advances in food and nutrition research volume 38

ADVANCES IN

FOOD AND NUTRITION RESEARCH

VOLUME 38

ADVISORY BOARD

DOUGLAS ARCHER Gainesville, Florida

JESSE F. GREGORY 111 Gainesville, Florida

SUSAN K. HARLANDER Minneapolis, Minnesota

DARYL B. LUND New Brunswick, New Jersey

BARBARA 0. SCHNEEMAN Davis, California

ADVANCES IN

FOOD A N D NUTRITION RESEARCH

VOLUME 38

Edited by

JOHN E. KINSELLA College of Agricultural and Environmental Sciences

University of California, Davis Davis, California

STEVE L. TAYLOR Department of Food Science and Technology

University of Nebraska Lincoln, Nebraska

ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper. @

Copyright 0 1995 by ACADEMIC PRESS, INC.

All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Academic Press, Inc. A Division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495

United Kingdom Edition published by Academic Press Limited 24-28 Oval Road, London NWl 7DX

International Standard Serial Number: 1043-4526

International Standard Book Number: 0-12-016438-8

PRINTED IN THE UNITED STATES OF AMERICA 95 96 9 7 9 8 99 OOQW 9 8 7 6 5 4 3 2 1

CONTENTS

CONTRIBUTORS TO VOLUME 38 ............................. vii PREFACE ................................................. ix

Hydrolytic and Transgalactosyiic Activities of Commercial 6-Galactosidase (Lactase) in Food Processlng

Lori F. Pivarnik, Andre G. Senecal, and Arthur G. Rand

I. Introduction ........................................ 1 11. Hydrolase Activity .................................. 22

111. Transgalactosylase Activity ........................... 59 IV. Summary and Research Needs ........................ 89

References ......................................... 90

Glass Transitions and Water-Food Structure interactions

Louise Slade and Harry Levine

I. Introduction ........................................ 103 11. Foundation of the “Food Polymer Science”

Approach .......................................... 106 111. Key Elements and Applications of the “Food Polymer

Science” Approach .................................. 138 IV. Research Needs: Outstanding Problems, Issues, and

Unanswered Questions .............................. 226 V. Conclusions and Future Prospects ..................... 233

References ......................................... 234

Corn Wet Milling: Separation Chemistry and Technology

David S. Jackson and Donald L. Shandera, Jr.

I. Introduction ........................................ 271 11. Corn: Structure and Types Used ...................... 273

V

vi CONTENTS

I11 . Steeping: Process and Equipment .................... 278 IV . Milling and Final Processing ........................ 287 V . Laboratory versus Commercial Milling . . . . . . . . . . . . . . . 288

VI . Research to Improve Wet Milling . . . . . . . . . . . . . . . . . . . 290 VII . End Products ..................................... 292

VIII . Summary ......................................... 296 References ....................................... 297

INDEX ................................................... 301

CONTRIBUTORS TO VOLUME 38

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

David S. Jackson, Department of Food Science and Technology, University of Nebraska-Lincoln, Lincoln, Nebraska 68583 (271)

Harry Levine, Nabisco, Fundamental Science Group, East Hanover, New Jersey 07936 (103)

Lori F. Pivarnik, Department of Food Science and Nutrition, Univer- sity of Rhode Island, Kingston, Rhode Island 02881 (1)

Arthur G. Rand, Department of Food Science and Nutrition, Univer- sity of Rhode Island, Kingston, Rhode Island, 02881 (1)

Andre G. Senecal, U S . Army Soldier Systems Command, Natick Research, Development, and Engineering Center, Natick, Mas- sachusetts 01 760 (1)

Donald L. Shandera, Jr., Department of Food Science and Tech- nology, University of Nebraska-Lincoln, Lincoln, Nebraska 68583 (271)

Louise Slade, Nabisco, Fundamental Science Group, East Hanover, New Jersey 07936 (103)

vii

This Page Intentionally Left Blank

PREFACE

As the new editor of Advances in Food and Nutrition Research, I am undertaking this task with decidedly mixed emotions. The untimely death of the previous editor and my friend, John Kinsella, was a loss to the entire food science and nutrition community. John had done a wonderful job in his few years at the helm of Advances in Food and Nutrition Research. The volumes that he edited are outstanding. In fact, John was the person most responsible for changing the name and scope of the series from Advances in Food Research to Advances in Food and Nutrition Research. Those of you who knew him (and even many who did not), know that John Kinsella was the perfect editor for an Advances series that covered the broad scope of food and nutritional sciences. John’s own research covered as much of that broad scope as any research that I have ever seen. The volumes that John edited for this series reflect the quality and foresight that were so evident in his own work. John Kinsella has left some mighty big shoes to fill with respect to all of his professional endeavors, including the editorship of Advances in Food and Nutri- tion Research. Although I am saddened to take this role in these circumstances, I will certainly do my utmost to continue the legacy of quality in these volumes that was so ably established by John Kinsella.

Volume 38 represents a transitional volume. The chapters “Glass Transitions and Water-Food Structure Interactions” by Louise Slade and Harry Levine and “Hydrolytic and Transgalactosylic Ac- tivities of Commercial P-Galactosidase (Lactase) in Food Process- ing” by Lori Pivarnik, Andre Senecal, and Arthur Rand were invited by John Kinsella. I am certain that you will agree that these chapters are outstanding and valuable insightful reviews for professionals in the field. They also ably reflect Dr. Kinsella’s career-long interest in the practical applications of food biochemistry. I appreciate the dedication of the authors of these two chapters in completing their writing assignments after the death of John Kinsella.

X PREFACE

The chapter “Corn Wet Milling: Separation Chemistry and Tech- nology” represents my first contribution as the new editor. Corn wet milling is a mature technology, but it has not been the subject of many recent reviews. I believe that it is often useful to review existing technologies in light of modern-day science. David Jackson and his graduate student, Donald Shandera, have done an excellent job with this assignment. I appreciate and acknowledge Dave Jackson’s efforts in putting together an outstanding review under a rather tight publication schedule after I assumed the editorship.

This last chapter also reflects a new opportunity for Advances in Food and Nutrition Research. The review on corn wet milling is a revised version of the literature review from Donald Shandera’s thesis. I suspect that there are many excellent literature reviews in the theses of graduate students which are never published. Advances in Food and Nutrition Research will consider for publication thesis literature reviews that are novel, insightful, and thorough in subject areas that are of substantial or potential importance to food and nutritional sciences. I look forward to receiving many good ideas for future reviews in Advances in Food and Nutrition Research.

I intend to maintain the broad focus on food and nutritional sciences established by John Kinsella. I welcome suggestions from readers on topics for future consideration.

STEVE L. TAYLOR

ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 38

HYDROLYTIC AND TRANSGALACTOSYLIC ACTIVITIES

FOOD PROCESSING OF COMMERCIAL p-GALACTOSIDASE (LACTASE) IN

LORI F. PIVARNIK, ANDRE G . SENECAL, AND ARTHUR G. RAND

Department of Food Science and Nutrition University of Rhode Island

Kingston, Rhode Island 02881 U.S. Army Soldier Systems Command

Natick Research, Development, and Engineering Cen ter Natick, Massachusetts 01 760

I. Introduction A. Nutrition B. Food Processing C. Waste Utilization

11. Hydrolase Activity A. Enzyme Mechanism for Activity B. Assay Methods and Enzyme Activity C. Properties of Commercial Enzymes D. Purity of Commercial Enzymes E. Activation and Inhibition F. Immobilization Mechanisms and Reactor Systems G. Ultrafiltration Bioprocess Reactors H. Future Potential for Commercial Enzyme Sources

A. Mechanism B. Assay Methods C. Properties D. Activity in Food E. Future Utilization

References

111. Transgalactosylase Activity

IV. Summary and Research Needs

I. INTRODUCTION

Food enzyme technology is the use of commercial sources of these bio- catalytic compounds in the processing of food. The enzymes used for this

1

Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.

2 LORI F. PIVARNIK etal.

purpose may be derived from plant, animal, or microbial sources, but generally do not require much refining to reach food-grade status. Since the 1950s, most commercial enzymes have been derived from microbial sources, replacing many of the plant- and animal-derived enzymes first used in food processing. One of the more successful commercial microbial source enzymes in foods in recent years has been P-D-galactosidase (EC 3.2.1.23), commonly known as lactase, because its main action is the catalysis of the hydrolysis of lactose (milk sugar). Lactase hydrolyzes the P-D-galactoside bond between D-galactose and glucose. Since the only source of lactose is milk, the principal application of this enzyme has been as an additive (commercial food-grade lactase) to milk and milk products; the consequent incubation predigests the lactose and converts milk sugar into the component monosaccharides. One of the main reasons to consider producing this change in milk is the resulting increase in sweetness and solubility of the sugar. This approach takes advantage of two unique properties of enzymes:

1. The narrow specificity guarantees that only the desired reaction occurs and maintains the nutritional integrity of the food.

2. The catalytic function means that a very small amount of enzyme is required to produce a very large effect.

A. NUTRITION

Milk, nature’s most nearly perfect food! This phrase has been used both to describe and to promote milk for years [Patton, 1969; National Dairy Council (NDC), 1977; Rosensweig, 19781. The basis for this statement about a single food is the vital role milk plays in sustaining life, particularly for newborn mammals. Milk and other dairy foods have contributed significantly to human diets as sources of the essential nutrients protein, riboflavin, calcium, vitamin D, niacin, vitamin BIZ, phosphorus, magnesium, vitamin A, zinc, and iodine (Phillips and Briggs, 1975; NDC, 1977; Savaiano and Kotz, 1988). The nutrients in milk have also found increasing roles as functional ingredients in an amazing variety of food products (Jonas, 1973; Jost, 1993; Allen, 1993).

1. Sugar Malabsorption

Problems with digestion and absorption of components in foods, such as the carbohydrates, is not a new issue (Townley, 1967). Early in this century, clinicians noted a correlation between dietary disaccharides and diarrhea. It has only been in the past 30 years, with the wide availability and dependence on milk and dairy foods in the food supply, that milk has created

COMMERCIAL P-GALACTOSIDASE IN FOOD PROCESSING 3

serious problems for those individuals described as lactose intolerant (Solo- mons, 1986). Perhaps 70% of the world’s population may develop gastrointestinal problems after ingesting milk products or foods containing dairy ingredients that incorporate sufficient amounts of the milk sugar lactose (Kretchmer, 1972; Rosensweig, 1978; Savaiano and Kotz, 1988). Similar to other mammals, after weaning, humans lose the majority of intestinal lactase (P-galactosidase) activity required for lactose digestion (Kretchmer, 1972). The majority of world population groups that have gastrointestinal problems with milk lose the most lactase and have genetic origins in zones of the world where milk has not been a component of the traditional diet (Kretchmer, 1972; Rosensweig, 1978; Simoons, 1981). The small proportion of the world’s adults (about 30%) who have adapted to dairy foods by maintaining ample lactase levels genetically originated in areas with a long tradition of milk consumption (Simoons, 1981; Savaiano and Kotz, 1988). In the United States, the lactose-intolerant population varies widely, depending on genetic origin (Holsinger and Kligerman, 1991). It is estimated that on average about one-third of the U.S. population has lactose digestion difficulty (Houts, 1988; Holsinger and Kligerman, 1991).

Low intestinal lactase levels result in the gastrointestinal discomfort experienced by lactose-intolerant individuals. Intestinal lactase, immobilized in the brush border of the small intestine surface, accomplishes both digestion and absorption of dietary carbohydrate (Gray, 1971,1975; Rosensweig, 1978). Lactase deficiency is a reduction in the sites available to complex with dietary lactose and remove sugar from the intestinal fluid. The lactose bound to the enzyme at the brush border undergoes hydrolysis and absorption. The extent of lactose intolerance symptoms depends on the level of active lactase remaining after the postweaning decline and the ratio of lactase to ingested lactose. Following consumption of milk or dairy foods, lactose not bound and digested in the small intestine travels to the large intestine, where fermentation by the microorganisms converts the sugar into acids and gas (Wolin, 1981). The result is the rapid onset of flatulence, pain, and diarrhea, the classic symptoms of lactose intolerance. This problem is much more prevalent than the adverse reactions individuals can have to other components in milk, as in allergies that involve the body’s immune system but do not impair the ability to digest lactose (Solomons, 1986; Matsuda and Nakamura, 1993).

The decrease in intestinal lactase levels with age, increase in symptom production, and resulting discomfort from ingesting lactose lead to the avoidance of dairy foods (Paige, 1981). Although these immediate effects of lactase deficiency represent natural consequences, the long-term impact has serious nutritional complications (Newcomer, 1981). A decline in consumption of dairy foods has been linked to poor nutritional status (Phillips


and Briggs, 1975). Phillips and Briggs found that adults who limit or omit milk from the diet run the risk of low intake of calcium and possibly other nutrients such as riboflavin, thiamine, phosphorus, and vitamin A. Dairy products contribute 42% of the riboflavin and up to 20% of the thiamine in the U.S. diet. Great concern has been focused on the loss of calcium in the diet since Birge et al. (1967) connected osteoporosis with intestinal lactase deficiency, which was affirmed in reviews of bone metabolism and biochemistry (Lutwak, 1975) and metabolic bone disease (Nutrition Re- views, 1979).

As dairy products account for just over half to perhaps three-quarters of the calcium intake by the US. population (Phillips and Briggs, 1975; Allen and Wood, 1994), there is cause for concern about adequate intake if people avoid milk. There has been increasing evidence that adequate dietary calcium not only helps prevent brittle bone disease (Chapuy et al., 1992), but may help mitigate the risk from several other serious and chronic diseases (CRH, 1994; Heaney and Barger-Lux, 1994). Current recommendations for calcium intake are 1000 mg/day for all adults and 1500 mg/ day for postmenopausal women (Heaney, 1993; CRH, 1994; Heaney and Barger-Lux, 1994). These levels can easily be met only if people consume milk-based foods. Calcium intake has also been linked to a role in blood pressure regulation and control of hypertension [National Academy of Sciences (NAS), 1989; Allen and Wood, 19941. Other evidence suggests that increased calcium in the U.S. diet could be effective in lowering total and low density lipoprotein cholesterol (Bell et al., 1992; Denke et al., 1993). It has been estimated that the effects of elevated dietary calcium levels on cholesterol and hypertension could reduce the risk of heart attack by more than 20% (CRH, 1994). A high calcium intake in excess of the recommended daily allowance (RDA), 1000 mg, can also be protective against colon cancer (NAS, 1989; Allen and Wood, 1994; CRH, 1994). High intakes of calcium, vitamin D, and milk products could reduce the risk of colon cancer 20 to 30% (Bostick et al., 1993). It thus seems clear that this extremely valuable nutrient must be made available to people at levels higher than those currently in existence. Most Americans consume too little of this valuable nutrient, and avoiding milk products prevents those individuals from obtaining the vast potential health benefits of calcium (Heaney and Barger-Lux, 1994).

2. Lactase

As most adults are lactose intolerant because they have insufficient quantities of lactase in their small intestine, these individuals tend to avoid milk because of the risk of serious discomfort. Development of commercial


lactase (P-galactosidase) has stimulated the production of lactose-reduced foods (Rand, 1981). This enzyme is able to hydrolyze lactose into the monosaccharides glucose and galactose, which are readily absorbed by the small intestine (Paige et al., 1975; Turner et al., 1976; Hourigan and Rand, 1977; Cheng et al., 1979; Reasoner et al., 1981). Short-term studies have established that lactose-intolerant individuals who consume low-lactose milk show improved absorption and reduced symptoms, and that the increase in osmolarity caused by the hydrolysis does not appear to have any side effects. The change in taste, resulting from the increased sweetness of the monosaccharides, has caused palatability problems. Cheng et al. (1979) demonstrated the acceptability of low-lactose milk over an extended period. Lactose-intolerant individuals are now able to benefit from the nutrients supplied by milk, thus reducing the danger of low intake of calcium, phosphorus, vitamin A, and riboflavin (Rand, 1981).

The impact of lactase on predigesting the lactose in milk and making the food available to more people raises some questions on the nutritional changes produced. It has been known since the 1930s that lactose has a stimulating effect on calcium absorption (NDC, 1977; Pansu et al., 1979). In fact, lactose has profound effects on the absorption and retention of many minerals, including iron, zinc, magnesium, and manganese (Bushnell and DeLuca, 1981). What happens when lactose is converted to glucose and galactose by enzyme treatment? It appears that the presence of the monosaccharides does not alter the absorption of nutrients (Kobayashi et al., 1975; Vega et al., 1992).

B. FOOD PROCESSING

1. Commercial Enzyme Sources

Lactase (P-galactosidase) is widely distributed in nature and has been found in plants, animals, and microorganisms (Pomeranz, 1964a; Mahoney, 1985; Agrawal et al., 1989). For use in food processing, however, the enzyme must be derived from sources that are classified “generally regarded as safe” (GRAS) by the Food and Drug Administration (FDA). Sources that have been regarded as food constituents have found easy acceptance as GRAS or in the Food Chemicals Codex (FCC) (Nelson, 1980). The ap- proved sources of lactase all fall in the category of fermentation-derived enzymes, primarily from yeasts and molds (Mahoney, 1985; Agrawal et al., 1989; Holsinger and Kligerman, 1991; Bigelis, 1993). Current production seems to be restricted to the molds Aspergillus niger and Aspergillus oryzae and the yeasts Kluy veromyces lactis, Kluyveromyces fragilis, and Candida pseudotropicalis (Kuntz, 1993). All enzymes from these sources and from


bacteria closely related to Bacillus stearothermophilis can be used to hydrolyze the lactose in milk and alleviate the problems with milk consumption by lactose-intolerant populations (Mahoney, 1985). These enzymes have varied properties that provide strengths and weaknesses for a variety of applications. A review of lactase products from commercial suppliers indicates that only sources derived from yeasts and molds are currently available for industrial application.

2. Low-Lactose Milk

The mold lactases can be used to produce low-lactose milk, but work best at higher temperatures, 50" to 60"C, and lower pH, 3 to 5. These enzymes can be used to hydrolyze lactose at temperatures close to that of pasteurization, which minimizes growth of more than 90% of common spoilage bacteria (Rand and Linklater, 1973; Nijpels, 1981; Rand, 1981). Figure 1 illustrates the possible use of A. niger lactase to process milk during pasteurization. This enzyme can be added to the milk while cold; it is then heat activated as the temperature is raised to 63°C. Hydrolysis of lactose occurs during the 30- to 60-minute holding period. The lactose

~ - 5 0

h

3 - 4 0 m

H m

2 3 0 1 2 0 1 w

3

1 0

TEMPERATURE ( " C ) - HYDROLYSIS ( 5 )

I (1 0 5 0 100 1 5 0

TIME (min)

FIG. 1. Processing of lactose-hydrolyzed milk with fungal lactase (enzyme concentration = 0.33% w/v). Reprinted with permission from Rand and Linklater, 1973.


in milk is reduced while pathogenic microorganisms are eliminated and growth of spoilage bacteria is minimized. This illustrates how lactose reduction of milk products and pasteurization can be combined employing thermostable lactases. These enzymes might even survive conventional high temperature short time (HTST) pasteurization (Rand, 1981); however, the pH optima for these enzymes do not correspond well to that of milk, and they would be better suited for use in processing whey and acidified products (Nijpels, 1981; Holsinger and Kligerman, 1991; Bigelis, 1993).

The yeast lactases operate best near neutral pH and at milder temperatures, 30" to 40"C, conditions that prove very effective for production of low-lactose milk. This commercial enzyme source permits lactose hydrolysis in milk before or after pasteurization. The enzyme can be added to pasteurized milk, which is then incubated near the optimum temperature, about 30" to 35"C, for about 2 to 5 hours, hydrolyzing 70 to 90% of the lactose (Hourigan and Rand, 1977; Guy and Bingham, 1978; Mahoney, 1985). This, however, introduces an additional process step and additional expense. An approach has been proposed that avoids this problem: a very small amount of food-grade filter-sterilized yeast lactase is injected into ultrahigh- temperature (UHT)-sterilized and packaged milk (Dahlqvist et al., 1977). This product can be stored at room temperature and undergoes nearly complete lactose hydrolysis in 7 to 10 days. The enzyme used for this process must be free of protease (and other hydrolases) to prevent deterioration of the milk during storage (Mahoney, 1985). In an alternative process, which has proved most successful in the United States, lactose hydrolysis is induced in milk before pasteurization by adding yeast lactase to cold milk (Hourigan and Rand, 1977; Guy and Bingham, 1978; Reasoner et al., 1981). This treatment can accomplish up to 90% lactose hydrolysis while milk is held overnight at 4" to 6"C, a normal step before processing, which minimizes the growth of spoilage microorganisms.

3. Processing by the Consumer

The capability for low-temperature processing of low-lactose milk with yeast lactase led to the concept of low-lactose milk processing by the consumer (McCormick, 1976; Kligerman, 1981; Holsinger and Kligerman, 1991). Initially, lactase could be purchased in a powdered form, as packets, to treat a quart of milk overnight in the home refrigerator. This unique marketing concept, developed by Lactaid, Inc., made low-lactose milk available to many consumers long before commercial reduced-lactose fluid milk was in distribution in many areas of the United States. Now, the enzyme for home processing comes in a sterile liquid form, which is much more stable and easy to use.


Another alternative method for producing low-lactose milk at home supplies lactase to the consumer as a dose form (DeAngelis et af., 1979). The initial product consisted of yeast lactase encapsulated in algin beads, partially dehydrated with sugar to yield a reduced water activity (aw), subcoated, and enteric-coated capsule as shown in Fig. 2. The final product was packaged in gelatin capsules. The enzyme is delivered to the site of lactose digestion and absorption, the small intestine. The efficacy of this lactase dose form was tested in a clinical trial, as shown in Fig. 3. The average symptoms of lactose intolerance per person were reduced by the administration of one or two capsules just prior to the consumption of

FIG. 2. Stages involved in lactase immobilization and development of a capsule dose form.


FIG. 3. Effect of lactase capsule dose form on milk intolerance symptoms with 240 ml of skim milk. Average symptom = response to 4 symptoms of bloating, cramps, flatulence, and diarrhea. Severity rated as 1 = mild, 2 = moderate, and 3 = severe.

240 ml of skim milk. Commercial forms of this product were introduced in the mid-1980s (Holsinger and Kligerman, 1991). Now tablets containing combinations of yeast and mold lactases are available for consumers to take with a meal. Thus, whenever lactose-intolerant individuals suspect they have ingested or intend to ingest foods containing lactose, they can mitigate the symptoms by taking one of these dose forms.

4. Lactase Product Inhibition

Utilization of low-lactose hydrolyzed milk has been hampered because of product inhibition, attributed to the formation of galactose (Rand and Linklater, 1973; Hourigan and Rand, 1977; Rand, 1981). Product inhibition has represented a major disadvantage to batch-type processing. Large quantities of expensive enzyme have been required to increase the degree of lactose conversion in the process, which has normally been conducted over a period of about 24 hours while the milk was refrigerated. Reduction of galactose during lactose hydrolysis would greatly increase enzyme efficiency. This concept can be visualized comparing the kinetic constants for galactose inhibition in the two major classes of food-grade lactases. Studies comparing the major data have been reviewed by Rand (1981) and Mahoney (1985) and are summarized in Fig. 4. The K,,, values differ for mold and yeast lactases, depending on the medium used for evaluation. A. niger lactase had better affinity for lactose in milk than in buffer, whereas K.

10 LORI F. PIVARNIK etul.

FIG. 4. Comparison of kinetic values and product inhibition for A. niger and K. luctis lactases in both buffer and skim milk.

lactis enzyme performed more efficiently in buffer than in milk; however, the affinity for galactose, as a competitive inhibitor, was always higher for the mold enzyme. Thus, the K,,,/Ki ratios reveal that K. factis lactase per- forms better than A. niger enzyme in both milk and buffer when lactose is the substrate. It was also clear that any method that could remove the galactose inhibitor would improve the performance of both enzymes, but particularly the yeast enzyme, which did not perform as well in the food system.

Enzymatic modification of galactose during lactose hydrolysis to reduce the inhibitor was first proposed by Finnie et af. (1979). P-Galactose dehydrogenase was coupled with lactase to specifically remove the inhibitor and convert the sugar to the corresponding lactone with an impact during reactions in buffer (Rand, 1981). This coupled enzyme approach was evaluated during the hydrolysis of lactose in milk at 4°C with three lactase enzymes, and the results are compared in Fig. 5. The most significant impact was observed with K. fiagilis lactase (KFL) (Fig. 5A). Hydrolysis of milk proceeded over 24 hours and the effects of coupled enzyme modification of galactose improved as the reaction proceeded. The effect of coupled


FIG. 5. (a) Hydrolysis of lactose in milk with K. fragih lactase (KFL) at 4°C and coupled with P-galactose dehydrogenase (P-GDH). (KFL = 0.03 unit/ml, P-GDH = 0.04 unit/ml, 0.17 mg/ml NAD.) (b) Hydrolysis of lactose in milk with K. la& lactase (KLL) at 4°C and coupled with P-galactose dehydrogenase. (KLL = 0.04 unit/ml, P-GDH = 0.04 unit/ml, 0.17 mdml NAD.) (c) Hydrolysis of lactose in milk with A. niger lactase (ANL) at 4°C and coupled with fl-galactose dehydrogenase. (ANL = 0.09 unit/ml, P-GDH = 0.04 uniuml, 0.17 mdml NAD.)

enzyme hydrolysis with K. factis lactase (KLL) was less dramatic (Fig. 5B). Addition of &galactose dehydrogenase did not improve the process until significant amounts of galactose had accumulated, but when conversion

12 LORI F. PIVARNIK eral.

exceeded 50%, the coupled system began to have a definite impact. The control hydrolysis reaction clearly began to slow down in efficiency when about half the lactose had been hydrolyzed to galactose; however, the coupled enzyme system continued to proceed at a relatively constant rate from 12 to 48 hours, showing minimum impact of galactose inhibition. Coupled enzyme hydrolysis with A. niger lactase (ANL) (Fig. 5C) pro- gressed at a much slower rate, as expected at refrigeration temperatures, and did not reach the conversion levels of the yeast lactases; however, it seems clear that the coupled enzyme system was slowly improving the performance of the mold enzyme over the control process. This approach demonstrates the potential to improve significantly the efficiency of lactase enzymes to food systems by reducing the inhibitory galactose product.

The effect of galactose on lactase activity during milk processing can be described as typical competitive inhibition (Rand, 1981; Mahoney, 1985). This means that the effects of inhibition can be overcome by keeping the substrate concentration much higher in the vicinity of the enzyme. This might be the way the mammalian small intestine deals with galactose inhibition of lactase during digestion. Intestinal lactase, immobilized on the apical surface of the intestinal mucosa, permits rapid hydrolysis and absorption of the sugars in individuals who are not lactase deficient (Gray, 1975). Intestinal hydrolysis of lactose is much more efficient than that in milk, as the reaction is controlled by an absorptive process that maintains higher concentrations of substrate near the enzyme and rapidly removes products. This approach to minimizing product inhibition of lactase enzymes was evaluated in a membrane reactor that localized K. luctis lactase to the lumen side of a hollow fiber in a pressure-induced regime (Maculan et al., 1978). The results from this study are presented in Fig. 6. It is clear that as the pressure increased from 2.5 to 10 psig, the flux or flow across the enzyme layer and the membrane also increased from 52 ml/h up to nearly 360 ml/h. The concentration of glucose in the permeate decreased as flux increased and contact time with the enzyme layer decreased, but the actual rate of glucose production increased from 180 mg/h at 2.5 psig to 244 mg/hr at 10 psig. Thus, the enzyme operated more efficiently when the substrate concentration was kept high in the vicinity of the enzyme, while displacing galactose and removing this competitive inhibitor through the membrane.

It is clear from these studies that any approach that reduces or removes galactose from the system during lactose hydrolysis has the potential to improve the efficiency of lactase enzymes. The implications for milk processing could result in the use of smaller amounts of enzyme, thereby reducing cost and saving time in processing.


FIG. 6. Effect of lactose flow rate as pressure on flux and glucose production for a localized lactase UF bioreactor. (K. luctis lactase = loo0 pg/cm2, temperature = 30°C.)

5. Lactase in Ice Cream

Some of the earliest applications of lactase in food processing occurred in frozen products like ice cream (Pomeranz, 1964b). The original goal was to eliminate crystallization of the lactose and sandiness, but these problems have been almost completely eliminated by the increased use of stabilizers. The reduction of lactose content could have another benefit: the ability to use increased amounts of whey and skim milk products in the formulation (Mahoney, 1985). It now appears that reduced-lactose ice cream mixes yield products with other improved properties and the potential for incorporation into reduced-calorie forms (El-Neshawy et al., 1988; Alagialis et al., 1990; Rand and Kleyn, 1991). Lactose hydrolysis has been shown to produce products of higher viscosity and whipping ability and products with higher overrun and improved organoleptic properties (El-Neshawy et al., 1988). Alagialis et al. (1990) began with a premium-style ice cream of 16% fat and >40% total solids, which served as the control. The results, presented in Fig. 7, demonstrate that with lactose hydrolysis by lactase, advantage can be taken of the increased sweetness and the amount of sucrose in the formula can be reduced as one step toward calorie reduction. The control premium ice cream at 13.4% sucrose was evaluated on sensory scores for flavor, bodykexture, melting, and color as 24.3 out of a maximum total of 25. When yeast lactase was added at 3.5% to the formula to achieve 96% hydrolysis of lactose, the sensory score decreased about 1 point, mainly


FIG. 7. Effect of lactose hydrolysis catalyzed by lactase on sucrose reduction as demonstrated by the sensory properties of premium ice cream. (Liquid lactase = 3.5%, Mix total solids = >40%, fat = 168.)

because of a lower flavor score, which was attributed to increased sweetness. Reduction of the sucrose by 18% (11% sucrose) actually improved the sensory score to nearly the same as that of the control, demonstrating the benefits of improved bulking properties and the need to use less sucrose. Further reductions in sucrose to 10, 8, and 6% produced a steady decline in the sensory properties, which were reflected primarily in the flavor and bodykexture scores as sweetness and bulking decreased. Rand and Kleyn (1991) evaluated the effects of lactase on the properties of nonfat ice milk of 31% total solids, as shown in Fig. 8. When formulated without lactase and frozen, this ice milk was evaluated as poor with respect to flavor, body/ texture, and overall sensory perception. When yeast lactase was added to the mix, at 0.3%, and allowed to react overnight before freezing, the flavor and bodykexture scores increased nearly 2.5 times and the overall sensory score substantially improved. The effects of lactase treatment on the properties of ice cream-type products was confirmed by a study from the Nutra- Sweet Company (Keller et al., 1991). It found that lactase had definite benefits in permitting the use of increased milk solids to replace sucrose. The enzyme treatment allowed the use of milk solids instead of alternative bulking agents for aspartame-flavored frozen dairy desserts.

6. Lactase in Fermented Milk

The lactose content of dairy mix intended for yogurt manufacture was shown by Goodenough and Kleyn (1976a) to average about 8.5%, which


SENSORY SCORE

FIG. 8. Effect of lactase on the sensory properties of nonfat ice milk. (K. lactis lastase = 0.3%, Mix total solids = 31%.)

decreased 35% during fermentation to about 5.75%. This concentration was slightly greater than the normal lactose content of whole milk. Thus, it was somewhat surprising when consumption of lactose as yogurt, and other fermented dairy products, was found to cause minimal intolerance symptoms in lactase-deficient populations (Gallagher et al., 1974; Kolars et al., 1984; Gilliland, 1985). The beneficial effects of better digestion and utilization of lactose have been attributed in yogurt to the presence of active cultures of Streptococcus thermophilus and Lactobacillus bulgaricus (Kolars et al., 1984; Kelley, 1984). An alternative approach to digestion problems caused by lactose was the use of sweet acidophilus milk, introduced in 1975 (Shearin, 1977). The beneficial effects on lactose digestion in humans of consuming nonfermented milk, following the addition of live Lactobacillus acidophilus cells, have been reported (Gilliland, 1985). Subsequently, these digestive benefits have been attributed more specifically to the presence of substantial amounts of active lactase from the cells of the microbial cultures, which contribute to lactose digestion after consumption of the food (Kilara and Shahani, 1974; Goodenough and Kleyn, 1976b; Onwulata et al., 1989; De Vrese et al., 1992). These studies attempted to measure

16 LORI F. PIVARNIK etaf.

lactase activity either in the microbial cells in the food or in the intestinal tract after consumption. This led to the understanding that the lack of digestive problems after consuming these foods is due to the active cells and enzymes; this has discouraged a trend toward heat processing these products (Kelley, 1984; Gilliland, 1985).

Lactase has also found applications in the processing of fermented dairy foods. Although the microorganisms used to ferment dairy foods have been adapted to effectively ferment lactose, the process of hydrolysis before use appeared to be a rate-limiting step (Nijpels, 1981; Mahoney, 1985). Manufacture of yogurt from low-lactose mix has shown several benefits. Hydrolysis of lactose may reduce the gelation time for yogurt by 15 to 20% (Nijpels, 1981; Hilgendorf, 1981), body and texture properties may be improved (Dariani et al., 1982), and the enhanced sweet taste can improve acceptability without increasing calories (Engel, 1973; Nijpels, 1981; Hil- gendorf, 1981). The process of lactose hydrolysis may also permit the development of other approaches to manufacturing fermented dairy products, such as combining enzymatic and microbial acid development for the bio- processing of yogurt (Tahajod and Rand, 1993). The effect of lactose hydrolysis by lactase on promoting enzymatic acidification of milk by glucose oxidase (GOX) for yogurt manufacture is shown in Fig. 9. This study demonstrated the impact of two yeast lactase concentrations on the production of glucose and the corresponding conversion to gluconic acid by GOX. Within 30 minutes, lactase at each concentration produced sufficient glucose for acidification of milk by GOX, which was added at that point, followed by hydrogen peroxide. The pH of milk declined to pH 5.6 to 5.7 within an additional 30 minutes, which was considered optimum for the addition of Lactobacillus bulgaricus to complete the conversion of yogurt. Lactose hydrolysis may also benefit the manufacture of fresh cheeses, like cottage cheese, resulting in a shorter gelation time, a firmer curd, and perhaps a 10% increase in yield (Nijpels, 1981). The process for making aged cheese, like cheddar, has also shown potential benefits from lactase treatment. The acidification step has been shown to occur at a slightly higher rate and may produce higher bacterial counts (Nijpels, 1981). The principal effect of lactase treatment has been shown to be a reduction in the expensive ripening time for these cheeses, due to acceleration of the aging process (Anony- mous, 1977; Olson, 1979).

7. Immobilized Lactase Systems

Whether it is used to solve nutritional problems (lactose intolerance) or to improve the product (lactose solubility, sweetness, functionality), enzymatic hydrolysis must be economically feasible. The additional cost of


pH 6.6-

6 .4 -

6 .2-

6.0-

5 .8-

6.6' PH

6 . 4 -

6 . 2 -

6.0-

5 .8 -

5 .6 -

2 0 4 0 6 0 8 0 Time (dn)

0 2 0 40 60 80

Time (min)

8 0 0

600

400

2 0 0

0

1000

8 0 0

6 0 0

4 0 0

2 0 0

0

GLUCOSE PRODUCTION t d d l )

GLUCOSE PRODUCTION t d d l )

LLJ - Glucose

FIG. 9. Effect of lactase concentration on glucose production and acid formation by glucose oxidase in skim milk at 37°C: (a) 0.25 mg yeast lactase/ml milk; (b) 0.38 mg yeast lactase/ml milk. (Glucose oxidase = 2.5 units/&, added at 30 min; 0.3 ml 50% Hz02 added in 0.1-1111 increments at 10-minute intervals.)


adding soluble, food-grade lactase to milk has aroused interest in the use of immobilized systems for recovery and reuse of the expensive enzyme component. Immobilization of lactase has been reported employing virtually all of the major systems available: adsorption, covalent bonding, gel entrapment, and containment by semipermeable membranes (Agrawal et al., 1989).

