAC 3

Y= 257.67 +0.01x1-+93.60x2-0.73x4-1.33x5+0.03x6-0.13x7

3.91

7

Y= 70.78-0.05x1+746.33x2-0.99x4-2.27x5+0.03x6+0.09x7

5.03

9

Y=-22.66+0.08x1+1072x2-0.06x4-2.271 213x5+0.01x6+0.76x7

5.14

15

Y=-303-0.16x1-2051.8x2-0.85x4+22.92x5-1.45x7

5.31

21

Y=-583.33-0.24x1+3030.90x2+1.64x4+33.72x5-0.01x6-2.11x7

5.48

From presented data the significant differences resulted in count of staphylococci population in tvarogs, depending on packing way and type of applied packaging. Differences between sizes of staphylococci populations determined in tvarogs packed into both types of packaging could be observed as soon as from the day 3 of product storage. Difference values in predictions obtained with the help of JMTPH computer program indicated the higher dynamics of changes in tvarogs packed into Cryovac.

215

Influence of Interactions Occurring Between Micro-Organisms…

In optimum model conditions and at temperature of 220C, the growth of staphylococci population shows significant dynamics. The increase in count of these bacteria by 3 logarithmic cycles can occur during 72 minutes. Simulation carried out with the Pathogen Modelling Program for the temperature 40C and pH 4.7 showed that staphylococci reduction time by 1 logarithmic cycle was equal to 11.3 days (6). This evidences the significant resistance of these bacteria to the influence of low temperature and acid in the environment. Computer simulations conducted with Pathogen Modelling Program showed that other microorganisms e.g. Salmonella spp., were characterised by different dying-out dynamics in comparison with staphylococci population at the temperature of 60C and pH 4.7. Reduction in number of rods by 1 and 2 logarithmic cycles occurred from 7 up to 14 hours respectively. It also should be noticed that presented predictions developed using Pathogen Modelling Program concerned the individual growth of micro-organisms without taking the presence of other micro-organisms into account. Predicting the bacteria growth in model conditions in many cases assumes the existence of microbiological rules obligatory for monocultures. From our previous research it results that the behaviour of both staphylococci and Listeria spp. rods in hermetically packed tvarogs was different than expected (Steinka et al. 1999b). Also the growth of staphylococci in conditions occurring in food can proceed with different dynamics than the one observed in model conditions, what was proved by research of Wallas et al. (1997) and Smittle et al. (1994). From data obtained by other researchers it resulted that temperature also did not constitute the factor that limited absolutely the growth of staphylococci. The slow increase in staphylococci count below temperature of 10oC was observed in different types of food by Lee Wong et al. (2002). The number of staphylococci cells determined in tvarogs with JMTPH computer program presented in this paper was similar to sizes of populations obtained by e.g. Belickov et al. (2001). The mentioned-above 1

4

authors isolated staphylococci from tvarogs at the level from 9x10 cfu/g up to 1.07x10 cfu/g. Research conducted by those authors proved the presence of staphylococci in other fermented dairy products such as sheep cheese, buttermilk and yoghurt, what can evidence the resistance of staphylococci to antagonistic influence of lactic acid bacteria. The above-menioned factors show that presented prediction of staphylococci changes obtained with the help of JMTPH computer program can be actual and can reflect the behaviour of population in conditions of hermetically packed tvarogs.

SURVIVAL RATE OF STAPHYLOCOCCI ON THE SURFACE OF TVAROGS AND THE SYNTHESIS OF ENTEROTOXIN DURING STORAGE One of the determinants of food safety, apart from the absence of vegetative forms of pathogenic bacteria, is the absence of toxins released into food. The hazard of uncontrolled synthesis is created by mostly such bacteria as Staphylococcus aureus. The intensity of enterotoxin production by Staphylococcus aureus strains is dependent on environmental conditions, among which the most significant are: temperature, environment pH value, redox potential and gaseous atmosphere (Adams 1995). The synthesis of staphylococcal enterotoxin is dependant on many factors such as: environment in which staphylococci population occurs, compositions of a medium, micro-organism count and the presence of inhibiting substance.

216

Izabela Steinka

As it results from the literature, despite the antagonistic influence of other micro-flora present in the environment (even lactic acid bacteria), in certain conditions the toxicogenesis can occur. In order to evaluate food safety, it is possible to apply microbiological predicting that enables approximate assessment of population growth. The existing computer programs help to determine lag phase, magnitude of population count reduction or the rate of its growth, however they not always take all environmental conditions into account. Therefore, it seems to be important to compare the growth and survival rate of populations in model conditions and in monoculture with data obtained in food. From research of Neumayer et al. (1990), it results that in a medium enriched with ingredients occurring in such vegetables as pea or bean, the bacteria growth occurs with significant dynamics in model conditions, however the concentration of produced enterotoxin is highest in a medium where pea is present. Enterotoxin accumulated therein in amount of 14-15 ng per mL of a medium. The presence of high concentrations of valine, cistine and arginine contributes to the synthesis of SEA by staphylococcus aureus in amount smaller than SEB, SEC (Bergdol et al. 1989). Moreover, the same research also showed that glucose had inhibiting influence on the production of SEB and SEC. Opinions as regards number of staphylococci necessary for enterotoxin production are varied. Halin–Dohnalek et al. (1989) indicated the level of 106 cfu/g in food environment of high fat content. In sterile milk, this number is equal to 10 6,5 cfu/mL (Fukijawa et al. 2006). Detectability of enterotoxin is not only dependant on the level of cells. Otero (1988) did not detect the presence of enterotoxin in cheese even at the level of 108 cfu. Whereas Delbar et al. (2006) detected enterotoxin after 24 hours in cheese, where number of staphylococci was equal to 5.55-5.6 log10cfu/g. This confirms our observations (Steinka 2004). While testing ready-to-serve food, Bahk et al. (2006) observed that as much as 29.73% of samples showed the presence of staphylococci at the level > 105 cfu/g, what could favour enterotoxin synthesis. From tests conducted on hermetically packed tvarogs and tvarog coming from unsealed packaging, it resulted that enterotoxin was formed more often in the latter. Research showed that a small supply of oxygen to the product favoured the enterotoxin synthesis (Steinka 2004). The presence of enterotoxic strains in food is observed in many different food products (table 24). Table 24. Contribution of toxicogenic strains among staphylococci present in dairy products Type of food Milk, cheese Milk from cows with mastitis

Data on the basis of Loir et al. 2003.

Contribution of toxicogenic strains 15.9 % 43% 72.8%

Source Rosec 1997 Cordobo 1999 Akineden 2001

Influence of Interactions Occurring Between Micro-Organisms…

217

From research of Otero et al. it results that aerobic conditions cause the increase of kinetics of staphylococci population growth, however they are not the reason for occurrence of enterotoxin in tested cheese. The significant factor influencing the production of enterotoxin is temperature. Literature data evidence that the lowest temperature in which enterotoxin synthesis is observed is equal to 70C (Adams et al. 1995). According to Halin-Dohnalek et al. (1989), most of staphylococci strains present in sour cream at the temperature of 370C synthesised enterotoxin as soon as after 18 hours. At 220C this time was prolonged up to 52 hours. From research of these authors, it resulted that no staphylococcus strain produced enterotoxin at 40C during 14 days. According to some authors, heating chicken products and then storing them at the temperature of 40C guaranteed elimination of enterotoxin (Pepe et al. 2006). In lactic acid cheese tested by Steinka et al. (2001c, 2004), enterotoxin should not occur due to the presence of lactic acid bacteria. From data obtained by Loir et al., it results that in laboratory conditions – the acetic acid added to the medium has the greater inhibiting influence on enterotoxin production than e.g. lactic acid. Whereas, basic pH constitutes a very significant factor determining the production of toxins SEB, SEC and SED (Loir 2003). This can be confirmed by research of Delbar et al. (2006), who observed the release of enterotoxin by bacteria present in cheese at the increase of pH values. The enterotoxin synthesis by Staphylococcus aureus is dependant on the presence of accompanying micro-flora in the environment (Oter et al. 1988, Noleto et al. 1987, Sameshima et al. 1998, Steinka 2004). Sameshima (1998) tested the influence of Lactobacillus cultures on staphylococci in fermented sausages. Inoculum was equal to 104 cfu for staphylococci and 107 cfu for Lactobacillus. Enterotoxin was detected during fermentation at each temperature, if the certain strain of lactic acid bacteria e.g. Lactobacillus acidophilus FERM P-15119 occurred in the presence of staphylococci. In the presence of L. rhamnosus and L. paracasei strains, the inhibition of toxin release was also satisfactory. Lactic acid bacteria belonging to starter cultures do not always show the identical ability to inhibit enterotoxin synthesis. It depends both on type of enterotoxin and the inhibiting strain. The influence of commercial starters on the growth of Staphylococcus aureus and the production of C1 and C2 enterotoxins in model conditions was investigated by Otero et al. 1988. Staphylococcus aureus FRI 137 strain producing enterotoxin C1 as well as FRI 361 and L2 strains producing enterotoxin C2 grew well both individually and in the presence of starters. This starter contained standard Lactococcus lactics spp. strains. Commercial starters showed a weak inhibiting influence on S. aureus only in late phases of growth of these bacteria. In contrast, the enterotoxin synthesis was strongly inhibited after 18 hours in as much as 89% for Staphylococcus aureus FRI 137. Presence of the starter inhibited release of toxin in 80% by Staphylococcus aureus FRI 3 and in 69% by staphylococci L2 strain. Enterotoxin C1 was both synthesised and accumulated during all phases of growth both in monoculture and in mixed population. Unfortunately, the growth of other strains caused reduction of its concentration after 24-36 hours (Otero et al. 1988).