Adsorption of lactase to an immobilizing matrix has been reported using a variety of materials including phenol-formaldehyde resins (Olson and Stanley, 1973; Stanley and Palter, 1973; Okos et al., 1978), alumina (Charles et al., 1975; Finocchiaro et al., 1980b), stainless steel (Charles et al., 1975), and hydrophobic attachment to zeolites (Zentgraf and Gwenner, 1992). Covalent bonding of lactase to various supports has been the subject of studies with cellulose sheets (Sharp et al., 1969), the most common matrix of porous glass beads (Woychick and Wondolowski, 1972; Wierzbicki et al., 1973; Weetall et al., 1974a,b Okos and Harper, 1974), gelatin (Sungur and Akbulut, 1994), solid-phase Ni( 11) chelate (Irazoqui and Batista-Viera, 1994), and corn grits (Siso and Doval, 1994). The use of gel entrapment for lactase immobilization has been the subject of studies with polyacrylamide slices (Bunting and Laidler, 1972) and polyacrylamide beads (Nilsson et al., 1972), which resulted in its application to production of low-lactose milk (Dahlqvist ef al., 1973). The gel entrapment process evolved to use of the more acceptable food-grade algin beads for lactase immobilization (DeAngelis et al., 1979, Jacober-Pivarnik and Rand, 1984) and the development of fiber-entrapped lactase for milk processing (Dinelli, 1972; Morisi et al., 1973). The entrapment process has also led to a unique approach to the use of immobilized enzymes for lactose hydrolysis by incorporating the lactase enzyme into the packaging material (J. A. Budny, personal communication, 1989). This approach has the potential to convert lactose in fluid milk products following processing and packaging.

The last method of lactase immobilization, which holds perhaps the most promise, is ultrafiltration (UF) and the containment of enzymes by semipermeable membranes. Membrane technology has become standard in the food industry, which means the equipment and technology have become widely available to take advantage of the size differences between enzyme and product (Cheryan, 1986; Cuperus and Nijhuis, 1993). The initial UF systems for immobilizing lactase employed continuous stirred tank reactors (CSTRS), in which the enzyme recirculated with the substrate, while the glucose and galactose products were removed through the membrane (Miller & Brand, 1980; Huffman-Reichenbach and Harper, 1982). The CSTR system was not very efficient for a product-inhibited enzyme process and has never achieved commercial success. Plug-flow reactor (PFR)-based systems consisting of a single pass showed much greater prom-

COMMERCIAL B-GALACTOSIDASE IN FOOD PROCESSING 19

ise for lactase-catalyzed conversion of lactose, because at least part of the enzyme-matrix would always be exposed to fresh substrate to minimize product inhibition and maximize activity (Charles et al., 1975; Finocchiaro et al., 1980a). The use of a lactase tubular reactor was introduced by Ngo et al. (1976); the enzyme was placed on the inner surface of nylon tubes and shown to be potentially efficient. This type of enzyme reactor was applied to UF membranes by employing hollow-fiber technology (Breslau and Kilcullen, 1975; Huffman-Reichenbach and Harper, 1982; Jones et al., 1988). Thus lactase UF bioreactors moved away from the CSTR mode and closer to the PFR type. The enzyme is added to the membrane on the sponge side (outside surface) of the hollow fiber, and lactase is kept in place essentially by adsorption and entrapment. If the enzyme is not permanently bound to the sponge layer, removal is achieved relatively easily by simply reversing flow during cleaning, and the membrane is immediately made ready for reloading. As already discussed, Maculan et al. (1978) demonstrated the potential of operating a UF reactor as an intestinal model, using pressure as the immobilizing force, as shown in Fig. 10. In this approach, the enzyme is layered on the lumen side of the hollow fiber and immobilized by pressure, allowing operation as a PFR with improved efficiency. This type of bioreactor was evaluated on skim milk UF permeate and found to operate efficiently at 5 psi and 30°C with a yeast lactase (Seneca1 and Rand, 1992a). The lactase UF bioreactor shows great potential for treatment of milk and whey permeates, and could be coupled into a system such as that

UF OUTER "SPONGE"

k LAYER

1 -Product

J U R E IMMOBILIZED

F- ENZYME

FIG. 10. Pressure-induced immobilization of lactase on the surface of a hollow-fiber ultrafiltration (UF) membrane.


proposed by Miller and Brand (1980). This would permit separation of milk into retentate and permeate fractions as shown in Fig. 11. The separation process would be coupled directly to hydrolysis of the lactose in the permeate stream by the UF bioreactor. Retentate and permeate streams could then be recombined to produce low-lactose milk products. The main advantage of this approach is that the bioreactor would be kept essentially free of protein fouling and microbial contamination, which would enhance its potential to operate continuously for extended periods.

C. WASTE UTILIZATION

1. Whey

The largest waste problem for the dairy industry is whey: sweet whey as the by-product of cheese or rennet casein with a pH greater than 5.5, and acid whey derived from the manufacture of cottage cheese or acid casein with a pH less than 5.0 (McDonough, 1977; Shahani et al., 1978; Short, 1978; Newton, 1991; Siso and Doval, 1994). Only about half of the total world whey production has been used as human or animal food; the rest continues to be discarded, causing serious environmental problems (Mc- Donough, 1977; Shahani et al., 1978, Siso and Doval, 1994). The major constituuent of all types of whey is lactose, followed by high-quality protein as the second ingredient (Short, 1978). Whey proteins have usually been recovered by UF and used in foods as protein concentrates (Brinkman, 1976; Shahani et al., 1978; Siso and Doval, 1994); however, protein recovery

FIG. 11. Ultrafiltration process for the separation of milk and subsequent hydrolysis of lactose in the permeate stream with a hollow-fiber lactase bioreactor.

COMMERCIAL fl-GALACTOSIDASE IN FOOD PROCESSING 21

does not solve the environmental problems caused by whey, as most of the difficulties are due to lactose, which remains in the permeate (Siso and Doval, 1994). Direct utilization of lactose has been difficult, because in food applications the amount of lactose that can be added is limited by low solubility and sweetening power (Okos et al., 1978; Short, 1978). Hydro- lysis of lactose into glucose and galactose can solve many of these problems and potentially increase its value as a food ingredient. Hydrolysis produces a sugar mixture with improved solubility, sweetening power four times that of the original lactose, digestibility acceptable to lactose-intolerant people, and value as a microbial substrate.

The use of lactase to transform waste lactose into an improved form for food use has been a promising application for almost 20 years. Mold lactase has long been advocated for use in acid whey treatment, whereas yeast lactases would be the choice for sweet whey. As this treatment involves addition of enzyme to a waste product, the economics have not favored the use of free, soluble enzymes (Short, 1978). Immobilized enzyme reactors have the potential to lower the overall cost of processing by attaching enzymes to low-cost materials and reusing the catalyst (Short, 1978; Finoc- chiaro et al., 1980b; Siso and Doval, 1994). The ability to hydrolyze the lactose in acid whey has been demonstrated using immobilized A. niger lactase bound to porous glass (Wierzbicki et al., 1973) or phenol-formaldehyde resin, which retained nearly full activity over 120 days of continued processing at 40°C (Okos et al., 1978). K. lactis lactase immobilized on alumina has been used to hydrolyze lactose in sweet whey for up to 48 hours at 40°C (Finocchiaro et al., 1980b). K. lactis lactase immobilized on Ni( 11)-iminodiacetate-agarose yielded 96% activity and reached 70% lactose conversion in whey permeate in 2 hours (Irazoqui and Batista- Viera, 1994). In a simpler approach, Siso and Doval (1994) immobilized lactase-rich K. lactis cells on corn grits, as an alternate low-cost support, and produced 80% hydrolysis of the lactose in sweet whey for up to 2 days at 37°C.

2. Lactose Syrup Production

Preparation of syrups from acid whey at 15% total solids using A. niger lactase has been reported (Wierzbicki et al., 1974). Commercialization of lactose hydrolysis in acidified whey with immobilized lactase has been described (Moore, 1980). This process yields a glucose-galactose syrup, which is concentrated by evaporation to 60 to 68% solids (Mahoney, 1985). These glucose-galactose mixtures are about four times as sweet as lactose and about 60 to 75% as sweet as sucrose. The syrups can be made microbio- logically stable with up to 75% total solids (Nijpels, 1981). Major potential


uses for these hydrolyzed lactose syrups have been described by Short (1978) and are summarized here.

a, Sweetener and Preservative. Syrups could be used in such foods as canned fruit, soft drinks, ice cream, and frozen yogurt. They could also find a role in the manufacture of toffees, fudge, and candy bars.

b. Enhancer. As a microbial substrate, these syrups could be used in beer, wine, and baked goods.

c. Browning Agent. The syrups could be used in various baked goods and confections.

3. Product Synthesis

Glycosidase enzymes, like lactase and invertase, are able to catalyze not only hydrolytic reactions, but also the reverse reaction for transferase activity (Zarate and Lopez-Leiva, 1990). This type of reverse reaction has been used for synthesis instead of hydrolysis with several enzymes, including chymotrypsin, subtilisin and urease (Butler et al., 1976). Synthesis of sucrose from starch catalyzed by the enzyme invertase from yeast and molds (Butler et aL, 1976) and the enzyme from banana pulp (Glass and Rand, 1982) has been studied. The mechanism of lactase .activity has demonstrated the capability to transfer galactose to any acceptor containing a hydroxyl group (Mahoney, 1985). This transferase activity results in the synthesis of oligosaccharides, which appear in highest concentration after 50 to 90% lactose hydrolysis has occurred. The initial lactose concentration and the product environment also seem to be important, as does the enzyme form: immobilized versus free (Zarate and L6pez-Leiva, 1990). The potential of this reaction as a possible mechanism for conversion of lactose to useful products has received little attention. Considering the large amount of waste lactose available from whey protein recovery, the future application of immobilized lactase may lie in conversion of the waste sugar into a range of products, from vitamins like ascorbic acid to compounds with pharmaceutical potential.

11. HYDROLASE ACTIVITY

Reduction of lactose in food products can be accomplished either by physical separation or by hydrolysis; however, physical methods (such as gel filtration and ultrafiltration eliminate vitamins and minerals in addition


to lactose. Therefore, the logical approach, without complete reformulation of the milk product, has been lactose hydrolysis. Although hydrolysis can be achieved with acid- or cation-exchange resins, enzymatic conversion with P-galactosidase has been preferable because it is highly specific (Mahoney, 1985) and the other components of the substrate solution remain unchanged (Nijpels, 1981). Hydrolase activity would be the major route of catalysis for P-galactosidase. As stated in the introduction, lactases have been used commerically for the hydrolysis of lactose in dairy products. The catalytic properties of P-galactosidase, in both soluble and immobilized forms, have been extensively reviewed (Pomeranz, 1964b; Shukla, 1975; Richmond et al., 1981; Gekas and L6pez-Leiva, 1985; Mahoney, 1985).

P-Galactosidase has been purified from a vast number of microbes including yeast fungi, and bacteria (Shukla, 1975; Richmond et al., 1981; Mahoney, 1985; Gekas and Lbpez-Leiva, 1985; Agrawal et al., 1989). P-Galactosidase from Escherichia coli has specifically been examined and serves as a model for the study of the overall enzyme characteristics; however, E. coli-derived lactase has not been considered safe for commercial use and is used only as a valuable analytical tool for enzymatic determinations. Therefore, the following discussions pertain to those organisms used in the preparation of commercial lactases.

A. ENZYME MECHANISM FOR ACTIVITY

Glycoside hydrolases are named according to the type of bond hydrolyzed (Nilsson, 1991). p-Galactosidase, or lactase (EC 3.2.1.23), hydrolyzes the Pl-4 glycosidic linkage between the galactose and glucose moieties of lactose by transferring galactose to water and simultaneously liberating glucose. Formation of an enzyme-galactosyl complex with simultaneous glucose liberation is the critical step in hydrolysis (Prenosil et al., 1987a). The overall hydrolytic reaction is given in Fig. 12. Although the catalytic

GLUCOSE

ENZYME + GALACTOSE <-I

H20

FIG. 12. General reaction for hydrolysis of lactose by P-galactosidase.


mechanisms of most glycosidases have not been elucidated (Nilsson, 1991), a specific mechanism of catalytic activity for E. coli P-galactosidase was originally proposed by Wallenfels and Malhotra (1961). This proposed mechanism for neutral-pH lactases, illustrated in Fig. 13, has been extensively reviewed and characterized (Shukla, 1975; Nijpels, 1981; Richmond

A 4-C-i-Galactopyranosyl-D-glucopyranose molecule at the active site of J-galactosidase enzyme

Jl-Galactosidase-galactose complex t glucose

FIG. 13. Proposed mechanism for lactose hydrolysis catalyzed by /3-galactosidase. From Shukla (1975). Reprinted with permission of CRC Press, Inc.


et al., 1981; Mahoney, 1985; Prenosil et al., 1987a; Zarate and Lbpez-Leiva, 1990). The hydrolysis reaction, as proposed by this model, corresponds to an SN2-like displacement mechanism. The sulfhydryl group, at the active site of lactase, acts as a general acid to protonate the galactosidic oxygen atom, and an imidazole group acts as a nucleophile and attacks at C-1 of the glycone. This reaction led to the proposed formation of a covalent bond intermediate enzyme-galactosyl complex. Galactose is then removed from the enzyme through the abstraction of a proton from water by the sulfhydryl anion, which acts as a general base, and assists in the attack of OH- at the C-1 position.

Wendorff and Admundson (1971) indicated that K. fragilis P-galactosidase appears to be a sulfhydryl enzyme, as it is inhibited by heavy metals, para-chloromercuribenzoate, and iodoacetate. Mahoney and Whitaker (1977) suggested that K. fragilis P-galactosidase, a neutral-pH enzyme, as is E. coli P-galactosidase, operated according to this same proposed mechanism based on inhibition experiments with sulfhydryl reagents and pH-activity studies. Some studies, however, have indicated the possibility that the sulfhydryl group, although important in maintaining the active site, does not participate in the actual catalytic mechanism (Mahoney, 1985). A second general mechanism postulated for glycosidases, has been implicated in the catalytic mechanisms of E. coli P-galactosidase, hen’s egg white lysozyme, influenza virus neuraminidase, and other glycosidases as illustrated in Fig. 14 (Sinnott, 1987; Nilsson, 1991; Vulfson, 1993). Simply, this proposed catalytic mechanism involves an enzyme carboxylate group, which acts as a general acid that catalyzes cleavage of the glycosyl-oxygen bond. A covalent glycosyl-enzyme intermediate has been proposed that was reached from both directions, through oxocarbonium ion-like transition states. The final catalytic step for hydrolysis involves the deprotonation of water by A- which increases the nucleophilicity of the acceptor, facilitating the formation of a new glycosidic bond forming free galactose. The active sites of fungal P-galactosidases have remained a mystery, and no direct evidence relating to their mechanisms is currently available.

B. ASSAY METHODS AND ENZYME ACTIVITY

Lactases, regardless of source, carry out the same hydrolysis and are inhibited by galactose, competitively, to different degrees; however, enzyme properties and, ultimately, catalytic efficiency and conditions for optimum activity differ. Enzyme activity is influenced not only by source, pH, and temperature but also by substrate. Catalysis is influenced by variation in the aglycone moiety (glucose, in the case of lactose), and various enzyme sources respond differently to substrate structure. Listings of various properties of


GLYCOSYL G R O U P TRANSFER I-

A% H'

I

qo-R Y I

FIG. 14. Second general mechanism postulated for glycosidases. From Sinnott (1987). Re- printed with permission of Royal Society of Chemistry, London.

microbial lactases can be found in Shukla (1975), Greenberg and Mahoney (1981), Wasserman (1984), Gekas and L6pez-Leiva (1985), and Agrawal et al., (1989).

A number of methods have been developed to determine P-galactosidase activity using either lactose or other P-galactosides having an aglycone moiety different from glucose. Each procedure has its own built-in advantages and restrictions. The simplest methods use colorimetric analysis to


measure the quantity of monosaccharides in the presence of disaccharides (Bouvy, 1974; Nickerson et al., 1976) or specific measurement of glucose with glucose oxidase (Dahlqvist, 1964). In addition to colorimetric determinations, glucose concentrations can be measured with biosensors that use glucose oxidasekatalase membranes. These methods illustrated are relatively rapid, specific, and highly sensitive.

Chemically modified substrates have been commonly applied for the determination of lactase activity because they are easy to use. Commercial enzyme manufacturers and researchers have characterized P-galactosidase from different commercial sources (Wendorff and Amundson, 1971; Uwaj- ima etal., 1972; Mahoney and Whitaker, 1977; Dickson et al., 1979; Mahoney and Adamchuk, 1980; Hussein et al., 1989) and have routinely defined lactase units using the artificial substrate ortho-nitrophenyl-P-D-galactoside (ONPG). The relatively simple and straightforward procedure involved has made ONPG the substrate of choice because of the high reaction velocities and lower K , values (Richmond et al,, 1981). ONPG is hydrolyzed by the enzyme at the nonreducing end of the P-galactosidic bond, producing galactose and the yellow o-nitrophenol (Bouvy, 1974). One unit of lactase activity equals the production of 1 pmole of o-nitrophenol per minute under the conditions of the assay.

Wallenfels and Malhotra (1961) first reported that lactases from different sources exhibited optimum activity on different substrates because of variations in enzyme-substrate affinity. Pomeranz (1964a) noted that although hydrolysis rates have been routinely defined by ONPG activity, the affinity of the enzyme for different substrates varied; therefore, ONPG units did not directly reflect the activity lactase probably has on lactose in food. In addition, the analysis of enzyme activity in food substrates was complicated by the existence of both hydrolytic and transferase activities (Mahoney, 1985). This means that estimates of lactase activities on food lactose based on simple extrapolation from an ONPG assay may not be accurate. Buffer- ing and artificial substrates may produce optimum activity, but these conditions do not properly indicate enzyme activity as it is reflected by real-life processing situations. Hourigan (1976) proposed that lactase activity should be studied with a milk system. K , values for ONPG and lactose of lactases from commercially available yeast and fungal sources are listed in Table I. In all cases cited, the K , value for lactose was higher than that for ONPG, thereby indicating lower affinity for the natural lactose substrate. Furthermore, investigators have shown that milk constituents also have an effect on lactose hydrolysis and the degree is source dependent. The E. coli enzyme, probably the most studied of all P-galactosidases, has been shown to have an overwhelming preference for ONPG substrate while losing 95% of its activity in milk (Morisi et al., 1973). Yeast-derived lactases

28 LORI F. PIVARNIK er al.

TABLE I COMPARISON OF K,,, VALUES OF VARIOUS MICROBIAL P-GALACTOSIDASES ON

ONPG AND LACTOSE SUBSTRATES

K m (MI Enzyme source ONPG Lactose References'

A. niger 2.02 85 1 2 A. oryzae 0.77 50 2,3 K. fragilis 2.5, 2.72 24, 13.9 4 s K. lactis 1.61, 1.25 16.8, 16.0, 24.3., 28.0 6-9

' (1) Greenberg and Mahoney, 1981a; (2) Greenberg and Mahoney, 1981b; (3) Park eral., 1979; (4) Mahoney and Whitaker, 1977; (5) Wendorff and Amundson, 1971; (6) Dickson er al., 1979; (7) Rand, 1981; (8) Forsman er al., 1979; (9) Hussein et al., 1989.

were affected to a lesser degree, with K. fragilis and K. luctis p-galactosidases losing 40% and 10 to 30% of their activity in milk when compared with ONPG (Morisi et ul., 1973; Mozaffar et al., 1985). Finally, A. niger enzyme (Rand, 1981) activity has been shown to increase in milk. Therefore, proper choice of operating conditions, suitable to the p-galactosidase used, is important in establishing maximum enzyme activity.

Jacober-Pivarnik and Rand (1984) developed a milk assay to reflect the probable activity of lactase during dairy processing. They compared the lactase units of lactase powders derived from K. luctis and K. frugilis in a pH 6.5 imidazole-buffered lactose solution and reconstituted NFDM, both approximately 5% lactose. The milk system was found superior to the buffered enzyme-lactose system (Fig. 15). The assay method employing buffered lactose substrate did not accurately reflect the true enzyme activity in a milk system. Seneca1 and Rand (1992b) determined the lactase activity of five commercial yeast enzymes in a modified NFDM assay and compared the data with the manufacturer's data on ONPG activity (Table 11). Al- though p-galactosidase from K. lacfis achieved similar enzyme activity, the data revealed appreciable differences in the activity of lactases from K. fragilis and Cundida species on NFDM and ONPG substrates. The data revealed that lactase activity, as reported by the manufacturer for an ONPG assay, did not reflect the activity that occurred in reconstituted NFDM. It appeared that the molecular complexity of milk may have interfered with the chemical attraction of K. fragilis- and Cundida-derived lactases to lactose, resulting in decreased enzyme activity when compared with ONPG. This, however, was not observed with lactases derived from K. lacfis. The reason for the difference in substrate affinities for various species of lactase remains unknown.

COMMERCIAL P-GALACTOSIDASE IN FOOD PROCESSING

lactase units/mg

NFDM

.I” 4 t

+ Buffered lactose

,,-’

0 I -_> -.<’

29

0 10 20 30 40 50 60 mg of enzymelml

FIG. 15. Enzyme activity of Maxilact 40,000 lactase powder, isolated from K. lactis, in skim milk and pH 6.5 imidazole-HCI buffer at 37°C. From Jacober-Pivarnik and Rand (1984). Reprinted with permission of Institute of Food Technologists.

TABLE I1 ACTIVITY UNITS OF COMMERCIAL YEAST LACTASES ON NFDM

AND ONPG

LAU“

Lactase NFDM ONPGb Activity (8) range

K. lactis 5300 5000 106 K. lactis 8400 8000 105 K. fragilis 1900 3000 63 C. pseudotropicalis 1400 2750 51 C. pseudotropicalis 1400 2500 56

Lactase activity units as given are per gram of enzyme per milli- liter.

* As reported by the manufacturer. Presented at the 1992 annual Institute of Food Technologists meet-

ing, New Orleans.

30 LORI F. PIVARNIK eial.

Mahoney (1985) indicated that because both hydrolytic and transferase activities are present, the degree of lactose hydrolysis cannot be established by evaluating a single monosaccharide unless the hydrolysis approaches 100%. It was perceived that the determination of glucose during hydrolysis could possibly overstate the actual amount of lactose remaining, as some of the glucose would be incorporated into oligosaccharides. Analyses using galactose could be inaccurate because of the even greater tendency to form compound sugars. To completely analyze the activity of the enzyme in a reaction mixture, separation of the components of the mixture is required. Chromatographic methods have been used for complete analysis of reaction mixtures; however, only high-performance liquid chromatography provided a simple and rapid method for the simultaneous determination of lactose, glucose, and galactose in reduced-lactose milk (Pirisino, 1983; Pivarnik, 1990).

A number of assay methods have been used to determine rates of hydrolysis of lactose-containing substrates by lactase (Table 111). The majority of these are methods previously cited in reference to enzyme activity that were modified to satisfy the conditions. A rapid method that does not require expensive equipment and can be used routinely to determine the extent of lactose hydrolysis in dairy products is cryoscopic analysis. This method measures changes that occur in solutions as the result of production of monosaccharides during lactose hydrolysis. The method relies on the freezing point of the sample being depressed by an amount depending on the molal degree of hydrolysis of the lactose (Zadow, 1984). Baer et al. (1980) demonstrated the linear relationship between freezing point and lactose hydrolysis in neutralized acid whey and 5% lactose solution. Also, lactose hydrolysis of dairy products can be analyzed directly without depro- teinization, which is required by the other methods of analysis. A disadvantage of the assay is the formation of oligosaccharides during hydrolysis and their impact on freezing point depression (FPD). Oligosaccharide production results in a smaller FPD than would be associated with the simple breakdown of lactose into glucose and galactose (Jeon and Saunders, 1986). Jeon and Saunders compared the percentages of hydrolysis of C. pseudotropicalis for 5% lactose solution in 0.025 Mpotassium phosphate buffer (pH 6.6), whey permeate, and milk, as measured by FPD and high-performance liquid chromatography. The percentage hydrolysis was substantially less for FPD and the degree of variability correlated to the concentration of oligosaccharide formed.

C. PROPERTIES OF COMMERCIAL ENZYMES

A variety of commercial lactase enzymes are available for use in food. These enzymes are derived from yeasts and molds and differ widely in


TABLE 111 METHODS FOR ASSAYING P-GALACTOSIDASE ACTIVITYILA~OSE HYDROLYSIS ON

COMMERCIAL SUBSTRATES

Source of commercial Assay method Substrate enzyme Reference

Colorimetric Milk and whey K. lactis Milk K . fragilis Milk A . niger Milk K. lactis Whey A . oryzae Milk and whey K. fragilis

Milk K. marxianus

Milk and whey K. lactis Milk K. lactis

Chromatographic Milk K . lactis

Milk K. lactis

Whey A . oryzadK. marxianus Milk K . lactis/K. fragilis

Glucose analyzer Whey A . oryzae

Milk K . lactis Milk K. IactidK. fragilis/

C. pseudotropicalis Cryoscopic Milk K. Iactis/A. oryzae

Milk and whey A . oryzadA. niged

Whey K. lactis K . lactis

K. fragilis

Guy and Bingham (1978) Paul and Mathur (1989) Rand and Linklater (1973) Hussein et al. (1989) Ozbaas and Kutsal (1990) Mahoney and Adamchuk

Mahoney and Wilder

Dahlqvist et al. (1977) Forsman et al. (1979) Kosikowski and Wierzbicki

Pirisino (1983) Sheth et af. (1988) Jackson and Jelen (1989) Jacober-Pivarnik and Rand

Pivarnik and Rand (1992) Seneca1 and Rand (1992a)

Palumbo et al. (1995) Shih-Ling et al. (1981)

(1980)

(1988)

(1973)

(1984)

Baer et al. (1980) Nijpels et al. (1981)

their properties, particularly with respect to optimum pH and temperatures for enzyme activity. Shukla (1975) reported that P-galactosidase preparations from different sources are not identical with respect to structure, size, and optimum conditions for hydrolysis. Enzymes derived from different sources have demonstrated dissimilar properties as a result of variations in enzyme-substrate affinity ( Jacober-Pivarnik and Rand, 1984; Agrawal et al., 1989). Temperature and pH optima have been found to vary depending on the source and even the particular commercial preparation (Gekas and Lbpez-Leiva, 1985).

Table IV lists technical data on the commercial P-galactosidases enzymes available during 1993 in the United States. This comparison establishes that the two main sources of P-galactosidase being used commercially are

TABLE IV MANUFACTURERS' 1993 TECHNICAL DATA FOR COMMERCIALLY AVAILABLE YEAST LACTASES

optimum Lactase Source Substrate Optimum pH temperature ("C) Activity units"

Enzeco fungal lactase Enzeco immobilized lactase Yeast lactase L 50,000 Fungal lactase 100,000 Lactozyme 3000L Lactase F Neutral lactase Biolactase Maxilact L2000 Maxilact LX.5000 Neutral lactase

A. oryzae ONPGbflactose 4.5 - 5.0 A. orzyae - 4.0-5.0 K. lactis Lactose 6.0

K. fragilis Lactose 6.5

K. lactk Lactose 6.0

A. oryzae ONPG 4.5-5.0

A. oryzae ONPG 4.5

A. oryzae - 4.5-5.0 K. lacris Whey powder 6.3-6.7 K. lactis Whey powder 6.3-6.7 K. lactis - 6.5

55 100,ooO FCUg

45 50,000 ONPG unitdg 50-55 100,000 FCUg

37 3,000 LAU/mL 55 14,000 FCC/g 45 8,000 ONPG units/g

55-60 30,000 FCUg

35-40 5,000 NLU/g

50 -2,000 ILU/g

35-40 2,000 NLu/g

37 3,100 ONPG units/mL

FCC, Federal Chemicals Codex Lactase Units; ILU, International Lactase Units; NLU, Neutral Lactase Units; LAU, lactase activity units.


K . luctis and A. oryzae. Previously, it had been reported that the sources of the two most widely used commerical lactases were K. luctis and A. niger (Nijpels, 1981). Mahoney (1985) indicated that A. niger lactase was preferred over A. orytae enzyme for commercial applications because it was more stable and had a lower pH optimum, making it more applicable for immobilization. A. oryzue lactase (optimum pH 4.5-5.0), however, appeared to be more suitable for acid whey hydrolysis than A. niger enzyme (optimum pH 3.5-4.0). In addition, A. oryzae lactase was much less susceptible to product inhibition by galactose (Mahoney, 1985). The individual enzyme chosen for commercial utilization depends on a number of elements: substrate, pH, operational temperature, and stability.

1. p H

Operational pH has been the primary characteristic determining the application of a given enzyme (Mahoney, 1985). Not only does pH control enzyme activity, but it also is important in maintaining lactase stability. Fungal P-galactosidases have their pH optima in the range 3.0 to 5.0 and have an operational range of 2.5 to 6.0. This pH profile fits well for processing acid whey, acid whey permeate, or fermented dairy products. The optimum pH for purified A. oryzue lactase in ONPG and buffered lactose (0.1 M sodium acetate buffer) was found to be 5.0 (Park et al., 1979). Ozbaas and Kutsal (1990) determined the optimum pH for purified A. oryzae lactase to be 4.5 using reconstituted whey powder as the substrate; however, fungal lactase preparations from A. niger have also been shown to effectively hydrolyze lactose in milk (Rand and Linklater, 1973). The yeast lactases demonstrate optimum activity in the neutral pH range. In addition, yeast lactases operate in a much more limited pH range (pH 6.0-8.5) than do fungal lactases. Experimental studies on purified K. lactis and K. frugilis lactases with different substrates have indicated the optimum pH range to be 6.5 to 7.0 (Wendorff and Admundson, 1971; Uwajima et ul., 1972; Dahlqvist et al., 1977; Guy and Bingham, 1978; Hussein et al., 1989). One study found that the optimum pH range for purified K. luctis enzyme was 7.0 to 7.5 and that lactase activity dropped off rapidly below pH 6.9 and above 7.5 (Dickson et ul., 1979). This deviation from other studies may have been due to the utilization of sodium phosphate buffer as the test medium. Both K. luctis and K. frugilis lactases have been shown to be slightly inhibited by sodium (Mahoney, 1985).

2. Temperature

Temperature is another important factor to consider when choosing a commercial source of P-galactosidase. Fungus-derived lactases have higher

34 LORI F. PIVARNIK er al.

optimum temperatures for both activity and enzyme stability, and are routinely used for hydrolysis of lactose at 50 to 55°C. Yeast lactases are rapidly denatured above 40°C. High concentrations of enzyme (although expensive) should be used when hydrolysis is conducted with yeast lactases in the optimum temperature range (3O-4O0C), to increase the rate of hydrolysis and, therefore, reduce microbial growth and oligosaccharide production, which are favored by these conditions (Mahoney, 1985). As a result, hydrolysis of milk products has been conducted at refrigeration temperatures (Guy and Bingham, 1978); however, the activity of the enzyme is severely impaired at these reduced temperatures, and a much longer period, generally overnight, is required to achieve lactose hydrolysis in excess of 70%.

Both the activity and the stability of 0-galactosidase have been determined at various temperatures. Hourigan (1976) reported that K. lactis lactase in batch processes attained an 80 to 90% conversion rate in 5 hours at 35°C and in 18 hours at 6°C. Forsman et al. (1979) observed that incubation time and enzyme concentration have an effect on the temperature requirements for optimum lactose conversion with K. lactis lactase. They showed that the optimum temperature for hydrolysis increases when the enzyme concentration is decreased. In addition, the authors suggested the possibility of a saturation effect that becomes more pronounced with an increase in both temperature and incubation time. Guy and Bingham (1978) measured the effect of temperature on K. lactis lactase in cottage cheese whey substrate to determine the range in which the enzyme becomes inactivated. Activity was noticeably reduced when the enzyme was heated above 50°C for 1 minute, with complete inactivation at 70°C. Rand and Linklater (1973) showed no increase in the rate of hydrolysis catalyzed by K. fragilis lactase above 37°C and the enzyme was unstable. Lactose hydrolysis by A. niger lactase was about 15% lower than that by yeast lactase at 37°C for the same reaction time; however, increasing the temperature above 37°C had a pronounced effect on fungal lactase, resulting in an increase in the hydrolysis rate up to 63°C. The fungal enzyme was found to be stable up to 3 hours, indicating its potential utilization for batch-type pasteurization of milk. Studies on purified A. oryzae lactase in various substrates have found the optimum temperature range for maximum hydrolysis to be 50 to 55°C (Park etal., 1979; Ozbaas and Kutsal, 1990). When the crude enzyme was tested in similar substrates to determine the effect of temperature on lactase activity (Park et al., 1979), a higher optimum temperature range (55-60°C) was achieved (Fig. 16). This indicates that the crude enzyme is more heat resistant than the purified enzyme and may be a worthwhile subject of further research. This topic is discussed in later sections.

COMMERCIAL P-GALACI'OSIDASE IN FOOD PROCESSING

120

100

80

60

40

20

0

35

Purified enzyme - + Crude enzyme

-

-

-

-

I I I I I I 1 1 I

3. Stability

The stability of P-galactosidase should be considered when choosing an enzyme for either batch or continuous processing. The thermal stability of an enzyme is especially important, as it can be used to establish optimum processing temperatures. Commercial preparations of liquid yeast lactase have been shown to be stable under refrigeration for 2 years (Dahlqvist et al., 1977; Senecal, 1991). The yeast lactases, however, have proven to be extremely heat labile, particularly when used at temperatures needed to accelerate processing and inhibit microbial growth. Thus, considerable time and effort have been spent to identify a lactase source with greater thermostability at neutral pH, for use in processing milk products. Fungal lactases, although they display relatively good thermostability, have reduced activity at the pH of milk and, therefore, require increased amounts of expensive enzyme. Furthermore, larger amounts of enzyme could be a source of off- flavors in lactose-hydrolyzed milk because of the increased amount of insoluble residue present in food. Finnie (1980) studied the thermostability of A. niger lactase in the continuous pasteurization of milk. The fungal lactase did not display thermostability above 80°C, which would be necessary for UHT sterilization or pasteurization. For batch processing of milk at 63"C, the enzyme maintained 50% of its activity; however, for continuous


HTST pasteurization at 72"C, about 30% recovery of lactase activity was possible.

Very little data are available on the stability of soluble P-galactosidases in dairy products (Mahoney, 1985). Few studies have established the stabilizing effect of various milk components on lactase, other than those in simple buffer systems. Milk and whey have been found to increase the stability of K. lactis (Dahlqvist et al., 1977) and K. lactis (marxianus) lactases at temperatures higher than those of simple buffer systems (Mahoney and Wilder, 1988). The increased stability is attributed to the stabilizing properties of the casein and whey proteins. It has been theorized that both proteins and polyols extend the stability of enzymes through the disruption of solvent-enzyme interactions, thus encouraging intramolecular hydrophobic bonding (Schmid, 1979; Klibanov, 1983). In the presence of lactose, glycerol was found to be less effective than caseinate, implying that the complexity of the substrate-dependent stabilization by protein was greater than the dehydration effect promoted by polyols (Mahoney and Wilder, 1988). Byeong-Seon and Mahoney (1994), in studies on the stabilization of S. thermophilus lactase by bovine serum albumin (BSA), proposed the formation of a dimer-BSA complex. They proposed that the native S. thermophilus enzyme exists mainly as a tetramer (Nl) with some active dimer (D4) present. Heating at 60°C results in the formation of a new, kinetically less stable dimer (D3) from the D4 dimer and, ultimately, inactive aggregates. Addition of BSA with heating results in the production of a BSA-dimer complex more stable than the original dimer or tetramer. Formation of the BSA-dimer complex appears to improve enzyme dimer stabilization against unfolding and aggregation.

Mahoney and Wilder (1988) established that the stability of P-galactosidase in milk is also influenced by salts and sugars. Concentrations of milk solids up to 25% resulted in an increase in the factors important for lactase stability. Thompkinson et al. (1991) observed that increasing the concentration of milk solids to 35 to 40% allowed the optimum temperature for K. lactis lactase incubation to be increased from 37" to 50°C. The overall stability of lactase was noticeably reduced by the removal of divalent cations. Wendorff and Admundson (1971) determined that Mn2+ is a necessary cofactor and may increase stability by helping to maintain the conformation of the secondary and tertiary structure of the K. fragilis enzyme molecule. Enzyme stabilization was found to occur only with sugars capable of binding to lactase (substrate-products) (Mahoney and Wilder, 1988). It was determined that the stability afforded by sugars was not simply caused by com- plexing, but that other, as yet undetermined, phenomena must be involved. In addition, the reaction products appeared to be as effective as the substrate in enzyme stabilization. Mahoney and Wilder (1988) found that the


individual addition of lactose or caseinates resulted in a fivefold increase in enzyme stabilization; however, the synergistic effect of combined lactose and caseinate addition resulted in a SO-fold increase in the stability of K. fragilis lactase. When sugars and proteins were present together, the binding of lactose or galactose may result in an enzyme conformation that is more effectively stabilized by the proteins. Amino acids have also demonstrated stabilizing effects on K. lactis (marxianus) lactase at 45°C (Surve and Maho- ney, 1994). All amino acids (except proline) were found to improve enzyme stability to some degree. Histidine was found to be the most effective amino acid, capable of stabilizing the lactase 9-fold and 40-fold in the presence of 5% lactose. It was determined that the a-amino group and N-1 ring nitrogen of histidine are required for enzyme stabilization.

Senecal (1991) investigated the stability of five commercial yeast lactases in skim milk and skim milk ultrafiltration permeate (SMUFP) at 20", 30", and 40°C. In general, half-life data indicated that the lactase enzymes were more stable in SMUFP at 30" and 40°C and in skim milk at 20°C (Table V). Lactases derived from K. lucris had the greatest overall thermostability of all the commercial yeast-derived products tested. In addition, the K. lucfis lactases tested each had much higher units of activity per gram. Studies have been conducted to determine the effect of enzyme concentration on stability (Mahoney and Wilder, 1988; Senecal, 1991). Results demonstrate that enzyme concentration, even 30-fold, has no significant effect on lactase stability (Tables VI, VII).

Milk has been shown to have a greater thermostable effect on lactase than either salts or whey (Mahoney and Wilder, 1988). As in previous studies, increased stabilization was attributed to the combined presence of caseins and lactose. Dahlqvist el ul. (1977) found K. lucfis lactase to be more stable in UHT-processed milk than in UHT-processed whey at 24°C.

TABLE V EFFECT OF MEDIA AND TEMPERATURE ON LACTASE HALF-LIFE IN SKIM MILK'

Skim milk SMUFP

Lactaseb 20°C 30°C 40°C 20°C 30°C 40°C

K. lactis 118 29 2.5 55 12 5 K. lactis 105 13 4.5 62 17 14 K. fragilis 84 6.5 <1 82 14 <1 C. pseudotropicalis 81 5 <1 67 11 <1 C. pseudotropicalis 108 7.5 <1 43 10 <1

Half-life is expressed in hours. Enzyme concentration from all microbial sources is 11.9 mglml.

38 LORI F. PIVARNIK e fa l .

TABLE VI EFFECT OF ENZYME CONCENTRATION ON

KLUYVEROMYCES LACTIS LACTASE STABILITY

IN MILK AT 45°C"

Enzyme concentration (ONPG units/ml)

0.66 1 .o 2.0 4.0

10 20 30 Average C.V. ( W ) b

Half-life (min)

Trial 1 Trial 2

85 - 86 65 85 63 82 74 90 66 94 74

66 87 T 4.3 68 t 4.8 4.9 7.0

-

" From Mahoney, R. R., and Wilder, T. Thermo- stability of yeast lactase (Kluyveromyces marxianus) in milk. J . Dairy Res. 55,423-433. Copyright (1988) by Cambridge University Press. Reprinted with the permission of Cambridge University Press.

Coefficient of variation.

The compositions of the whey-derived substrates in these studies were different from those in the skim milWSMUFP study (Senecal, 1991). SMUFP was processed through a 50,000-MW-cutoff hollow-fiber mem-

TABLE VII EFFECT OF ENZYME CONCENTRATION ON STABILITY OF LACTASE IN

SKIM MILK ULTRAFILTRATION PERMEATE

Half-life (h) Concentration

Lactase (mg/ml) 20°C 30°C

K. fragilis 1.9 82 14 32.1" 79 15

C. pseudotropicalis 11.9 67 11 44.0" 53 12

C. pseudotropicalis 11.9 43 10 44.0' 37 11

" Concentrations based on theoretical values for amounts to be equivalent to K. lactis lactase.


brane, similar to those commonly used in the dairy industry. The final product contained concentrations of proteins and salts greater than those used in the previous studies. The increased concentrations of these constituents may have been contributing factors to lactase stability at higher temperatures. In addition, soluble a-lactalbumin and nonprotein nitroge- nous compounds may have had a greater thermostable effect than casein, because of their greater ability to react with the enzyme-substrate complex. Mahoney and Wilder (1988) found K. lacfis (marxianus) lactase to have greater stability in synthetic milk salts with added caseinate than in milk. These investigators postulated that the increase in stability may have been due to solubility of protein as a stabilizer. In milk, casein would be primarily in the form of micelles so the concentration of soluble protein was reduced.

In addition to substrate composition, pH has played an important role in the overall stability of P-galactosidase. Seneca1 (1991) observed that changes in substrate pH, over time, for skim milk and SMUFP were different at 20" and 30°C. SMUFP was much more susceptible to changes in pH due to the onset of spoilage. These pH changes exceeded the optimum range for lactase and may have contributed to the reduction in stability of the enzymes in SMUFP at 20°C. At 30°C the enzymes might have attained greater half-life values if spoilage could have been delayed. Enzymatic activity was found to decrease rapidly for K. lacris lactase below pH 5.9 (Guy and Bingham, 1978) and for K. frasiris lactase below pH 5.5 (Wendorff and Amund- son, 1971; Mahoney and Whitaker, 1977). Dahlqvist ef al. (1977) observed that stability of K. lactis lactase declined rapidly if bacterial growth occurred in milk.

Palumbo ef al. (1992) evaluated the storage stability of commercially available K. lacfis and A. oryzae lactases dry blended into milk powders produced with 2% vegetable oils. The objectives were to determine lactase stability and to establish processing parameters to formulate a reduced- lactose shelf-stable milk powder for military applications. A 6-month storage trial of milk powders was conducted at different temperatures. K. lacfis lactase proved to be relatively stable when stored under nitrogen at 20°C or below; however, when stored at 45°C in dry milk, the enzyme lost significant activity (25% within 1 month). P-Galactosidase derived from A. oryzae displayed much greater stability, retaining more than 90% of its activity even when stored at 45°C furthermore, it demonstrated the potential for incorporation into milk powders. It appears that K. lactis lactase can only be added to milk powders that would be stored in cool places.

D. PURITY OF COMMERCIAL ENZYMES

Previous sections have alluded to the dissimilarities in the properties of lactase enzymes derived from different sources. An important property


that has received very little attention in the literature is the purity of the commercial preparation, especially with respect to the presence of other enzymes (Le., proteases). Although proteases have not been a concern with oral lactase preparations, they could have a severe impact on the stability of lactase in continuous reactor systems. In addition, unless inactivated prior to packaging, protease could lead to undesirable changes in dairy products during storage.

The lactases used by researchers are either individually purified from cell-free extracts or obtained from commercial sources. Commercial lactases are obtained by the separation and purification of whole-cell extracts. Meth- ods for the isolation and purification of 0-galactosidase are summarized elsewhere (Shukla, 1975; Richmond et al., 1981; Agrawal et ul., 1989). Early research was conducted on highly purified enzymes produced in the laboratory; however, references in the literature have implicated commercial sources of p-galactosidases as being far from pure and containing enzymes such as protease (Mahoney and Wilder, 1988,1989). Protease activity present in commercial lactases may have influenced the results of thermostability studies (Mahoney and Wilder, 1989). Dickson et al. (1979) found that even under optimal storage conditions, K. luctis enzyme stability varies from preparation to preparation. It was assumed that K. luctis lactase stability is probably a function of protease contamination. This hypothesis was arrived at because of the loss of lactase activity with the decrease in molecular weight of the monomer and increase in lower-molecular-weight proteins. Results indicate that even a little protease activity can inactivate 25 to 50% of the K. lactis enzyme after 4 months at 4°C in glycerol phosphate buffer.

Protease present in lactase preparations may also cause undesirable flavor changes as a result of milk deterioration during storage (Mahoney, 1985). This is especially important if lactase is added to UHT-processed milk after processing. Dahlqvist et al. (1977) devised a method for adding p- galactosidase to UHT-processed milk. The lactase, which is readily inactivated by UHT processing, is added in much smaller amounts than normally required for batch processing, at the end of the UHT process. The commercial enzymes require further sterilization through a 0.22-pm filter to prevent bacterial contamination and a decrease in the stability of the UHT- processed milk.

Protease activity in commercial lactase preparations has also been implicated in the operational changes of a continuous membrane bioreactor (Senecal, 1991). Increasing the concentration of K. luctis lactase from 2000 to 3000 mg per 0.06 m2 of hollow-fiber membrane resulted in an unexpected increase in flux. Protein analysis of the permeate established that 18.7% more nitrogen-containing material was penetrating the UF membrane. The


nitrogen-containing material was believed to originate from the protective protein bed added to increase enzyme stability by separating the enzyme from the polysulfone membrane material. Increasing the lactase concentration may have resulted in an increase in protease enzymes, which had a reducing effect on the concentration polarization layer over time.

Despite the evidence that commercial lactase preparations contain protease, only one supplier has actually listed protease activity in their product specifications for K. fragilis lactase. Others have indicated that their preparations contain very small amounts of protease. Therefore, depending on the intended application, commercial preparations should be assayed for protease activity by known analytical methods.

E. ACTIVATION AND INHIBITION

Table VIII lists the ionic activators of K. luctis and K. frugilis lactases. In general, heavy metals are inhibitory, whereas potassium, sodium, magnesium, and manganese stimulate the enzyme. It should be noted that some activators are more effective than others and some become inhibitory at high concentrations or in combination with other ions. Ammonium ion alone activates K. frugilis lactase, but when combined with adequate amounts of potassium, it inhibits the enzyme (Wendorff and Admundson, 1971). Therefore, the original literature should be consulted. In addition, the type of buffer used can also influence the effects of ions. For example, Mahoney and Whitaker (1977) found that although zinc stimulates K. frugilis 6-galactosidase in phosphate buffer, it completely inhibits it in Tris buffer. The magnitude of the effect of these ions also differs depending on whether the enzyme is used in whey, milk, or buffered lactose. Finally, Dahlqvist et ul. (1977) and Mahoney and Adamchuk (1980) found that milk proteins help activate and stabilize K. luctis and K. fragilk lactases, respectively. A. niger and A. oryzae P-galactosidases do not appear to be dependent on activating ions (Greenberg and Mahoney, 1981). As previously stated in the Introduction, product inhibition has been found to be a limiting disadvantage in obtaining the maximum effect of the p- galactosidase enzyme in commercial applications. Galactose has been cred- ited as a competitive inhibitor. Rand and Linklater (1973) observed that galactose reduces the hydrolysis of A. niger lactase in milk by about 50% compared with glucose. Forsman et al. (1979) found that galactose inhibition occurs after the first 10 minutes of incubation and reaches a saturation level at 60 mM galactose. In their UHT system the saturation level was reached in about 2 hours, indicating the importance of product inhibition in reactions that are extended. p-Galactosidases from different organisms have displayed dissimilar effects of product inhibition. Wendorff and

42 LORI F. PIVARNlK eral.

TABLE VIII

INHIBITORS OF P-GALACTOSIDASE FROM

EFFECTS OF IONS AS ACTIVATORS OR

MICROBIAL SOURCES

Enzyme source

Ion K. lactis K. fragilis

-0

+ - Ag Ca2+ c1- ne cu2+ - co2+ ne + Fe2+ ne Hg2+ - K+ + + Na+ + Mg2+ + + Mn2+ + + Ni2+ P042- + Zn2+ - Referencesb 1,2 3 P

-

-

+

+, Activating effect; -. inhibiting effect; ne, no effect.

(1) Dickson er al., 1979; (2) Guy and Bingham, 1978; (3) Mahoney and Whitaker, 1977; (4) Wend- orff and Amundson, 1971.

Amundson (1971) discerned that K. frugilis lactase is inhibited by high concentrations of both glucose and galactose. Furthermore, glucose was found to be more inhibitory than galactose in their experiments. It was concluded that glucose is a noncompetitive inhibitor, and galactose inhibition is competitive.

Recently, it has been reported that galactose had a noncompetitive inhibitory effect on fungal P-galactosidase from A. oryzue (Shukla and Chaplin, 1993). This study, conducted with a wide range of substrate and inhibitor concentrations, has been found to be contradictory to earlier investigations. The position taken by the authors was that the results of previous studies may not have been correctly interpreted and that a closer inspection of l/v versus inhibition concentration demonstrated noncompetitive rather than competitive inhibition. They identified the importance of these findings (noncompetitive versus competitive inhibition) on the design of enzyme reactors. Reduction of competitive enzyme inhibition in a reactor can be


achieved through the addition of substrate. However, noncompetitive enzyme inhibition can only be reduced by raising the enzyme concentration or through the removal of inhibitor. Further studies under diversified reaction conditions and lactase sources may be required for complete understanding of the different mechanisms involved in lactase inhibition.

F. IMMOBILIZATION MECHANISMS AND REACTOR SYSTEMS

Many excellent articles have reviewed the immobilization methods available and the sources of lactase that have been investigated using a variety of techniques and support carriers (Shukla, 1975; Finocchiaro et al., 1980b; Greenberg and Mahoney, 1981; Richmond ef al., 1981; Gekas and L6pez- Leiva, 1985; Thompkinson, 1989; Bodalo et al., 1991). In addition, Finocchi- aro et al. (1980b) and Gekas and L6pez-Leiva (1985) assembled extensive tables on lactase immobilization systems to include source, carrier, and reactor type. The immobilized reactor systems most frequently cited were batch, fluidized bed, and packed bed. Table IX lists methods that have been studied for P-galactosidase immobilization that use commercially available lactase sources. Methods that have been developed recently or have had limited review are discussed further.

Algin immobilization of lactases has received limited attention. Food- grade algin is considered a simple method for entrapment (Kierston, 1981), is inexpensive, and is available in both sodium and potassium forms (Na and K are activator ions for microbial sources of P-galactosidase). Algin encapsulation is accomplished by crosslinking with calcium chloride. Jacober-Pivarnik and Rand (1984) reported successful encapsulation of K. lactis lactase in both sodium and potassium forms and system performance for lactose hydrolysis. Lactase entrapped in potassium alginate resulted in twice as much lactose hydrolysis as lactase entrapped in sodium alginate. The alginate beads were easily recovered and washed to avoid microbial growth.

Immobilization of whole cells has been studied because it is less expensive and eliminates the need for enzyme purification (Champluvier et al., 1988a,b). Champluvier et al. (1988a) studied the preparation and properties of P-galactosidase confined in cells from Kluyverornyces sp. to develop a treatment that could be combined with the various methods available for whole-cell immobilization. Treatment with glutaraldehyde was conducted to stabilize the confinement of the lactase to the cell. The combination of chloroform-ethanol treatment and immobilization with 0.4% glutaraldehyde resulted in a biocatalyst that could be stored for 14 months under

TABLE IX IMMOBILIZED LACTASE SYSTEMS

Methodlimmobilization Source Reference" Carrier agent Substrate Remarks

K. la&

K. lacris/frgih and K. marxianus

K. lactidfragilis

Aspergillus sp.

B. circulans A. oryzae

K. fragilis

K. lacfis

1 K/Na alginate

2 Whole cell

3 Whole cell

4 Microbial pellets

5 PVC silica gel ribbed membrane

6 corn grits

7 Nylon-6 microbeads

Ca-alginate beads

ChloroformlEtOH permeabilization confinement by glutaraldehyde

Polycations used to promote adhesion to solid supports

Immobilization of mycelium-associated fl-galactosidase

Covalently attached using PEI and glutaraldehyde

Covalently attached using glutaraldehyde

PEI and gluteraldehyde

NFDM

Buffered lactose

Buffered lactose

Buffered lactose

Skim milk

Whey

Skim milk

Potassium showed activating effect in immobilized system

Biocatalyst used and recovered 7 times without loss of activity

50% initial activity after 25 days of continuous operation

50% of mycelium-associated lactase activity after 1300 h

Determined kinetic parameters of rates of expressions

50% hydrolysis with packed-bed reactor, reaction rate not limited by diffusion

reactor without plugging 75% hydrolysis in a spin-basket

(1) Jacober-Pivarnik and Rand, 1984; (2) Champluvier et aL, 1988a; (3) Champluvier ef aL, 1988b; (4) RBuey ef a[., 1992; (5) Bakken ef aL, 1990, 1991, 1992; (6) Siso ef al., 1994; (7) Ortega-Lopez ef aL, 1993.


refrigeration with only 20% deactivation and could be used and recovered seven times without apparent activity loss.

Champluvier et al. (1988b) also studied the immobilization of whole yeast cells by adhesion of cells to a support. Cells were first treated with chloroform-ethanol and glutaraldehyde and immobilized to glass, polycarbonate, or polystyrene supports using chitosan to promote cell adhesion. After 25 days of continuous operation at 30°C, a microreactor of the immobilized lactase retained 50% of the initial activity. RCczey et al. (1992) investigated the in situ whole-cell immobilization of mycelium-associated P-galactosidase through mycelial pellet formation in 15 Aspergiffzu and Peniciflium strains. The immobilized cells were produced through the growth of the fungal strains into pellets in shaker flasks, in an air-lift fermenter. The pellets were tested in repeated batch hydrolysis of lactose solutions (50 g/liter). Pellets produced from the fungal strain A . pheonics QM329 has the least amount of enzyme leakage during hydrolysis. It was estimated that 50% of the mycelium-associated lactase activity remained after 1300 hours of hydrolysis.

Batsalova et al. (1987) immobilized a fungal P-galactosidase in polyvinyl alcohol and studied the properties of the immobilized preparation in 4% lactose solution and whey substrate. The immobilized enzyme exhibited properties similar to those of the free enzyme and could be used repeatedly without any considerable loss of activity. The immobilized system initially suffered from enzyme leakage, which may have been due to gel pore size configuration. Treatment with 0.2% glutaraldehyde for 60 minutes or with 0.4% glutaraldehyde for 30 minutes was found to greatly reduce leakage without any adverse effect on hydrolysis. Combination of casein with the glutaraldehyde treatment resulted in greater retention of immobilized enzyme and an improved reusability. K. fragifis lactase has been covalently linked by glutaraldehyde to chemically modified corn grits for utilization in packed-bed bioreactors (Siso et al., 1994). Corn grits were found to be very stable and to possess good mechanical properties as well as being an inexpensive enzyme support. Deproteinized whey from milk was hydrolyzed up to 50% within 3 hours of operation with negligible diffusional limitations in the packed-bed bioreactor. In addition, the immobilized system was recycled up to five times (4 hours each time) before significant reductions in enzyme activity occurred.

A novel chemical reactor for the hydrolysis of lactose has been tested with the immobilization of A. oryzae and B. circufans lactases onto a ribbed membrane of polyvinyl chloride (PVC) and silica (Bakken et af., 1990,1991, 1992). The fungal lactase is immobilized onto the PVC-silica membrane by adding polyethyleneimine (PEI) and glutaraldehyde. The immobilized membrane is modeled mathematically for lactose hydrolysis of skim milk


in an axial-annular flow reactor (Fig. 17). The reactor is found to avoid some of the problems associated with other reactor configurations, such as plugging, microbial contamination, and pressure drop (Bakken er af., 1990). Ortega-Lopez er af. (1993) developed a novel spin-basket reactor using K. lacris lactase immobilized on nylon-6 microbeads. The reactor was capable of hydrolyzing 75% of the lactose in skim milk (28.6% total solids) within 7 minutes at 34°C and did not experience any of the plugging associated with packed-column reactors.

Despite all the studies on the development of immobilized lactose hydrolysis systems, very little information has been gathered on their commercial utilization. Why? The manufacturers may not disclose proprietary information, the systems may not be cost effective for wide-scale dairy applications, or the installation and operation of these systems may not be cost effective. Developers of commercial immobilized enzyme reactors for the continuous hydrolysis of milk have had to face many difficulties, such as the pH of milk, which encourages microbial growth; the presence of milk proteins, which can foul reactors; and neutral-pH yeast lactases, which are not very stable (Mahoney, 1985). Immobilized lactase reactors developed for whey hydrolysis have found more success because of the greater stability of the

FEED RESERVOIR

WATER BATH

TEMPERATURE CONlROL

FIG. 17. Process schematic for the axial-annular flow reactor. From Bakken, A. P., Hill, C., Jr., and Amundson, C. H. Use of novel, immobilized @-galactosidase reactor to hydrolyze the lactose, constituent of skim milk. Eiotechnol. Eioeng. 36,293-309. Copyright 6, 1990 by John Wdey & Sons, Inc. Reprinted by permission of John Wiley & Sons, Inc.


fungal lactases and their activity at low pH, which inhibits microbial growth. Table X lists the lactase reactor systems that have been used in commercial applications. Until recently, Snamprogetti was the only immobilized lactase system known to hydrolyze lactose in milk on an industrial scale (Gekas and L6pez-Leiva, 1985). The system has been operated in a recycle reactor with K. l a d s lactase immobilized by entrapment in cellulose acetate fibers (Finocchiaro et al., 1980a). The milk is first sterilized by UHT processing prior to lactose hydrolysis to reduce microbial contamination. At present, the Enzyme Development Corporation (New York) has made commercially available an immobilized fungal lactase for the hydrolysis of milk and whey. Lactase derived from A. oryzae is covalently bound to macroporous amphometric ion-exchange resin, with a multifunctional protein crosslinking agent. This system was initially developed for pilot plant application by Sumitomo of Japan. The immobilized A. oryzae system has been used in a column reactor for the continuous processing of milk and whey substrates (Fig. 18). The half-life data for the continuous column reactor were obtained at different concentrations of lactose for skim milk, whey, and lactose substrates (Table XI). The data demonstrate that the half-life for the column reactor increases with higher lactose concentrations, except when skim milk is used as the substrate. This improvement in reactor half-life may be due to the increase in the concentration of calcium in the skim milk substrate. Calcium ion concentrations have been found to have an inhibiting effect on remaining A. oryzae lactase activity (Enzyme Development Corp.).

Hydrolysis of lactose in whey has found greater commercial success, as demonstrated in Table X. The most successful systems to date have used immobilized lactase from A. niger. One of these is the system developed by the Corning Glass Company (Corning, NY) that immobilizes lactase on porous glass beads. The beads are packed into a cylindrical column for the continuous processing of whey (Fig. 19). The column is operated at 50°C with a residence time of 15 to 25 minutes, and the final product has an estimated 80% hydrolysis rate (Moore, 1980; Mahoney, 1985). The second commercial process using P-galactosidase from A. niger was developed by Valio Dairy in Finland. The enzyme is immobilized by absorption onto a phenol-formaldehyde resin and fixed in place by glutaraldehyde crosslinking (Olson and Stanley, 1973). For continuous processing, the immobilized lactase resin is packed into a fixed-bed columnar reactor and operated in a plug-flow mode (Mahoney, 1985).

G. ULTRAFILTRATION BIOPROCESS REACTORS

The third method of lactose hydrolysis is an alternative method of enzyme immobilization using ultrafiltration. As previously discussed in the Intro-

P W

TABLE X COMMERCIALLY AVAILABLE TECHNOLOGIES FOR IMMOBILIZING LACTASE

System Application

Snamprogetti

Corning Glass

Connecticuthhigh Universities

Valio Laboratory

Gist-Brocades Rohm GmbH

Sumitomo

Amerace Corporation

Fiber (cellulose triacetate)-entrapped yeast for lactose reduction in milk, batch process

A. niger lactase bound to silica beads to hydrolyze acid whey; UF permeate; fixed bed

A. niger lactase adsorbed to porous alumina carrier; fluidized bed; processing of whey UF permeate

A. niger lactase adsorbed to phenol-formaldehyde resin; Duolite ES-762; processing of whey UF permeate

Immobilization of Maxilact enzyme; processing of milk A. oryzue lactase covalently bound to Plesxsym LA-1;

processing of acid whey and milk A. oryzne lactase covalently bound to a macroporous

amphoteric ion-exchange resin, processing of whey and milk A. oryzae lactase covalently bound to a microporous PVC-silica

sheet; processing of whey

Industrial, Centrale Lattern di Milano (Italy)

Semi-industrial; ULN Con& (France), Dairy

Pilot plant (USA)

Industrial Keymenlaakse Dairy (Finland)

Pilot plant (Holland) Pilot plant (Germany)

Pilot plant (Japan)

Pilot plant (USA)

Crest (UK), Kroger (USA)

Reprinted from Gekas, V. and Upez-Leiva, M. H. (1985). Hydrolysis of lactose: A literature review. Process Biochem. 20,(2) 2-12, Copyrights 1985, with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington 0x5, lGB, UK.

HEAT EXCHANGER I -A DISINFECTANT ALKALI

TANK TANK

COLUMN

FEED PUMP

FIG. 18. Flow chart of column reactor using immobilized A. oryzae enzyme system. Manufactured by Sumitomo Chemical. For future information, please contact Enzyme Development Corporation.

50 LORI F. PIVARNIK et al.

TABLE XI EVALUATION OF HALF-LIFE IN COLUMN REACTION'

Substrate Concentrations Temperature Half-life

Whey (pH 4.5) 7% 50 45 40 30

12% 50 45 40 30

4.5

4.5

Skim milk (pH 6.65) 10%

18%

Lactose (pH 4.5)

50 45 40 30

50 45 40 30

4.5

4.5

5% 50 40

7% 50 40

12% 50 40

250 (260)-680 h 500 (520)-1780 h 4880 h 1000 2 200 days 3000 2 1000 days 410 (435)-750 h 820(870)-1950 h 6090 h 1600 2 200 days 2400 2 loo0 days

870 (190)-1820 h 2920 h 9000h 760 2 200 days 2300 ? 500 (days) 1180 h 1790 h 8300 h 1300 2 300 (days) 2000 2 500 (days)

3330-3750 h

3440-3880 h 1500 2 100 (days)

1600 2 100 (days)

1750 2 150 (days) 3760-4240 h

Lowering reaction temperature increases the half-life in column operation. Manufactured by Sumitomo. Chemical. For further information, please contact Enzyme Development Corporation.

duction, there are two ways of using UF membranes as membrane bioreactors: as a CSTR that continuously separates the biocatalyst from the reaction mixture, or as a hollow-fiber bioreactor that immobilizes the biocatalyst on one side of the membrane.

The second approach used for the development of UF bioreactors (hollow-fiber enzymatic reactors) allows the lactase to be immobilized under mild conditions without altering its kinetic behavior as might be experienced through chemical coupling onto solid supports (Jones et al., 1988). Breslau and Kilcullen (1975) discussed alternative approaches to diffusion,

WHEY/MILK u TEMPERATURE

WATER --------.-o CONTROL

DILUTE T i ACETIC

HYDROCHLORIC pH

PROTEIN ULTRAFILTRATION T.----() ACID CONTROL - I

PERMEATE

DEMINERALIZATION

J. ASH

.1 V

FEED TANK PH

ADJUSTMENT

HEAT I I M

EXCHANGER

. . . . IMMOBI

ENZY

ACID

SANITATION

I I COLUMN-/

LACTOSE SWEETENER

FIG. 19. Coming process for hydrolysis of lactose in ultrafiltered milk or whey. From Moore (1980). Reprinted with permission from Cahners Publishing Company.

52 LORI F. PIVARNIK et af.

the primary method that had been used for contacting substrate with enzyme in hollow-fiber reactors. These procedures took into account the unique properties of hollow fibers. Three modes of operation were proposed with the enzyme immobilized in the shell side of the hollow-fiber membrane reactor: UF, backflushing, and recycling. Huffman-Reichenbach and Harper (1982) studied the retention of A. oryzue lactase immobilized on the shell side of hollow-fiber membranes operated in the UF, recycle, backflush, and static modes; leakage averaged 5,30,40, and 7%, respectively. Loss in lactase activity was 50% after 2 hours in the UF mode and less than 10% for the other three modes of operation. Jones et ul. (1988) conducted experimental and theoretical studies on a backflush hollow-fiber enzymatic reactor (HFER), using buffered lactose (Fig. 20). Four lactases derived from A. niger, K. luctis, K. frugilis, and E. coli were tested for stability and specific activity. The bacterial lactase was found to have the lowest specific activity and retained less than 10% of its initial activity after 3 days of storage at 37°C. The yeast-derived lactases demonstrated substantial losses in stability and had lower specific activity than the A. niger lactase. Only the fungal lactase retained 100% of its original activity under the conditions studied.

Maculan (1979) studied the application of a localized UF membrane lactase bioreactor modeled after mammalian intestinal mucosa. Lactases derived from K. luctis, A. niger, and a 50: 50 combination of both were immobilized on the lumen side of a polysulfone UF membrane and tested for optimum conditions of hydrolysis and stability with 5% buffered lactose as substrate. K. luctis lactase was rather quickly denatured in the polysulfone hollow-fiber membrane. Addition of bovine serum albumin stabilized enzyme activity by shielding the lactase from the membrane material (Table XII). Studies involving hollow-fiber enzyme reactors reported in the literature have used buffered lactose solutions as substrate. These systems were important in establishing operating conditions, but have little correlation to actual industrial application.

Seneca1 and Rand (1993) tested a modified version of the UF lactase bioreactor for the hydrolysis of lactose in skim milk permeate (Fig. 21). Selection of hollow-fiber membrane porosity allows the use of the inherent larger-molecular-weight proteins, present in the skim milk permeate, to form a protective barrier against polysulfone membrane denaturation of the immobilized K. luctis lactase. Formation of the protein bed on the lumen side of a HlP30 hollow-fiber membrane is followed by loading with lactase derived from K. luctis to form a dynamic secondary membrane. The system is operated at a constant pressure, preserved by a pressure reservoir, and controlled with a pressure switch, to ensure laminar-flow conditions and reduce the possibility of disrupting the protein/lactase layer. Operating


b

a I 1 I I CARTRIDGE

- ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... .......

P ' ENZYME SOLUTION

F 7 ULTRAFI LTE R

I I

PRODUCT I PREFILTER

FEED TANK

FIG. 20. (a) Experimental setup for enzyme loading. (b) Backflush hollow-fiber enzymatic reactor system. From Jones, C. K. S. , Yang, R. Y. K., and White, E. T. (1988). A novel hollow- fiber reactor with reversible immobilization of lactose. AZChE J. 34, 293-304. Reproduced by permission of the American Institute of Chemical Engineers. Copyright 6 1988 AIChE. All rights reserved.

conditions that were tested for effects on flux and hydrolysis were temperature, pressure, and enzyme concentration. Results indicate that reactor conditions of 5 psi and 30°C with an enzyme load of 2000 mg of K. luctis lactase per 0.06 m2 hollow-fiber membrane lead to the best performance of the bioreactors tested. The bioreactor has a half-life of approximately 19 hours and is capable of a greater than 40% rate of lactose hydrolysis (Table XIII). It appears that the protein layer, deposited on the lumen of the hollow-fiber membrane to protect the lactase from denaturation, eventually reduces the half-life of the bioreactor by introducing bacterial contamination over time. Permeate samples at 17 to 21 hours had pH readings around 5.8. K. luctis activity has been shown to drop rapidly below


TABLE XI1 COMPARISON OF LACTASE HOLLOW-FIBER ENZYME BIOREACTORS LOADED ON THE

LUMEN SIDE OF THE MEMBRANE

Permeate glucose Glucose production Reactor type (mg”/.) Half-life (h)

K. lactis 345 180 13.6

A. niger 277 234 57.0 A. niger/K. lactis 324 230 14.6 K. lactisbovine albumin 500 240 26.0

K. lactis technical grade 40 37 4.5

pH 5.9 (Guy and Bingham, 1978). The enzyme load for maximum efficiency has been optimized at 3000 mg (16,000 lactase units) per 0.06 m2, as shown in Fig. 22. This lactase membrane bioreactor demonstrated the potential to accomplish 84% hydrolysis of the lactose in skim milk UF permeate in one pass. It operated continuously for more than 40 hours with skim milk UF permeate, as shown in Fig. 23. The reactor operated at steady state for nearly 15 to 20 hours, with a flux that ranged between 9.5 and 13 liters/h/m2 and glucose production that averaged nearly 70% hydrolysis of the lactose.

A major concern in studies of lactase immobilization with hollow-fiber reactors is their tendency to suffer enzyme leakage. This phenomenon occurs even when the molecular weight cutoff of the membrane used is much smaller than the enzyme molecular weight. The problems associated

FIG. 21. Schematic diagram of the modified localized ultrafiltration enzyme bioreactor. Presented at the 1993 Institute of Food Technologists Annual Meeting, Chicago.


TABLE XI11 COMPARISON OF DIFFERENT UF LACTASE BIOREACTORS WITH RESPECI TO FLOW

RATE AND GLUCOSE PRODUCTION"

Pressure Temperature Glucose Glucose Flux 7% Half-life (Psi) ("C) (mg 9%) ( m o ) (liters/m*/h) Hydrolysis (h)

2.5 20 2342 411 0.30 41.5 47 2.5 30 2404 911 0.63 44.1 13 2.5 40 2087 1470 1.18 38.1 11 5.0 30 1716 2390 2.34 31.3 16 5.0 30" 2021 2264 1.77 39.2 19

Bioreactors were operated with lo00 mg of lactase. Values are averages over 6 hours of

Bioreactor operated with 2000 mg of lactase. operation for each bioreactor.