218

Izabela Steinka

Extremely important research for understanding the character of staphylococci toxicogenesis was conducted by Noleto et al. 1987, who evaluated production of staphylococcal enterotoxin by Staphylococcus aureus in the presence of other pathogenic bacteria such as: Bacillus cereus and Escherichia coli, Streptococcus faecalis, Pseudomonas aeruginosa. All staphylococci strains showed the growth and production of enterotoxin in the presence of enterococci. In tvarogs tested by Steinka (2004), count of enterococci was significant, what could be the reason for stimulation of staphylococci to synthesise toxin. From research it resulted that in model conditions, the metabolism ability of staphylococci was also determined by the type of medium. Application of different medium types allowed observing that this phenomenon occurred only when preponderance of staphylococci over enterococci count was detected in the environment. Other behaviour of staphylococci was observed in the presence of Bacillus sp. Staphylococci showed growth in the presence of rods in both types of medium, however enterotoxin was produced only when their inoculum was 10 versus 104 in relation to Bacillus sp. Quantitive ratio of both populations necessary for enterotoxin production was dependent on type of staphylococci strain used for medium inoculation. And so, e.g. FRI 196E strain produced toxin in both types of applied medium even when inoculums of both bacteria were equal. Staphylococci did not produce enterotoxin only when Escherichia coli remained preponderant in number in both medium types. In order to observe enterotoxin synthesis, the count of staphylococci had to be significantly higher than number of rods (10 versus 104). Other types of relationships were observed in the presence of Pseudomonas aeruginos. Staphylococci did not produce enterotoxin when staphylococci inoculum exceeded 10 versus 3 4 10 or 10 . From the research of Noleto, it resulted that enterotoxin was produced only when staphylococci number was equal to or greater than number of accompanying species (Noleto 1987). A very important condition favouring enterotoxin synthesis is also the presence of inhibiting substances in the environment (Gonzales –Fandos et al. 1994). Epidemiologic data suggest that the enterotoxin B (SEB) is seldom observed in food, whereas enterotoxin of type A (SEA) and C (SEC) are predominant, and they are responsible for more than 70% of contamination of food samples. Staphylococcal enterotoxins SEA and SED are more often produced in food of basic pH, even at low number of cells. However, even among staphylococci isolated from fermented milk, the significant amount belongs to toxicogenic strains. From research of Steinka (2004), it results that concentration of determined enterotoxin was varied, and was higher in tvarogs packed with vacuum system then in tvarogs packed into Cryovac films. Lack of packaging hermetic properties favoured the enterotoxin synthesis. In samples of vacuum-packed tvarog coming from depressurised packaging, the 2-3-fold higher enterotoxin concentration was observed in comparison with hermetic packaging. 0 0 Enterotoxin was detected in tvarogs stored in the temperature from 6 to 8 C for the period of 7 and 14 days. Enterotoxin presence was not observed in samples taken for tests on the day of introducing a product into the market.

Influence of Interactions Occurring Between Micro-Organisms…

219

However, the observed case of the increase in enterotoxin concentration in samples coming from the same lot between day 7 and day 14 of product storage can suggest the probable existence of this toxin synthesis. The possibility of growth of Staphylococcus aureus strains producing enterotoxin in fermented dairy products arouses controversy. As it results from the literature, despite the antagonistic influence of other micro-flora present in the environment (even lactic acid bacteria), in certain conditions the toxicogenesis can occur. The necessity to evaluate the probability of staphylococci survival and the staphylococcal enterotoxin synthesis in hermetically-packed products during storage at low temperature is the reason for developing models determining the risk. For instance, the growth of staphylococci after 14 days in tvarogs stored in hermetic packaging could not be described with linear models, because they showed a low degree of equation matching. NPA/PE = 214.86 + 1.3269•N7 r2

0.057

(16)

NCryovac = 236.63 – 0.0068•N7 r2

0.131

(17)

where NPA/PE14 – staphylococci count in tvarogs packed into PA/PE laminated after 14 days NCryovac 14 – staphylococci count in tvarogs packed into Cryovac laminated after 14 days N7 – staphylococci count in tvarogs after 7 days of storage Staphylococcal enterotoxin synthesis depending on storage time, hermetic properties and packaging type as well as the ability of coagulase synthesis is presented in table 25. Table 25. Models of staphylococcal enterotoxin synthesis in dependence on the ability of coagulase synthesis and number of staphylococci in tvarogs Equation form Ep7=-84.7853+0.425N0

r2 0.865

Ep7=79.4059+0.3342N0+56.6643K7 Erp7=211.122-0.134925N0

0.867

2 Erp14=300.907-230.4(K14) Eb14=38.7367+74.0549K0-250.317K14+257.751K7-0.3713N14-0.0165N7

0.460 0.920

Erb14=-21.2696+3.21196N14 +1.74943N7 – 17002.1K14

0.856

0.176

Steinka 2004, N- staphylococci count; 0,7,14 tvarog storage time E- staphylococcal enterotoxin; K- coagulase; Indexes: p - vacuum, b – non-vacuum, r – depressurised.

As a result of conducted research, staphylococcal enterotoxin was observed in the insignificant percentage (4.5%) of tested samples coming both from tvarogs packed with

220

Izabela Steinka

vacuum system and those packed into Cryovac. Application of multifactor regression analysis as well as data transformation allowed expressing the probability of occurrence of enterotoxin in hermetically packed tvarogs with the help of linear and polynomial equations. When enterotoxin was present in tvarogs packed into Cryovac, the statistically significant relationships was also observed between the probability of enterotoxin occurrence and the staphylococci population count during storage, as well as synthesis of coagulase by these bacteria (table 25). Equation describing occurrence of enterotoxin after unsealing the nonvacuum packaging had quadratic form, what significantly evidences the influence of environment on enterotoxin synthesis. Equation describing this relationship showed the existence of weak connection between synthesis of enterotoxin and coagulase by Staphylococcus aureus in these conditions. This can indicate the crucial importance in evaluation of risk. In optimum model conditions, the growth of staphylococci count and the production of enterotoxin show significant dynamics. At temperature of 220C, the growth of staphylococci by 3 logarithmic cycles should occur after 72 minutes. According to Pathogen Modelling Program, in aerobic conditions, at pH change and temperature drop to 10-120 C, changes in staphylococci population count by 1 logarithmic cycle were not observed before 10-21 days. Shorter time (by 11 days) of staphylococci breeding predicted with this program for anaerobic conditions of growth show a significant meaning of redox potential for the growth of these bacteria. Vernozy-Rozynand et al. (1998) suggest that enterotoxin production is a slow process. In ripening cheese, the presence of enterotoxin was not observed in fresh curd, but only after 21 days of ripening. The ability of staphylococcal enterotoxin synthesis does not have to be connected with synthesis of coagulase, because there exist enterotoxic coagulase-negative Staphylococcus aureus strains that produce enterotoxin, as well as other species of enterotoxic staphylococci such as: Staphylococcus haemolyticus, Staphylococcus warneri, Staphylococcus saprophyticus or Staphylococcus epidermidis.

METHODS OF OPTIMISING QUALITY OF TVAROGS Before implementation of HACCP system into food production, tvarogs were characterised by low microbiological quality before storage. The traditional packing system connected with non-automated technological line can still arouse quality problems. The conducted numerous attempts to optimise the quality of cottage cheese and tvarogs (Kornacki et al. 1999, Kornacki et al. 2002 Molska 1992, Rosenthal et al. 1996 Herve et al. 1998, Stanton et al. 1998, Neugebauer et al. 2005) indicated not only the problem of obtaining lactic acid cheese of high stability level, but they also emphasised the significant influence of micro-flora on achieving desired sensory features of these products. Microwaves and ionising radiation are the directions of optimising microbiological quality of cottage cheese proposed by Hevre et al. 1998 and Rosenthal et al. 1995. Kornacki et al. 1999 suggested enriching starters used for tvarog production with addition of such bacteria as: Streptococcus salivarius spp. termophilus, Lactobacillus delbrueckii bulgaricus, Bifidobacterium bifidum, Lactobacillus acidophilus. Similar suggestions were made by Neugebauer et al. in relation to cottage cheese (Neugebauer et al. 2005).

Influence of Interactions Occurring Between Micro-Organisms…

221

Kornacki et al. (2002) showed that application of traditional butter starters used for production of tvarog made of raw material of low microbiological quality did not result in reduction in number of coliforms and psychrotrophs, and even the growth of proteolytic micro-organisms count in ready product was observed. Therefore, it is necessary to find a method for optimising the quality of these products. Optimisation of tvarog quality can be carried out with several methods: • • •

Application of additives of biocide character Modification of packing system Application of safety predictions for instance with the help of a computer program

Optimisation with Biocides Plant additives are often added to dairy products such as yoghurts, cottage cheese or ripening cheese (Ahmed et al. 2002, Beckmang et al. 1996, Grega et al. 1999, 2001). They include pieces of fruit, chives, onion, garlic, paprika, tomatoes, cucumbers, horseradish and herbs. However, the data prove that some of the spices and herbs modify the metabolism of starter cultures. This is the reason for changes in organoleptic properties of products manufactured using those starters (Arora et al. 1999). Various research concerning the behaviour of staphylococci in the presence of different substances of plant origin has been conducted until now. For instance, some authors evaluated the possibility of enterotoxin synthesis at variable concentration of garlic in the environment (Gonzales-Fandos et al. 1994). Enterotoxins SEA and SEB were detectable at garlic concentration not exceeding 1%, whereas SED was produced by staphylococci even at the level of 2%. Generally, no additives are applied in tvarogs. Some additives of flavour character are used as a supplement in cottage cheese. Some of them have also biostatic properties. This concerns for example garlic, onion and herbs. However, from our research (Steinka 2006a) it results that the presence of garlic in these products gives them a specific bitter taste. Searching for other plant additives e.g. mixture of rowan and aloe in model testing performed on semi-products (lactic acid curds) showed varied influence on the facultative anaerobic bacteria and fungi, which can be present in tvarogs (Steinka 2005c). Table 26 presents changes in count of staphylococci, enterococci and yeast, depending on plant additive. Table 26. The influence of plant additives on facultative anaerobic micro-flora Micro-organisms Staphylococcus aureus

Tvarog curd Aloe arborescens Y=0.13x+4.3

Sorbus aucuparia Y=-0.2x+4.63

Enterococcus faecalis Candida sp.