Presented at the 1993 annual Institute of Food Technologists meeting, Chicago.

with enzyme leakage are a result of the wide pore size distributions of commonly available UF membranes (Cheryan, 1986). Leakage has been controlled by choosing membrane pore sizes 5 to 10 times smaller than the

30 0 lo00 2000 3000 4000 5000

ENZYME CONCENTRATION (mg)

FIG. 22. Effect of K. luctis lactase concentration on lactase-catalyzed hydrolysis with a hollow-fiber UF membrane bioreactor operated in the intestinal mode at 30°C and 5 psi (substrate = 5% buffered lactose).


FIG. 23. Continuous operation of the localized membrane lactase bioreactor at 30°C and 5 psi, showing the flux and glucose production from skim milk UF permeate (lactase = 3000 mt/0.6 mZ).

molecular weight of the enzyme. No leakage was experienced in studies conducted with 10,000 MW permeable membranes (Maculan, 1979; Jones et al., 1988); however, significant enzyme leakage occurred when XM50 membranes (Huffman-Reichenbach and Harper, 1982; Jones et al., 1988) and HlP50 membranes (Maculan, 1979) were used. Enzyme leakage was not detected during operation with K. luctis lactase loaded on the lumen side of an HlP30 hollow-fiber membrane (Senecal, 1991). Reactor flow rates have also been found to have a significant effect on enzyme leakage. At 200 mumin there was about 1% leakage from a PM30 membrane; however, at 6000 ml/min, there was an average of 29% enzyme leakage (Huffman-Reichenbach and Harper, 1982).

H. FUTURE POTENTIAL FOR COMMERICAL ENZYME SOURCES

Although development of the current commerical sources of p-galactosidase has allowed the consumption of dairy products by lactose-intolerant individuals, a great deal of work must still be performed to improve optimum enzyme performance, to reduce cost and increase the rate of hydrolysis. It has already been shown that commercial lactases perform differently under similar environmental conditions. In addition, industrial applications that use lactases differ in their operating temperature, pH, ionic environ-


ment, and substrate. The ideal enzyme would be P-galactosidase with the proper activity, specificity, and stability to satisfy the conditions of each individual application. Current research has been dedicated to the development of stable, lower-cost, and productive reactor systems. New lactase sources that perform in low- or high-temperature environments have been studied as means of improving existing continuous reactor systems.

Commerical lactases are produced by the extraction, separation, and partial purification of whole cells. This costly procedure has led to the study of immobilization of whole cells of suitable sources, such as L. bulguricus, K . luctis, and K. frugilis (Mahoney, 1985). Utilization of the whole cell has demonstrated both advantages and disadvantages. The advantages are that the whole cell has been shown to have greater stability than the purified enzyme (Park et ul., 1979; Mahoney, 1985). The disadvantages are that diffusional restrictions can lead to lower enzyme activity and other active enzymes (i.e., lipase or protease) could affect end-product flavor (Ma- honey, 1985). Chawan el ul. (1993) studied the stability of liposome- encapsulated P-galactosidase in milk. Their objective was to determine the stability and feasibility of encapsulating P-galactosidase in liposomes to reduce the flavor changes that occur in milk during hydrolysis. These systems would allow lactase to be added to milk without changing the product composition or requiring oral supplements. Results indicate negligible hydrolysis of lactose in milk containing liposomes with E. coli, but significant (25%) hydrolysis in milk containing liposomes with A. flhvus. Liposomes or other microencapsulation technologies can lead to enzyme systems that can be tailored to meet the requirements of the different available commercial lactases. In addition, the encapsulated enzymes can be adapted for utilization and release in either liquid or dry food systems. Important stabilizers and activators can be added with the lactase enzyme to improve both stability and functionality in different industrial applications. Research that incorporates encapsulated enzyme systems should include bioavailability studies, especially if the intended use is for activity within the intestine. Microencapsulation is a growing field that has tremendous potential for future applications in enzyme technology.

Enzyme modification is another potential area of research that could lead to improved commercial lactases. The diversity already exhibited by different sources of P-galactosidase could be expanded even further through enzyme modifications. These modifications could be effected through chemical or genetic methods that must be identified and require further extensive studies. Genetic methods have grown rapidly with the advent of increasingly sensitive analytical equipment. Genetic manipulation could lead to the development of a wide range of lactases, each engineered for a particular commercial application. The increase in the use of the thermostable en-

58 LORI F. PIVAFWIK etal .

zymes a-amylase and neutral protease from thermophilic organisms is, in part, due to gene cloning of the thermophiles into mesophilic production strains (Zamost et al., 1991).

1. Future Enzyme Sources

A great many sources of the enzyme P-galactosidase have been identified in the literature (Gekas and L6pez-Leiva, 1985; Mahoney, 1985); however, only lactases derived from yeasts and molds and recognized as safe are currently used for industrial applications. Over the years, extensive studies have been conducted and technologies developed with the purpose of improving the stability and functionality of these commercial lactases. Still, these commercial enzymes lack important properties, such as thermostability at the pH of milk products, that are required to render them cost effective in a continuous commerical process. Fungal lactases have proven to be stable but costly in the commercial hydrolysis of whey. The industrial application of thermostable enzymes allows the process to operate at higher temperatures, resulting in shorter processing times and reduced chances of microbial growth. Therefore, studies have been continued on potential lactase sources in an attempt to discover new "safe" lactases that would function ideally in normal commercial processing.

Studies have been conducted on P-galactosidase derived from thermophilic lactic acid bacteria such as Streptococcus thermophillus (Greenberg et al., 1985; Smart and Richardson, 1987; Chang and Mahoney, 1989). This lactase was attractive because it is active at temperatures above 50"C, it is a natural constituent in fermented milks, and competitive product inhibition by galactose is weak. Even though the enzyme is more stable in milk than in buffer, operation at temperatures above 50°C is limited (half-life of 146 minutes at 60°C) (Chang and Mahoney, 1989). Other thermophilic bacteria that have received attention were derived from Bacillus. Bacillus sterarothermophilus lactase, which has slightly greater thermostability than S. ihermophillus lactase (Mahoney, 1985), was studied extensively by Grif- fiths and Muir (1978) in purified, whole-cell, and immobilized whole-cell forms. The enzyme in all forms was found to be very stable at 55" and 60°C. The whole cell was more stable than the purified extract above 65°C; however, all forms lost significant activity at higher temperatures. The only disadvantage to date has been a reported yield 10 times lower than what was currently attainable from yeast (Mahoney, 1985). There has been no evidence of an improvement of enzyme yield at this time. Bacillus circulans lactase has been investigated because of its pH optimum of 6.0 and its greater thermostability compared with yeast lactases. Applications for


GRAS status have been submitted to the Food and Drug Administration (Bakken et al., 1992).

Psychotrophic microorganisms would be another potential source of p- galactosidase for dairy processing; however, little is known about their types, physiology, and safety. Loveland et al. (1994) isolated three possible members of the Arthrubacter genus with p-galactosidase activity. Lactase activity of whole cells was optimum at 20"C, and the cells were labile when incubated above 40°C. Lactases active at refrigeration temperatures would be beneficial in hydrolyzing lactose in dairy products at low temperature, thus reducing product spoilage from microbial contamination.

Fermented milk products have been identified as alternative dairy products for lactose-intolerant individuals who want to avoid the symptoms of lactose malabsorption (Marteau et al., 1990 DeVrese et al., 1992). Kefir, a cultured milk product produced by inoculation of cow's milk with kefir grains, contains P-galactosidase activity primarily as a result of the presence of lactobacilli. Itoh et al. (1992) isolated cell-free extracts of Lactobacillus kejiranufaciens from kefir grains. The enzymes had an optimal temperature of 50°C and a pH of 6.5 and were rapidly denatured above 55°C. In addition, the lactase was inhibited by both glucose and galactose; however, galactose inhibition was weaker than in most other lactase sources.

Finally, Gonzalez and Monsan (1991) purified p-galactosidase from the fungus A. funsecaeus grown on wheat bran. They were able to obtain a yield of 81% without protease contamination. The A. fonsecaeus lactase was found to have greater thermostability than a commerical A. oryzae lactase. A. funsecaeus lactase retained 100% activity for more than 1 week at 50°C and lost 50% of the initial activity in 36 hours at 65°C. In comparison, the enzyme activity for A. uryzae lactase decreased 23% after 8 hours at 50°C and 50% after 10 minutes at 65°C.

Ill. TRANSGALACTOSYLASE ACTIVITY

Of commercially available enzymes that command food industry interest, 90% are hydrolytic enzymes-proteases, carbohydrases, and lipases (Mon- san et al., 1989). Enzymatic hydrolysis of carbohydrates has found many applications in the food industry. The economic impact of glycosidases account for a third of the worldwide enzyme market (Vulfson, 1993). Like many hydrolases, however, the glycosidases can also catalyze the synthesis of simple and more complex saccharides, resulting in novel oligosaccharides and glycosides (Monsan et al., 1989; Cote and Tao, 1990; Nilsson, 1991; Bucke, 1993; Vulfson, 1993).


Since the 1950s and early 1960s, investigators have been studying and reporting on the transgalactosylase activity of P-galactosidases derived from fungal, bacterial, and yeast sources using a variety of lactose-containing substrates (Aronson, 1952; Pazur, 1953,1954; Roberts and McFarren, 1953; Roberts and Pettinati, 1957; Pazur et al., 1958; Wallenfels and Malholtra, 1961). Subsequently, studies have shown that the number, concentration, and type of saccharides formed depend on the enzyme source, nature and concentration of substrate, pH, temperature, ion activators/inhibitors, type of process (free versus immobilized), and degree of lactose conversion (Shukla, 1975; Zarate and L6pez-Leiva, 1990). Until recently, the primary reason for investigation of the transgalactosylation reaction of P-galactosidases has been nutritional concerns regarding the potential indigestibility of newly formed saccharides (Asp et aZ., 1980; Kwak and Jeon, 1986; Burvall et al., 1980; Richmond et al., 1981; Mahoney, 1985; Zarate and L6pez-Leiva, 1990) and the knowledge that allolactose, a major disaccharide formed during this enzymatic process, is involved in stimulation of the lac operon of E. coZi (Jobe and Bourgeois, 1972; Huber et al., 1976; Cote and Tao, 1990). In addition, formation of high levels of oligosaccharides in concentrated lactose solutions, such as whey, could lead to crystallization of these sugars during storage because of low solubility (Mahoney, 1985).

Interest in and recognition of the oligosaccharide moieties of glycoproteins and glycolipids in biological processes led to efforts to synthesize them (Smart, 1991; Thiem, 1993). Chemical synthesis has not been practical because of the laborious nature of conventional methods, the low yield of desired product, the multiple reaction steps, and the need to use expensive/ hazardous chemicals (Hedbys et al., 1984; Nilsson, 1990; Nilsson and Fernandez-Mayoralas, 1991; Vulfson and Law, 1993). Glycohydrolases, however, are inexpensive, are easily obtained, and tolerate handling and immobilization well (Thiem, 1993). Therefore, the development of alternative biosynthetic strategies employing enzymes has become attractive (Nils- son, 1988a; Vulfson and Law, 1993).

The characterization of galactosyl (and glycosyl) enzymatic transfer reactions only became of interest with the recognition of their potential applications in the synthesis of pharmaceutical, medical, and immunological products, the formation of biologically active compounds, and the synthesis of food ingredients (Nilsson, 1987, 1991; Vulfson, 1993; Vulfson and Law, 1993; Monsan et al., 1989; Hedbys et aZ., 1984, 1989; Cote and Tao, 1990; Zarate and L6pez-Leiva, 1990; Thiem, 1993). The application of transgalactosylase activity to the production of a wide variety of products has enormous potential and can serve as an alternative to the more complex chemical syntheses of oligosaccharides.


A. MECHANISM

I . Transgalactosylation versus Reverse Hydrolysis

There are two options for the utilization of the stereospecific synthesis potential of P-galactosidase: reverse hydrolysis and transgalactosylation (Nilsson, 1987, 1991; Monsan et al., 1989; Cote and Tao, 1990; Vulfson, 1993). Although not favored under physiological conditions, reverse hydrolysis reactions can be achieved by creating conditions, such as very concentrated sugar solutions, whereby the thermodymanic equilibrium is modified. P-Galactosidase can form saccharides by reverse hydrolysis via a condensa- tion reaction and release of a water molecule (Monsan et al., 1989; Cote and Tao, 1990; Nilsson, 1991). Although this has been done successfully by several authors (Ajisaka et al., 1987a,b, 1988), there are drawbacks. First, synthesis is limited because of unfavorable equilibrium conditions that also require long reaction times (Monsan et al., 1989). As reaction time is slow because of the limited activity in the concentrated sugar solutions needed to accomplish synthesis, higher enzyme concentrations are needed. Hahn-Hagerdahl et al., (1987) showed that the very low water activity that accompanies concentrated sugar solutions is inhibitory to the synthetic (and hydrolytic) activity of P-galactosidase and maximum enzyme activity cannot be obtained. Furthermore, the enzyme is more specific to galactosides than to monosaccharides; therefore, not only are enzyme requirements higher, but pure enzyme preparations must be used and yields are lower (Nilsson, 1991; Vulfson, 1993).

Transgalactosylation or “transhydrolysis” is the synthesis mode addressed in this review. Because of the higher specificity, product formation using the transgalactosylic properties of P-galactosidase results in higher yields of compounds in shorter periods. Furthermore, crude enzyme preparations may be used (Nilsson, 1991) and, as will be discussed later, regioselectivity can be manipulated. The mechanism of P-galactosidase activity was outlined in the discussion of its hydrolytic characteristics. Figure 12 indicates that hydrolase activity is achieved by the transfer of galactose (glycone) from glucose to a water molecule. If the acceptor (aglycone) were other hydroxyl-containing compounds, synthesis would occur and galactosides would form (Shukla, 1975; Burvall et al., 1979; Mahoney, 1985; Prenosil et al., 1987a; Monsan et al., 1989; Cote and Tao, 1990; Zarate and L6pez-Leiva, 1990; Nilsson, 1991; Vulfson, 1993). Glycosidic bonds are formed via transfer of galactose to other acceptors, such as sugars, alcohols, amine-containing sugars, and nitrophenyl-containing sugars, at different positions, thus forming various di-, tri-, and other oligosaccharides. The enzyme exhibits preference for formation of a P(1-6) linkage when sugars


are used as the acceptors, but other glycosidic bonds have been formed depending on the microbial source of the enzyme (discussed later). In the presence of lactose, without addition of aglycone or acceptor molecules such as monosaccharides, the major disaccharide formed as a result of galactosyl transfer activity using P-galactosidase is allolactose, created by the P( 1-6) bond between galactose and glucose (6-O-P-~-galactopyranosyl- D-galactose). Another commonly formed disaccharide is galactobiose [p- D-galactose-( 1-6)-~-galactose]. The general principles for transgalactosylation were illustrated by Nilsson (1991) (Fig. 24). The donor glycoside (i.e., lactose) is hydrolyzed, an enzyme-galactosyl complex is formed, and glucose is liberated. The transgalactosylation product is formed via transfer of galactose to an acceptor molecule. The energy necessary for this synthesis comes from the splitting of the glycosidic bond stored as a covalent enzyme- derivative intermediate (Monsan et aL, 1989), in this case, an enzyme-galactosyl complex. The efficiency of this mechanism of galactosyl transfer can be enhanced by using increased concentrations of the desired donor and acceptor molecules, decreasing available water (without becoming inhibitory), and removing final product (Monsan et al., 1989; Vulfson, 1993.

Huber et al. (1976) studied the transgalactosylase activities of P-galactosidase from E. coli on lactose and were the first to postulate a mechanism for activity. These researchers reported two types of transgalactosylase activities: direct and indirect. In the direct mode, the glucose moiety is not released from the active site of the enzyme after hydrolysis and the galactose is directly transferred to the glucose acceptor. Indirect activity involves the loss of glucose from the enzyme prior to galactose transfer and the binding of an acceptor in the second phase of the reaction. The authors postulated that the indirect mode of galactosyl transfer is the predominant route to the formation of higher-molecular-weight oligosaccharides and that, initially, direct transfer is the main mode of allolactose production. Free glucose

E Gal=H20---)Gal-OH 11 H l b

-ROH Gal&OR + EH EH GalE-OR 1_ E-Gal

41

I1 HoA I .

E*Gal.HOA GalB=OA

FIG. 24. General principles of transgalactosidation. Galp-OR, donor (lactose); EH, enzyme; HOA, acceptor; Galp-OA, transgalactosidation product. Reprinted from Nilsson (1991). pp. 135, by courtesy of Marcel Dekker, Inc.

COMMERCIAL 6-GALACTOSIDASE IN FOOD PROCESSING 63

becames a significant acceptor for formation of allolactose at concentrations greater than 0.01 M.

Kinetic models must incorporate both hydrolysis and transgalactosidation activities to accurately reflect enzymatic action (Zarate and Lbpez-Leiva, 1990). A simplified model (Fig. 25) describing both hydrolysis and transfer reactions was developed by Prenosil et al. (1987a). The model illustrates the numerous possibilities for P-galactosidase activity in the presence of lactose, glucose, galactose, and other oligosaccharides. Yang and Tang (1988) proposed a kinetic model for A. niger-derived lactase hydrolysis and transgalactosidation that fulfills a material balance for initial substrate and final product concentrations. These authors came to the conclusion that a kinetic model is plausible. Prenosil et al. (1987a), however, reviewed proposed models for p-galactosidase action on lactose and concluded that although models provide qualitative information on the reactions, the complexity of the equations and the number of kinetic constants involved make classic modeling extremely difficult.

Regardless of the mechanism proposed, it has been widely accepted that /3-galactosidase, although specific for the glycone or galactose part of lactose, has less selectivity for the acceptor or aglycone portion. Griffith (1991) incubated K. lactis lactase with galactose and found poor disaccharide production, leading to the conclusion that transgalactosylation activity is not a viable alternative to chemical synthesis; however, this study actually measures reverse hydrolysis and not transgalactosylation. Research on

QalaCtose+E 4-b E.Qa-4-Ql 4-> E.Qa.01 - S.Oa+H1O

FIG. 25. Transgalactosylase and hydrolase activities of p-galactosidase. From Prenosil, J. E., Stuker, E., and Bourne, J. R. Biorechnol. Bioeng. 30, 1019. Copyright 0 1987 John Wiley & Sons, Inc. Reprinted by permission of John Wdey & Sons, Inc.


transgalactosylase activity of yeast lactases has shown that saccharides are not formed when glucose or galactose is present, alone, in low concentrations, up to 10% (Tikhomirova et al., 1980; Pivarnik, 1990). The efficiency of galactosyl transfer is maximized when lactose is present in the reaction mixture and is in the “hydrolytic” mode (Pivarnik, 1990; Pivarnik and Rand, 1991). Furthermore, the resulting products and their concentrations are critically impacted by the microbial source of the enzyme, the acceptor molecules available, and experimental conditions. Thus, it is impossible to make overall conclusions concerning the value of transgalactosylase activity. Two review articles (Prenosil et al., 1987a, and Zarate; L6pez-Leiva, 1990) discuss oligosaccharide formation during lactose hydrolysis by a variety of microbial lactases. Subsequent discussion in this review focuses on synthesis parameters with commerically available, food-grade P-galactosidases from A. niger, A. oryzae, K. la&, K. fragilis, and C. pseudotropicalis.

B. ASSAY METHODS

The methodology to assay the transgalactosylic activity of j3-galactosidase varies with the enzyme source and the products to be measured. One consequence associated with transgalactosylase activity that occurs during lactose hydrolysis concerns the erroneous estimation of lactase hydrolytic activity units due to the incorporation of monosaccharides into oligosaccharide production (Mahoney, 1985). Table XIV illustrates the numerous combinations of times, temperatures, pH, and substrates that have been used to study product formation due to transgalactosylase activity from commercially available P-galactosidases. The products formed, which are addressed later in this review, vary with the starting substrates and microbial source of the enzyme. The complete analysis of the oligosaccharides formed requires separation procedures, such as paper chromatography (PC), column chromatography, thin-layer chromatography (TLC), high- performance liquid chromatography (HPLC), and gas-liquid chromatography (GLC). Furthermore, identification of new products and the glycosidic linkages formed by the enzyme requires techniques such as mass spectroscopy (MS) and nuclear magnetic resonance (NMR). Generally, it has been shown that higher concentrations of lactose result in larger amounts of product; however, Vaheri and Kauppinen (1978) and Pivarnik and Rand (1991) both found that there are optimum lactose concentrations for the formation of products if acceptor molecules are present in the reaction mixtures. In addition, the presence of the aglycone molecule is one mechanism that leads to the more controlled synthesis of desired products. Although Table XIV shows the conditions under which product synthesis was followed and monitored, very little has been done to

COMMERCIAL B-GALACITOSIDASE IN FOOD PROCESSING 65

delineate actual “units” of P-galactosidase activity, from any microbial source, devoted to synthesis of di-, tri-, and/or oligosaccharides under any conditions of activity.

Huber et ul. (1976) reported kinetic data obtained from E. coli lactase, and their study was the first definitive research concerning velocity measurements. Information relating to the assay of units for lactase derived from commercial sources has, however, been lacking. Pivarnik (1990) and Pivar- nik and Rand (1991) have shown that under controlled conditions of substrate and enzyme concentrations, transgalactosylase activity units for K. luctis P-galactosidase can be determined (Fig. 26). The linear relationship of allolactose formation over time means that units of transgalactosylase activity can be calculated as micromoles produced per minute under these conditions. As illustrated, the velocity is constant up to 10 and 15 minutes when 2.5 and 5.0% lactose substrate is used, respectively. Table XV shows that units can be calculated, uniformly, for the linear portion of the curves. When the units of transgalactosylase activity are defined as micromoles of allolactose produced per minute per milligram of enzyme, 0.42 unit of activity was calculated in this study versus 7.1 to 7.2 hydrolytic units (pmole glucose produced/min/mg). Therefore, under the conditions of this assay, it is possible to determine that the transgalactosylase activity constituted approximately 6% of the enzyme activity.

C. PROPERTIES

1. p H

The impact of pH on enzyme activity is, of course, critical to achieving optimal enzyme action on a substrate. Huber et ul. (1976) showed that the pH profile for optimal production of oligosaccharides due to transgalactosylase activity in the presence of lactose is significantly different from that due to hydrolase activity. The curves representing transgalactosylase activity were clearly “skewed” toward higher pH values when compared with curves for glucose and galactose production. Other investigators working with food-grade lactases under various experimental conditions, however, found the pH effect to be variable.

Tikhomirova et al. (1980) investigated the effect of pH on transgalactosylase activity of lactase derived from K. frugilis and found that both the hydrolytic and transgalactosylic actions were optimal at pH 7.0. Figure 27 illustrates the effect of pH on the transgalactosylic and hydrolytic activities of K. luctis P-galactosidase. Although the optimum pH appears to be 7.0 for hydrolytic activity, the pH required for the transgalactosylic mechanism in the presence of lactose, glucose, and galactose is not as clearly delineated.

TABLE XIV TRANSGALACTOSYIASE ASSAY CONDlTIONS OF VARIOUS COMMERCIALLY AVAILABLE /%GALACIVSIDASES

Activating Temperature Incubation Analytical Enzyme source Substrate PH ion ("C) time (h) method" Reference

A. niger

K. fragilis K. hctis A. niger A. oryzae K. fragilis A. niger K. hctis

K. kactis

K. fiagilis

K. kactis A. niger

K. kaciis

A. oryzae

Acid whey 0

2-20% lactose 530% fructose

4-35% lactose

30% lactose

5 2 0 % lactose UHTSTmilk@

5% lactose 20% lactose

5,1096 lactose 2 10% glucose or galactose

UF whey 5, 10,25%

lactose 4.56% lactose as

Skim Inilk Phen yl-p-

galactosides

4.5

7.2

7.0 4.0 6.8

6.8

7.0

6.817.0 4.5

7.0

5.0

55

37

40

24

37

30

4,30 30,55

40

10 20

5

1-5

2-48

0-24 0-14 days

16

0.5-24

12-48 5-24

-

720 min 20 min

Carbodcelite Wierzbicki and Kosikowski (1973) TLC GC Vaheri and Kauppinen (1978)

PC Toba and Adachi (1978) GC PC Burvall ef aL (1979) SPEC

Sephadex Asp et al. (1980) G-15, PC, GC-MS

PC, GC, SPEC

Tikhomirova ef al. (1980)

GC Giec et aL (1981)

HPLC Nakanishi et aL (1983)

Sephadex Ooi ef al. (1984) G-25

HPLC

A. oryzae

C. pseudotropicalis K. lactis K. lactis

C. pseudotropicalis K. lactis C. pseudotropicalis

A. niger A. oryzae K . lactis K. fragilis A. niger A. oryzae K. lactis K. lactis

A. oryzae

A. oryzae

30% lactose

5.20% lactose

4.56% lactose 4.56,12%

NFDM HTST Inilk

5% lactose, whey, whole milk

2-30% lactose

2 5 2 5 % lactose Lactose + Rubusoside 2.5,5% lactose +

0-5% glucose + 0-5% galactose

ONPG + xylop yranoside derivatives

ONPG + L-serine

4.8

6.6

6.6

-

6.6

opt.

opt. 6.0

6.5

6.5

5.0

37

4 37 25, 37

1, 7, 37

37

opt.

50 40

37

35

20.37

8

0-24 0-4 6, 10

0.5-14 days 0.25-4 0-4

-

0-8 1-15

5 min 0-120 min

2

30 + rnin

PC, GC, NMR

HPLC

PC SPEC

HPLC

HPLC

HPLC

HPLC TLC, HPLC NMR HPLC

NMR,GC

Dowex, Sephadex G-10, NMR, HPLC

Toba et al. (1985)

Jeon and Mantha (1985)

Mozaffar et oL (1985)

Kwak and Jeon (1986)

Jeon and Saunders (1986)

Prenosil et al. (198%)

Yang and Tang (1988) Kitahata et ul. (1989)

Pivarnik (1990) Pivarnik and Rand (1991)

Lopez et al. (1991)

Sauerbrei and Thiem (1992)

9 (continues)

TABLE XIV (Continued)

Activating Temperature Incubation Analytical Enzyme source Substrate PH ion ("C) time (h) method" Reference

A. oryzae Lactitol 5.0 - 50 4 TLC, HPLC, Yanahira et af. (1992) NMR, GC-MS

A. oryzae 20% lactose + 4.5 - 40 60 min HPLC, FAB- Kitahata et aL (1992) K. fragilis 40% 6.5 MS K. Inctis branched 6.0

C. pseudotropicalis Chocolate/plain - - 4,7,10 0-20 - Huang and Rosenberg (1993)

cyclodextrins

Inilk * 3.95, 7.5% glucose

TLC, thin-layer chromatography; GC, gas chromatography; PC, paper chromatography; SPEC, spectrophotometry; MS, mass spectroscopy; HPLC, high-performance liquid chromatography; NMR, nuclear magnetic resonance; FAB, Bame absorption spectrometry; opt, optimum conditions.


11 - 10 - 2.5 % Lactose

A 50JLactose

1

0 10 2 0 30 4 0

Time (minutes)

FIG. 26. Effect of lactose concentration on allolactose production catalyzed by K . lactis lactase. Experimental conditions: 2.5% glucose, 2.5% galactose, 0.001 M MgCI2, and 17.5 hydrolytic units reacted at pH 6.5 and 37°C. Presented at the 1991 annual Institute of Food Technologists meeting, Dallas.

Allolactose production was not significantly different between pH 6.5 and 7.0. Vaheri and Kauppinen (1978) studied the effect of pH on the synthesis of lactulose from lactose (donor) and fructose (acceptor) using K. fragilis P-galactosidase. They reported that the optimum pH was 7.2 for both hydrolysis of lactose and formation of lactulose, and the ratio of lactulose formed to lactose hydrolyzed was unaffected by pH. Lopez etal. (1991), however, reported an increase in the optimal synthesis of P-galactopyranosyl- D-xylose disaccharides catalyzed by A. oryzae p-galactosidase when the pH was increased from 5.0 to 6.5. As illustrated in Tabel XIV, studies following the transgalactosylase activity of commerically available lactases


TABLE XV KLUYVEROMYCES ucris p-GALACTOSIDASE

VELOCITY FOR THE PRODUCTION

OF ALLOLACI'OSE

Adjusted velocity (pmole/min)

Time (min) 2.5% Lactose 5.0% Lactose

1 2 5 6 8 10 15 30 60 120

0.82 1.07 1.04 1.01 0.92 0.93 0.68 0.34 0.16 0.03

- 1.06 - - 0.99 1.04

Both systems contained 2.5% glucose, 2.5% galactose, and 17.5 hydrolytic units of p-galactosidase activity were reacted at pH 6.5 at 37°C.

have been conducted at or near the optimum pH for the microbial source. On the basis of limited information, it appears that the transgalactosylation may occur optimally at the same or similar pH, about pH 7.0, for yeast- derived lactases. This pH optimum may also correspond to the value for hydrolytic activity. The same conclusions may not be valid for fungus- derived lactases. More research is necessary to evaluate properly the effect of pH, and, more important, each experimental condition must be evaluated separately.

2. Temperature and Reaction Time

As with all enzymatic reactions, both temperature and reaction time have an effect on the pattern and/or concentration of di-, tri-, and oligosaccharides synthesized by 6-galactosidases. Generally, as temperature is increased to an optimum, the transgalactosylase reaction rate increases. Al- though the resulting di-, tri-, and/or oligosaccharides reach their maximum concentrations sooner at higher temperatures, the tendency to form the same galactosyl transfer products remains, even at lower temperatures (Vaheri and Kauppinen, 1978; Giec et al., 1981; Mozaffar et al., 1985; Kwak and Jeon, 1986; Yang and Tang, 1988; Lopez et al., 1991). The commercially available fungal lactases A. oryzae and A. niger produce saccharides differing little in concentraction when moderate or optimum temperatures are

COMMERCIAL j3-GALACTOSIDASE IN FOOD PROCESSING 71

.) Allolactose 4 %Hydrolysis

-40

-30 (I)

!z E -

$ i#

-20

- 10

0 ; I I i 0 5 6 r 8

PH

FIG. 27. Allolactose production and lactose hydrolysis by K. lactis j3-galactosidase as a function of pH in a 5-minute assay at pH 6.5 and 37°C using 2.5% lactose, 2.5% glucose, and 2.5% lactose. Presented at the 1991 annual Institute of Food Technologists meeting, Dallas.

increased by 10" to 20"C, although some results indicate that the trisaccharides vary (Lopez et af., 1991; Giec et af., 1981; Prenosil et af., 1987b). Yang and Tang (1988), using an immobilized system containing A. niger lactase, found that higher temperatures promote oligosaccharide formation. Re- search conducted with yeast p-galactosidase on various lactose-containing substrates showed that at very low temperatures (4"C), there is both a delay (as expected) and a reduction in peak product formation when compared with formation at higher temperatures (Kwak and Jeon, 1986; Giec et af., 1981). Sugars formed under low-temperature conditions, however, have been shown to remain an extended period before being hydrolyzed by the enzyme (Kwak and Jeon, 1986). When product formation at lower

72 LORI F. PIVARNIK etal .

temperatures was evaluated over the longer time frame, the results appeared to show that total sugar formation is similar to that at the higher temperature peak. Vaheri and Kauppinen (1978) showed that the maximum temperature for transgalactosylase activity from K. fragilis shifts from the hydrolytic maximum of 36°C to 43"-53"C.

Reaction time has a significant effect on both the peak formation of transgalactosylase products and the maximum recovery of these products. Table XVI illustrates the relationship between lactose hydrolysis and maximum product synthesis. Transgalactosylase activity due to yeast-derived lactases results in maximum formation of oligosaccharides at higher degrees of lactose hydrolysis than that due to Aspergillus-derived lactases. All investigators have shown that the products formed appear to be related to the lactose hydrolyzed until lactose hydrolysis levels off and product hydrolysis begins. The decrease in available lactose in a reaction mixture will result in a shift in lactase activity, favoring the hydrolysis of the transgalactosylic products formed during transfer reactions (Monsan et ul., 1989). Figure 28 is a representative profile of the transgalactosidation products resulting from use of K. lactis P-galactosidase during lactose hydrolysis. Under assay conditions described by Pivarnik (1990) and Pivarnik and Rand (1991), of

TABLE XVI PERCENTAGE LACTOSE HYDROLYSIS AT MAXIMUM FORMATION OF

COMPOUNDS VIA TRANSGALACTOSYLASE ACTIVITY

OF P-GALACTOSIDASES

Enzyme Lactose hydrolysis (8) source at maximum synthesis Reference

K . fragilis 80 Vaheri and Kauppinen (1978) K . lactis 65-76 Burvall et al. (1979) K . fragilis 60 Tikhomirova et al. (1980) K. lactis 95-99 Giec et al. (1981)

K . lactis 60-85 Nakanishi et al. (1983) K . lactis 80-90 Mozaffar et al. (1985) K. lactis 82-85 Jeon and Mantha (1985) A. niger 47' Prenosil et al. (198%)

A. niger 65-91

A. oryzae 44-53 K. fragilis 75 K. lactis 68 A. niger 40 Yang and Tang (1988) K. lactis 70-85 Pivarnik (1990)

Pivarnik and Rand (1991)

All values estimated from graphic presentation.


80

0 Lactose

0 MonosaceherMes

0 Disaccharide A

Allolactose (C)

0 1 0 2 0 3 0 40 5 0 6 0 7 0 8 0 9 0 100 110 1 2 0 130

Time (minutes)

FIG. 28. Profile of substrate utilization and product formation during hydrolysis catalyzed by K. lactis P-galactosidase. Experimental conditions: 2.5% lactose, 2.5% glucose, 2.5% galactose, 17.5 hydrolytic units, 0.001 MgCI2, and reacted at pH 6.5 and 37°C. Presented at the 1991 Annual Institute of Food Technologists Meeting, Dallas.

two disaccharides synthesized with yeast lactase transgalactosylase activity, one appears at 6.4% lactose hydrolysis, between 0 and 1 minute, and the other after 5 minutes, when lactose hydrolysis reaches 38.5%. Both compounds peak to a maximum concentration that remains constant and, ultimately, are hydrolyzed at different rates. A rapidly formed compound may be more susceptible to secondary hydrolysis, and the ratio of product isomers would depend on the time of reaction termination (Nilsson, 1991). Therefore, the impact reaction time has on product formation and, ultimately, hydrolysis depends on the specific product(s) formed, assay conditions, availability of acceptor moleculedsubstrate, and microbial lactase source.