Y=3.11x-0.34 Y=0.94x+5.29

Y=-0.3x+3.07 Y=-0.76x+6.99

Steinka 2005d, x-value of control sample.

222

Izabela Steinka

The effort to find the appropriate conditions of adding aloe during production of lactic acid curds encountered significant difficulties (Steinka 20002b, 2003b, 2003d). The attempts to optimise the quality of ready lactic acid cheese with aloe aerosol were also made (Steinka 2003c). In tvarog samples taken for tests during refrigerating storage, the differences were noticed between sizes of micro-organism populations in products with aloe extract additive and those where aloe aerosol was not added. The average values of population counts indicated the stimulating action of aloe aerosol in relation to enterococci and yeast. From the fourth day of storage of tvarogs with aloe additive, the increase in staphylococci count was observed. The growth of yeast in tvarogs treated with aloe aerosol was observed during the entire period of lactic acid cheese storage. Whereas, the effectiveness of aerosol action was observed in relation to mould and staphylococci. During storage of tested products, the inhibition of mould and staphylococci counts were detected in samples with aloe extract additive. It was also observed that the presence of aloe resulted in the inhibition of growth of Lactococcus sp. population during the first days of product storage. From day 4 until day 7 of product storage, the growth of lactic acid bacteria in tvarogs with additive was observed, and then the reduction in number of these bacteria was noticed between day 7 and day 14 of tvarog storage. The results of statistical analysis showed a high correlation (r 0.9697) between the enterococci contamination level present in tvarogs not subjected to action of aerosol and stored in packaging and the addition of aloe extract (table 27). Table 27. Linear correlation between micro-organism populations present in control tvarogs and in tvarogs with aloe aerosol additive Type of micro-organisms Enterococcus sp.

Equation of linear correlation Ea= 13811+1.0894E

r

2 r

0.969

0.940

Staphylococcus aureus

Sa=13.199+0.79701S

0.792

0.627

Yeast

Da=154600+0.71205D

0.708

0.500

Mould

GSa=191500+0164523GS

0.385

0.148

Lactococcus sp.

La =2971000+0.03452L

0.114

0.012

Steinka 2003b, Ea - Enterococcus sp. in tvarogs with aloe additive; Sa- Staphylococcus aureus in tvarogs with aloe additive; Da- yeast in tvarogs with aloe additive; GSa – staphylococci in tvarogs with aloe additive; La - Lactococcus sp. in tvarogs with aloe additive; E- Enterococcus sp; S- Staphylococcus aureus; D- yeast; L- Lactococcus sp.

High coefficient of determination r2 0.9403 showed that enterococci count in tvarogs without additive was similar to number observed in tvarogs with aerosol additive. From statistic analysis it resulted that the variance of enterococci count only in 6% could be determined by the presence of aloe. Whereas, 37% of variance of staphylococci count in tvarogs with aloe additive could result from the influence of aloe.

223

Influence of Interactions Occurring Between Micro-Organisms…

Basing on coefficient of determination, it was also proved that mould and yeast showed varied susceptibility to aloe influence. In the case of yeast, more than 50% of variance of these fungi count resulted from the action of aloe aerosol, whereas 85% of variance of mould could result from the presence of aloe aerosol and the influence of this additive on fungi populations. Changes in number of bacteria and fungi in tvarogs stored with aloe additive, depending on storage time and the behaviour of other micro-organism populations present in tvarogs could be expressed with polynomial equations of the following form: Y = a1x1 + a2x2 + a3x3 + a4x4 …

(18)

which are presented in table 28. Table 28. The influence of storage time and the interactions among micro-organisms on the population of micro-organisms in tvarogs with aloe aerosol additive Type of microorganisms

Equation of linear correlation

r

R2

Enterococcus sp.

Ea= 29479.82-2466.85t+1.1E-9.05S0.09D+0.16GS-0.01L

0.983

0.962

Staphylococcus aureus

Sa=164.8455-14.2526t+0.0019E +0.7548S0.0002D-0.0001GS

0.802

0.593

Yeast

Da=134169.7-3827.7t-0.3E+3.3S+D-0.1GS

0.735

0.476

Mould

GSa=64808.98-1018.67t1.55E+36.38S+0.09D +0.88 GS-0.02L

0.807

0.603

Steinka et al. 2003b.

The presented coefficients of determination of equations describing changes of enterococci, staphylococci and yeast populations, taking interactions among micro-organisms and action of aloe into account, were similar in both types of tested tvarogs (tables 27 and 28). Coefficients of determination defined for those equations differed from each other by 2.18%, 3.41% and 2.55% respectively, indicating the insignificant influence of factors other than aloe on changes of these micro-organisms populations (tables 27 and 28). In the case of mould, differences in coefficients of determination of derived equations showed a great influence of interactions occurring among micro-organisms present in products and the time of tvarog storage on the level of fungi in tvarogs treated with aloe.

224

Izabela Steinka

In the case of mould population in tvarogs with aloe additive, the inhibition of its growth was the combined effect of application of aloe, hermetic packaging, low temperature as well as interactions occurring among the micro-organisms present in the product. Table 29 presents the influence of lactic acid bacteria on the growth of individual groups of micro-organisms in stored tvarogs as well as the influence of interactions among the microorganisms and the aloe aerosol on growth of these populations. These relationships were expressed with quadratic equations (Second Order Polynomial). Table 29. The influence of storage time, additive of aloe and lactic acid bacteria on changes of secondary micro-flora population in stored tvarogs

Type of microorganisms

Enterococcus sp

Equation form Tvarogs stored without aloe aerosol additive E=29195+4361,14t-0,002L180,297 t2+3,203e-4tL+1,037e10L2

Staphylococcus aureus

S=

1377,034-248,149t-8,123e5L+11,44t2+1,755e-5tL-1,059e12L2

Yeast

D=-35801,83+1,001e5t +0,005L-5082,894t2+0,008tL6,035e-10L2

Mould

GS=22261,7+71775,96t-0,006L1966,29t2+0,002tL+2,338e-10L2

Tvarogs stored with aloe aerosol additive Ea= 24412,87-36566,1t +0,03La+2653,61t2+0,017tLa -3,923e-9La2 Sa= 1202,67-232,937t-5,122e-5La +11,263t2 +1,223e-5tLa-1,014e12L 2 a Da=72940,75+1,006e5t+0,085La -6253,274t2+0,025tLa-7,192e-9 La2 GSa=-7391,507+35196,96t +0,037La-702,675t2+0,011t La -3,179e-9 La2

Steinka 2003b Ea - Enterococcus sp. in tvarogs with sloe additive; Sa- Staphylococcus aureus in tvarogs with sloe additive; Da- yeast in tvarogs with sloe additive; GSa – staphylococci in tvarogs with sloe additive; La - Lactococcus sp. in tvarogs with sloe additive; E- Enterococcus sp; S- Staphylococcus aureus; D- yeast; L- Lactococcus sp.; t- storage time.

From obtained data, it also results that the influence of lactic acid bacteria on the growth of secondary micro-flora was influenced by aloe additive (table 29).

Influence of Interactions Occurring Between Micro-Organisms…

225

Second order polynomial equations describing the growth of staphylococci, enterococci and mould in stored tvarogs differed among each other, if tvarogs were sprayed with aloe aerosol. Only surface response models determining the influence of lactic acid bacteria on yeast showed a very similar shape and direction of action in the case of tvarogs both without and with aloe aerosol additive. Changes of Lactococcus sp. populations under the influence of aloe could be expressed by the equation of the following form: La= -1.248e6 + 1.008e6t + 1.194L – 64381.85t2 – 0.043tL – 3.768e-9L2

(19)

where: La – number of Lactococcus sp. in tvarogs with aloe additive L – number of Lactococcus sp. in control tvarogs t - storage time of tvarogs Negative value of coefficient a in the equation indicated the direction of changes of Lactococcus sp. population count under the influence of aloe, suggesting inhibition of growth of lactic acid bacteria populations in tvarogs. The bactericidal properties of aloe in relation to many bacteria are commonly known. In food products its application is limited due to the presence of significant amount of aloin, which excessive amount creates the hazard of excretory system hyperaemia and nephropathogenesis. While aloin level in food products is not advantageous (special regulations control the level), the small amount of aloe additive to food influences favourably the intestine peristaltic motion as well as functioning of immunological system. There exist data indicating the possibility to lower the adverse effect of aloin in the presence of lactic acid bacteria (Steinka 2002b). Addition of small aloe amount can be applied in fermented food with the help of lactic acid cultures. From previously conducted research (Steinka 2001d,e 2002b), it resulted that the presence of aloe in the form of a pulp influenced significantly the growth of certain types of bacteria and fungi in lactic acid curd, as well as it affected animal organisms. The influence on micro-organisms was varied, depending on type of micro-organisms, bacteria species, presence of other micro-organisms and their count in tested environment (Steinka 2001d,f). It was showed that the additive of aloe in the form of aerosol into tvarogs of low microbiological quality lowered the staphylococci count during 14 days of product storage in hermetic packaging. Moreover, the aerosol stabilises the level of mould in these products. However, this method of aloe application shows stimulation influence on the growth of enterococci and yeast population in stored tvarogs.

226

Izabela Steinka

Aloe aerosol also cannot be applied for optimising the quality of tvarogs of high contamination level with enterococci and yeast. To sum up, optimising the quality of tvarogs of a signification level of contamination with facultative anaerobic bacteria, using aloe aerosol and vacuum re-packing is possible, but it requires more research to be conducted.

Optimising the Quality through Modification of the Packing System This type of optimisation involves packing in clean air atmosphere. From various research, it results that air in dairy plant can be the reason for contamination of ready products. The literature data report that department of fermented drink and cheese production can show a higher level of contamination with micro-organisms in comparison with other departments (Ren et al. 1992, Salustiano et al. 2003). Salustiano et al. detected the presence of staphylococci in dairy plants (table 30). Table 30. Contamination of air in different departments of dairy plant Processing area

Mesophilic Bacteria CFU/m3 110-600

Aerobic

Yeast and moulds

Total coliform CFU/m3

70-160

0.00-0.66

Milk acceptance section Cheese

100-920

90-610

0.33-1.66

Yoghurts

100-320

100-940

0.00-0.66

Own study on the basis of Salustiano et al. 2003.

The reasons for product contamination can be both traditional production technology, in which pressing and draining can constitute the critical control points. In automated production technologies, these stages do not create any hazard. According to FIL IDF, the air quality of the highest purity class should be characterised by OLD value not greater than 10 cfu/m3 while both fungi and yeast as well as pathogenic bacteria should not be present at all. The lowest air class (i.e. D) allows the presence of ca. 10000 aerobic mesophilic bacteria and more than 100 fungi and 10 pathogenic bacteria. Research conducted by Olbromska et al. 2005 showed that contact with personnel, machines, air and most of all with packaging is often the reason for secondary contamination determining the safety of dairy products. However, research of Steinka et al. 1998 did not show the high contamination level of packaging materials applied for packing lactic acid cheese and cottage cheese (table 31).

Influence of Interactions Occurring Between Micro-Organisms…

227

Table 31. Microbiological quality of materials for packing tvarogs Packaging type

Mould

Yeast

cfu/25cm2 Parchment paper

11

0

PA/PE laminate

30

0

Frischaltenfoliae PE

44

22

Aluminium foil

33

0

Styrofoam tray

70

0

Steinka et al. 1998.

In the case of tvarogs, modification of the packing system is hard due to the consistency and delicate structure of these products. Applying modification of traditional packing system and substituting it with Styrofoam trays does not guarantee high quality of tvarogs as well. Products packed in this way and wrapped with thin frischaltenfolie showed high count of 6 5 fungi. After storage, the level of yeast reached 1.1•10 cfu/g, while mould 4.3•10 cfu/g and these counts were higher than in the case of tvarogs packed into other types of packaging.

Photo 1. Tvarogs packed onto Styrofoam trays wrapped with thin PA/PE foil.

It is possible to protect the product against changes of sensory properties resulting from significant fungi count by avoiding the non-hermetic packaging, or storing the product for a short period of time - not longer than 4 days. There also exists the possibility to modify packaging materials. At the present, the project of creating laminate of modified composition is at development stage. Among modifying elements the followings are proposed: starch and two biocides.

228

Izabela Steinka

Optimisation with the Help of Prediction Models Among the possibilities of optimising the quality and safety of tvarogs, there is also microbiological predicting. The behaviour of multi-species populations in stored tvarogs as well as change of basic packaging properties can be evaluated with Twarogi JMTPH computer program (Steinka 2003a). For evaluation of staphylococcal enterotoxin synthesis, the probabilistic model was developed taking storage time, staphylococci and yeast populations counts in tvarogs into account. Development of computer program called TEG required application of Boolean expression, which helped to precise the series of conditions that have to be fulfilled in tvarogs for the occurrence of enterotoxin synthesis.

E = [(GN = 0) ∧ (GP ≥ 4)] ∨ {((GN + GP ) ≥ 5) ∧ [(GP > 3.5) ∨ (GN > 3.4 + 0.1t )

∨ (GP / D > 0.6) ∨ (GN / D > 0.7) ∨ ( GP − GN /(GP + GN ) ≤ 0.2)]}

(20)

where: GN - coagulase negative Staphylococcus aureus count, GP - coagulase positive Staphylococcus aureus count, D - yeast count, t - time In a computer program developed in Delhi 7.1, the staphylococci count, storage time and size of predominant population in tvarogs were also taken into account. The result of program operation is obtaining the answer that negates or confirms the presence of staphylococcal enterotoxin in a products (present-absent). The example of computer simulation is presented in table 32 below. Table 32. Predicting the presence of staphylococcal enterotoxin with TEG computer program Staphylococcus aureus CP count 2.78

Staphylococcus aureus CN count

Yeast count

Storage time of tvarogs

3.44

5.58

14

Result of simulation of enterotoxin presence Present

2.32

2.11

3.89

7

Absent

Steinka et al. 2007.

Influence of Interactions Occurring Between Micro-Organisms…

229

CONCLUSION Tvarogs constitute the important of many nutrients. Their nutritious and taste values evidence that they are significant diet components for both adults and children. The microbiological quality, taking safety of products into account, has been constantly improving. The interactions occurring among tvarog allochthonous micro-flora influence the growth of all micro-organisms present in a product. The dynamics of growth of individual types of these micro-organisms can be evaluated using JMPTH computer program. The packaging hermetic properties determines significantly the direction of changes of allochthonous micro-flora under the influence of lactic acid bacteria. The packing system as well as hermetic properties and type of packaging influence the growth of staphylococci count in a product. The probabilistic model is suitable for assessing the presence of enterotoxin in products, depending on the level of staphylococci and yeast. There also exists a possibility to optimise the quality of lactic acid cheese, using biostatic plant additives or through modification of the composition of packaging material.

REFERENCES [1]

Adams M.R., Moss M.O. (1995). Food Microbiology. London. The Royal Society of Chemistry. [2] Adams M.R., Moss M.O. (2000). Food Microbiology. London. The Royal Society of Chemistry. pp 112, 141, 211. [3] Ahmed A.K., Johnson K.A. (2000). Horticultural development of Australian native edible plants. Austral. J. Botany, 48, 417-426. [4] Alves R.M., Sarantopoulos C.I., van Dender A.G.F., Faria J. (1996) Stability of sliced Mozzarella cheese in modified atmosphere, J. Food Prot., 59, 8, 838-844. [5] Arora D.S., Kaur J. (1999). Antimicrobial activity of spices. Int. J. Antimicrobial Ag., 12, 257-262. [6] Ashenafi M. (1990). Microbial quality of ayib, a traditional Ethiopian cottage cheese, Int. J. Food Microbiol., 10, 3-4, 263-268. [7] Bahk G.J., Hong C.H., Oh D.H., Ha S. D., Park K. H., Todd E. C. (2006). Modelling the level of contamination of Staphylococcus aureus in ready–to-eat kimbab in Korea, J. Food Prot., 69, 6, 1340-1346. [8] Baranyi J., Ross T., McMeekin T.A., Roberts T.A. (1996). Effects of parametrization on the performance of empirical models used in predictive microbiology, Food Microbiol., 13,83-91. [9] Baumgartner A., Kuefer M., Rohner P. (2001). Occurrence and antibiotic resistance of Enterococci in Various Ready–to-Eat. Foods Activ. Fur Lenbensmittelhygiene 52, 124,16-19. [10] Beckmanng., Koszegl D., Sonnensche in B., Limbeck R.: On the microbial status of herbs and spices. Fleischwirtchaft 70, 240-243, 1996.

230

Izabela Steinka

[11] Belickova E., Tkacikova L., Nass H. T., Vargova M., Ondrasovic M., Ondrasovicova O., Obsitikova Toth L. (2001). Staphylococci plate counts in foods of milk origin. Vet. Med. –Czech. 46, 1, 24-27 [12] Benkeroum N., Sandine W.E. (1988). Inhibitory action of nisin against Listeria monocytogenes. J. Dairy Sci. 71, 3237 -3245. [13] Bergdol M.S. (1995) Importance of staphylococci that produce nanogram quantites of enterotoxin. Zentralbl. Bacteriol., 282,1, 1-6. [14] Bertola N. A., Bevilcaqa A., Zaritzky N. (2001). Reological Behaviour of White and Cream Cheese. Int. Congress on Engineering and Food Technomic Publishing Company, Inc Lancaster,573-577. [15] Bolton L.F., Frank J.F. (1999). Defining the growth/no growth interface for Listeria monocytogenes in Mexican –style cheesebased on salt, pH and moisture content. J. Food Prot., 62, 601-609. [16] Bozukart H., Erkman O. (2001). Predictive modeling of Yersinai enterocolitica inactivation I Turkish feta during storage. J Food Engeneering 47, 2, 81-87. [17] Brocklehurst T. F., Lund B. M. (1988). The Effect of pH on the Initiation Growth of Cottage Cheese Spoilage Bacteria, 6, 43 - 49. [18] Brody A.L. (1989) Controlled/Modified Atmosphere Vacuum Packaging of Foods. Food and Nutrition Press. Inc.,179. [19] Brouilaud-Delarte A., Maire M., Colette C., Mattei C., Lahellec C. (1997). Predictive microbiology of dairy products: influence of biological factors affecting growth of Listeria monocytogenes. J AOAC Int., 80,4,913-919. Md 403 [20] Buchanan R, L., Klawitter L, A. (1992). The effect of incubation temperature initial pH and sodium chloride on the growth kinetics of Escherichia coli 0 157 : H7. Appl. Environ. Microbiol, 58, 2513 - 2516. [21] Cais D., Wojciechowski J. (1996). Zmiany wybranych cech jako ciowych serów twarogowych w trakcie ich przechowywania. Przegl. Mlecz. 6, 177-179. [22] Castilejo -Rodriquez et al. A.M., Garcia Gimeno R.M., Zurera-Cosano G., Barco – Alcala E., Rodriquez –Perez M.R.(2002) Assessment of mathematical models for predicting Staphylococcus aureus growth in cooked meat products. J.Food Prot., 65, 4, 659-665. [23] Centeno J.A., Menende S., Harmida Ma. Rodriquez – Otero J.L. (1999). Effects of the Addition of Enterococcus faecalis in Cebreiro Cheese Manufacture, International J. Food Microbiol., 48, 97-111. [24] Champagne C.P., Laing R.R., Roy D., Mafu A.A., Grifith M.W. (1994). Psychrotrophs in dairy products: their effects and their control. Crit. Rev., Food Sci. Nutr., 34, 1, 1-30. [25] Chen et al. J.H., Hotchkiss J.H. (1993). Growth of Listeria monocytogenes and Clostridium sporogenes in cottage cheese in modified atmosphere packaging. J. Dairy Sci., 76, 972-977. [26] Chojnowski W., uraw J., Sell J., Jęsiak Z. (1997). Materiale form 4th National Scientific Session. Postęp w technologii, technice i organizacji mleczarstwa. Olsztyn, 204-205. [27] Crofcheck C.L., Payne F.A., Nokes S.E.(1999). Predicting the cutting time of cottage cheese using light backscatter measurements. Transaction of the ASAE,42, 4, 1039-1045.