There could be a nutritional implication for low-lactose milk products. If milk is hydrolyzed only 75 to 85% using yeast-derived lactases, as is

TABLE XVII EFFECT OF LACTOSE CONCENTRATION ON THE PRODUCTION OF OLIGOSACCHARIDES BY TRANSGALACEIDATION

Lactose Oligosaccharide Enzyme concentration concentration Temperature source (96) Matrix (%I (“C) Reported as Reference

A. niger 4 >4-35

s. h t i s 5 10 20

10 s. ktis 5

15 25

s. lnctis 5 15

S. fragilis 5

A. niger 5 15 25

Acid whey

Buffer

Buffer

Cheese Whey Permeate Deionized

cheese whey permeate

Cheese whey permeate

1-2 24 % Total sugars Wierzbicki and Kosikowski (1973) inc. w/wnc. 5 8.5

55 % Total sugars Burvall ez al. (1979)

12.5

17.0 ND, N D P 430 % Total sugars Giec et al. (1981)

6 30 % Initial lactose Tikhomirova et aL (1980)

5.W, 6.88 9.26, 16.01 3.21, 4.94, 4 3 % Total sugars Giec ez al. (1981) 6.24, 9.28

2.11, ND 3035 % Total sugars Giec e? a1 (1981) 5.94,4.02

10.44. 8.86

A. niger

s. loctis s. lacris

K . lactis K. lactis

A. oryzae

A. niger

15 25

4.56 4.56

12 5 5

20 2 5 7.5

10 15 30 2.5 5

10 15 20 30

Deionized cheese whey permeate

Skim milk Skim milk

HTST milk Buffer

Buffer

Buffer

4.45, 6.11 30,55 96 Total sugars 7.26, 7.74

7.0b 40 % Initial lactose 14, 20, 23 25, 37.25 Q Initial lactose

8.87 7 % Lactose hydrolyzed 11.3 37 % Initial lactose 15.3-16.5

56 50 % Total sugar 11 15 19 22 29 2b 50

5 8

11 16

Giec ef al. (1981)

Nakanishi ef al. (1983) Mozaffar et al. (1985)

Kwak and Jeon (1986) Jeon and Mantha (1985)

Prenosil et al. (198%)

8 Total sugar Prenosil et al. (198%)

a ND, not detected. Determined from data presented as graphs.


usually done commercially (Prenosil et al., 1987b), research has shown that at least 5% of the total sugars could occur as oligosaccharides (Burvall et al., 1979). The presence of the oligosaccharides may eliminate some of the beneficial effects of low-lactose milk for lactose-intolerant individuals. An increased reaction time could help alleviate any potential problems.

3. Substrate Concentration

a. Lactose. Transgalactosidation has been found, by many investigators, to be enhanced by increasing the concentration of lactose (Mozaffar et al., 1985; Mahoney, 1985; Prensoil et al., 1987a; Zarate and L6pez-Leiva, 1990; Monsan et al., 1989). Table XVII illustrates the effect of increasing initial lactose concentrations on the transgalactosylase activity of P- galactosidase from commercial sources. A similar trend was observed for immobilized P-galactosidase (Prenosil et al., 1987b; Yang and Tang, 1988). Although it is evident that increasing the concentration of donor results in greater production of di-, tri-, and/or oliogosacchrides, direct comparisons between studies are difficult. The bases on which the results were calculated and reported were different and experimental parameters (time, temperature, enzyme concentration, microbial source, substrate matrix) also varied. Lactose concentration also has an impact on the types of transgalactosylase products formed: more tri- and tetrasaccharides are produced as lactose levels are increased. In addition, the values reported in Table XVII show a “relative” change, but may not reflect the maximum concentrations actually obtained. Investigators may have reported oligosaccharide concentrations that were determined at a point in the reaction where the products were also being hydrolyzed.

b. Donor plus Acceptor Molecules. Although the use of glycosidases is attractive for synthetic applications, the yield and regioselectivity of desired products have been low (Nilsson, l987,1988b, 1991; Lopez et al., 1991; Lopez and Fernandez-Mayoralas, 1992). The variety of (1-6) linkages and, to a lesser extent, (1-2), (1-3), and (1-4) isomers has hampered the use of glucosidases and galactosidases for the synthesis of functional and biologically active carbohydrates (Nilsson, l987,1988b, 1991). Many investigators have used milk, whey, or buffered lactose as a medium to study the transgalactosylase activity of P-galactosidase and have reported a variety of results. The addition of monosaccharides to the system, in the presence of the donor (i.e., lactose), to maximize the formation of desired compounds is now receiving more attention. The presence and nature of the aglycone molecule can lead to more controlled synthesis and yield of desired products and the formation of oligosaccharides can be manipulated (Nilsson, 1987,


1988b, 1991; Monsan et al., 1989; Nilsson and Fernandez-Mayoralas, 1991; Lopez and Fernandez-Mayoralas, 1992; Bucke, 1993).

Of the work cited above, delineation of the impact of acceptor molecules on activity of transgalactosylases derived from commercial sources has been limited. In 1952, Aronson reported the effects of adding sugars-glucose, galactose, xylose, glycerol-to a reaction mixture containing lactase from K. fragilis and 2% lactose. Although the results of the experiments were qualitative, the author concluded that addition of acceptor molecules would result in the formation of additional products. More recent work has quantitatively evaluated the formation of saccharides due to the action of commercially available lactases on substrate mixtures of donor and acceptor molecules. The formation of lactulose, cardiac glycosides, /3-xylopyranosides, branched cyclodextrins, and lactitol-derived oligosaccharides has been reported (Vaheri and Kauppinen, 1978; Ooi ef al., 1984; Lopez ef al., 1991; Kitahata et al., 1992; Yanahira et al., 1992). All of these compounds have potential food ingredient, nutritional, or medical uses.

Optimizing the transgalactosylase reaction of commercial lactases in the presence of donor and acceptor molecules has received little attention. Vaheri and Kauppinen (1978) showed that although the absolute yield of lactulose increased with increasing concentrations of lactose, the ratio of lactulose to hydrolyzed lactose did not change when lactose concentrations increased above 10%. Pivarnik (1990) and Pivarnik and Rand (1991) studied K. lactis P-galactosidase to delineate the effects of adding glucose and/or galactose to lactose on the formation of allolactose. Table XVIII shows the relative impact of the monosaccharides on the production of allolactose. Increased concentrations of both lactose and glucose had an impact on product formation. The study showed that the presence of a desired acceptor molecule, in increasing concentrations, could result in a higher percentage of desired product while minimizing undesirable compounds. Table XIX illustrates the total di- and trisaccharide production as a percentage of the total sugars in the reaction mixture. It would appear that regardless of the lactose concentration used, once hydrolysis reaches 57 to 58%, the percentage of galactosyl transfer products formed is approximately the same, 9%. Combining lower concentrations of lactose with higher concentrations of acceptor molecules could maximize formation of desired end products and minimize formation of undesirable oligosaccharides.

4. Effect of Commercial Source

Many investigators have studied, both qualitatively and quantitatively, the structures of oligosaccharides formed during lactose hydrolysis catalyzed by transgalactosylase. The number and type of di-, tri-, and oligosac-


TABLE XVIII

KLUYVEROMYCES LACTIS /3-GALACTOSIDASE WITH VARYING

REACTION TIME AT PH 6 .5 AND 37°C

RELATIVE TRANSGALACTOSYLASE ACTIVITIES OF

SUBSTRATE COMPONENT COMBINATIONS AFTER A 5-MINUTE

Initial substrate level (8) Relative activity of allolactose

Lactose Glucose Galactose production (a).

0.0 2.5 2.5 0 2.5 2.5 2.5 100 2.5 0 0 256 2.5 2.5 0 106 2.5 2.5 0.5 115 2.5 2.5 5.0 112 2.5 5.0 0 160 2.5 5.0 1 .o 131 2.5 5.0 2.5 132 2.5 5.0 5.0 148 2.5 1.0 5.0 82 5.0 2.5 2.5 155 5.0 0 0 59 5.0 5.0 0 247

~ ~~

Based on observed values and the standard assay (100%). Value obtained at 15 minutes; therefore, compared with 15-mi-

nute standard assay data to obtain relative activity.

charides produced varied with the yeast, fungal, and bacterial sources studied and the conditions of the experimental design and detection. The disaccharide allolactose [P-D-Gal( 1-6)-~-Glu] has been reported to be the major product formed via transgalactosylase activity by many of the authors cited in this review. Galactobiose [P-~-Gal(l-6)-~-Gal] is another promi- nent disaccharide produced under conditions of lactose hydrolysis. The trisaccharides that have been reported in significant amounts are galactolac- tose, galactoallolactose and galactogalactobiose. Review articles by Prenosil et al. (1987a) and Zarate and L6pez-Leiva (1990) present comparative data on structures delineated by a variety of researchers. Therefore, this review briefy describes the impact of the microbial source on oligosaccharide formation during lactose hydrolysis.

There is a lack of agreement in the literature concerning the identification and enumeration of di-, tri-, and oligosaccharides formed during galactosyl transfer for all commercially available lactases. Yeast-derived lactases have been studied extensively. Aronson (1952), Pazur (1954), and Pazur et al. 1958) reported the formation of 4 to 5 saccharides when K. fragilis P-


TABLE XIX DISACCHARIDES AND OLIGOSACCHARIDES FORMED BY KLUYVEROMYCES LACTIS

LACTASE AS A PERCENTAGE OF THE TOTAL SUGARS PRESENT IN THE ASSAY AT A

GIVEN TIME INTERVALO

2.5% Lactoseb 5.0% Lactose*

Time Di Tri % 8 Di Tri % %

( i n ) (mg) (mg) Total Hydrolysis (mg) (mg) Total Hydrolysis

- - - - 2 0.73 NDC 1.2 15.2 5 3.50 ND 6.2 38.2 3.94 0.97 4.0 25.3

10 5.24 ND 8.7 57.2 4.81 1.89 5.6 49.0 15 5.56 ND 8.7 70.9 8.47 2.23 8.9 58.6 30 5.67 ND 8.6 86.2 60 5.76 ND 8.1 93.7

120 1.17 ND 1.5 95.9

- - - - - - - - - -

Lactose left + monosaccharides formed + di- and trisaccharides formed. Both systems contained 2.5% glucose, 2.5% galactose, and 17.5 hydrolytic units of fl-

galactosidase activity. ND, not detected.

galactosidase was used during lactose hydrolysis; however, Roberts and McFarren (1953) reported that the number of products changed with higher lactose concentrations and identified 10 distinctly different di- and higher oligosaccharides. Roberts and Pettinati (1957) and Toba and Adachi (1978) identified 11 and 12 oligosaccharides, respectively, due to galactosyl transfer for K. fragilis lactase. Tikhomirova et al. (1980) reported the production of only four: one trisaccharide and three disaccharides. Although many investigators have reported that production of oligosaccharides via C. pseudotropicalis and K. lactis activity is limited (only 5 or 6 oligosaccharides) (Asp et al., 1980; Jeon and Mantha, 1985; Kwak and Jeon, 1986; Pivarnik, 1990; Pivarnik and Rand, 1991), Giec et al. (1981) reported 12 different saccharides produced by the action of K. lactis lactase (6 disaccharides and 6 trisaccharides).

Generally, mold-derived lactases, A. niger and A. oryzae, have been found to produce both a larger number of total transgalactosylase products and a higher proportion of trisaccharides when compared with yeast lactases (Giec ef al., 1981). Originally, Wierzbicki and Kosikowski (1973) described the formation of five oligosaccharides resulting from A. niger lactase activity on lactose in acid whey. As many as 10 to 16 different di-, tri-, and oligosaccharides have since been reported to have been formed during lactose hydrolysis in the presence of A. niger lactase (Toba and Adachi,


1978; Giec et af., 1981); and as many as 20 different oligosaccharides appeared when A. oryzae P-galactosidase was used (Toba e? aZ., 1985). Glyco- sidic bonds consisting of p(1-6), p(1-3), and p(1-2) linkages and di-, tri-, tetra-, penta-, and hexasaccharides have been reported.

The impact of microbial source has also been evident when transgalactosylation has been used to form oligosaccharides from a variety of donor and acceptor molecules. Six trisaccharides with a variety of glycosidic bonds, including p(1-5) and p(1-1) to a small degree, were formed from lactitol by a transgalactosylation reaction catalyzed by A. oryzae lactase (Yanahira et aZ., 1992). Transgalactosidation compounds produced from lactose and rubusoside by the action of A. oryzae lactase contained p(1-6) and p(1-4) linkages (Kitahata et af., 1989). Vaheri and Kauppinen (1978) showed that K. fragilis, K . Zactis, A. niger, and A. oryzae lactases all catalyze the formation of lactulose from lactose and fructose p( 1-4), but formation proceeded at variable rates. Other glycosidic isomers were also formed. Although p- galactosidase derived from A. oryzae has been reported to successfully form p-xylopyranosides, the degree and velocity are not as great as with p-galactosidases from other sources (Lopez et af., 1991).

Although all commercially available lactases studied are capable of forming transgalactosidation products during lactose hydrolysis, it has been shown that not all lactases perform galactosyl transfer reactions in the presence of all acceptors. Both K. Zactis and K. fragilis lactases do not catalyze a transgalactosylation reaction to branched cyclodextrins (Kitahata et af., 1992); however, (I-galactosidase derived from A. oryzae successfully produces transgalactosylation derivatives, although not to the same degree as lactases from other microbial sources. In addition, K. Zactis p- galactosidase does not form rubusoside derivatives, whereas A. oryzae lactase has some transgalactosylase activity under the same experimental conditions (Kitahata et al., 1989). From the limited work, it would appear that yeast-derived lactases have a more restrictive transgalactosylase activity, and all food-grade lactases must be evaluated if optimum production of novel compounds is to be achieved.

The potential of the transgalactosylase-catalyzed reaction lies in the production of novel, functional products from a variety of donor and acceptor molecules. Research has shown that choice of the lactase source has a critical impact on final product type and concentration.

5. Specificity

As stated previously, transgalactosylase activity results in the formation of di- and trisaccharides which appear to be common to most of the micro-


bial lactases. Some sources of lactase, however, have been reported to be more restrictive than others with respect to the selectivity of saccharides formed. Although the p(1-6) linkage is generally regarded as the major route of galactosyl transfer, other glycosidic bonds have been formed. E. coli-derived P-galactosidase has been shown to produce galactosyl transfer products in a (1-6)>(1-4)>(1-3) order of preferred bond formation (Wal- lenfels and Maholtra, 1961; Nisizawa and Hashimota, 1970; Nilsson, 1991). Though a variety of glycosidic bonds are possible, some sources of lactase clearly are more restrictive with respect to the selectivity of bonds formed. K. lactis-derived P-galactosidase has transgalactosylation activity with a high specificity for the formation of p(1-6) galactosidic linkages (Asp et al., 1980). Toba et al. (1985) elucidated the structures of oligosaccharides produced by A. oryzae lactase and found that both p(1-3) and p(1-4) linkages were formed, in addition to p(1-6) bonds. Both A. niger and K. fragilis lactases were found to form disaccharides that contained both p(1-2) and p(1-3) bonds; however, this was not considered the major route of galactosyl transfer and p(1-6)>p(1-3) was the predominant order of synthesis (Toba and Adachi, 1978).

Conducting the transgalactosidation reaction in low-water media may also help manipulate enzyme specificity and product recovery (Vulfson, 1993). Addition of organic solvents to the medium has been studied with successful results (Vulfson et al., 1990; Sauerbrei and Thiem, 1992; Vulfson and Law, 1993).

The structure of acceptor molecules, or aglycones, can influence the specificity of the transgalactosylase reaction. The regioselectivity of glycosidase-catalyzed transglycosidations can be manipulated by changing the structure and/or anomeric configuration of the aglycone (Nilsson, 1987, 1990, 1991; Lopez and Fernandez-Mayoralas, 1992). Size and hydrophobicity of the acceptor aglycone may also have an impact on the specificity to form certain glycosidic linkages (Nilsson, 1991). The yield and regiospec- ificity of transglycosidation reactions have been controlled simply by modi- fying one or more of the hydroxyl groups of the acceptor molecule, other than the reducing end (Nilsson and Fernandez-Mayoralas, 1991). Lopez and co-workers (1991), using A. oryzae-derived P-galactosidase for the formation of p-xylopyranosides, showed that the aglycone acceptor molecules had a significant impact on the yield and types of products formed. Sauerbrei and Thiem (1992) studied products formed by A. oryzae p- galactosidase in a reaction system containing nitrophenyl galactosides as donor and acceptor. Under similar experimental conditions, ortho- nitrophenyl and para-nitrophenyl galactosides were converted to compounds containing different glycosidic bonds: p(1-6)/p( 1-4) and /3(1-2)/


P( 1 a), respectively. Most of the work on specificity manipulation, however, has not been done on commercially available, food-grade P-galactosidases.

6. Activation and Inhibition

Information on the activation and inhibition of the transgalactosylase activity of lactases has been limited. Huber et aZ. (1976) first showed the impact of MgZ+ ions on E. coli lactase-catalyzed production of allolactose in a buffered system. A dramatic 70% decrease in activity was observed when the ion was eliminated from the assay mixture. Potassium ions have also proven to be a powerful activator, in the presence of magnesium, for P-galactosidase derived from S. thermophilus (Smart, 1991). The effect of magnesium ions on K. Zactis P-galactosidase (Table XX), however, has been shown to be less severe, with approximately 30% of activity loss in a buffered system when these ions were eliminated (Pivarnik, 1990; Pivarnik and Rand, 1991). This study also reported relatively little impact of potassium (K+) ions on galactosyl transfer. Although potassium has an enormous impact on the hydrolytic activity of this enzyme (Jacober-Pivarnik and Rand, 1984), it would appear that it plays no role, either activation or inhibition, on the transgalactosidase activity of K. Zactis lactase under the experimental conditions of this study. Vaheri and Kauppinen (1978) reported on the effects of ions on K. fragilis activity for the formation of lactulose. Incorporation of magnesium, potassium, or manganese ions increased lactulose formation more than lactose hydrolysis. Calcium had a

TABLE XX EFFECTS OF POTASSIUM AND MAGNESIUM IONS

ON THE TRANSGALAIXOSYLASE-CATALYZED

PRODUmION OF ALLOLAIXOSE AFTER A 5- MINUTE REACTION TIME

Relative activity (8)

Added KCl ( M ) With Mg Without Mg

0 100" 71.2* 0.05 94.1 63.7 0.106 100 70.5 0.502 106 63.5

Assay-standard comparison using 2.5% lactose, 2.5% glucose, 2.5% galactose, and 17.5 hydrolytic units of lactase.

Magnesium added as MgC12 at 0.001 M.

COMMERCIAL p-GALACTOSIDASE IN FOOD PROCESSING 83

S

8 -

r 7 ,

h

P 6 -

s 6 5 3 -

v

i E 5:

8 4 -

'B a 2 -

1 -

0

slight inhibitory effect, and cadmium profoundly inhibited the transfer reaction.

Ions are not the only possible source of lactase inhibition and activation. Pivarnik and Rand (1991) showed that galactose inhibition, common in hydrolytic reactions to varying degrees, does not seem to affect the transgalactosylase activity of K. luctis lactase for the production of allolactose, when a significant quantity of glucose is added during lactose hydrolysis (Fig. 29).

These studies illustrate that the transgalactosylase activities of all p- galactosidases must be evaluated separately under all unique conditions of desired product formation. Maximum transgalactosylase activity can be achieved only if the proper P-galactosidase is chosen for a specific purpose.

e

8 8

I I I 1

0 . 0 1 .o 2 . 0 3.0 4.0

8 2.5KGlucore 0 5%Glucow

(Both wlth 2.5% Lactose)

Galactose Concentration In Assay (YO)

FIG. 29. Effect of changes in glucose and galactose concentrations on allolactose production catalyzed by K. luctis lactase. Presented at the 1991 Annual Institute of Food Technologists Meeting, Dallas.


D. ACTIVITY IN FOOD

Researchers have demonstrated that formation of high concentrations of oligosaccharides during lactose hydrolysis can cause intestinal discomfort in healthy humans. Unlike many microbial lactases, human small intestine P-galactosidase has strict substrate specificity and has been shown to have limited hydrolytic activity on galactosyl products with p(1-6) linkages (Bur- vall etal., 1980). Therefore, most galactosyl transfer products, formed during lactose hydrolysis, would pass through the small intestine undigested and into the large bowel, where bacterial degradation would result in gastrointestinal discomfort (Zarate and Lbpez-Leiva, 1990). Although small amounts of transgalactosylation products would form in lactose-hydrolyzed milk, incomplete digestion by individuals, particularly those already com- promised by lactase deficiency, could result in symptoms of lactose intolerance. Furthermore, lactose hydrolysis in concentrated milk or whey products would result in a much greater production of oligosaccharides and a greater possibility of intestinal distress (Burvall et al., 1979). For these reasons, the presence of undesirable galactosyl transfer products in dairy products should be evaluated.

Commercially available, low-lactose milk or whey products have not been specifically studied with respect to the degree of oligosaccharides formed; however, other dairy products have been evaluated. Although microbial p-galactosidase activity in yogurt (Streptococcus thermophilus and Lactobacillus bulgaricus) has been shown to enhance lactose digestibility in lactose-intolerant individuals (Kolars et al., 1984; Lin et al., 1989; Onwulata et al., 1989), these lactases have been demonstrated to form oligosaccharides (Toba et al., 1981). Galactosyl transfer reactions of p-galactosidase from S. thermophilus have been shown to be significant (Greenberg and Maho- ney, 1983; Smart, 1991), with both S. thermophilus and L. bulgaricus lactases producing the highest yield of galactooligosaccharides compared with other p-galactosidase-generating bacteria (Kobayasi et al., 1990). Allolactose and galactobiose were first isolated from commercial yogurts by Toba et al. (1982). The concentrations were low, ranging from 0.03 to 0.09%. The investigators concluded that these compounds were formed by p- galactosidase from the lactic acid bacteria. The oligosaccharide content of yogurt has been shown to be affected by storage and fermentation time. The amount of transgalactosylic products formed in yogurts in a 4-hour incubation period was higher than the amount formed in a 6- or 10-hour incubation period (Toba et al., 1983). When low-lactose yogurt was manufactured, the amounts of saccharides formed due to galactosyl transfer were 4 to 19 times higher than the amount obtained with a control yogurt (0.3-1.7%) (Toba et al., 1986). Therefore, transgalactosylase activity can


be found in other fermented dairy products, particularly when additional lactase is added to the culture to form a low-lactose product.

E. FUTURE UTILIZATION

The importance of galactosides in biological processes, therapeutic applications, and food processes has dictated that simple, inexpensive, and rapid methods of synthesis be evaluated and optimized. Research has shown that utilization of p-galactosidase for synthesis of novel oligosaccharides and glycosides has enormous potential; however, the feasibility of synthesis via transgalactosidation requires more information and study. Characterization and control of galactosyl (and glycosyl) transfer reactions have received attention because of the potential applications of the oligosaccharides that can be produced. Glycoproteins, which contain sugar residues with unique characteristics, are involved in the determination of blood group type, play a role in bacterial and viral infections, are responsible for cell-cell recognition, and are involved in cell receptor sites (Cantacuzene et al., 1991). Galactosides can be used as food additives for a variety of functional purposes, such as humectants, sweeteners, stabilizers, and solubilizers.

Of particular interest is the formation of galactosyl oligosaccharides for the production of Bifidobacterium growth promoters. Although the health benefits of Bifidobacterium spp. are still being questioned and studied (Hoo- ver, 1993), transgalactosidation products of p-galactosidases are perceived as potential growth promoters for BiJdobacterium in the lower intestine of infants (Smart, 1991). Bifidogenic growth factors include the oligosaccharides that are the major components of human breast milk and have been known to accelerate the growth of Bifidobacterium (Garza et al., 1987; Kobayashi et al., 1990). Specifically, p-galactosyl (1-6) lactose, lactulose, and amine-containing galactosides have been shown to stimulate certain Bifidobacterium stains (Monsan et al., 1989; Modler et al., 1990). Other researchers have shown that many galactosyl oligosaccharides are used by bifidus flora (Zarate and L6pez-Leiva, 1990). Dumortier etal. (1994) showed that p-galactosidase from B. biJdium exhibited transgalactosylase activity that resulted in 10 different galactooligosaccharides in the presence of lactose. The tri-, tetra-, and pentasaccharides had predominant linkages of p(1-3) and maximum transgalactosylase activity accompanied a 60% lactose conversion. Dumortier and co-workers determined the effects of metals and pH and found the location of transgalactosylase activity to be in the outer membrane region of the cells, indicating the importance of the presence of galactooligosaccharides to the growth of the organism. In addition, lactose derivatives formed through transgalactosylase catalysis not only have enormous future applications in the production of a variety of useful


products, but could take advantage of the large lactose waste problem that accompanies whey disposal.

1. Food and Other Applications

Future food, pharmaceutical, and medical applications of the transgalactosidation process lie in the synthesis of unique and functional compounds, in a controlled manner, through the addition of donor and acceptor molecules. Tables XXI and XXII illustrate the current status of production of novel galactooligosaccharides for potential practical applications. Patent requests have been associated primarily with the formation of galactosides for the promotion of B$dobacterium growth, except for that of Nilsson (1989), who directs the use of galactooligosaccharide formation to therapeutic and diagnostic uses. Table XXII lists the wide variety of compounds synthesized by the transgalactosylase mechanism of P-galactosidases of different origins, not all commercially available and/or food grade. Al- though all the lactases illustrated are not acceptable for use, this review has shown that P-galactosidases from a variety of microbial sources can accomplish similar tasks given the correct experimental conditions. There- fore, all these products could have enormous potential applications if appropriate enzyme sources are studied.

2. Future Enzyme Sources

In addition to commercially available, food-grade lactases from yeasts and molds, P-galactosidases from other food-grade microbial sources could have potential uses. S. thermophilus and L. bulgaricus are both potential sources of lactase that have significant transgalactosylase activity but have not been extensively studied. Bacillus subtilis KL88, a psychotrophic, food- grade bacteria, has been shown to form galactooligosaccharides in dairy products (Rahim and Lee, 1991) and has potential as a future source of P- galactosidase for commercial applications of transgalactosidase activity. Of particular interest are the hydrolase and transgalactosylase activities associated with P-galactosidase derived from Bacillus circulans. B. circulans lactase has particularly strong transgalactosylase activity and wide acceptor specificity (Kitahata et al., 1991). In instances where commercially available, food-grade P-galactosidases had limited or no transgalactosylase activity for the production of a novel compound (Le., branched cyclodextrins, Kita- hata et al., 1992), B. circulans P-galactosidases catalyzed the synthesis of a significant amount of galactosyl transfer products. Mozaffar and co-workers (1984, 1985, 1986, 1987, and 1989) have been studying P-galactosidase-1 and P-galactosidase-2 derived from B. circulans in the free and immobi-

TABLE XXI PATENTS USING P-GALACTOSIDASE ACTIVITY FOR PRODUCTION OF DESIRABLE GALACTOOLIGOSACCHARDES FOR USE IN FOOD PRODUCTS

Inventor Year Title Country Patent No.

Kobayashi et al. 1988 Method for producing milk containing galactooligosaccharide European 88312328.3 Okonogi et al. 1988 Enzymatic manufacture of P-~-galactopyranosyl-( 1-6)-p-~- Japan JP 6394,987

Nilsson 1989 Enzymatic synthesis of complex oligosaccharides International WO 8909,275 Kobayashi et al. 1990 Method for producing milk containing galactooligosaccharide United States 4, 944, 952 Oki et al. 1990 Vinegar containing galactooligosaccharides Japan JP 04218,362

galactopyranosyl-( 1-4)-~-fructose from lactulose

TABLE XXII COMPOUNDS FORMED BY TRANSGALACTOSYLASE ACTIVITY OF P-GALACTOSIDASES

Compound Potential use Enzyme source Reference"

Lactulose (4-0-fl-~-galactopyranosyl-o- fructose)

6-O-fl-~-galactopyranosyl-2-acetoamido- 2-deoxy-~-galactose

Glycosides of gitoxigenin Isoraffinose (6-0-fl-galactosyl sucrose) Monoacyl galactoglycerides G@1-3GlcNAc G@1-3GkNAcfl-Set

Rubusoside derivatives

8-Galactosyl-serine 8-Galact osyl-xylopyranosides

4-O-~-galactosyl-maltopentose Galactosyl branched cyclodextrins

8-Galactosyl-serine

Lactitol oligosaccharides

Humectant Bifidobacterium growth promoter

Cell surface receptor

Cardiac glycosides Bifidobacterium growth promoter Surfactants Bifidobacterium growth promoter Blood group determinant

Sweetner

Glycoprotein component I n vivo evaluation of intestinal

Bifidobacterium growth promoter Emulsifier, stabilizer, masking

odors, production of powders from viscous or oily compounds

Glycoprotein component

Bifidobacterium growth promoter

lactase activity

K. fragilis A. niger K. lactis A. oryzae E. coli

A. oryzae E. coli E. coli Bovine testes E. coli Bovine testes A. oryzae B. circulans P. multicolor E. coli E. coli E. coli A. oryzae B. circulans A. oryzae B. circulans P. multicolor

A. niger E. coli A. oryzae

Vaheri and Kauppinen (1978)

Hedbys et al. (1984)

Ooi et al. (1984) Suyama et al. (1986) Bjorkling and Gotfredson (1988) Hedbys et aL (1989) Hedbys et al. (1989)

Kitahata et al. (1989)

Cantacuzene et aL (1991) Lopez et al. (1991) Lopez and Fernandez-Mayoralas (1992) Kitahata et aL (1991) Kitahata et al. (1992)

Sauerbrei and Thiem (1992)

Yanahira et al. (1992)


lized forms and have shown significant transgalactosylase activity even at low lactose concentrations. Experimental conditions optimized for P-galactosidase from one microbial source may not accommodate another. The potential future use of transgalactosidation processes must include evaluation of any feasible source of microbial P-galactosidase under a variety of experimental conditions.

IV. SUMMARY AND RESEARCH NEEDS

The functionality and stability of commercial and potential commercial sources of P-galactosidases for hydrolytic processes must be improved to use the enzymes under conditions that are safe from microbial contamination. To realize this goal, research is required to develop lactases that are resistant to normal processing temperatures, such as that of pasteurization. The continuing development of new microencapsulation technologies could result in a variety of heat-resistant enzymes and more research is needed. Encapsulated materials and geometries should be investigated for effects on lactase functionality as well as increased thermal stability. We may not be that far from realization of the goal. Finnie (1980) demonstrated that A. niger lactase has a potential application in HTST pasteurization of milk with 30% recovery. Studies investigating technologies for microencapsulation of lactases from thermostable microorganisms, such as A. niger, could result in increased enzyme stability during HTST processing of milk products. Research into genetic alterations or combinations of enzyme characteristics may also improve the functionality and stability of lactases. Genetic cloning of genes from molds into yeasts could result in a P-galactosidase that is stable at traditional processing temperatures yet functional at the pH characteristic of dairy products.

The transgalactosylase activity of lactases has enormous potential for conversion of lactose into useful products; however, because of the extreme diversity of microbial lactase activities, each commercially viable lactase must be evaluated individually, under a variety of experimental conditions, if optimum production of novel compounds is to be achieved. Other food- grade, microbial P-galactosidases, not usually used for their lactase capabilities, may have applications in this area and should be examined. Further- more, if theses future procedures are to be reproducible, methods for assaying transgalactosylase activity must be optimized and used routinely. Research into methods to increase specificity through modification of substrate molecules should be expanded. Low-lactose milk and whey products should be analyzed to quantitate the extent of transgalactosidase reaction products present and to ascertain their effect on the digestion of these


commodities. Finally, the potential therapeutic, food product, and medical and pharmaceutical applications are almost unlimited; however, selectivity and isomeric control must be achieved to make transgalactosidation reactions an alternative to other mechanisms of synthesis.

ACKNOWLEDGMENTS

Contribution No. 2658 of the College of Resource Development, University of Rhode Island, with support from the Rhode Island Agricultural Experiment Station. The research was also partially supported by the Shadow Medical Research Foundation, Sigma Xi, and the US. Army Soldier Systems Command, Natick Research, Development, and Engineer- ing Center.

REFERENCES

Agrawal, S., Garg, S. K., and Dutta, S. M. 1989. Microbial P-galactosidase production, properties and industrial applications-A review. Indian J. Dairy Sci. 42(2), 251-262.

Ajisaka, K., Nishida, H., and Fujimoto, H. 1987a. The synthesis of oligosaccharides by the reversed hydrolysis reaction of beta-glucosidase at high substrate concentration and at high temperature. Biotechnol. Lett. 9(4), 243.

Ajisaka, K., Nishida, H., and Fujimoto, H. 1987b. Use of an activated carbon column for the synthesis of disaccharides by use of reversed hydrolysis activity of beta-galactosidase. Biotechnol. Len. 9(6), 387.

Ajisaka, K., Fujimoto, H., and Nishida, H. 1988. Enzymic synthesis of disaccharides by use of the reversed hydrolysis activity of beta-D-galactosidases. Carbohydr. Res. 180,35.

Alagialis, A., Cosgrove, C. J., and Rand, A. G. 1990. Formulation of premium-style ice milk, low in lactose, sugar and fat. Inst. Food Technol. Annu. Meet., Program Abstr. 51, p. 280.

Allen, A. H. 1993. Dairy-based ingredients: A whey to enhance function. Food Prod. Des.

Allen, L. H., and Wood, R. J. 1994. Calcium and phosphorus. In “Modern Nutrition in Health

Anonymous. 1977. Accelerated cheesemaking. Dairy & Ice Cream Field, March, pp. 66-6611. Aronson, M. 1952. Transgalactosidation during lactose hydrolysis. Arch. Biochem. Biophys.

39,370. Asp, N. G., Burvall, A., Dahlqvist, A., Hallgren, P., and Lundblad, A. 1980. Oligosaccharide

formation during hydrolysis of lactose with Saccharomyces luctis lactase (Maxilact): Part 2. Oligosaccharide structures. Food Chem. 4(4), 147.

Baer, R. J., Frank, J. F., and Morrison, L. 1980. Freezing point measurement of lactose hydrolysis in acid whey and lactose solutions. J. Assoc. Of$ Anal. Chem. 63,587-590.

Bakken, A. P., Hill, C., Jr., and Amundson, C. H. 1990. Use of novel immobilized P-galactosidase reactor to hydrolyze the lactose constituent of skim milk. Biotechnol. Bioeng. 36,

Bakken, A. P., Charles, G. H., Jr., and Amundson, C. H. 1991. Use of novel immobilized p- galactosidase reactor to hydrolyze the lactose constituent of skim milk. 11. Multiresponse nonlinear regression analysis of kinetic data. Appl. Biochem. Bioeng. 2W29.741-756.

Bakken, A. P., Charles, G. H., Jr., and Amundson, C. H. 1992. Use of novel immobilized p- galactosidase (Bacillus circulans). Biotechnol. Bioeng. 39,408-417.

3(7), 27-42.

and Disease,” 8th ed. Lea & Febiger, Philadelphia.

293-309.


Batsalova, K., Kunchev, K., Popova, Y., Kozhukharova, A., and Kirova, N. 1987. Hydrolysis of lactose by P-galactosidase immobilized by polyvinylalcohol. Appl. Microbiol. Biotech- nol. 26,291-296.

Bell, L., Halstenson, C. E., Halstenson, C. J., Macres, M., and Keane, W. F. 1992. Cholesterol- lowering effects of calcium carbonate in patients with mild to moderate hypercholesterol- emia. Arch. Intern. Med. 152, 2441-2444.

Bigelis, R. 1993. Carbohydrates. In “Enzymes in Food Processing” (T. Nagodawithana and G. Reed, eds.), 3rd ed. Academic Press, San Diego, CA.

Birge, S. J., Keutmann, H. T., Cuatrecasas, P., and Whedon, G. D. 1967. Osteoporosis, intestinal lactase deficiency and low dietary calcium intake. N . Engl. J . Med. 276, 44-448.