Influence of Interactions Occurring Between Micro-Organisms…

231

[28] Delbes C., Alomar J., Chougui N., Martin J.F., Montel M.C. (2006). Staphylococcus aureus growth and enterotoxin production during the manufacture of uncooked, semihard cheese from cows’ raw milk. J. Food Prot. 69, 9, 2161-2167. [29] Dixon, N.M., Kelly D. B. (1988). The inhibition by CO2 of the growth and metabolism of microorganisms. J. Appl. Bacteriol., 67,109. [30] Elliot S.C., Vuillemard J.C., Emond J.P. (1998). Stability of shredded Mozzarella cheese under modified atmospheres. J. Food Sci. 63, 6, 1075-1080. [31] Farrag S. A., Marth E. H. (1992) Interactions between Listeria monocytogenes and other psychrotrophic bacteria i dairy foods: review. Food Aust., 44, 281-284, 286. [32] FDA /CFSAN. 2001. Microbiological Safety of Controlled and modified atmosphere Available from: packaging of fresh and fresh-cut produce. www.cfsan.fda.gov/comm/ift3-6.html. 30.03.07 [33] Fedio W. M., Macleod A., Ozimek L. (1994). The Effect of Modfied Atmosphere Packing on the Growth of Microorganisms in Cottage Cheese. Milschwissenschaft 49,11, 622-629. [34] Feehery F. E., Ross E.W., Doona C.J. (2004). Quasi chemical modeling of Staphylococcus aureus growth-decline kinetics in intermediate moisture meats as functions of humectans, anti-microbials, and acidulants. IFT Annual Meeting, 12-16.07. Las Vegas, Session 30, Food Microbiology: Meat, poultry and seafood. [35] Fukijawa H., Morozumi S. (2006) Modelling Staphylococcus growth and enterotoxin production in milk.Food Microbiol. 23, 3, 260-267. [36] George S. M., Lund B. M., Brocklehurst T. F. (1988). The effect of pH and temperature initiation of growth of Listeria monocytogenes. Letters in Appl. Microbiol.6, 153 – 156. [37] Gonzales –Fandos E., Garcia-Lopez M.L., Sierra M.L., Otero A. (1994). Staphylococcal growth and enterotoxins (A-D) and thermonuclease synthesis in the presence of dehydrated garlic. J. Appl. Bacteriol., 77,5, 549-552. [38] Grega T., Sady M., Wszołek M. (2001). Ocena jako ci jogurtów z dodatkiem ziarna amarantusa (Amaranthus cruenthus). Przegl. Mlecz., 5, 223-226. [39] Halin –Dohnalek M.I., Marth E. H. (1989). Growth and production of enterotoxin A by Staphylococcus aureus in cream. J. Dairy Sci., 72, 9, 2266-2275. [40] Herve A.G., Tang J., Luedecke L., Feng H. (1998). Dielectric properties of Cottage Cheese and Surface Treatment Using Microwaves. J. Food Engineer. 37,389-410. [41] Hicks S. J., Lund B., M. (1991). The survival Listeria monocytogenes in cottage cheese. J. Appl. Bact., 70, 308. [42] Holly A., von Holzapfel W.H. Dykes G.A. (1992). Bacterial Populations Associated with Viena Sausage Packaging. Food Microbiology 9 (1), 45-53. [43] Honer C. (1988). Closing the freshness gap with CAP and MAP. Dairy Foods 89, (3) 53. [44] Hsiao CH., P., Siebert K. J. (1999). Modeling the Inhibitory Effects of Organic Acids on Bacteria. Int. J. Food Microbiol., 47,189-201. [45] Ingham S., Losinski J.A., Dropp B., Vivio L.L., Buege D.R., Research note: Evaluation of Staphylococcus aureus growth potential in ham during a slow-cooking process: use of predictions derived from the U.S. Department of Agriculture Pathogen Modeling Program 6.1 Predictive model and an inoculation study. J. Food Prot., 2004, 67, 7, 1512-1516.

232

Izabela Steinka

[46] Jackson T. C., Acuff G.R, Vanderzant C., Sharp T.R., Savell J.W. (1992). Identification and Evaluation of Volatile Compounds Associated with Vacuum and Modfied Atmosphere Packaged Beef Strip Loins. Meat Sci., 13, 175-179. [47] Jordan K. N., Oxford L., O Bryan C. P. (1999). Survival of Low - pH stress by Escherichia coli O157:H7. Appl. and Environm. Microbiol, 3048-3055. [48] Kimura B., Toshiyama T., Fuji T., J. (1999). Carbon Dioxide Inhibition of Escherichia coli and Staphylococcus aureus on pH – adjusted Surface in Model System. Food Sci., 64,2, 367-370. [49] Koło yn-Krajewska D. (2003). Higiena produckji ywności. Wydawnictwo SGGW. Warszawa. [50] Kornacki K., Maciejska A., Klebukowska L. (1999). Modyfikacja szczepionek do produkcji nie dojrzewających serów twarogowych. 7th Scientific Session ‘Postęp w technologii technice i organizacji mleczarstwa’. Olsztyn 15-16.02.1999, 237-242. [51] Kornacki K., Klebukowska L., Pruszyńska M. (2002). Wpływ rodzaju szczepionek bakterii fermentacji mlekowej na jako ć mikrobiologiczną twarogów., 8th Scientific Session ‘Postęp w technologii technice i organizacji mleczarstwa’, Olsztyn 21-22.02. 2002, 375-384. [52] Kujawski M., Rymaszewski J., Cichosz G., Łaniewska –Moroz, Fetliński A. (1994). Wpływ bakterii fermentacji propionowej na tworzenie cech smakowo-zapachowych sera i twarogów. Przegląd mleczarski 10, 263-266. [53] Larssen A.G., Knochl S. (1997). Antimicrobial activity of food-related Penicillium sp. Against pathogenic bacteria in laboratory media and a cheese. J. Appl. Microbiol., 83, 111-119. [54] Le Loir Y., Baron F., Gautier M. (2003). Staphylococcus aureus and food poisoning. Genet. Mol. Res. 2, 1, 63-76. [55] Lee Wong A., Bergdol M.S. (2002). Staphylococcal food poisoning, 231-248. clavier [56] Linquist R., Sylven S., Vagshoim I. (2002) Quantitative microbial risk assessment exemplified by Staphylococcus aureus in unripened cheese made from raw milk. Int. J. Microbiol, 78,1-2, 155-170. [57] Maniar A.B., Marcy J.E., Bishop J.R., Duncan S. E. (1994). Modified Packaging to Maintain Direct - Set Cottage Cheese. J. Food Sci., 59, 6, 1305. [58] Masa-Calpe (1996). Microbiological quality of cheeses: importance of good handling practices. Aliment. 270, 69-72. [59] McCann T.L., Eifert D., Gennings C., Schilling M.W., Carter W.H. (2003). A predictive model with repeated measures analysis of Staphylococcus aureus growth data. Food Microbiol., 20,2, 139-147. [60] Moir C.J., Eyles M.J., Devey J.A. (1993). Inhibition of pseudomonas in cottage cheese by packaging in atmospheres containing carbon dioxide. Food Microbiol., 10,345-351. [61] Molska I., Kielak I., Lapińska H. (1992). Wpływ termizacji i przechowywania mleka na jako ć mikrobiologiczną i trwało ć sera twarogowego. Przem. Spo . 2,42-46. [62] Neugebauer K.A., Gilliand S.E. (2005). Antagonistic action of Lactobacillus delbrueckii spp. Lactis RM2-5 toward spoilage organisms in cottage cheese. J. Dairy Sci. 88,4, 1335-1341. [63] Neumayer L., Kramer J. (1990). Production of enterotoxin A and thermonuclease by Staphylococcus aureus in legumes. Int. J. Food Microbiol., 10, 3-4, 225-233.

Influence of Interactions Occurring Between Micro-Organisms…

233

[64] Nguyen-the C, Carlin F. (1994). The microbiology of minimally processed fresh fruits and vegetables. Crit. Rev. Food Sci. Nutr., 43, 4, 371-401. [65] Nicolai B.M., Van Impe J.F., Verlinden B., Martens T., Vandewalle J., De Baerdsmsaeker J. (1993). Predictive Modelling of Surface Growth of Lactic Acid Bacteria in Vacuum-Packed Meat. Food Microbiol. 10, 229-238. [66] Noleto A.L., Malburg J. L. M., Bergdoll N.M.S. (1987). Production of staphylococcal enterotoxin in mixed cultures. Appl. Environ. Microbiol., 53, 10, 2271-2274. [67] Olborska K., Lewicki P.P. (2005). Zagro enia w procesie pakowania produktów mleczarskich. 5th Scientific Conference on Quality and Safety of Food. Bia³obrzegi, 17-18.11, 44-45. [68] Otero A., Garcia M.C., Garcia M.L., Prieto M., Moreno B. (1988). Behaviour of Staphylococcus aureus strains, producers of enterotoxins C1 or C2, during the manufacture and storage of Burgos cheese. J. Appl., Bacteriol., 64, 2, 117-122. [69] Panfil-Kuncewicz H., Kuncewicz A., Bohdziewicz K., Kornacki K., K³êbukowska L., (1997b). Trwałość przechowalnicza twarogów pakowanych w pró ni i w modyfikowanej atmosferze. MateriałyXXVI Sesji KTiCh PAN 1997b, 85-88. [70] Panfil-Kuncewicz H., Kuncewicz A., Szpendowski J. (1997a). Pakowanie twarogów w atmosferze dwutlenku węgla. Materiały VI Sesji Naukowej ‘Postęp w technologii, technice i organizacji mleczarstwa’ Olsztyn, 294-295. [71] Pathogen Modeling Program www.arserc.gov/mfs/PATHOGEN.HTM [72] Pepe O., Blaiotta G., Bucci F., Anastasio M., Aponte M., Villani F. (2006). Staphylococcus aureus and staphylococcal enterotoxin A in breaded chicken products: detection and behaviour during the cooking process. Appl. Environ. Microbiol., 72,11, 7057-7062. [73] Pereira J.L., Salzberg SP., Bergdol M.S. (1991). Production of staphylococcal enterotoxin D in foods by low-enterotoxin–producing staphylococci. Int. J. Microbiol, 14,1, 19-25. [74] Philips C.A. (1996). Review: Modified atmosphere packaging and its effects on the microbiological quality and safety of produce. Int. J., Food Sci Technol., 31, 463-479. [75] Piccinnin D., Shelef L.A. (1995). Survival of Listeria monocytogenes in Cottage cheese. J. Food Protect., 58,2,128-131. [76] Pieczonka W. (1993). Jako ć organoleptyczna twarogów kozich uzyskanych z mro onego mleka lub skrzepu. Przegląd Mlecz. 10, 284-286. [77] Poirazi P., Leroy F., Georgalaki M.D., Aktypis A., De Vuyst L., Tsakalidou E. (2007). Use of Artificial Networks and a Gamma-concept –based approach to model growth of and bacteriocin production by Streptococcus macedonicus ACA-DC 198 under simulated conditions of kasseri cheese production. Appl. Environ., Microbiol., 73, 3, 768-776. [78] Post L.S., Lee D.A., Solberg M., Furgang D., Specchio J. (1988). Development of Staphylococcal Toxin and Sensory Deteroration during Storage. J. Food Sci., 53(2), 383-387. [79] Ren T.J., Frank F.J. (1992). A survey of four fluid milk processing plants for airborne contamination using various sampling methods. J. Food Protect. 55, 38-42

234

Izabela Steinka

[80] Rosenthal I., Rosen B., Berstein S. (1996). Surface pasteurization of Cottage cheese. Milschwissenschaft 51,4,198-201 [81] Ryster E. T., Marth E. H., Doyle M. P. (1985). Survival of Listeria monocytogenes during manufacture and storage cottage cheese. J. of Food Protect., 48, 746 –750. [82] Salustiano V.C., Andrade N.J., Brandao S.C.C., Azeredo R.M.C., Lima S.A.K.L. (2003). Microbiological air quality of processing areas in a dairy plants as evaluated by the sedimentation technique and a one-stage air sample. Braz. J. Microbiol., 34, 255259. [83] Sameshima T., Magome C., Tekeshita K., Arihara K., Itoh M., Kondo Y. (1998). Effect of intestinal Lactobacillus starter cultures on the behaviour of Staphylococcus aureus in fermented sausage. Int. J. Food Microbiol., 41, 1, 1-7. [84] Schaffener D.W., Vora P., Duffy S. (2001). Initiative modeling survival of Escherichia coli and Staphylococcus aureus with probability distribution functions. Society for Risk Analysis. www. riskworld.com/abstract/2001 [85] Severini C., Bressa F., Romani S., Dalla Rosa M. (1998). Physical and Chemical Changes in Vacuum Packaged Permigano Reggiano Cheese during Storage at 25,2 and 0

[86]

[87]

[88] [89] [90] [91] [92] [93]

[94] [95]

– 25 C. Journal of food Quality 21, 355-367. Sims G.R., Glenister D.A., Brocklehurst T.F., Lund B.M. (1989). Survival and growth of ford poisoning bacteria following inoculation into cottage cheese varieties. Int. J., Food . Microbiol., 9, 3, 173-195. mietana Z., wiegoń J., Szpendowski J., Bohdziewicz K. (1997). Propozycja techniczno-technologiczna produkcji tradycyjnych serów twarogowych. Materiały VI Sesji Naukowej ‘Postęp w technologii, technice i organizacji mleczarstwa’ Olsztyn, 198-200. Smittle R.B., Flowers R.S. (1994). Predictive Modeling and Microbiological Food Studies SCOPE 9, 1-5,7. Stańczak B., Szczawiński J., Pęcinek J. (2000). Survival of Listeria monocytogenes in cottage cheese. Med.Wet., 56, 4, 251-254. Stanton C., Gardiner G., Lynch P.B., Collins J.K., Fitzgerald G., Ross R.P. (1998) Probiotic Cheese. Int. Dairy J. 8, 491-496. Steinka I., Przybyłowski P. (1998). Jako ć mikrobiologiczna kwasowych serów twarogowych a metody pakowania. Przem. Spo . 11, 47- 49. Steinka I., Stankiewicz J. (1999a). System pakowania twarogów - aspekty higieniczne. ywność, 21 (4), 106-116. Steinka I., Stankiewicz J. (1999b). Influence of Vacuum Packing and Low Temperatures on the Level of Pathogenic Psychrotrophs in Cottage Cheese. Joint Proceedings WSM Gdynia, Hochschule Bremerhaven, 21, 63-67, 1999b. Steinka I. (1999c). Wpływ hermetycznego pakowania na jakość twarogów i twaro ków. Zeszyty Naukowe WSM w Gdyni 30, 53-62. Steinka I., Pukszta T. (2000). Influence of Mycota in Cottage Cheese on Establishing the Product Quality. Joint Proceedings. WSM Gdynia, Hochschule Bremerhaven, 13, 84-92.

Influence of Interactions Occurring Between Micro-Organisms…

235

[96] Steinka I. (2001a). Importance of Packaging Tightness for the Quality of Lactic Acid th

Cheese Packed in Plastic Films, Proceedings of the 13 IGWT Symposium Commodity Science in global quality perspective, Maribor, 2-8 09. 2001, 751-755. [97] Steinka I. (2001b). The effect the Interaction between Microorganisms Present in the Vacuum Packed Lactic Acid Cheese has on the Microbiological Quality. Joint Proceedings, WSM Gdynia- Hochschule Bremerhaven, 14, 41-48. [98] Steinka I., Kurlenda J. (2001c). Ocena zagro eń mikrobiologicznych związanych z systemem i hermetyką pakowania twarogów. Medycyna Weterynaryjna, 57, (10),757761. [99] Steinka I. (2001d). Próba zastosowania aloesu w fermentowanych przetworach mleczarskich. ywienie Człowieka i Metabolizm Suplement, 785-792. [100] Steinka I. (2001e). Possibility to Apply Aloe Vera Solution to Optimize the Quality of the Lactic Acid Cheese Curd, Dairy Conference, Cairo 3-8.11.2001, 235-242. [101] Steinka I. (2001f) Assessment of the efficacy of Aloe vera added to Propionibacterium rd

growth medium. 3 International Symposium Propionibacteria, 8-11.07. 2001, Zurich, 35. [102] Steinka I., Kurlenda J. (2002a). Determination of Microbiological Quality Indicates of Vacuum-Packed Cottage Cheese and Prediction of their Changes during Storage. Polish Journal of Food and Nutral Sciences 11(2), 199-206. [103] Steinka I. (2002b). Toxic Test Possibilities Testing the New Supplement to Lactic Acid Cheese. Joint Proceedings, WSM Gdynia, Hochschule Bremerhaven, 15, 61-66. [104] Steinka I. (2003a). Wpływ interakcji opakowanie–produkt na jako ć mikrobiologiczną hermetycznie pakowanych twarogów. Wydawnictwo AM Gdynia, 1-135. [105] Steinka I., Kukulowicz A. (2003b). Jako ć i bezpieczeństwo twarogów związane ze sposobem pakowania produktów. ywienie Człowieka i Metabolizm, 30, 3/4, 888-892. [106] Steinka I., Kukulowicz A. (2003c). Próba optymalizacji jako ci twarogów za pomocą aerozolu aloesowego. Brom. Chem. Toksykol. 36, 4, 341-346. [107] Steinka I. (2003d). Sposób antygronkowcowej suplementacji skrzepu twarogowego P346068, Biuletyn Urzędu Patentowego, 17(721), 3. [108] Steinka I. (2004) Ocena prawdopodobieństwa występowania gronkowców i enterotoksyny gronkowcowej w twarogach pakowanych w laminaty z tworzyw sztucznych, Roczn. PZH, 55, 1, 89-98. [109] Steinka I. (2005a). Wyznaczanie jako ci higienicznej kwasowych serów twarogowych wytwarzanych w ró nych warunkach. Brom. i Chem. Toksykol., Suplement, 373-376. [110] Steinka I. (2005b). Wpływ jako ci mikrobiologicznej twarogów na fazę wodną w warunkach pakowania hermetycznego. Roczniki PZH 56, 3, 275-281. [111] Steinka I., Blokus A. (2005c). Próba uwzględnienia interakcji między drobnoustrojami w procesie prognozowania jako ci twarogów i ich opakowań. Joint Proceedings Wydawnictwo Akademia Morska, Gdynia, XVIII, 14-21. [112] Steinka I. (2005d) Comparison of biostatic action of Sorbus aucuparia and Aloe arborescens under model conditions and in food. The IX International Congress Actual Problems of creation of new medicinal preparations of nutral origin Phytopharm 2005, St. Petersburg June 22-25, 381-383. [113] Steinka I, Nowaczyk M. (2006a). Wpływ dodatków ro linnych na wybrane cechy twaro ków smakowych. Handel Wewnętrzny 6, 422-427.