Bjorkling, F., and Godtfredsen, S. E. 1988. New enzyme catalyzed synthesis of monoacyl galactoglycerides. Tetrahedron 44( lo), 2957.

Bodalo, A., Gomez, E., Gomez, J. L., Bastida, J., Maximo, M. F., and Diaz, F. 1991. A comparison of different methods of beta-galactosidase immobilization. Process Biochem. 26,349-353.

Bostick, R. M., Potter, J. D., Sellers, T. A,, McKenzie, D. R., Kushi, L. H., and Folsom, A. R. 1993. Retention of calcium, vitamin D, and dairy food intake to incidence of colon cancer among older women. Am. J . Epidemiol. 137(12), 1302-1317.

Bouvy, F. A. M. 1974. Whey production and utilization in Europe: Applications of lactase- treated whey and other dairy products. Proc. Whey Prod. Conf: Chicago.

Breslau, B. R., and Kilcullen, B. M. 1975. Hollow fiber enzymatic reactors an engineering approach. Int. Conf: Enzyme Eng., Jrd, Portland OR, pp. 3-8.

Brinkman, G. E. 1976. New ideas for the utilization of lactose-Principles of lactose manufacture. J . SOC. Dairy Technol. 29(2), 101-107.

Bucke, C. 1993. Oligosaccharide synthesis using glycosides. J. Chem. Technol. Biotechnol.

Bunting, P. S. , and Laidler, K. J. 1972. Kinetic studies on solid-supported P-galactosidase. Biochemistry 11, 4477-4483.

Burvall, A., Asp, N. G., and Dalqvist, A. 1979. Oligosaccharide formation during hydrolysis of lactose with Saccharomyces Iactis lactase (Maxilact): Part 1. Quantitative aspects. Food Chem. 4,243.

Burvall, A., Asp, N. G., and Dahlqvist, A. 1980. Oligosaccharide formation during hydrolysis of lactose with Saccharomyces lactis lactase (Maxilact). Part 3: Digestibility by human intestinal enzymes in vitro. Food Chem. 5, 189.

Bushnell, P. J., and Deluca, H. F. 1981. Lactose facilitates the intestinal absorption of lead in weanling rats. Science 211(1), 61-63.

Butler, L. G., Squires, R. G., and Kelly, S. J. 1976. The enzyme-catalyzed synthesis of sucrose from starch. In “Enzyme Technology and Renewable Resources,” NSF/RA/760180. Na- tional Science Foundation, Washington, DC.

Byeong-Seon, C., and Mahoney, R. R. 1994. Thermal denaturation of P-galactosidase from Streptococcus rhermophilus and its stabilization by bovine serum albumin: An electropho- retic study. Biotechnol. Appl. Biochem. 19, 169-178.

Cantacuzene, D., Attal, S., and Bay, S. 1991. Synthesis of (Y and P-glycopyranosyl-serine derivatives by enzymic transglycosylation. Biomed. Biochim. Acta 50,231-236.

Champluvier, B., Kamp, B., and Rouxhet, P. G. 1988a. Preparation and properties of P- galactosidase confined in cells of Kluyveromyces sp. J. Enzyme Microb. Technol. 10,

Champluvier, B., Kamp, B., and Rouxhet, P. G. 1988b. Immobilization of P-galactosidase retained in yeast: Adhesion of the cells on a support. J. Enzyme Microb. Technol. 27,

56,411-412.

611-617.

464-469.


Chang, B.-S., and Mahoney, R. P. 1989. Purification and thermostability of beta-galactosidase (lactase) from an autolytic strain of Streptococcus salivarius subsp. thermophilus. J . Dairy Res. Xi, 117-127.

Chapuy, M. C., Arlot, M. E., Duboeuf, F., Brun, J., Crouzet, B., Arnaud, S., Delmas, P. D., and Meunier, P. J. 1992. Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J . Med. 327,1637-1642.

Charles, M . , Coughlin, R. W., Paruchuri, E. K., Allen, B. R., and Hasselberger, F. X. 1975. Enzymes immobilized on alumina and stainless steel supports. Biotechnol. Bioeng. 17,

Chawan, C. B., Penmetsa, P. K., Veeramachaneni, R., and Rao, D. R. 1993. Liposomal encapsulation of beta-galactosidase: Effect of buffer molarity, lipid composition and stability in milk. J. Food Biochem. 16,349-357.

Cheng, A. H. R., Brunsor, O., Esponoza, J., Fones, H. L., Monckeberg, F., Chichester, C. O., Rand, A. G., and Hourigan, J. A. 1979. Long term acceptance of low-lactose milk. Am. J . Clin. Nutr. 32, 1989-93.

203-210.

Cheryan, M. 1986. “Ultrafiltration Handbook.” Technomic Publ. Co., Lancaster, PA. Cote, G. L., and Tao, B. Y. 1990. Oligosaccharide synthesis by enzymatic transglycosylation.

CRH. 1994. “Calcium: Ultrafiltration Handbook.” Technomic Publ. Co., Lancaster, PA. Cuperus, F. P., and Nijhuis, H. H. 1993. Application of membrane technology to food process-

Dahlqvist, A. 1964. Method for assay of intestinal disaccharides. Anal. Biochem. 7, 18-25. Dahlqvist, A., Matiiasson, A., Matiiasson, B., and Masbach, K. 1973. Hydrolysis of 0-galactosi-

dase using polymer-entrapped lactase. A study towards producing lactose-free milk. Biotechnol. Bioeng. 15,395-402.

Dahlqvist, A., Asp, N. G., Burvall, A., and Rausing, H. 1977. Hydrolysis of lactose in milk and whey with minute amounts of lactase. J. Dairy Res. 44, 541-548.

Dariani, D. N., Frank, J. F., and Loewenstein, M. 1982. Manufacture of low lactose yogurt by simultaneous lactose hydrolysis and bacteriol fermentation. Cult. Dairy Prod. J.

DeAngelis, M. A., Lausier, J. M., and Rand, A. G. 1979. Development of a lactase dose form by microencapsulation. Inst. Food Technol. Annu. Meet., Program Abstr. 39, p. 129.

Denke, M. A., Fox, M. M., and Schulte, M. C. 1993. Short-term dietary calcium fortification increases fecal saturated fat content and reduces serum lipid in men. J. Nutr. 123,

De Vrese, M., Keller, B., and Barth, C. 1992. Enhancement of intestinal hydrolysis of lactose by microbial beta-galactosidase (EC 3.2.1.23) of Kefir. Br. J. Nutr. 67, 67-75.

Dickson, R. C., Dickson, L. R., and Markin, J. S. 1979. Purification and properties of an inducible p-galactosidase isolated from the yeast Kluyveromyces lactis. J. Bacteriol. 12,

Glycoconjugate J. 7, 145.

ing. Trends Food Sci. Technol. 4,277-282.

17(2), 18-22.

1047-1053.

51-61. Dinelli, D. 1972. Fibre-entrapped enzymes. Process Biochem. 7,9-12. Dumortier, V., Brassart, C., and Bouquelet, S. 1994. Purification and properties of p-D-

galactosidase from Bijidobacterium bijidium exhibiting a transgalactosylation reaction. Biotechnol. Appl. Biochem. 19,341-354.

El-Neshawy, A. A., Abdel, Baky, A. A,, Rabie, A. M., and Metwally, S. A. 1988. Organoleptic and physical properties of ice cream made from hydrolysed lactose reconstituted d k . Food Chem. 27,83-93.

Engel, W. 1973. Increasing the sweetness of yogurt without adding calories. Dairy Znd. News 69(10), 10, 16.


Finnie, K. J. 1980. Continuous processing of low lactase milk to minimize lactase product inhibition. Master of Science Thesis, University of Rhode Island, Kingston.

Finnie, K. J., Hourigan, J. A., and Rand, A. G. 1979. Reduction of product inhibition during enzymatic lactose hydrolysis. Inst. Food Technol. Annu. Meet., Program Abstr. 39, p. 101.

Finocchiaro, T., Olson, N. F., and Richardson, R. 1980a. Use of immobilized lactase in milk systems. Adv. Biochem. Eng. l5,71-88.

Fmocchiaro, T., Richardson, T., and Olson, N. F. 1980b. Lactase immobilized on alumina. J. Dairy Sci. 63(2), 215-222.

Forsman, E., Heikonen, M., Kiviniemi, L., Kruela, M., and Linko, P. 1979. Kinetic investigations of the hydrolysis of milk lactose with soluble Kluyveromyces lactis P-galactosidase. Milchuwissenschaft 34,618-621.

Gallagher, C. R., Molleson, A. L., and Caldwell, J. H. 1974. Lactose intolerance and fermented dairy products. J. Am. Diet. Assoc. 65, 418-419.

Garza, C., Schlanler, R. J., Butte, R. J., and Motil, K. J. 1987. Special properties of human milk. In “Clinics in Perinatology” (R. Lawrence, ed.), pp. 11-31. Saunders, Philadelphia.

Gekas, V., and L6pez-Leiva, M. H. 1985. Hydrolysis of lactose: A literature review. Process Biochem. 20,2-12.

Giec, A., Gruchala, L., and Wasowicz, E. 1981. Formation of oligosaccharides during lactose hydrolysis with beta-galactosidase in cheese whey permeates. Acta Aliment. Pol. 7,

Gilliland, S. E. 1985. Influence of bacterial starter cultures on nutritional value of foods: Improvement of lactose digestion by consuming foods containing lactobacilli. Cult. Dairy Prod. J. M(2), 28-33.

Glass, R. W., and Rand, A. G. 1982. Alginate immobilization of banana pulp enzymes for starch hydrolysis and sucrose interconversion. J. Food Sci. 47(6), 1836-1839.

Gonzalez, R. R., and Monsan, P. 1991. Purification and some characteristics from Aspergillus fonsecaeus. Enzyme Microb. Technol. 13,349-352.

Goodenough, E. R., and Kleyn, D. H. 1976a. Qualitative and quantitative changes in carbohydrates during the manufacture of yogurt. J . Dairy Sci. 59,45-47.

Goodenough, E. R., and Kleyn, D. H. 1976b. Influence of viable yogurt microflora on digestion of lactose by the rat. J. Dairy Sci. 59, 601-606.

Gray, G. M. 1971. Intestinal digestion and maldigestion of dietary carbohydrates. Annu. Rev. Med. 22,391-404.

Gray, G. M. 1975. Carbohydrate digestion and absorption. N. Engl. J. Med. 292(23), 1225-1230. Greenberg, N. A., and Mahoney, R. R. 1981a. Immobilization of lactase (beta-galactosidase)

for use in dairy processing: A review. Process Eiochem. 2,2-8,49. Greenberg, N. A., and Mahoney, R. R. 1981b. Rapid purification of beta-galactosidase (Asper-

gillus niger) from commercial preparation. J. Food Sci. 46, 684. Greenberg, N. A., and Mahoney, R. R. 1983. Formation of oligosaccharides by P-galactosidase

from Streptococcus thermophilus. Food Chem. 10,195-204. Greenberg, N. A., Wilder, T., and Mahoney, R. R. 1985. Studies on the thermostability of

lactase (Streptococcus thermophilus) in milk and sweet whey. J. Dairy Res. 52(3), 439-449. Griffith, L. S. 1991. Ce(IV) analysis of carbohydrates. 11. Transgalactosylation products of

beta-galactosidase from Maxilact. In “Dissertation Abstracts International.” University of Alberta, Canada.

Griffiths, M. W., and Muir, D. D. 1978. Properties of a thermostable beta-galactosidase from a thermophilic bacillus: Comparison of the enzyme activity of whole cells, purified enzyme and immobilized whole cell. J . Sci. Food Agric. 29, 753-761.

Guy, E. J., and Bingham, E. W. 1978. Properties of P-galactosidase of Saccharomyces lactis in milk and milk products. J. Dairy Sci. 61, 147.

199-208.


Hahn-Hagerdal, B., Nilsson, M., and Anderson, E. 1987. Synthesis of oligosaccharides and peptides using hydrolytic enzymes at decreased water activity. Ann. N.Y. Acad. Sci 50,193.

Heaney, R. P. 1993. Thinking straight about calcium. N. Engl. J . Med. 328(7), 503-505. Heaney, R. P., and Barger-Lux, M. J. 1994. Low calcium intake: The culprit in many chronic

diseases. J. Dairy Sci. 77, 1155-1160. Hedbys, L., Larsson, P.-O., and Mosbach, K. 1984. Synthesis of the disaccharide 6-O-beta-~-

galactopyranosy~-2-acetamido-2-deoxy-~-ga~actose using immobilized beta-galactosidase. Biochem. Biophys. Res. Commun. 8,l.

Hedbys, L., Johansson, E., Mosbach, K., Larsson, P.-O., Gunnarsson, A., Svensson, S., and Lonn, H. 1989. Synthesis of Galbetal-3GlcNAc and Galbetal-3GlcNAcbeta-SEt by an enzymatic method comprising the sequential use of beta-galactosidases from bovine testes and Escherichia coli. Glycoconjugate J. 6,161.

Hilgendorf, M. J. 1981. Optimization of fungal lactase levels in yogurt manufacture. Cult. Dairy Prod. J. 16(1), 5-7.

Holsinger, V. H., and Kligerman, A. E. 1991. Applications of lactase in dairy foods and other foods containing lactose. Food Technol. 45(1), 92-95.

Hoover, D. G. 1993. Bifidobacteria: Activity and potential benefits. Food Technol. 47(6),

Hourigan, J. A. 1976. Kinetic studies of enzymatic hydrolysis of lactose. Ph.D. Thesis, Univer-

Hourigan, J. A., and Rand, A. G. 1977. The production and use of low-lactose milk. Proc.

Houts, S . S . 1988. Lactose intolerance. Food Technol. 42(3), 110-113. Huang, J., and Rosenberg, M. 1993. Lactose reduced chocolate milk-Lactose hydrolysis and

oligosaccharide formation. Znst. Food Technol. Book Abstr. Huber, R. E., Kurz, G., and Wallenfels, K. 1976. A quantitation of the factors which affect

the hydrolase and transgalactosylase activities of beta-galactosidase (E. coli) on lactose. Biochemistry l5(9), 1994-2001.

Huffman-Reichenbach, L. M., and Harper, J. W. 1982. Beta-galactosidase retention by hollow fiber membranes. J. Dairy Sci. 65, 887-898.

Hussein, L., Elsayed, S., and Foda, S. 1989. Reduction of lactose in milk by purified lactase produced by Kluyveromyces lactis. J. Food Prot. 52,30-34.

Irazoqui, G., and Batista-Viera, F. 1994. Immobilization of P-galactosidase (K. lactis) on solid phase Ni( 11)-chelate. Biotechnol. Tech. 8(4), 233-238.

Itoh, K., Toba, T., Itoh, T., and Adachi, S. 1992. Properties of beta-galactosidase of Lactobacil- lus kefiranofaciens K-1 isolated from kefir grains. Lett. Appl. Microbiol. 15,232-234.

Jackson, E. H., and Jelen, P. 1989. Batch hydrolysis of lactose in concentrated whey systems. J. Food Sci. 54(4), 1086-1087.

Jacober-Pivarnik, L., and Rank, A. G., Jr. 1984. Use of a milk assay to evaluate the effects of potassium on commercial yeast lactases. J. Food Sci. 49, 435.

Jeon, I. J., and Mantha, V. R. 1985. High performance liquid chromatography analysis of oligosaccharides formed during beta-galactosidase action on lactose. J. Dairy Sci. 68,581.

Jeon, I. J., and Saunders, S. R. 1986. Effect of oligosaccharide formation on the cryoscopic measurements of enzymatic hydrolysis of lactose of dairy products. J. Food Sci. 51(l), 245.

Jobe, A,, and Bourgeois, S. 1972. lac repressor-operator interaction. VI. The natural inducer of the lac operon. J. Mol. Biol. 69, 397.

Jonas, J. J. 1973. Utilization of dairy ingredients in other foods. J. Milk Food Technol.

Jones, C . K. S., Yang, R. Y. K., and White, E. T. 1988. A novel hollow-fiber reactor with

120-124.

sity of Rhode Island, Kingston.

Nutr. SOC. Aust. 2, 72.

36(6), 323-332.

reversible immobilization of lactase. AZChE J. 34,293-304.


Jost, R. 1993. Functional characteristics of dairy proteins. Trends Food Sci. Technol. 4(9),

Keller, S. E., Fellows, J. W., Nash, T. C., and Shazer, W. H. 1991. Use of lactase enzyme in bulking-agent-free frozen dessert sweetened with aspartame. Ins?. Food Technol. Annu. Meet., Program Abstr. 52, p. 352.

Kelley, L. L. 1984. Lactose intolerance-Could yogurt be the answer? Cult. Dairy Prod. J.

Kierston, M. 1981. The use of calcium alginate gel for solids separation and diffusional chromatography of biological materials. Biotechnol. Bioeng. 23(4), 707.

Kilara, A., and Shahani, K. M. 1974. P-Galactosidase activity of cultured and acidified dairy products. J . Dairy Sci. 57, 592.

Klibanov, A. M. 1983. Stabilization of enzymes against thermal inactivation. Adv. Appl. Microbiol. 29, 1-25.

Kitahata, S., Ishikawa, H., Miyata, T., and Tanaka, 0. 1989. Production of rubusoside derivatives by transgalactosylation of various P-galactosidases. Agric. Biol. Chem. 53(11),

Kitahata, S., Fujita, K., Hara, K., Hara, K., and Hashimoto, H. 1991. Enzymic synthesis of 4-0-P-galactosyl-maltopentose by Bacillus circulans P-galactosidase. Agric. Biol. Chem.

Kitahata, S., Fujita, K., Takagi, Y., Hara, K., Hashimoto, H., Tanimoto, T., and Koizumi, K. 1992. Galactosylation at side chains of branched cyclodextrins of various P-galactosides. Biosci. Biotechnol. Biochem. 56(2), 242-245.

Kligerman, A. E. 1981. Development of lactose-reduced products. In “Lactose Digestion: Clinical and Nutritional Implications” (D. M. Paige and T. M. Bayless, eds.), pp. 252-256. Johns Hopkins Univ. Press, Baltimore.

Kobayashi, A., Kawai, S., Ohbe, Y., and Nagashima, Y. 1975. Effects of dietary lactose and lactase preparation on the intestinal absorption of calcium and magnesium in normal infants. Am. J . Clin. Nutr. 28, 681-683.

Kobayashi, Y., Kan, T., and Terashima, T. 1988. Method for producing processed milk containing galactooligosaccharides. Eur. Pat. Appl. 88312,328.3.

Kobayashi, Y., Kan, T., and Terashima, T. 1990. Method for producing processed milk containing galactooligosaccharide. US. Pat. 4,944,952.

Kolars, J. C., Levitt, M. D., Aouji, M., and Savaiano, D. A. 1984. Yogurt-an autodigesting source of lactose. New Engl. J . Med. 310(1), 1-3.

Kosikowski, F. V., and Wierzbicki, L. E. 1973. Lactose hydrolysis of raw and pasteurized milks by Saccharomyces lactis lactase. J. Dairy Sci. 56,146-148.

Kretchmer, N. 1972. Lactose and lactase. Sci. Am. 227,70-78. Kuntz, L. A. 1993. Dairy enzymes: The biochemical benefit. Food Prod. Des. 3(7), 59-72. Kwak, H. S., and Jeon, I. J. 1986. Effect of various conditions on the formation of oligosaccha-

rides in milk treated with beta-galactosidase. J. Dairy Sci. 69,2785. Lin, W., Savaiano, D. A., and Harlander, S. K. 1989. A method for determining P-galactosidase

activity of yogurt cultures in skim milk. J . Dairy Sci. 72, 351-359. Lopez, R., and Fernandez-Mayoralas, A. 1992. Controlling yield and regioselectivity in enzy-

matic synthesis of P-D-ga~actopyranosyl-~-D-xy~opyranosides. Tetrahedron Lett. 33(37),

Lopez, R., Fernandez-Mayoralas, A., and Martin-Lomas, M. 1991. EnzymaticP-galactosidation of 8-xylopyranosides. Biotechnol. Lett. 13( lo), 705-710.

Loveland, J., Gutshall, K., Kasmir, J., Prema, P., and Brenchley, J. E. 1994. Characterization of psychotrophic microorganisms producing beta-galactosidase activities. Appl. Environ. Microbiol. 60(1), 12-18.

283-288.

19,(2), 12-14.

2923-2928.

55,(9), 2433-2434.

5449-5452.


Lutwak, L. 1975. Metabolic and biochemical considerations of bone. Ann. Clin. Lab. Sci.

Maculan, T. 1979. “Ultrafiltration Membrane Immobilization of Lactase and the Processing of Low Lactose Milk.” Master’s thesis, University of Rhode Island, Kingston, RI.

Maculan, T. P., Hourigan, J. A., and Rand, A. G. 1978. Applications of a localized enzyme membrane reactor for the hydrolysis of lactose. J. Dairy Sci. 61, Suppl. 1, 114.

Mahoney, R. R. 1985. Modification of lactose and lactose-containing dairy products with beta- galactosidase. In “Developments in Dairy Chemistry: Lactose and Minor Constituents” (F. Fox, ed.), pp. 69-108. Elsevier, New York.

Mahoney, R. R., and Adamchuk, C. 1980. Effect of milk constituents on the hydrolysis of lactose by lactase from Kluyveromyces fragilis. J. Food Sci. 45, 962-968.

Mahoney, R. R., and Whitaker, J. R. 1977. Stability and enzymatic properties of beta-galactosidase from Kluyveromyces fragilis. J. Food Biochem. 1, 327-350.

Mahoney, R. R., and Wilder, T. 1988. Thermostability of yeast lactase (Kluyveromyces marxianus) in milk. J. Dairy Rex 55, 423-433.

Mahoney, R. R., and Wilder, T. 1989. Stabilization of lactase (Escherichia coli) by milk components and related compounds. J. Food Sci. 54,899-901.

Marteau, P., Flourie, B., Pochart, P., Chastang, C., Desjeux, J.-F., and Rambaud, J.-C. 1990. Effect of the microbial lactase (EC3.2.1.23) activity in yoghurt on the intestinal absorption of lactose: An in vivo study in lactase-deficient humans. Br. J. Nutr. 64,7179.

Matsuda, T., and Nakamura, R. 1993. Molecular structure and immunological properties of food allergens. Trends Food Sci. Technol. 4(9), 289-293.

McCormick, R. D. 1976. A nutritious alternative for the lactose-intolerant consumer. Food Prod. Dev. 6, 17-18.

McDonough, F. E. 1977. Whey solids utilization and salvage systems. Cult. Dairy Prod. J.

Miller, J. J., and Brand, J. C. 1980. Enzymatic lactose hydrolysis. Food Technol. Aust. 32(3). Modler, H. W., McKellar, R. C., and Yaguchi, M. 1990. Bifidobacteria and bifidogenic factors.

Can. Inst. Food Sci. Technol. J. 23(1), 29-41. Monsan, P., Paul, F., Remaud, M., and Lopez, A. 1989. Novel enzymatic synthesis of oligosac-

charides. Food Biotechnol. 3(1), 11. Moore, K. 1980. Immobilized enzyme technology commercially hydrolyzes lactose. Food Prod.

Dev. 14(1), 50-54. Morisi, F., Pastore, M., and Viglia, A. 1973. Reduction of lactose content of milk by entrapped

P-galactosidase. I. Characteristics of P-galactosidase from yeast and Escherichia coli. J. Dairy Sci. 56(9), 1123-1127.

Mozaffar, Z., Kazuhiro, N., Matsuno, R., and Kamikubo, T. 1984. Purification and properties of beta-galactosidases from Bacillus circulans. Biol. Chem. 48(12), 3053.

Mozaffar, Z., Nakanishi, K., and Matsuno, R. 1985. Formation of oligosaccharides during hydrolysis of lactose in milk using beta-galactosidase from Bacillus circulans. J. Food Sci. 50, 1602.

Mozaffar, Z., Nakanishi, K., and Matsuno, R. 1986. Continuous production of galacto-oligosaccharides from lactose using immobilized beta-galactosidase from Bacillus circulans. Appl. Microbiol. Biotechnol. 25, 224.

Mozaffar, Z., Nakanishi, K., and Matsuno, R. 1987. Effect of glutaraldehyde on oligosaccharide production by beta-galactosidase from Bacillus circulans. Appl. Microbiol. Biotechnol. 25, 426.

Mozaffar, Z., Nakanishi, K., and Matsuno, R. 1989. Production of trisaccharide from lactose using beta-galactosidase from Bacillus circulans modified by glutaraldehyde. Appl. Micro- biol. Biotechnol. 31, 59.

5,(3), 185-194.

U(1), 8-11.


Nakanishi, K., Matsuno, R., Torii, K., Yamamoto, K., and Kamikubo, T. 1983. Properties of immobilized beta-D-galactosidase from Bacillus circulans. Enzyme Microb. Technol. 5, 115.

National Academy of Sciences (NAS) 1989. Minerals. i n “Diet and Health: Implications for Reducing Chronic Disease Risk.” National Academy Press, Washington, DC.

NDC 1977. The role of dairy foods in the diet. Dairy Counc. Dig. 48(3), 13-18. Nelson, J. H. 1980. Enzyme producers, food processors, and the regulators. In “Enzymes:

The Interface Between Technology and Economics” (J. P. Danehy and B. Wolnak, eds.). Dekker, New York.

Newcomer, A. D. 1981. Immediate symptomatic and long-term nutritional consequences of hypolactasia. In “Lactose Digestion: Clinical and Nutritional Implication” (D. M. Paige and T. M. Bayless, eds.), pp. 124-133. Johns Hopkins Univ. Press, Baltimore.

Newton, S. 1991. Lactose leaps into non-food markets. Dairy Foods 92(4), 93-94. Ngo, T. T., Narinesingh, D., and Laidler, K. J. 1976. Hydrolysis of D-galactosides in an open

tubular lactase reactor. Biofechnol. Bioeng. 18, 119-127. Nickerson, T. A., Vujicic, I. F., and Lin, A. Y. 1976. Colorimetric estimation of lactose and

its hydrolytic products. J. Dairy Sci. 59(3), 386-390. Nijpels, H. H. 1981. Enzymes in food processing. In “Lactases and Their Application”

(G. G. Birch, ed.), pp. 89-104. London. Nilsson, H., Mosbach, R., and Mosbach, K. 1972. The use of bead polymerization of acrylic

monomers for immobilization of enzymes. Biochim. Biophys. Acta 268,253-256. Nilsson, K. G. I. 1987. A simple strategy for changing the regioselectivity of glycoside-catalysed

formation of disaccharides. Carbohydr. Rex 167, 95-103. Nilsson, K. G. I. 1988a. Influence of various parameters on the yield and regioselectivity

of glycosidase-catalyzed formation of oligosaccharide glycosides. Ann. N. Y. Acad, Sci.

Nilsson, K. G. I. 1988b. A simple strategy for changing the regioselectivity of glycoside- catalysed formation of disaccharides. Part 11. Enzymic synthesis in situ of various acceptor glycosides. Carbohydr. Res. 108, 53-59.

Nilsson, K. G. I. 1989. Enzymic synthesis of complex carbohydrates. International Pat. WO 89/09,275.

Nilsson, K. G. I. 1990. Enzymic synthesis of HexNAc-containing disaccharide glycosides. Carbohydr. Res. W,79-83.

Nilsson, K. G. I. 1991. Enzymatic synthesis of complex carbohydrates and their glycosides. in “Applied Biocatalysis” (H. W. Blanch and D. S. Clark, eds.), Vol. I, pp. 117-177. Dekker, New York.

Nilsson, K. G. I., and Fernandez-Mayoralas, A. 1991. a-D-Galactosidase-catalysed synthesis of partially protected a-liked digalactopyranosides. Biofechnol. Lef t 13(10), 715-720.

Nisizawa, K., and Hashimoto, Y. 1970. Glycoside hydrolases and glycosyl transferases. In “The Carbohydrates” (W. Pigman, ed.), p. 241. Academic Press, New York.

Nutrition Reviews. 1979. Clinical nutrition: Metabolic bone disease as a result of lactase deficiency. Nutr. Rev. 37(3), 72-73.

Oki, J., Kobashi, K., Hashimoto, K., and Kubota, A. 1990. Vinegar containing galactooligosaccharides. Jap. Pat. JP 04218,362.

Okonogi, S., Tomita, M., Tomimura, T., Tarnura, Y., and Mizota, T. 1988. Enzymic manufacture of P-D-galactopyranosyl-( l-6)-P-~-galactopyranosyl-( 1 -4)-~-fructose from Iactulose. Jap. Pat. J P 63/94, 987.

Okos, E. S., and Harper, W. J. 1974. Activity and stability of P-galactosidase immobilized on porous glass. J. Food Sci. 39, 88-93.

542,383-389.


Okos, M. R., Grulke, E. A., and Syverson, A. 1978. Hydrolysis of lactose in acid whey using 8-galactosidase adsorbed to a phenol-formaldehyde resin. J. Food Sci. 43,566-571.

Olson, A. C., and Stanley, W. L. 1973. Lactase and other enzymes bound to a phenol-formaldehyde resin with glutaraldehyde. J. Agric. Food Chem. 21.

Olson, N. F. 1979. Acceleration of cheese ripening-potential for hydrolyzing lactose in cheese milk. Dairy & Ice Cream Field, June, pp. 90e-90f.

Onwulata, C. I., Rao, D. R., and Vankineni, P. 1989. Relative efficiency of yogurt, sweet acidophilus milk, hydrolyzed-lactose milk, and commercial lactase tablet in alleviating lactose maldigestion. Am. J. Clin. Nurr. 49, 1233.

Ooi, Y., Hashmoto, T., Mitsuo, N., and Toshio, S. 1984. Enzymic synthesis of chemically unstable cardiac glycosides. Tetrahedron Lett. 25(21), 2241.

Ortega-Lopez, J., Morales-Ramos, L. H., Montes, M. C., and Magana-Plaza, I. 1993. Lactose hydrolysis by immobilized beta-galactosidase on nylon-6-A novel spin-basket reactor. Biotechnol. Tech. 7(11), 775-780.

Ozbaas, T., and Kutsal, T. 1990. Hydrolysis of lactose in whey powder solutions by Aspergillus oryzae lactase. J. Food Sci. Technol. 27,195-197.

Paige, D. M. 1981. Lactose malabsorption in children: Prevalence, symptoms and nutritional considerations. In “Lactose Digestion: Clinical and Nutritional Implications” (D. M. Paige and T. M. Bayless, eds.), pp. 151-161. Johns Hopkins Univ. Press, Baltimore.

Paige, D. M., Bayless, T. M., Huang, S. S., and Wexler, R. 1975. Lactose hydrolyzed milk. Am. J. Clin. Nutr. 28, 818-822.

Palumbo, M. S., Smith, P. W., Strange, E. D., Van Herkken, D. L., Tunick, M. H., and Holsinger, V. H. 1995. Stability of B-galactosidase from Aspergillus oryzae and Kluyveromyces lactis in dry milk powders. J. Food Sci. 60(1), 117-119.

Pansu, D., Bellaton, C., and Bronner, F. 1979. Effect of lactose on duodenal calcium-binding protein and calcium absorption. J. Nutr. 109, 508-512.

Park, Y. K., DeSanti, M. S. S., and Pastore, G. M. 1979. Production and characterization of 8-galactosidase from Aspergillus oryzae. J . Food Sci. 44,100-103.

Patton, S . 1969. Milk Sci. Am. 221, 58-68. Paul, S. C., and Mathur, B. N. 1989. Standardization of processing parameters for enzymic

Pazur, J. H. 1953. The enzymatic conversion of lactose into galactosyl oligosaccharides. Science

Pazur, J. H. 1954. The mechanism of enzymatic synthesis of galactosyl oligosaccharides.

Pazur, J. H., Tipton, C. L., Budovich, T., and Marsh, J. M. 1958. Structural characterization

Phillips, M. C., and Briggs, G. M. 1975. Milk and its role in the American diet. J. Dairy Sci.

Pirisino, J. F. 1983. High performance liquid chromatographic determination of lactose, glucose, and galactose in lactose-reduced milk. J. Food Sci. 48,742-744.

Pivarnik, L. F. 1990. Characterization of hydrolase and transgalactosidase activities of commercial, food-grade yeast lactases. Ph.D. Dissertation, University of Rhode Island, Kingston.

Pivarnik, L. F., and Rand, A. G. 1991. Transgalactosidase activity of Kluyveromyces lactis derived beta-galactosidase. Inst. Food Technol. Annu. Meet., Abstr., p. 579.

Pivarnik, L. F., and Rand, A. G. 1992. Assay conditions effect on P-galactosidase activity from Kluyveromyces laciis. J. Food Sci. 57,1020-1021.

Pomeranz, Y. 1964a. Lactase (galactosidase): I. Occurrence and properties. Food Technol.

hydrolysis of lactose on a pilot scale. Indian J. Dairy Sci. 42,643-644.

117,355.

J. Biol. Chem. 208, 439.

of products of enzymatic disproportion of lactose. J. Am. Chem. SOC. 80, 119.

58,1751-1763.

18(5), 88-93.


Pomeranz, Y. 1964b. Lactase (P-galactosidase): 11. Possibilities in the food industries. Food Technol. 18,96-103.

Prenosil, J. E., Stuker, E., and Bourne, J. R. 1987a. Formation of oligosaccharides during enzymatic lactose. Part 1. State of the art. Biofechnol. Bioeng. 30, 1019.

Prenosil, J. E., Stuker, E., and Bourne, J. R. 1987b. Formation of oligosaccharides during enzymatic lactose hydrolysis and their importance in whey hydrolysis process. Part 2. Experimental. Biofechnol. Bioeng. 30, 1026.

Rahim, K. A., and Lee, B. H. 1991. Specificity, inhibitory studies and oligosaccharide formation by P-galactosidase from psychrotrophic Bacillus subtilis KL88. J. Dairy Sci. 74,1773-1778.

Rand, A. G. 1981. Enzyme technology and the development of lactose-hydrolyzed milk. In “Lactose Digestion: Clinical and Nutritional Implications” (D. M. Paige and T. M. Bayless, eds.), pp. 219-230. Johns Hopkins Univ. Press, Baltimore.

Rand, A. G., and Kleyn, D. H. 1991. Comparison of the functional effects of substitutes for fat and sugar in low lactose ice milk. Inst. Food Technol. Annu. Meet., Program Abstr. 52, p. 705.

Rand, A. G., and Linklater, P. 1973. The use of enzymes for the reduction of lactose levels in milk products. Aust. J . Dairy Technol. 63-67.

Reasoner, J., Maculan, T. P., Rand, A. G., and Thayer, W. R. 1981. Clinical studies with low- lactose milk. Am. J. Clin. Nufr. 34, 54-60.

Reczey, K., Stalbrand, H., Hagerdal-Hahn, B., and Tjerneld, F. 1992. Mycelia-associated p- galactosidase activity in microbial pellets of Aspergillus and Penicillium strains. Appl. Microbiol. 38,393-397.

Richmond, M. L., Gray, J. I., and Stine, C. M. 1981. Beta-galactosidase: Review of recent research related to technological application, nutritional concerns, and immobilization. 64,1759.

Roberts, H. R., and McFarren, E. F. 1953. The chromatographic observation of oligosaccharides formed during the lactase hydrolysis of lactose. J. Dairy Sci. 36, 620.