236

Izabela Steinka

[114] Steinka I. (2006c). Ocena jako ci twarogów dostępnych w handlu przed nowelizacją wymagań mikrobiologicznych. Jako ć towarów i usług w obrocie. Wydawnictwo WSG Bydgoszcz , Volume II, 167-174. [115] Steinka I. Blokus A. (2007). Szacowanie wielko ci populacji zdolnej do syntezy enterotoksyny gronkowcowej w twarogach, unpublished data [116] Steinka I., Parisi S. (2006b). The influence of cottage cheese manufacturing technology and packing method on the behaviour of micro-flora. Joint Proceedings, AM Gdynia, Hochschule Bremerhaven, 19, 30-37. [117] Stewart C. M., Cole M.B., Legan J.D., Slade L., Vanden M.H., Schaffener D.W. (2002). Staphylococcus aureus growth boundaries: moving towards mechanistic predictive models based on solute-specific effects, Appl. Environ. Microbiol., 68,4, 1864-1871. [118] Tamagnini L.M., de Sousa G.B., Gonzales R.D., Reveill J., Budde C.E. (2005). Behaviour of Yersinai enterocolitica and Salmonella typhimurium in Crottin goat’s cheese. Int. J. Food Microbiol., 99,2, 129-134. [119] Taub I. A., Feeherry F.E., Ross E.W., Kustin K., Doona C.J. (2003). A quasi –chemical kinetics model for the growth and Heath of Staphylococcus aureus In intermediate moisture bread. J. Food Sci., 68, 8, 2530-2537. [120] Turhan M., Kaletunc G. (1992). Modeling of salt diffusion In white cheese during longterm brining. J. Food Sci., 57,5, 1992. [121] Tyszkiewicz I. (1992) Przechowywanie mięsa w atmosferze gazów ochronnych, Gospodarka Mięsna 8, 20-22. [122] Usajewicz I. (1995) Oddziaływanie paciorkowców rodzaju Enterococcus na wzrost oraz aktywno ć metaboliczną bakterii grupy coli i Clostridium spp. Zeszyty Naukowe Akademii Rolniczo Technicznej. Olsztyn [123] Usajewicz I. (2000). Prze ywalność S. ureus, S. galinorum i E.coli w kefirze oraz w mleku ukwaszonym przez kultury L. plantarum i L. casei SHIROTA, Materiały Naukowe Szkoła letnia ‘Bakterie fermentacji mlekowej- klasyfikacja, metabolizm, genetyka, wykorzystanie’. Kazimierz Dolny, 59. [124] Vernozy-Rozynand C., Meyrand A., Mazuy A., Mazuy C., Delignette-Muller M.L., Joubert G., Perrins G., Lapeyere C., Richard Y. (1998). Behaviour and enterotoxin production by Staphylococcus aureus during the manufacture and ripening raw goats milk lactic cheeses. J. Dairy Res. 65, 273-281. [125] Walls I., Scott V.N. (1997). Validation of Predictive Mathematical Models Describing the Growth of Listeria monocytogenes. J. Food Protect. 60, (9), 1142-1145. [126] Walls I., Scott V.N.,Dane B. T. (1996). Validation of predictive mathematical models describing growth of Staphylococcus aureus. J. Food Prot., 59, 1, 11-15(5). [127] Westal S., Filtenborg O. (1998). Spoilage Yeasts of Decorated Soft Cheese Packed in Modfied Atmosphere. Food Microbiology 15, 243-249. [128] Zagory D. (1995). Principles and practice of modified atmosphere packaging of horticultural commodities. Lancaster, PA, Technomic Publishing Co Inc. 175-204. [129] Ziółkowski T., Staniewska K., Panfil-Kuncewicz H. (2003). Metody pakowania a bezpieczeństwo i trwało ć twarogów i serów dojrzewających. Przegl. Mlecz., 7, 269272. [130] Zurera –Cosano G., Castillejo-Rodriguez A.M., Garcia-Gimeno R. M., Rincon –Leon F. (2004). Performanc e of response surface and Davey model for prediction of

Influence of Interactions Occurring Between Micro-Organisms…

237

Staphylococcus aureus growth parameters under different experimental conditions. J. Food Prot., 67, 6, 1138-1145.

In: Focus on Food Engineering Research and Developments ISBN: 978-1-60021-898-9 Editor: Vivian N. Pletney, pp. 239-305 © 2007 Nova Science Publishers, Inc.

Chapter 3

THE DEVELOPMENT OF ENGINEERING TECHNOLOGY TO IMPROVE THE QUALITY OF PRODUCTION OF TROPICAL FRUIT IN DEVELOPING COUNTRIES B. Jarimopas1, P. Sirisomboon2, R. Sothornwit3 and A. Terdwongworakul4 1,3,4

Faculty of Engineering at Kamphaengsaen, Kasetsart University, Kamphaengsaen, Nakohn Pathom, Thailand 2 Department of Agricultural Engineering, Faculty of Engineering, King Mongkut’s Institute of Technology Ladkrabang, Bangkok, Thailand

ABSTRACT Many developing countries are rich in agricultural and food resources but are unable to maximize the export income they earn from them because they lack value-adding technology. In other words, developing countries typically must sell their products in cheap unfinished form to nations which possess the technology that adds profitability to these goods. Accordingly, if developing countries wish to earn more revenue for the improvement of their people’s employment and education, they must develop food engineering technology alongside other food science technologies. These efforts at technological self-improvement should be supported by the developed countries as the reduction of the knowledge and income gaps between the industrialized and developing worlds will do much to further global peace and happiness. The desired trend for food engineering research is to focus on developing engineering technology that will help to improve tropical fresh produce quality. This chapter discusses three facets of this trend. The first aspect concerns the physical properties of tropical fruit and vegetables, which consist of post-harvest loss, physical characteristics, mechanical properties, firmness, friction, and non-destructive quality grading techniques relating to mangoes, mangosteen, durian, sweet tamarind, guava, tangerines, snake egg plants, white long radish and lime. The second aspect concerns innovations in machinery and devices used with mangosteen, durian, young coconut, dry

240

B. Jarimopas, P. Sirisomboon, R. Sothornwit et al. over-mature coconut and baby corn. State of the art design, operating principles and key performance tests of tropical fruit machinery and inventions will be reviewed. The third aspect concerns packaging technology, particularly that which is directed towards the extension of the shelf life of the aforementioned tropical fresh produce. There are three current realities which inform this book. They are as follows: that there is a high incidence of post-harvest loss and a corresponding magnitude of shortage in research and development work on tropical fresh produce; that the global flow of information is increasing while agricultural labor is becoming scarcer and more expensive; and that tropical produce engineering technology must be thoroughly understood. Accordingly, we make two recommendations: for producer countries to instigate a dramatic increase in the research and development that they conduct into tropical fresh produce, and in the support that they provide for this research; and that the research trend should cover all economic tropical fruit and vegetable goods grown in the producer countries and all aspects of engineering technology that they use, with a particular emphasis on developing computerized non-destructive techniques for quality assurance.

INTRODUCTION If the task of science is to understand the composition of nature, the goal of engineering is to employ that scientific understanding in the quest to create new things. In other words, engineers who wish to improve the production of tropical fresh fruit and vegetables first must understand the natural behavior of tropical fresh produce. Only when they possess this knowledge can they develop innovations in production processes, devices and machinery. The next goal for engineers is to improve the distribution and preservation processes; that is, they must seek to improve the packaging that holds produce together and protects it from the adverse conditions of handling, transport and environment. Indeed, the drive to improve packaging has become of paramount importance due to the modern recognition of its capacity to secure standard qualities of freshness, uniformity, flawlessness and attractive appearance. Great opportunities exist today to increase export sales of tropical fruit and vegetables. This is implied by the comparatively low level of sales that currently exist. To clarify, the United States Department of Agriculture (2004) reports that exports to the world market of four temperate region fresh fruit and vegetables (apples, pears, potatoes, tomatoes) in 2002 were valued at more than USD5250 m., while exports by Thailand in 2006 of tropical fresh fruit and vegetables produced only USD172 m. in income (Customs Department, 2007). Thus, in effect, the huge markets of the US, the EU, Japan and China are challenging developing countries to improve their fruit and vegetable production technology, especially their engineering technology. As noted before, improvements in engineering technology should include developments in knowledge of the natural behavior of tropical fresh produce and the creation of innovative packaging technology for this produce. The natural behavior of tropical fresh produce relates to its physical properties. Innovations in this area occur when developments are made in primary processing and postharvest machinery, and in devices which facilitate the production and consumption of tropical fresh produce. Packaging technology relates to packaging for distribution and for extension of shelf life of produce.

Development of Engineering Technology to Improve the Quality of Production… 241 However, at present, the investment into the research and development of tropical fresh produce is comparatively very low. A Google Scholar search in April 2007 found that 79% more research (≅ 233) is conducted into apples, peaches and pears than the leading tropical fresh fruit (durian, longan, mangosteen, mango, young coconut, pineapple, pumelo, rambutan, rose apple, dragon fruit) in terms of their physical properties and associated nondestructive techniques. Meanwhile, 92.3% more research in the same areas is conducted into tomatoes, potatoes and carrots (≅ 298) than into the dominant tropical fresh vegetables (egg plant, snake egg plant, Chinese radish, chilli, Chinese cabbage, cabbage, kale, water morning glory). Accordingly, there is a serious shortage of knowledge in these aspects of engineering technology with regard to tropical fruit and vegetables. This chapter reviews and discusses the current body of knowledge relating to the natural properties of tropical produce and engineering innovations associated with these products, and then suggests a future trend for research and development which it is hoped will help to redress the serious imbalances described above.

1. PHYSICAL PROPERTIES OF TROPICAL FRESH PRODUCE IN DEVELOPING COUNTRIES 1.1. Post-Harvest Loss and Physical Characteristics of Selected Tropical Fresh Fruit Longan is one of Thailand’s major export fruits. The fruit are usually presented in bunches and, after manual harvesting, undergo sorting, handling, packaging and transportation. The first step of this post-harvest process is to separate twin fruit and broken branches and leaves from the gross product, a procedure which typically results in the retention of 78.5-87.0% of marketable fruit (Jarimopas, 1985). The usual effect of handling, packaging and transportation of the fruit is that some mechanical damage is caused in the form of fruit rupture and berry-dropping, with average losses running at 9.0% in two stacking containers and 14.8% in four stacking containers. The containers are cylindrical bamboo baskets which are piled directly on top of each other without cushioning. Mechanical damage is lowest in containers placed at the tops of stacks and is highest in containers placed at the lowest levels, possibly due to the higher compressive loads exerted by the upper containers. Another major export fruit in Thailand is mangosteen. This fruit has been investigated for post-harvest loss occurring at grading-buying points in the orchards and at retailer locations. Table 1 shows the various kinds of loss recorded from samples taken at the grading-buying points. The relevant fruit were stored at ambient temperatures until ripe, and were inspected and analyzed. The sampling survey was carried out in two provinces - Chantaburi in the east, and Chumpon in the south - both of which are major mangosteen production areas in Thailand. Samples were collected three times: at the beginning, the middle, and the end of the harvesting season. The greatest loss (48.6%) was due to rough surface, which greatly reduces the value of the fruit. The second greatest loss (33.9%) was due to internal defects (although it should be noted that mangosteen might have more than one defect (Pushpariksha and Jarimopas, 2006a)).

242

B. Jarimopas, P. Sirisomboon, R. Sothornwit et al.

To determine the post-harvest loss at the retailing stage, sampling was performed at representative sites of the most typical vending locations: at supermarkets, open markets, and mobile retailers (those who sell fruit from utility vehicles such as pick-up trucks). The sampled fruit was kept at room temperature for 3-4 days before being inspected and analyzed. Losses recorded in rank of frequency were due to rough surface (83.8%), hard rind (33.4%), translucent flesh and gummosis (26.9%), and decay (6.9%). The incidence of hard rind fruit, which is caused by mechanical injury, increased tremendously from the wholesaler to the retailer stages due to unsatisfactory packaging and inefficient handling and transit procedures. To clarify, mangosteen in Thailand which are intended for local consumption (rough fruit typically are not exported) usually are packed with minimal protection and care into paperlined reusable plastic containers before being transported by small trucks from the eastern and southern provinces to Bangkok. However, despite the high quantity of rough fruit found at local retailers, cracked fruit are generally absent because they have been culled out by the wholesalers. The quantity of fruit with internal disorders (translucent flesh and gummosis) tends also to be relatively low at retail point of sale because fruit with these defects tend to decay after they leave the wholesalers and accordingly are sorted out by the retailers (Pushpariksha et al., 2006). Post-harvest loss of rose apple fruit was ascertained with respect to two variables: variety (the Thongsamsri and Toonklao strains) and transport destination (retailers and wholesalers). Samples from retail locations were drawn from three mobile vendors, large open markets, and popular supermarkets around Bangkok and provincial cities. Wholesale samples were taken from three large fruit markets in Bangkok. The rose apples were manually harvested, packed and transported by trucks to all the above vending locations. Post-harvest loss was quantified in terms of abrasion and bruising damage. Two parameters of the damage evaluation were: Average damage per fruit (D x ) =

Total area of each damage type on fruit surface Total fruit of the package

Average damage percentage per package ( D y ) =

Number of damaged fruit in a package ×100 Total fruit of the package

(1)

(2)

With regard to post-harvest damage of the premium Thongsamsri variety found at wholesalers, the average damage due to bruising was 0.45 cm2/fruit (Dx) or 23.3% (Dy), while that due to abrasion was 0.66 cm2/fruit (Dx) or 72.2% (Dy). At the retailers, the average damage due to bruising was 1.45 cm2/fruit (Dx) or 56.8% (Dy), while abrasion was 0.95 cm2/fruit (Dx) or 78.1% (Dy). With regard to the Toonklao variety (a popular variety which is cheaper and less sweet than the Thongsamsri strain), damage was found to be higher at retail point of sale locations than at wholesalers. The major damage was bruising and abrasion. The average bruising and abrasion were 0.22 cm2/fruit (Dx) or 26.7% (Dy) , and 0.45 cm2/fruit (Dx) or 58.3% (Dy), respectively at the wholesalers while at retailers, average bruising and abrasion were 0.61 cm2/fruit (Dx) or 61.7% (Dy), and 1.21 cm2/fruit (Dx) or 90.0% (Dy), respectively (Toomsaengthong et al., 2006). We turn now to studies conducted into maturity grading and sizing of fresh tropical fruit. To evaluate the maturity of two Thai mango cultivars (Nam Dokmai and Chok Anan), Kittawee and Jarimopas (2006) attempted to measure the specific gravity (SG) of samples every two days for 40 days starting from when the fruit were immature until they were over-

Development of Engineering Technology to Improve the Quality of Production… 243 ripe by using the technique suggested by Mohsenin (1996). Good correlation was found between SG and maturity (based on time T after fruit set) for the Nam Dokmai variety while poor correlation occurred with reference to the Chok Anan strain. The regression equation for the Nam Dokmai analysis was SG = 0.87+0.0016T (R2 = 0.90). Sizing of produce is in general based on its dimensions. Sizing is not only useful for packaging, but also adds value to the produce (Jarimopas, 2006). Pushpariksha and Jarimopas (2006) studied physical characteristics of mangosteen by measuring the following variables of newly harvested fruit of four sizes (large, medium, small and undersize): weight, maximum diameter Dmax, minimum diameter, volume, diameter of calyx circumscribing circle Dc, and height with and without calyx. Table 2 shows the physical characteristics of fresh mangosteen related to size. The dimension ratio (DR), which was the ratio between Dc and Dmax, was proposed as a parameter to identify undersize fruit. It was found that the DR of the marketable mangosteen (excluding those that were undersized) was greater than 1. Therefore, DR could be a suitable parameter for undersize sorting. Since the weight of mangosteen stored at ambient temperatures dropped 14% in 2 weeks while their diameter decreased only 0.3%, mangosteen sizing by dimension would probably return more consistent results than sorting by weight.

1.2. Mechanical and Textural Properties Studies have been conducted into the mechanical and textual properties of tropical fruit such as mango and sweet tamarind and vegetables including snake egg plant and white long radish. Parameters of these properties includes rupture force, rupture deformation, the slope of the force-deformation curve, the Poisson’s ratio, modulus of elasticity, and firmness. Friction properties will also be reviewed and discussed in this section. A discussion of the technical experience of testing and measurement of the physical properties is also included. Table 3 (Chaiyapong and Jarimopas, 2006) shows the mechanical properties with respect to the rupture force FR, rupture deformation DR, and firmness expressed by the slope of the force-deformation curve S (Jarimopas and Kittawee, 2007) of the Thai popular mature mango “Nam Dokmai”. The experiment was performed with slow compression of the fruit by a 4mm plunger driven by the universal testing machine UTM (INSTRON 5569) at a loading rate of 20 mm/min. The compression test was controlled by ripeness (unripe, ripe) and fruit orientation (5 points of loading application – figure 1). FR, DR and S were significantly affected by ripeness and fruit orientation at the significance level of 5%. The top edge and the bottom edge of the mature and unripe mango statistically exhibited the highest and the lowest FR, respectively. Specifically, the FR of the bottom edge and the head of the mature and unripe mango were insignificantly different, with the greatest and the lowest slope S recorded at the top edge and the tail of the fruit, respectively. This implies the maximum firmness of mango occurs at the top edge and the minimum firmness at the tail. When mango is ripe, the ripening process has changed flesh cells to totally soluble solids, which results in reduction of the strength of the maximum FR of the ripe fruit to one-eleventh of that of the unripe fruit. The maximum FR of the ripe mango in this experiment occurred at the cheek (the most convex part of the fruit) while the FR of every point of load application was statistically indifferent, except at the fruit head where FR was the least. Firmness in terms of the slope of the ripe mango dropped to about one eleventh of that of the unripe fruit, while

244

B. Jarimopas, P. Sirisomboon, R. Sothornwit et al.

firmness of every load application point was statistically indifferent. However, the slope could not differentiate the maturity of Nam Dokmai and Chok Anan mangoes (Jarimopas and Kittawee, 2007). The Poisson’s ratio μ and modulus of elasticity E of the mango were also determined by means of uniaxial compression with the INSTRON 5569 until 50% of rupture force. The sample was cylindrical mature Nam Dokmai flesh measuring 15 mm in diameter by 30 mm in length. The resulting μ and E was 0.24±0.05 and 3.39±0.3 MPa, respectively (Chaiyapong and Jarimopas, 2006). Sirisomboon et al. (in press) have conducted a preliminary study which intended to add to the progress towards the design of a firmness tester suitable for mango maturity classification. They have designed an experiment which consisted of two parts: probe selection, followed by evaluation of the selected probe in mango maturity classification using a texture analyzer. In the first part, mango samples (Mangifera indica L. variety Namdokmai) at three different stages of maturity (120 fruit each at 60, 70 and 80% maturity, comprising a total sample size of 360 fruit) were tested. These different maturity stages were classified by farmers through sensory evaluation. A texture analyzer (TA-XT2i, Stable Micro System, UK) with 4 probes including a 5, 10, and 15 mm diameter spherical stainless steel probe and a 75- mm diameter circular flat aluminum plate probe were used to measure the firmness of the mango samples. Thirty mango fruit at each stage of maturity were subjected to a compression test at a maximum force of 3 N and probe speed of 0.2 mm/s (as no bioyield point appeared during the procedure, this can be considered a non-destructive test). Each probe was compressed on the cheek of one side of the fruit. The result showed that the 5-mm diameter spherical stainless steel probe provided the best performance due to its minimum value of standard deviation, coefficient of variation and variance. According to the Duncan’s test the means of the firmness values tested by the probe at different stages of maturity were significantly different (p13.0 N) and the third category (