Roberts, H. R., and Pettinati, J. D. 1957. Concentration effects in the enzymatic conversion of lactose to oligosaccharides. Agric. Food Chem. 5(2), 130.

Rosensweig, N. S. 1978. “A Review of Dietary Lactose and Its Varied Utilization by Man.” The Whey Products Institute, Chicago.

Sauerbrei, B., and Thiem, J. 1992. Galactosylation and glucosylation by use of P-galactosidase. Tetrahedron Lett. 33(2), 201-204.

Savaiano, D. A., and Kotz, C. 1988. Recent advances in the management of lactose intolerance. Contemp. Nutr. 13(9), 10.

Schmid, R. D., 1979. Stabilized soluble enzymes. Adv. Biochem. Eng. l2,41-118. Senecal, A. G. 1991. Fat replacement and lactose reduction processes for production of a

nutritionally improved low fat milk. Ph.D. Thesis, University of Rhode Island, Kingston. Senecal, A. G., and Rand, A. G. 1992a. Stability of beta-galactosidase and immobilization

on an ultrafiltration membrane for the continuous processing of low lactose milk. Proc. Natick Sci. Symp. 4, 277-291.

Senecal, A. G., and Rand, A. G. 1992b. Stability of commercial yeast lactases in skim milk and skim milk ultrafiltration permeate. Inst. Food Technol. Annu. Meet., New Orleans, LA.

Senecal, A. G., and Rand, A. G. 1993. Lactose hydrolysis of milk ultrafiltration permeate with a localized enzyme membrane bioreactor. Inst. Food Technol. Annu. Meet., Chicago.

Shahani, K. M., Mathur, B. N., and Kilara, A. 1978. Utilization of whey as human food. Cult. Dairy Prod. J. l3(2), 7-12.

Sharp, A. K., Kay, G., and Lilly, M. D. 1969. The kinetics of p-galactosidase attached to porous cellulose sheets. Biofechnol. Bioeng. 11,363-380.


Shearin, B. 1977. Marketing experiences with sweet acidophilus low-fat milk. Cult. Dairy Prod. J . l2(2), 26-27.

Sheth, H., Jelen, P., and Shah, N. 1988. Lactose hydrolysis in ultrafiltration-treated cottage cheese whey with various whey protein concentrations. J. Food Sci. 53(3), 746-748.

Shih-Ling, Y. C., Frank, J. F., and Loewenstein, M. 1981. Estimation of lactose hydrolysis by freezing point measurements in milk and whey substrates treated with lactases from various microorganisms. J. Assoc. Off: Anal. Chem. 64(6), 1414-1419.

Short, J. L. 1978. Prospects for the utilization of deproteinated whey in New Zealand-A review. N. Z. J. Dairy Sci. Technol. 13, 181-194.

Shukla, H., and Chaplin, M. 1993. Noncompetitive inhibition of beta-galactosidase (A . oryzae) by galactose. Enzyme Microb. Technol. 15(4), 297-299.

Shukla, T. P. 1975. Beta-galactosidase technology: A solution to the lactose problem. CRC Crit. Rev. Food Technol. 5, 325.

Simoons, F. J. 1981. Geographic patterns of primary adult lactose malabsorption. A further interpretation of evidence for the old world. In ”Lactose Digestion: Clinical and Nutri- tional Implications” (D. M. Paige and T. M. Bayless, eds.), pp. 23-48. Johns Hopkins Univ. Press, Baltimore.

Sinnott, M. L. 1987. Glycosyl group transfer. In “Enzyme Mechanisms” (M. I. Page and A. Williams, eds.), p. 263. Royal Society of Chemistry, London.

Siso, M. I. G., and Doval, S. S. 1994. Kluyveromyces lactis immobilization on corn grits for milk whey hydrolysis. Enzyme Microb. Technol. 16,303-310.

Siso, M. I. G., Freire, A., Ramil, E., Belmonte, E. R., Torres, A. R., and Cerdan, E. 1994. Covalent immobilization of beta-galactosidase on corn grits. A system of lactose hydrolysis without diffusional resistance. Process Biochem. 29, 7-12.

Smart, J., and Richardson, B. 1987. Molecular properties and sensitivity to cations of beta- galactosidase from Streptococcus thermophilus with four enzyme substrates. Appl. Micro- biol. Biotechnol. 26,177-185.

Smart, J. B. 1991. Transferase reactions of P-galactosidase from Streptococcus thermophilus. Appl. Microbiol. Biotechnol. 34,495-501.

Solomons, N. W. 1986. An update on lactose intolerance. Nutr. News 49(1), 1-3. Stanley, W. L., and Palter, R. 1973. Lactase immobilization on phenolic resins. Biotechnol.

Bioeng. l5,597-602. Sungur, S., and Akbulut, U. 1994. Immobilization of beta-galactosidase on gelatin by gluteral-

dehyde and chromium (111) acetate. J. Chem. Technol. Biotechnol. 59,303-306. Surve, S . S., and Mahoney, R. R. 1994. Kinetic stabiIization of Kluyveromyces marxianus

P-galactosidase by histidine and other amino acids. Biotechnol. App. Biochem. 20 (Part l ) , 55-65.

Suyama, K., Adachi, S., Toba, T., Sohma, T., Hwang, C.-J., and Itoh, T. 1986. Isoraffinose (6-P-galactosylsucrose) synthesized by the intermolecular transglycosylation reaction of Escherichia coli P-galactosidase. Agric. Biol. Chem. 50(8), 2069-2075.

Tahajod, A. S., and Rand, A. G. 1993. Bioprocess combinations for manufacturing cultured dairy products. Cult. Dairy Prod. J. 28(1), 10-14.

Thiem, J. 1993. Chemoenzymatic synthesis of oligosaccharides. Biofutur 12553-59. Thompkinson, D. K. 1989. Immobilized systems for 8-galactosidase. Indian J. Dairy Sci.

Thompkinson, D. K., Mathur, B. N., and Rajore, R. B. 1991. Process alternatives for lactose hydrolysis. Indian Dairyman 43, 394-402.

Tikhomirova, A. S., Tsereteli, A. K., Kulikova, A. K., Letunova, E. V., and Bezborodov, A. M. 1980. Transglycosylating effect of enzyme preparations of beta-galactosidases. Appl. Biochem. Microbiol. 16(1), 49-54.

42,787-791.


Toba, T., and Adachi, S. 1978. Hydrolysis of lactose by microbial beta-galactosidases. Forma- tion of oligosaccharides with special reference to 2-0-beta-~-galactopyranosy~-~-g~ucose. J. Dairy Sci. 61,33.

Toba, T., Tomita, Y., Itoh, T., and Adachi, S. 1981. Beta-galactosidase of lactic acid bacteria: Characterization by oligosaccharides formed during hydrolysis of lactose. J . Dairy Sci. 64, 185.

Toba, T., Watanabe, A., and Adachi, S. 1982. Allolactose and 6-O-beta-galactopyranosyl-o- galactose in commercial yogurt. J. Dairy Sci. 65, 702.

Toba, T., Watanabe, A., and Adachi, S. 1983. Quantitative changes in sugars, especially oligosaccharides, during fermentation and storage of yogurt. J. Dairy Sci. 66, 17.

Toba, T., Akitsugu, Y., and Adachi, S. 1985. Oligosaccharide structures formed during the hydrolysis of lactose by Aspergillus oryzae beta-galactosidase. Food Chem. 16, 147.

Toba, T., Arihara, K., and Adachi, S. 1986. Quantitative changes in oligosaccharides during fermentation and storage of yogurt inoculated simultaneously with starter culture and beta-galactosidase preparation. J. Dairy Sci. 69, 1241.

Townley, R. R. W. 1967. Disordered disaccharide digestion and absorption and its relation to other malabsorptive disorders in childhood. Bordon Rev. Nutr. Res. 28(2), 33-48.

Turner, S. J., Daly, T., Hourigan, J. A., Rand, A. G., and Thayer, W. R. 1976. Utilization of a low-lactose milk. Am. J . Clin. Nutr. 29, 739-744.

Uwajima, T., Yagi, H., and Terada, 0. 1972. Purification, crystallization and some properties of P-galactosidase from Saccharomyces fragilis. J . Agric. Biol. Chem. 36, 570-577.

Vaheri, M., and Kauppinen, V. 1978. The formation of lactulose (4-0-beta-galactopyranosyl- fructose) by beta-galactosidase. Acta Pharm. Fenn. 87, 75.

Vega, Y. M. Puchal, A. A., and Buddington, R. K. 1992. Intestinal amino acid and monosaccharide transport in suckling pigs fed milk replacers with different sources of carbohydrate. J . Nufr. l22(12), 2430-2439.

Vulfson, E. N. 1993. Enzymatic synthesis of food ingredients in low-water media. Trends Food Sci. Technol. 4,209-215.

Vulfson, E. N., and Law, B. A. 1993. Enzymatic synthesis and modification of carbohydrates. J. Chem. Technol. Biotechnol. 56,409.

Vulfson, E. N., Patel, R., and Law, B. A. 1990. Alkyl-P-glucosidase synthesis in water- Organic and two-phase system. Biotechnol. Lett. 12(6), 397-402.

Wallenfels, K., and Malhotra, 0. M. 1961. Galactosides. Adv. Carbohydr. Chem. 16, 239. Wasserman, B. P. 1984. Thermostable enzyme production. Food Technol. 38(2), 78. Weetall, H. H., Havewala, N. B., Pritcher, W. H., Detar, C. C., Vann, W. P., and Yaverbaum,

S. 1974a. The preparation of immobilized lactase and its use in the enzymatic hydrolysis of acid whey. Biotechnol. Bioeng. 16, 689-696.

Weetall, H. H., Havewala, N. B., Pritcher, W. H, Detar, C. C., Vann, W. P., and Yaverbaum, S. 1974b. Preparation of immobilized lactase. Continued studies of enzymatic hydrolysis of acid whey. Biotechnol. Bioeng. 16,295-313.

Wierzbicki, L. E., and Kosikowski. F. V. 1973. Formation of oligosaccharides during beta- galactosidase action on lactose. J. Dairy Sci. 56(11), 1400.

Wierzbicki, L. E., Edwards, V. H., and Kosikowski, F. V. 1973. Immobilization of microbial lactases by covalent attachment to porous glass. J . Food Sci. 38, 1070-1073.

Wierzbicki, L. E., Edwards, V. H., and Kosikowski, F. V. 1974. Hydrolysis of lactose in acid whey by lactase bound to tubular reactors. J. Food. Sci. 39, 374-378.

Wendorff, W. L., and Amundson, C. H. 1971. Characterization of beta-galactosidase from Saccharomyces fragilis. J. Milk Food Technol. 34,300-306.

Wolin, M. J. 1981. Fermentation in the rumen and human large intestine. Science 213, 1463-1468.


Woychik, J. H., and Wondolowski, M. V. 1972. Covalent bonding of fungal P-galactosidase to glass. Biochim. Eiophys. Acta 289,347-351.

Yanahira, S. , Suguri, T., Yakabe, T., Ikeuchi, Y., Hanagata, G., and Deya, E. 1992. Formation of oligosaccharides from lactitol by Aspergillus oryzae @-D-galactosidase. Carbohydr. Res.

Yang, S . T., and Tang, I. C. 1988. Lactose hydrolysis and oligosaccharide formation catalyzed

Zadow, G. J. 1984. Lactose: Properties and uses. J. Dairy Sci. 67,2654-2679. Zamost, B. L., Nielsen, H. K., and Starnes, R. L. 1991. Thermostable enzymes for industrial

Zarate, S., and Lbpez-Leiva, M. H. 1990. Oligosaccharide formation during enzymatic lactose

Zentgraf, B., and Gwenner, C. 1992. Fixation of thermostable P-galactosidase on zeolites with

232,151-159.

by beta-galactosidase, Ann. N. Y. Acad. Sci. 542,417.

application. J. Ind. Microbiol. 8,71-82.

hydrolysis: A literature review. J. Food Pror. 53(3), 262-268.

controlled hydrophobicity. Acra Biotechnol. l2(6), 527-530.

ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 38

GLASS TRANSITIONS AND WATER-FOOD STRUCTURE INTERACTIONS

LOUISE SLADE AND HARRY LEVINE

Nabisco Fundamental Science Group

East Hanover, New Jersey 07936

I. Introduction 11. Foundation of the “Food Polymer Science” Approach

A. Glasses and Glass Transitions B. Evolution of the “Food Polymer Science” Approach C. Ts and Methods of Its Measurement in Foods D. Effect of Plasticization on Tg E. “Water Dynamics” and “Glass Dynamics” F. State Diagrams

111. Key Elements and Applications of the “Food Polymer Science” Approach A. Effect of Molecular Weight on TB B. Plasticization by Water C. The Dynamics Map D. WLF Theory and WLF Kinetics E. Thermosetting of Amorphous Polymers

IV. Research Needs: Outstanding Problems, Issues, and Unanswered Questions V. Conclusions and Future Prospects

References

I. INTRODUCTION

In the decade of the 1980s, the value of a polymer science approach to the study of the glassy state phenomenon and glass transitions, and of their importance to water-food structure interactions, in food materials, products, and processes was increasingly recognized by a growing number of food scientists and technologists (Table I). Why did this happen, and

103

Copyright Q 1995 by Academic Press. Inc. AU rights of reproduction in MY form reserved.

104 LOUISE SLADE AND HARRY LEVINE

TABLE I REFERENCES TO STUDIES OF GLASSY STATE PHENOMENON AND GLASS TRANSITIONS

IN FOODS AND FOOD MATERIALS BY RESEARCHERS AND GROUPS MOST ACTIVE IN

FIELD IN 1980s"

Researchers References

Blanshard

van den Berg

Karel and Roos

Lillford

~

Franks Franks et al. (1977, 1991), Franks (1982, 1983a,b, 1985, 1986a,b, 1989, 1990, 1991a,b,c, 1992a,b, 1993a,b), Finegold et al. (1989), Hatley et al. (1989), Franks and Grigera (1990), Franks and Hatley (1990, 1992, 1993), Franks and van den Berg (1991), Hatley (1991), Hatley and Franks (1991), Hatley et al. (1991), Hatley and Mant (1993), Ramanujam er al. (1993), Suzuki and Franks (1993)

Blanshard (1986, 1987, 1988), Blanshard and Franks (1987), Marsh and Blanshard (1988), Blanshard et al. (1991), Arvanitoyannis et al. (1992, 1993), Blanshard and Mitchell (1992), Kalichevsky and Blanshard (1992a-e), Kalichevsky et al. (1992a-d, 1993a,b), Tian and Blanshard (1992a,b,c), Arvanitoyannis and Blanshard (1993a,b)

van den Berg (1981, 1986, 1991, 1992), van den Berg et a/. (1992, 1993)

Karel and Flink (1983), Karel (1985, 1986, 1989, 1990, 1991a,b,c, 1992,1993), Roos (1987, 1992a-d), Karel and Langer (1988), Paakkonen and Roos (1990), Roos and Karel (1990, 1991a-h, 1992, 1993), Karel and Saguy (1991), Shimada et al. (1991), Aguilera et al. (1992), Anglea et al. (1992a,b), Buera and Karel (1992), Buera et al. (1992), Jouppila and Roos (1992), Karathanos et al. (1992), Karmas et al. (1992, 1993), Labrousse et al. (1992), Levi and Karel (1992a, 1993a,b), Karel et al. (1993a,b)

Ablett et al. (1986, 1988, 1992a,b,c, 1993), Edwards eta/. (1987), Lillford (1988), Attenburrow et al. (1989, 1990, 1992), Ablett and Lillford (1991), Davies et al. (1991), Izzard et al. (1991), Lillford el al. (1992), Attenburrow and Davies (1993)

Hutchinson et al. (1989), Orford et al. (1989, 1990), Morris (1990), Noel et al. (1990, 1991, 1992, 1993), Ollett and Parker (1990), Smith (1990, 1992a,b,c), Warburton et al. (1990, 1992), Whittam et al. (1990, 1991), Cairns et al. (1991a,b), Ollett et al. (1991, 1993a,b), Ring and Whittam (1991), Botham et aL (1992), Noel and Ring (1992), Parker and Smith (1992), Whittam (1992), Donald er al. (1993), Kirby et al. (1993)

Morris, Ring, and Smith Ring et al. (1987), I'Anson et al. (1988,1990),

(continues)

GLASS TRANSITIONS 105

Maurice and Biliaderis

Schenz Tomka

Slade and Levine

TABLE I (Cont inued)

Researchers References

Reid Reid (1985, 1990, 1992a,b), Lim and Reid (1991, 1992), Hsu and Reid (1992), Reid and Hsu (1992), Reid et al. (1992, 1993a,b), Taylor el a/. (1992), Kerr et al. (1993)

Hoseney et al. (1986, 1992), Yost and Hoseney (1986), Doescher et al. (1987), Zeleznak and Hoseney (1987a,b), Faubion and Hoseney (1989), Hoseney and Rogers (1990,1993), Hoseney (1991,1992), Dong (1992)

Simatos and Karel (1988), Le Meste and Duckworth (1988), Blond (1989, 1992, 1993), Simatos et al. (1989), Le Meste and Simatos (1990), Le Meste et al. (1990, 1991a,b, 1992), Blond and Colas (1991), Blond and Simatos (1991). Le Meste and Huang (1991), Matthiesen et al. (1991), Simatos and Blond (1991, 1993), Aynie et al. (1992a,b), Huang (1992, 1993a,b), Huang et al. (1992, 1993), Simatos (1994)

1986a,b,c), Biliaderis and Galloway (1989), Biliaderis (1990, 1991a,b, 1992a,b), Biliaderis and Seneviratne (1990a,b), Biliaderis and Zawistowski (1990), Caldwell e ta / . (1990), Biliaderis and Tonogai (1991), Michniewicz er al. (1992), Goff et al. (1993)

Hoseney

Simatos, Le Meste, and Huang

Maurice et al. (1985, 1991), Biliaderis et al. (1985,

Schenz et al. (1984, 1991, 1992) Wittwer and Tomka (1984), Tomka (1986, 1991), Sala

and Tomka (1992a,b), Willenbucher el al. (1992) Slade (1984), Slade and Levine (1984, 1985, 1987a,b,

1988a-d, 1989, 1990, 1991a,b, 1992a-c, 1993a-c), Slade et al. (1987, 1989, 1993), Levine and Slade (1986, 1988a-d, 1989a-d, 1990, 1991, 1992a,b, 1993), Levine et al. (1991, 1992), Cole et al. (1983, 1984)

~ ~ ~~~ ~~

a Modified from Levine and Slade (1992b).

why do many current workers in the field consider this polymer science approach to be a significant advance in food research?

In this chapter, we attempt to answer these questions by reviewing the so-called “food polymer science” approach and various recent examples of how it has been used to define structure-function relationships in food products and processes, and to assess and understand the effects of glass transitions and water plasticization on food quality, safety, and stability. We review some theoretical principles from the field of synthetic polymer


science that have been shown to be applicable to studies of the glassy state phenomenon in foods, and illustrate those applications by means of a broad overview focusing primarily on the most recent experimental studies by a number of groups that have been especially active in this rapidly growing field of research. In so doing, we attempt to convince readers, especially those previously unfamiliar with the food polymer science approach, of the importance of such an approach to food R&D. It is our hope that this review is sufficiently comprehensive and self-contained to enlighten those readers not previously acquainted with its subject matter. At the same time, we intend that it should build on our several earlier reviews (Levine and Slade, 1988a, 1989b, 1992b; Slade et al., 1989; Slade and Levine, 1991a,b, 1993b,c), by emphasizing critical coverage of the most up-to-date studies, applications, and remaining (and/or resulting) questions concerning glass transitions and water-food structure interactions, in order to satisfy those readers already familiar with this subject. All readers are referred to those previous reviews for discussions of various earlier studies and applications exhaustively detailed there but not here.

II. FOUNDATION OF THE “FOOD POLYMER SCIENCE” APPROACH

The research approach known as “food polymer science” (Slade/Levine references in Table I) emphasizes the fundamental and generic similarities between synthetic polymers and food molecules, and has provided a new interpretive (rather than theoretical) and experimental framework for the study of food systems that are kinetically constrained. Based on established structure-property relationships from the field of synthetic polymer science (Table 11), this innovative approach was developed to unify structural aspects of foods, viewed as kinetically metastable, completely amorphous or partially crystalline, homologous polymer systems, with functional aspects, dependent on mobility and conceptualized in terms of “water dynamics” and “glass dynamics.” These unified concepts have been widely applied to explain and predict functional properties of food materials during processing and product storage (Slade/Levine references in Table I). Key elements of this interpretive approach to investigations of structure-function relationships in food systems, with demonstrated relevance to moisture management and water relationships (Slade and Levine, 1991a), include recognition of the following:

1. The behavior of foods and food materials as classical polymer systems, and that the behavior is governed by dynamics rather than energetics


TABLE I1 S E L E a E D REFERENCES ON STRUCTURE-PROPERTY RELATIONSHIPS FROM SYNTHETIC

POLYMER SCIENCE LITERATURE”

(A) Textbooks Flory (1953). Wunderlich (1973, 1976, 1980, 1990), Cowie (1973), Nielsen (1977), Ferry

(1980), Rowland (1980), Tun (1981), Sears and Darby (1982), Billmeyer (1984), Sperling (1986), Elliott (1990), Gray (1991), Matsuoka (1992)

(B) Journal Articles and Book Chapters Kauzmann (1948), Gordon and Taylor (1952), Williams et al. (1955). Walton (1969),

Brydson (1972). Sharples (1972), Hardy et al. (1973), Haward (1973). Roberts and White (1973), Flory (1974), Petrie (1975), Johari (1976, 1991), Bar0 et al. (1977), Buchanan and Walters (1977), Couchman and Karasz (1978), Fuzek (1980). Johnson et al. (1980), Moy and Karasz (1980). Starkweather (1980). Bair (1981, 1985), Batzer and Kreibich (1981), Keinath and Boyer (1981), Maurer (1981), Prime (1981), Shalaby (1981), Wunderlich (1981), Fried et al. (1982), Gaeta et al. (1982), Angell (1983, 1988), Eisenberg (1984), Ellis et al. (1984), Graessley (1984), Jin et al. (1984), Boyer et al. (1985), Matsuoka et al. (1985), Robertson (1985), Alfonso and Russell (1986), Blum and Nagara (1986), Domszy et al. (1986). Hiltner and Baer (1986), Mandelkern (1986), Cheng and Wunderlich (1987), Huang et al. (1987), Aubin and Prud’homme (1988), Burchard (1988), Ellis (1988), Sichina (1988), Cheng (1989), Murthy (1989), Murthy el al. (1989,1993), Ehlich and Sillescu (1990), Angell et al. (1991, 1992), Cassel and Twombly (1991), Pissis and Apekis (1991), Roland and Ngai (1991, 1992), Chin et al. (1992), Lomellini (1992), Xenopoulos et al. (1992), Allen (1993). Compan et al. (1993), Ngai and Roland (1993)

Modified from Levine and Slade (1992b).

2. The importance of the characteristic temperature, Tgr at which the glass-rubber transition occurs, as a physicochemical parameter that can determine processibility, product properties, quality, stability, and safety of food systems

3. The central role of water as a ubiquitous plasticizer of natural and fabricated amorphous food ingredients and products 4. The effect of water as a plasticizer on Tg, and the resulting non-

Arrhenius, diffusion-limited behavior of amorphous polymeric, oligomeric, and monomeric food materials in the rubbery liquid state at T > Tg

5. The significance of nonequilibrium glassy solid and rubbery liquid states (as opposed to equilibrium thermodynamic phases) in most “real world” food products and processes, and their effects on time-dependent structural and mechanical properties related to quality and storage stability.

A. GLASSES AND GLASS TRANSITIONS

A glass is operationally defined as an amorphous (i.e., noncrystalline) solid (Haward, 1973). In fact, a glass is actually an undercooled liquid of


such high viscosity [q > 1010-1014 Pa s (Ferry, 1980; Wunderlich, 1981; Sperling, 1986)] that it exists in a metastable, mechanical solid state, in which it is capable of supporting its own weight against flow due to the force of gravity (Haward, 1973). A glass forms when a typical liquid, with a disordered molecular structure, is cooled to a temperature generally about 100°C (for many pure liquids) below its equilibrium crystalline melting temperature (T,) or freezing point, at a cooling rate sufficiently high to avoid crystallization of the liquid (Ferry, 1980). This solidification process, known as vitrification (Luyet, 1960), results in the “freezing in” or immobilization of the disordered structure of the liquid state (Wunderlich, 1981), such that the resulting glassy solid is spatially homogeneous, but without any long-range lattice order, and is incapable of exhibiting any long-range, cooperative relaxation behavior (e.g., translational mobility) on a practical time scale (Ferry, 1980; Wunderlich, 1990).

A glass transition in amorphous systems is a temperature-, time- (or frequency-), and composition-dependent, material-specific change in physical state, from a glassy mechanical solid to a rubbery viscous liquid (capable of flow in real time) (Ferry, 1980; Wunderlich, 1990). In terms of thermodynamics, the glass transition is operationally defined as a second-order transition (Sperling, 1986; Wunderlich, 1990) (as opposed to a first-order transition, e.g., crystalline melting) and denoted by (1) a change in slope of the volume expansion (a first derivative of the free energy); (2) a discontinuity in the thermal expansion coefficient; and (3) a discontinuity in the heat capacity (a second derivative of the free energy) (Kauzmann, 1948; Wunder- lich, 1990). The glass transition is also operationally defined, based on mechanical properties, in terms of a mechanical relaxation process such as viscosity (Ferry, 1980). Figure 1 shows that as the temperature ( T ) is lowered from that of the low-viscosity liquid state above T,, where familiar Arrhenius kinetics apply, through a temperature range from T, to Tg, a completely different, nonlinear form of kinetics, with an extraordinarily large temperature dependence (Angell, 1988; Franks and Grigera, 1990; Angell et al., 1991, 1992), becomes operative (Roberts and White, 1973). Then, at a temperature where cooperative mobility becomes limiting, a state transition occurs, typically manifested as a three orders-of-magnitude change in viscosity, modulus, or mechanical relaxation rate (Eisenberg, 1984; Sperling, 1986; Noel etal., 1990,1991). At this glass transition temperature, the viscosity of a liquid is Pa s (1013p) (Koide et al., 1990 Noel et al., 1991), the structural relaxation time [as determined calorimetrically, e.g., by differential scanning calorimetry (DSC)] for such a liquid is =200 seconds (Angell, 1988 Angell et al., 1991, 1992), and, equivalently, the relaxation frequency is Hz (Murthy ef al., 1993). A “mechanical” glass transition can be defined by combinations of temperature and deformation


‘ “ I Rubber I

0.5 1 .o 1.5 2.0

TlnK

FIG. 1. Viscosity as a function of reduced temperature (T,,,/T) for glassy and partially crystalline polymers. Reprinted from Levine and Slade (1988a) with permission of Cambridge University Press. Copyright 1988 Cambridge University Press. Adapted from Franks (1982).

frequency for which, as temperature decreases, sufficiently large numbers of mobile units (e.g., small molecules or backbone-chain segments of a macromolecule) become cooperatively immobilized (in terms of large-scale rotational and translational motions) during a time comparable to the experimental period (Buchanan and Walters, 1977; Angell, 1988; Green and Angell, 1989; Hosea et al., 1990), such that the material becomes a mechanical solid capable of supporting its own weight against flow. Perhaps the most important distinction between dimensionally extended (alpha) relaxations, which give rise to the glass transition as translational motions become constrained at Tg, and small-scale (beta and gamma) relaxations (Scandola et al., 1991; Botham et al., 1992; Noel et al., 1992), for which small-scale rotational motions do not become constrained as T falls below Tg, is the cooperative nature of alpha relaxations (Johari, 1976; Ferry, 1980; Wunderlich, 1981). Arrhenius kinetics become operative once again in the glassy solid at T < Tg (Noel et al., 1991), but the rates of all diffusion- limited processes are much lower in this high-viscosity solid state than in the liquid state at T > Tg (Levine and Slade, 1988a; Karmas et al., 1992). In fact, the difference in average relaxation times between the two Arrhenius regimes (at T < Tg and T > Tm) is typically more than 14 orders of magnitude (Slade and Levine, 1988b).

B. EVOLUTION OF THE “FOOD POLYMER SCIENCE” APPROACH

The genesis of a polymer science approach to the study of glasses and glass transitions in foods dates back at least to 1966 and a seminal


review by White and Cakebread on the glassy state and Tg in certain sugar- containing food products. They recognized (1) the importance of the glassy state, and of the glass transition temperature and its location relative to the temperature of storage (either ambient or subzero), in a variety of aqueous food systems, including but not limited to boiled sugar candies, and (2) the critical role of water as a plasticizer of food glasses and the quantitative Tg-depressing effect of increasing content of plasticizing moisture, whereby Tg of a particular glass-forming solute-water mixture depends on the corresponding content of plasticizing water (W,) in that glass at its Tg (Levine and Slade, 1986, 1988a; Slade and Levine, 1988a). White and Cakebread (1966; Cakebread, 1969) were apparently the first food scientists to allude to the broader implications of nonequilibrium glassy and rubbery states to the quality, safety, and storage stability of a wide range of glass- forming aqueous food systems. Evidently, outside a small community of candy technologists, (1) the work of White and Cakebread, (2) the even earlier (but somewhat better known by food technologists) work of Ma- kower and Dye (1956) on the kinetics of crystallization of amorphous sucrose and glucose (from the water-plasticized, rubbery state at room T > Tg), (3) the more recent work by other candy technologists such as Herrington and Branfield (1984) on the physicochemical properties of sugar glasses, and (4) the broader relevance of this body of work to the field of food science and technology went largely unnoticed until the early 1980s.

Since that time, many other workers (Tables I and IIIA) have helped to advance, with increasing momentum since 1984, concepts and approaches based on recognition of the importance of, and application of the principles underlying, nonequilibrium glassy solid and rubbery liquid states in foods. The current situation of accelerating activity and interest in this area is illustrated by the fact that 16 different chapters in a recent book on water relationships in foods (Table IIIB) included discussions of glasses and glass transitions in food systems. Another strong indication of the emergence of this subject is the fact that a new graduate course, “Advanced Topics in Food Science: Glass Transitions, Tg, A,, and the Physical Properties of Foods,” was offered for the first time in the summer of 1990 by Professors T. Labuza and E. Davis at the University of Minnesota’s Food Science Department. Still another indication of the recent prominence of this subject is the fact that (1) an entire 4-day-long international conference on “The Science and Technology of the Glassy State in Foods” was held as one of the well-known “Easter Schools” at Nottingham University (UK) in April 1992) (Table IIIE), to mark the completion of phase I of the ground- breaking ACTIF (Amorphous and Crystalline Transitions in Foods) research program at Nottingham; (2) a symposium on “Glass Transitions in Cereal-Based Foods” (Table IIIC) and another on “Developments in


TABLE 111 REFERENCES TO RECENT PUBLICATIONS DEALING WITH THE PRACTICAL

SIGNIFICANCE OF THE GLASSY STATE PHENOMENON AND GLASS TRANSITION TO THE

PROCESSING, QUALITY, SAFETY, AND STABILITY OF FOOD AND BIOLOGICAL

INGREDIENTS AND PRODUCTS'

(A) General Soesanto and Williams (1981), Flink (1983), Saleeb and Pickup (1985, 1989), Chan er al. (1986), Hardman (1986), Atkins (1987). Gosline (1987), Russell (1987a), Could and Christian (1988). Lineback and Rasper (1988), Quinquenet et al. (1988), Skarra et al. (1988), Takahashi et al. (1988), Zobel (1988, 1992a,b), Zobel et af. (1988), BeMiller (1989, 1993), Cairault et al. (1989), Colonna et al. (1989). Doublier (1989), Eliasson (1989, 1992), Fujio and Lim (1989), Russell and Oliver (1989), Shi and Seib (1989, 1992), Stepto and Dobler (1989). Ahlneck and Zografi (1990), Hosea et al. (1990), Johnston-Banks (1990), Johnson et al. (1990), Kararli and Catalan0 (1990), Kararli et al. (1990), Lillie and Gosline (1990), Mita (1990), Oksanen and Zografi (1990, 1993), Pikal (1990a,b, 1993), Pikal and Shah (1990). Pikal et al. (l990,1991a,b), Roozen and Hemminga (1990, 1991). Roper and Koch (1990), Rubin et al. (1990), Schiraldi (1990), Stauffer (1990, 1992), Takushi et al. (1990), Van Scoik and Carstensen (1990), Williams et af. (1990), Yoshida et al. (1990, 1992), Belton (1991), Chang and Baust (1991a,b,c), Chinachoti er al. (1991a), Cocero and Kokini (1991). Davis (1991, 1992a,b), Donhowe et al. (1991), Fujio et al. (1991), Gaines (1991, 1993), Graf and Saguy (1991), Hartel and Shastry (1991), Hegenbart (1991, 1993), Karathanos er al. (1991), Karger and Ludemann (1991), Lai and Kokini (1991), Larsson (1991), Larsson and Eliasson (1991), Liu and Lelievre (1991a,b, 1992a,b), Liu et al. (1991), Ma and Harwalkar (1991), MacDonald and Lanier (1991), Nakamura et al. (1991), O'Brien (1991), Parak and Nienhaus (1991), Patil (1991), Raemy and Lambelet (1991), Roozen et al. (1991), Roser (1991a,b), Scandola et al. (1991), Shukla (1991), Sochava et al. (1991), Spratt et al. (1991), Tanner et al. (1991). Tolstoguzov (1991, 1992), Young and Scholl (1991), Bagley (1992), Berail et al. (1992), Best (1992), Bizot et al. (1992), Boskovic et al. (1992), Bruin (1992), Caldwell et al. (1992), Chang and Randall (1992), Chinachoti (1992a, 1993), Cocero et al. (1992a), Colaco et al. (1992), Colonna and Buleon (1992), de Graaf et al. (1992), Eerlingen and Delcour (1992), Finley et al. (1992), Gaines et al. (1992a-c, 1993a,b), German et al. (1992), Goff (1992). Gontard et al. (1992a,c), Gudmundsson (1992), Gudmundsson and Eliasson (1992), Hanover (1992), Harrison et al. (1992), Hartel (1992, 1993), Jovanovich et al. (1992), Jurgens et al. (1992). Kaletunc and Breslauer (1992), Karathanos and Saravacos (1992), Katsuta et al. (1992), Kawai et al. (1992), Kim and Setser (1992), Kim and Walker (1992), Kokini (1992), Kokini et al. (1992a,b), Kresin and Rau (1992), Labuza and Baisier (1992), Lawton (1992a,b), Lim et al. (1992), Magoshi et al. (1992), Megret et al. (1992), Milczarek et al. (1992), Nelson and Labuza (1992b), Newman (1992), Peleg (1992), Pissis et al. (1992), Rajagopalan and Seib (1992), Rao et al. (1992). Roy et al. (1992), Sahagian and Goff (1992, 1993), Sapru and Labuza (1992b), Saunders et al. (1992), Schroeter and Hobelsberger (1992), Shah and Ludescher (1992), Shogren (1992), Shogren et al. (1992), Sochava and Belopolskaya (1992), Sochava and Smirnova (1992), Stute (1992), Summers (1992), te Booy et al. (1992), Tolstoguzov and Nesmeyanov (1992), Watase et af. (1992), Zasypkin et al. (1992), Zhang and Jackson (1992), Anonymous (1993a,b), Appelqvist et af. (1993), Carpenter ef al. (1993), Crowe and Crowe (1993), Crowe et af. (1993), Eerlingen ef al. (1993). Eliasson and Larsson (1993), Ford and

(continues)


TABLE I11 (Continued)

Dawson (1993), George (1993), Hancock and Zografi (1993), Haynes and Locke (1993), Haynes et al. (1993), Lawton and Wu (1993), Mancini (1993), Mestres et al. (1993), Nelson (1993), O’Donnell (1993), Ohshima et al. (1993), Peppas (1993), Rohde et af . (1993), Roser and Colaco (1993), Seow and Teo (1993), Seow and Thevamalar (1993), Shalaev and Kanev (1993), Yano (1993). Zanoni et al. (1993)

(B) Chapters in “Water Relationships in Foods” (Plenum Press, New York 1991) Franks (1991a). Blanshard et al. (1991), van den Berg (1991), Karel and Saguy (1991), Lim

and Reid (1991), Hoseney (1991), Simatos and Blond (1991), Le Meste et af . (1991b), Maurice et al. (1991), Schenz et al. (1991), Biliaderis (1991a), Slade and Levine (1991b), Given (1991). Nishinari etal. (1991), Tomka (1991). Wasylyk and Baust (1991)

(C) “Glass Transitions in Cereal-Based Foods” Symposium and Other Papers at AACC Annual Meeting (1992)

Amemiya and Menjivar (1992), Botham et al. (1992), Cocero et al. (1992b), Hoseney et al. (1992), Huang et af. (1992), Kalichevsky and Blanshard (1992c), Lelievre (1992b), Lillford et al. (1992), Slade and Levine (1992a), Smith (1992b)

(D) Glass Transitions in Natural Plant Tissues Burke (1986), Hirsh (1987), Koster and Leopold (1988), Williams and Leopold (1989),

Vertucci (1990), Bruni and Leopold (1991, 1992), Dereuddre et al. (1991), Koster (1991), Lin et al. (1991), Williams (1991), Leopold et al. (1992), De Bry (1993), Koster et al. (1993), Williams et af . (1993)

(E) Chapters in T h e Glassy State in Foods” (Nottingham University Press, Loughborough, 1993)

Allen (1993), Karel et al. (1993a), Slade and Levine (1993c), Reid et al. (1993a), Kalichevsky et al. (1993a), Hemminga et al. (1993), Noel el al. (1993), Ablett et al. (1993), Roos and Karel (1993), van den Berg et al. (1993), Lillie and Gosline (1993), Gidley et af. (1993), Attenburrow and Davies (1993), Levine and Slade (1993), Donald et al. (1993), Simatos and Blond (1993), Peleg (1993), MacInnes (1993), Bolton et al. (1993)

Poster Papers at Easter School Conference (1992) Schenz et al. (1992), Aynie er al. (1992a), Kalichevsky and Blanshard (1992d,e), Labuza

(1992), Livings et al. (1992), Nesvadba (1992a), Parker and Smith (1992), Sala and Tomka (1992a,b), Willenbucher et al. (1992). Tian and Blanshard (1992c), Sapru and Labuza (1992a)

(F) Papers at I m ’92 Annual Meeting (1992) Attwool et al. (1992), Chinachoti (1992a), Chuy and Labuza (1992), Cocero and Kokini

(1992), Hallberg and Chinachoti (1992), Hsu and Reid (1992), Huang (1992), Jeffery (1992), Leung et af. (1992), Levi and Karel (1992a), Madeka and Kokini (1992b), Nelson et al. (1992), Reid et al. (1992), Roos (1992b), Taylor et al. (1992)

(G) Papers at ISOPOW-V Conference (1992) Blond (1992), Nelson and Labuza (1992a), Reid (1992a), Slade and Levine (1992b),

Blanshard and Mitchell (1992), Angel1 et al. (1992), Kokini et al. (1992d), MacInnes (1992), Nesvadba (1992b), Jouppila and Roos (1992), Chinachoti (1992b), Gontard et af. (1992b), Aynie et af. (1992b)

(continues)


TABLE 111 (Continued )

(H) Papers at CoFE ’92 Meeting (1992) Labuza and Nelson (1992), Anglea ef al. (1992b), Roos (1992~). Madeka and Kokini

(1992a). Lim and Reid (1992)

(I) Chapters in “Developments in Carbohydrate Chemistry” (AACC, St. Paul, 1992) Zobel (1992a), Lelievre (1992a). Biliaderis (1992a), Sullivan ef al. (1992), Thompson (1992)

(J) Papers at ICEF 6 Congress (1993) Karel (1993), Cocero ef al. (1993), Roos and Jouppila (1993), Huang (1993a), Nesvadba

(1 993)

(K) Papers at Im ’93 Annual Meeting (1993) Barrett et al. (1993), Cherian and Chinachoti (1993), Cocero and Kokini (1993), Huang

(1993b), Karmas et al. (1993), Kim and Taub (1993), Levi and Karel (1993a), Madeka and Kokini (1993), McCurdy ef al. (1993). Nelson and Labuza (1993), Shah and Ludescher (1993), Wang and Chinachoti (1993)


Carbohydrate Chemistry” (Table 1111) were held at the AACC 77th Annual Meeting in Minneapolis in September 1992; and (3) many papers on the glassy state and glass transitions were also presented in 1992 and 1993 at the following: IFT ’92 Annual Meeting in New Orleans (Table IIIF); CoFE-AIChE 1992 Summer National Meeting in Minneapolis (Table IIIH); ISOPOW-V Meeting on the Properties of Water in Foods in Valencia (Spain) in November 1992 (Table IIIG); ICEF-6 Congress in Chiba (Japan) in May 1993 (Table IIIJ); and IFT ’93 Annual Meeting in Chicago (Ta- ble IIIK).

Recognition of several key elements (mentioned earlier) of the food polymer science approach and their relevance to the behavior of a broad range of different types of foods [e.g., intermediate-moisture (IMF), low- moisture, frozen, starch-based, gelatin-, gluten-, and other protein-based foods] and corresponding model systems increased markedly during the 1980s (Slade and Levine, 1991a). Various studies have illustrated the perspective afforded by using this conceptual framework and demonstrated the technological utility of this new approach to understand and explain complex behavior, design processes, and predict product quality, safety, and storage stability, based on fundamental structure-property relationships of food systems viewed as homologous families of partially crystalline glassy polymer systems plasticized by water (Slade/Levine references in Table I). Referring to the food polymer science approach, John Blanshard, in the proposal for the ACTIF research program (personal communication, 1987), stated that “it is not often that a new concept casts fresh light across a


whole area of research, but there is little doubt that the recognition of the importance of the transition from the glassy to the crystalline or rubbery state in foodstuffs, though well known in synthetic polymers, has opened up new and potentially very significant ways of thinking about food properties and stability.” In a lecture on historical developments in industrial polysaccharides, James BeMiller (1989) echoed Blanshard’s words by re- marking that a key point regarding the future of polysaccharide research and technology is “the potential, already partly realized, in applying ideas developed for synthetic polymers to polysaccharides; for example, the importance of the glassy state in many polysaccharide applications.” More support for the food polymer science approach came in two recent columns in Cereal Foods World, one on food technology, entitled “A New Develop- ment in Carbohydrate Research,’’ in which Triveni Shukla (1991) said “a recent breakthrough in fundamental research on starches [translated from basic concepts of synthetic polymers] concerns glass transition behavior and the glass transition temperature, Tg,” and the other on methodology, entitled “The “Nobel” Science of Foods,” in which Eugenia Davis (1991) said “we are seeing an explosion of polymer science application to food polymers in the last 10 years. For example, the study of glass transitions (Tg) in formulated and fabricated foods as they go from the glassy to the rubbery state is found in virtually every food science-related journal. Calorimetric, rheological, and spectroscopic methods are used. As a result, it has become imperative that we incorporate such basic scientific information into our courses and research programs.” As if to support Davis’ and Shukla’s observations, John O’Brien (1991) commented on “the rapid advances in starch chemistry and technology. . , the manipulation of glass transition temperatures, and controlled crystallization/gelatinization” in a recent editorial in Trends in Food Science and Technology on “Characteriz- ing Food and Its Ingredients.” The food polymer science approach and the importance of glass transitions in foods have even become subjects of increasing discussion in several recent food trade magazine articles (Hegenbart, 1991,1993; Best, 1992; Anonymous, 1993a; Mancini, 1993; O’Donnell, 1993).

C. Tg AND METHODS OF ITS MEASUREMENT IN FOODS

The technological importance of the glass transition in amorphous polymers and of the characteristic temperature at which it occurs ( Tg) is well known in synthetic polymer science (Ferry, 1980; Rowland, 1980; Sears and Darby, 1982; Eisenberg, 1984). Especially in the last several years, a growing number of food scientists from both academia and industry have increasingly recognized the practical significance of the glass transition as a physicochemical event [i.e. a change of state but not a change of phase


(Allen, 1993)] that can govern food processing, product properties, quality, safety, and stability (Tables I and 111). This recognition of the importance of the glass transition in foods [as well as in related biological materials (Table IIID), such as plant cell walls, seeds (with emphasis on the glassy state, produced by sucrose and various other sugars at low moisture, as the mechanism of control of the dormant state), and other natural plant tissues, and pharmaceutical materials] has gone hand-in-hand with an increasing awareness of the inherent nonequilibrium nature of most food products and processes (Table IV). Many examples of food systems whose behavior is governed by dynamics far from equilibrium and of practical problems of food science and technology posed by their nonequilibrium nature have been described (e.g., Slade/Levine references in Table I).

The polymer science-based interpretive approach to studies of structure- function relationships in food systems emphasizes insights gained by an appreciation of the fundamental similarities between synthetic amorphous polymers and glass-forming aqueous food materials with respect to their thermal, mechanical, and structural properties (Slade and Levine, 1991a). Such properties, for synthetic polymers as well as food materials, are most commonly measured by thermal analysis methods such as DSC and dynamic mechanical [thermal] analysis (DMA [DMTA]) (Turi, 1981; Wunderlich, 1990; Harwalkar and Ma, 1990). DSC and DMA have become established methods for characterizing glass transitions in completely amorphous or partially crystalline polymer systems (Fuzek, 1980), and have been the ones of choice in most recent investigations of glasses and glass transitions in food systems. However, in many other recent experimental studies, other methods (listed in Table V), complementary to DSC and DMA, have also been used to measure glass transitions in food ingredients and products and/or aspects of molecular mobility and diffusivity related to the effects of glass transitions in aqueous food glasses and rubbers. Probably the most powerful instrumental approach to the study of glass transitions in foods would involve combined use of several of these different but complementary experimental techniques (Noel et al., 1990). For example, in recent studies by Kalichevsky et al. (1992a-c) of the glass transition of amylopectin (the branched polymer of starch), amylopectin-sugar mixtures, wheat gluten protein, and gluten-sugar mixtures at low moisture contents, DSC, DMTA, NMR, and Instron analyses were utilized. Similarly, Shogren (1992) combined DSC, DMTA, and NMR measurements to study the effect of moisture content on physical aging of extruded corn starch glasses; Scandola et al. (1991) combined DSC, DMTA, and TDEA measurements to investigate the glass transition of dry dextran, pullulan, and amylose; Sahagian and Goff (1992) used DSC, TMA, and NMR to determine the influence of xanthan gum on the thermomechanical behavior of a frozen sucrose solu-


TABLE IV SELECTED REFERENCES TO RECENT PUBLICATIONS THAT HAVE HIGHLIGHTED

THE NONEQUILIBRIUM NATURE OF FOOD (OR PHARMACEUTICAL)

PRODUCTS AND PROCESSES‘

Examples of nonequilibrium products

and processes

Microbiological activity

Sorptionldesorption hysteresis

Graininess and iciness in ice cream, reduced survival of frozen enzymes and living cells, reduced activity and shelf stability of freeze-dried proteins

Caking and other diffusion- limited, physical or chemical “collapse” processes in amorphous powders

References

Slade and Levine (1985, 1988a,b, 1991a), Gould and Christian (1988), Sapru and Labuza (1992a,b), Bolton et al. (1992)

van den Berg (1981,1986, 1991), Hardman (1986), Levine and Slade (1988a, 1989b), Lillford (1988), Cairault et al. (1989), Slade et al. (1989), Myers-Betts and Baianu (1990), Oksanen and Zografi (1990), Paakkonen and Roos (1990), Ablett and Lillford (1991), Franks (1991a,b), Pikal er al. (1991a). Slade and Levine (1991a), Reid (1992b), Yano (1993)

Cole et al. (1983, 1984), Levine and Slade (1986, 1988c,d, 1989c,d, 1990, 1992a), Blanshard and Franks (1987). Hirsh (1987), Franks (1989, 1990, 1991c, 1992b, 1993a,b), Franks and Hatley (1990, 1992, 1993), Le Meste and Simatos (1990), Noel et al. (1990), Pikal (1990a,b, 1993). Pikal and Shah (1990), Pikal er al. (1990, 1991a,b), Reid (1990, 1992a), Belton (1991), Blond and Colas (1991), Chang and Baust (1991a,b,c), Franks and van den Berg (1991), Franks et al. (1991), Hatley (1991), Hatley and Franks (1991), Lim and Reid (1991, 1992), Roos and Karel (1991f,g), Berail et al. (1992), Best (1992), Blond (1992, 1993), Caldwell et al. (1992), Chang and Randall (1992), Goff (1992), Karel (1992), Nesvadba (1992a), Reid et al. (1992), Sahagian and Goff (1992, 1993), van den Berg (1992), van den Berg et al. (1992, 1993), Carpenter et al. (1993), Crowe and Crowe (1993), Crowe er al. (1993), Ford and Dawson (1993), Goff et al. (1993), Huang et al. (1993), Kerr et al. (1993), Ohshima et al. (1993). Williams et al. (1993)

Tsourouflis et al. (1976), To and Flink (1978), Downton et al. (1982), Flink (1983), Karel and Flink (1983), Karel (1985, 1986, 1989, 1990, 1991c), Levine and Slade (1986, 1988b, 1989a. 1992b), Roos (1987, 1992b-d), Karel and Langer (1988), Simatos and Karel (1988), Wallack and King (1988), Ahlneck and Zografi (1990), Paakkonen and Roos (1990), Roos and Karel (1990, 1991a,b,c,e, 1992, 1993), Van Scoik and Carstensen (1990), Karel and Saguy (1991), Levine ef al. (1991,1992), Shimada et al. (1991), Aguilera er al. (1992). Anglea et al. (1992b),

(continues)


TABLE IV (Continued)

Examples of nonequilibrium products

and processes References

Cooking of cereals and grains (including, e.g., by extrusion)

Expansion of bread or collapse of cake during baking, effects of flour and sugar on cookie baking

Staling of baked products

Effects of sugar-water glasses and rubbers on texture and storage stability of cookies

Recipe requirements for gelatin desserts

Best (1992), Bruin (1992), Buera and Karel (1992), Chuy and Labuza (1992), Karathanos et al. (1992), Karmas ef al. (1992, 1993), Labrousse er af. (1992), Levi and Karel (1992a, 1993a,b), Nelson and Labuza (1992a,b), te Booy ef al. (1992), van den Berg (1992), Hartel (1993), Hegenbart (1993), Huang (1993a), Karel et al. (1993a,b), Nelson (1993), Peleg (1993), Slade and Levine (1993~)

Slade and Levine (1984, 1987b, 1988b,c,d, 1989), Reid and Charoenrein (1985), Burros er af. (1987), Paton (1987), Mestres et al. (1988), Levine and Slade (1989b). Knutson (1990), Lai and Kokini (1991), Larsson and Eliasson (1991), Liu and Lelievre (1991a,b, 1992a), Liu ef al. (1991), Colonna and Buleon (1992), Rajagopalan and Seib (1992), Tolstoguzov (1992), Warburton et al. (1990, 1992), Zobel (1992a), BeMiller (1993), Donald ef al. (1993), Seow and Teo (1993), Seow and Thevamalar (1993)

Blanshard (1986), Slade et af. (1989, 1993), Levine and Slade (1989b, 1993), Slade and Levine (l990,1993b,d), Hoseney and Rogers (1993)

Slade (1984), Miles et al. (1985b), Hoseney (1986), Slade and Levine (1987b, 1988d, 1989), Slade er af. (1987), Skarra et nl. (1988), Clark et al. (1989), Gidley and Bulpin (1989). Biliaderis (1991b, 1992a,b), Cairns er al. (1991b), Cameron and Donald (1991), Patil (1991), Chinachoti (1992b), Gudmundsson (1992), Hallberg and Chinachoti (1992). Michniewicz ef al. (1992), Rao ef al. (1992), Shogren (1992), Zhang and Jackson (1992), Hoseney and Rogers (1993)

Slade and Levine (1990, 1991a,b, 1993b,d), Amemiya and Menjivar (1992), Gaines et af. (1992a-c, 1993a,b), Gaines (1991, 1993), Levine and Slade (1993), Slade et al. (1993)

Slade and Levine (1987a), Levine and Slade (1988a)



tion; Cocero and Kokini (1991) used combined mechanical spectrometry and DSC measurements in a study of the glass transition, at low moisture contents, of the glutenin component of gluten; Rubin et al. (1990) investigated vitrification of honey, using DSC and NMR, Johnson et al. (1990) examined the interactions of starch and sugar-water, using ESR and DSC, Ollivon (1991; Quinquenet etal., 1988) employed DSC and TDEA measurements to analyze solute-water interactions in aqueous glasses and rubbers of various sugars and polyhydric alcohols; Huang et al. (1992) used TMA, TSC, and dielectric spectroscopy measurements to study the glass transition, at low moisture contents, of starch, gluten, and white bread; Ollett and Parker (1990) showed that both viscosity and DSC measurements yielded comparable Tg values for undercooled melts of fructose or glucose; and Noel et al. (1991) showed the same for low-moisture maltose-water mixtures. Results of such studies have graphically demonstrated “the fact that the glass transition is not observed at a unique temperature, but is related to the frequency [or time scale (Gosline, 1987; Slade and Levine, 1988c, 1991a; Orford el al., 1989; Hosea et al., 1990, Koide et al., 1990; Noel et al., 1990,1991; Pissis and Apekis, 1991; Ring and Whittam, 1991; Nelson, 1993)] of the measurement technique” (Kalichevsky et al., 1992a) and that “the various techniques are sensitive to different degrees of molecular mobility” (Kalichevsky et al., 1992a).

Based on a polymer science approach, various DSC, DMA, and other results have been used to demonstrate that product quality and stability often depend on maintaining foods in kinetically metastable, dynamically constrained, time-dependent glassy andlor rubbery states, and that these nonequilibrium physical states determine the time-dependent thermomechanical, rheological, and textural properties of foods (Slade/Levine references in Table I).

D. EFFECT OF PLASTICIZATION ON Tg

Plasticization, and its modulating effect on the temperature location of the glass transition, is another key technological aspect of synthetic polymer science (Sears and Darby, 1982). A polymer science approach to the analysis of both model and real food systems involves recognition of the critical role of water as an effective plasticizer of amorphous polymeric, oligomeric, and monomeric food materials (Karel, 1985; Levine and Slade, 1988a, 1989b; Slade et al., 1989). It has become well documented that plasticization by water depresses the Tg [and the melt viscosity (Smith, 1990) and elastic or Young’s modulus (Ablett et al., 1986; Hutchinson et al., 1989; Smith, 1990, 1992a,c; Davies et al., 1991; Ollett et al., 1991; Kalichevsky et al., 1992a,c; Le Meste et al., 1992; Schroeter and Hobelsberger, 1992)] of completely


TABLE V REFERENCES TO STUDIES EMPLOYING METHODS OTHER THAN DSC OR DMA TO

INVESTIGATE GLASS TRANSITIONS AND THEIR EFFECTS IN FOOD INGREDIENTS

AND PRODUCTS'

Method References

Thermomechanical analysis (TMA)

Thermodielectrical analysis (TDEA)

Thermal stimulated current/ relaxation map analysis (TSCI RMA) spectroscopy

Electron spin resonance (ESR)

Nuclear magnetic resonance (NMR)

Fourier-transform infrared (FI'IR)

Brillouin scattering spectroscopy Phosphorescence spectroscopy Fluorescence microscopy Dielectric spectroscopy

Biliaderis et al. (1986a,c), Caldwell et al. (1990), Biliaderis (1990, 1991a), Le Meste and Huang (1991), Le Meste et al. (1991a, 1992), Maurice et al. (1991), Schenz et al. (1991), Chang and Randall (1992), Davis (1992b), Dong (1992), Goff (1992), Hoseney et al. (1992), Huang et al. (1992), Sahagian and Goff (1992, 1993), Goff et al. (1993), Huang (1993a,b), Reid et al. (1993b)

Schiraldi (1990), Ollivon (1991), Scandola et al. (1991), Botham et al. (1992). Hsu and Reid (1992), Noel et al. (1992), Reid et al. (1993b)

Matthiesen et al. (1991), Bruni and Leopold (1992), Huang et al. (1992), Leopold et al. (1992). Megret et al. (1992), Pissis et al. (1992)

Le Meste and Duckworth (1988), Simatos and Karel (1988), Johnson et al. (1990), Le Meste and Simatos (1990), Le Meste et al. (1990, 1991b), Roozen and Hemminga (1990,1991), Bruni and Leopold (1991, 1992), Roozen et al. (1991), Leopold et al. (1992), Hemminga er al. (1993)

(1988), Rubin et al. (1990), Ablett and Lillford (1991), Belton (1991), Chinachoti et al. (1991a), Given (1991), Karger and Ludemann (1991), Pika1 et al. (1991a), Tanner et al. (1991), Chinachoti (1992a), Harrison et al. (1992), Kalichevsky et al. (1992a-d, 1993a), Kalichevsky and Blanshard (1992c), Kawai et al. (1992), Lillford et al. (1992), Newman (1992), Sahagian and Goff (1992), Shogren (1992), Tian and Blanshard (1992b), Willenbucher et al. (1992), Cherian and Chinachoti (1993), Gidley et al. (1993), Oksanen and Zografi (1993)

Young and Scholl (1991), Tian and Blanshard (1992a)

Hosea et al. (1990) Shah and Ludescher (1992, 1993) Nelson (1993) Huang et al. (1992), Haynes and Locke (1993),

Chan et al. (1986), Quinquenet er al. (1988).

Ablett et al. (1986, 1993). Simatos and Karel

Haynes et al. (1993)

(continues)


TABLE V (Continued)

Method References

Mossbauer spectroscopy Mechanical spectrometry

Parak and Nienhaus (1991) Gosline (1987), Masi (1989). Lillie and Gosline

(1990, 1993), Cocero and Kokini (1991, 1993), Roos and Karel (1991f). Anglea et al. (1992a), Davis (1992a), de Graaf et al. (1992), Kokini (1992). Sala and Tomka (1992a), Barrett et al. (1993), Karel et al. (1993b), MacInnes (1993), Simatos and Blond (1993)

Mita (1990), Madeka and Kokini (1992a, 1993), Yano (1993)

Ollett and Parker (1990), Noel et al. (1991), Cocero and Kokini (1992), Cocero et al. (1992b)

Attenburrow et al. (1990, 1992), Davies ef al. (1991), Ollett et al. (1991), Amemiya and Menjivar (1992), Cocero et al. (1992a,b), Kalichevsky and Blanshard (1992b), Kalichevsky et al. (1992a-d, 1993b), Lillford ef al. (1992), Smith (1992a), Attenburrow and Davies (1993), Hegenbart (1993), Kirby et al. (1993), Rohde et al. (1993)

van den Berg (1981, 1986, 1991), Slade and Levine (1985, 1991a), Levine and Slade (1988a), Slade er al. (1989), Oksanen and Zografi (1990, 1993), Arvanitoyannis et al. (1993), Hancock and Zografi (1993)

Dynamic rheometry

Viscometry

Analysis by Instron (or other physical testing apparatus)

Gas permeability or vapor sorption


amorphous or partially crystalline food ingredients and products, and that this Tg depression may be advantageous or disadvantageous to ingredient and product processing, functional properties, and storage stability (Slade and Levine, 1991a; Levine and Slade, 1992b). Recently, there has been expanding interest in the importance of the effect of water as a plasticizer of many different food materials and biopolymers (Table VI).

A unified conceptual approach to research on the glassy state phenomenon and glass transitions in food polymer systems, based on principles translated from synthetic polymer science, has enhanced our qualitative understanding of structure-function relationships in a wide variety of food ingredients and products (Slade/Levine references in Table I). Many workers have recently applied a synthetic polymers or “materials science” (see references of Lillford and co-workers at Unilever in Table VIIA) approach to characterize the glass transition, melting, crystallization, annealing, or


TABLE VI

BIOLOGICAL MATERIALS~

REFERENCES TO STUDIES OF EFFECTS OF WATER AS A PLASTICIZER OF FOOD AND

Starch, amylose, amylopectin: Kainuma and French (1972), van den Berg (1981, 1986, 1991, 1992), Slade (1984), Slade and Levine (1984, 1987b, 1988c,d, 1989, 1991a,b, 1992c, 1993a,b), Wittwer and Tomka (1984), Biliaderis ef al. (1985, 1986a,b,c), Maurice et al. (1985), Ablett et al. (1986), Blanshard (1986, 1987, 1988), Yost and Hoseney (1986), Chungcharoen and Lund (1987), Ring et al. (1987), Russell (1987a), Slade et al. (1987), Tanner et al. (1987,1991). Zeleznak and Hoseney (1987a,b), Levine and Slade (1988a, 1989b, 1992b). Lillford (1988), Lineback and Rasper (1988), Marsh and Blanshard (1988), Skarra et al. (1988), Zobel (1988, 1992a), Zobel et al. (1988), Attenburrow et al. (1989, 1992), Biliaderis and Galloway (1989), Colonna et al. (1989), Hutchinson et al. (1989). Lund (1989), Orford et al. (1989), Russell and Oliver (1989), Shi and Seib (1989, 1992), Stepto and Dobler (1989), Biliaderis (1990, 1991a,b, 1992b), Biliaderis and Seneviratne (1990a,b), Biliaderis and Zawistowski (1990), Ghiasi and Skarra (1990), I’Anson et al. (1990), Johnson et al. (1990), Knutson (1990), Morris (1990). Noel et al. (1990), Roper and Koch (1990), Schiraldi (1990), Smith (1990, 1992a,c), Warburton et al. (1990, 1992), Whittam et al. (1990, 1991), Cairns et al. (1991b), Chinachoti et al. (1991a,b), Garbow and Schaefer (1991), Given (1991), Karathanos et al. (1991), Lai and Kokini (1991), Larsson (1991), Larsson and Eliasson (1991), Liu and Lelievre (1991b. 1992a,b), Ollett er al. (1991, 1993a,b), Patil (1991), Roos and Karel (1991d). Scandola et al. (1991), Tomka (1991), Bizot et al. (1992), Colonna and Buleon (1992), Dong (1992), Eerlingen and Delcour (1992), Eliasson (1992), Gudmundsson (1992), Gudmundsson and Eliasson (1992). Hallberg and Chinachoti (1992), Huang et al. (1992), Jovanovich et al. (1992), Kaletunc and Breslauer (1992), Kalichevsky and Blanshard (1992a-d), Kalichevsky et al. (1992a,b, 1993a), Karathanos and Saravacos (1992), Kim and Walker (1992), Kokini et al. (1992c), Le Meste et al. (1992), Lelievre (1992a), Lillford et al. (1992), Lim et al. (1992). Michniewicz et al. (1992), Noel and Ring (1992), Parker and Smith (1992), Rajagopalan and Seib (1992), Roos (1992d), Sala and Tomka (1992a,b), Schroeter and Hobelsberger (1992), Shogren (1992), Shogren et al. (1992), Thompson (1992), Tian and Blanshard (1992b), Tolstoguzov (1992), Willenbucher et al. (1992), Zasypkin et al. (1992), Zhang and Jackson (1992). Arvanitoyannis et al. (1993), Attenburrow and Davies (1993), BeMiller (1993), Donald ef al. (1993), Eliasson and Larsson (1993), Gidley et al. (1993), Huang (1993b), Kirby et al. (1993), Seow and Teo (1993), Seow and Thevamalar (1993)

Starch hydrolysis products: Cole et al. (1983, 1984), Flink (1983), Karel and Flink (1983), Saleeb and Pickup (1985, 1989), Levine and Slade (1986, 1988a,b,c, 1989a,b,c, 1990, 1991, 1992b), Karel and Langer (1988), Orford et al. (1989), Levine et al. (1991, 1992), Lim and Reid (1991, 1992), Ring and Whittam (1991), Roos and Karel (1991b,d), Roozen and Hemminga (1991), Slade and Levine (1991a,b), Roos (1992b), Ablett et al. (1993), Hemminga ef al. (1993), Nelson (1993), Ollett ef al. (1993a)

Low-MW sugars: White and Cakebread (1966), Cakebread (1969), MacKenzie (1977), Soesanto and Williams (1981), Franks (1982,1985, 1986a, 1989, 1990, 1991a, 1992b, 1993a,b), Herrington and Branfield (1984), Schenz et al. (1984, 1991). Karel (1985, 1986, 1989, 1992). Slade and Levine (1985, 1988a,b, 1991a,b, 1992c, 1993b), Chan er al. (1986), Blanshard and Franks (1987), Roos (1987, 1992a,b,d), Karel and Langer (1988), Levine and Slade (1988a-d, 1989a-d, 1990, 1992a,b), Simatos and Karel (1988), Blond (1989, 1993), Finegold et al. (1989), Green and Angel1 (1989), Orford et al. (1989, 1990),

(continues)


TABLE VI (Continued)

Simatos et al. (1989). Williams and Carnahan (1989, 1990), Williams and Leopold (1989), Ahlneck and Zografi (1990), Caldwell et al. (1990), Franks and Grigera (1990), Franks and Hatley (1990, 1992, 1993), Le Meste and Simatos (1990), Noel et al. (1990, 1991), Reid (1990), Roos and Karel (1990, 1991a,c-h, 1992, 1993). Roozen and Henminga (1990, 1991), Rubin et al. (1990), Blond and Simatos (1991), Bruni and Leopold (1991, 1992), Franks and van den Berg (1991), Franks ef al. (1991), Hatley et al. (1991), Izzard et al. (1991), Karel and Saguy (1991), Koster (1991). Le Meste and Huang (1991), Levine et al. (1991, 1992), Matthiesen et al. (1991), Maurice et al. (1991), Shimada et al. (1991), Simatos and Blond (1991), Ablett et al. (1992a,b,c, 1993), Berail et al. (1992), Chinachoti (1992a), Huang (1993a,b), Jouppila and Roos (1992), Karmas et al. (1992), Labrousse et al. (1992), Levi and Karel (1993b), Nesvadba (1992b), te Booy et al. (1992), van den Berg (1992), Arvanitoyannis and Blanshard (1993b), Karel et al. (1993a), Nelson (1993), van den Berg et al. (1993), Williams et al. (1993)

Polyhydric alcohols: Reid (1985), Levine and Slade (1988a-d, 1989a,b,c, 1990, 1992b), Quinquenet et al. (1988), Slade and Levine (1988b, 1991a), Franks and Grigera (1990), Franks and van den Berg (1991), Koster (1991), Ablett et al. (1992a), Jouppila and Roos (1992), Roos (1992a)

Nonstarch polysaccharides: Kararli and Catalan0 (1990), Paakkonen and Roos (1990), Yoshida et al. (1990, 1992), Hegenbart (1991), Nishinari et al. (1991), Scandola et al. (1991), Slade and Levine (1991a,b), Ablett et aL (1993), Appelqvist et al. (1993), Gidley et al. (1993), Yano (1993)

(1987), Edwards et al. (1987), Levine and Slade (1988a, 1989b, 1992b), Lillford (1988), Attenburrow et a/. (1989, 1990, 1992), Fujio and Lim (1989), Slade et al. (1989), Hoseney and Rogers (1990), Stauffer (1990). Davies et aI. (1991). Garbow and Schaefer (1991), Given (1991), Hoseney (1991,1992). Slade and Levine (1991a, 1992c, 1993b), Aynie et al. (1992a,b), Cocero et nl. (1992a,b), de Graaf et al. (1992), Dong (1992), Gontard et al. (1992a,b,c), Hallberg and Chinachoti (1992), Huang et al. (1992), Kalichevsky et al. (1992c,d, 1993a). Kalichevsky and Blanshard (1992a,c), Kokini (1992), Kokini et al. (1992d), Le Meste et al. (1992), Lillford et al. (1992), Michniewicz et al. (1992), Saunders et al. (1992), Summers (1992), Tian and Blanshard (1992b), Attenburrow and Davies (1993), Huang (1993b), Lawton and Wu (1993)

Glutenin: Cocero and Kokini (1991), Cocero et aL (1992a,b), de Graaf et al. (1992), Kalichevsky et al. (1992c), Kokini et al. (1992d)

Gliadin: Cocero et al. (1992a,b), de Graaf et al. (1992), Kalichevsky et al. (1992c), Kokini er al. (1992d), Madeka and Kokini (1992a,b)

Gelatin: Jolley (1970), Yannas (1972), Borchard et al. (1980), Marshall and Petrie (1980), Slade and Levine (1984, 1987a, 1991a), Tomka (1986), Levine and Slade (1988a), Slade et al. (1989), Kalichevsky and Blanshard (1992~)

Collagen: Yannas (1972), Batzer and Kreibich (1981), Levine and Slade (1988a), Slade et al. (1989), Slade and Levine (1991a)

Elastin: Kakivaya and Hoeve (1975), Hoeve and Hoeve (1978,1980), Hoeve (1980), Atkins (1987), Gosline (1987), Levine and Slade (1988a), Slade et al. (1989), Lillie and Gosline (1990, 1993), Slade and Levine (1991a)

Zein: Cocero et al. (1992a,b), Kokini et al. (1992d), Lawton (1992a,b), Madeka and Kokini (1992a,b), Magoshi et al. (1992)

Gluten: Slade (1984), Ablett et al. (1986, 1988), Hoseney et al. (1986), Doescher et al.

(continues)


TABLE VI (Continued)

Other proteins or polypeptides: Le Meste and Duckworth (1988), Le Meste et al. (1990, 1991b), Levine and Slade (1988a), Slade et af. (1989). Pikal (1990a, 1993), Pikal and Shah (1990), Pikal er al. (1990), Belton (1991), Franks (1991c), Fujio et al. (1991), Slade and Levine (1991a), Sochava et al. (1991), Tolstoguzov (1991, 1992), Aguilera et af. (1992), Angel1 et al. (1992), Kalichevsky and Blanshard (1992a.c). Milczarek et al. (1992), Pissis er al. (1992), Roy et al. (1992), Sochava and Smirnova (1992), Tolstoguzov and Nesmeyanov (1992), Zasypkin et af. (1992), Kalichevsky et af. (1993b), Karel et al. (1993a). Lillie and Gosline (1993)

Lysozyme: Bone and Pethig (1982, 1985), Poole and Finney (1983, 1984), Finney and Poole (1984). Morozov and Gevorkian (1985), Franks (1988), Levine and Slade (1988a), Lillford (1988), Slade er af. (1989), Myers-Betts and Baianu (1990), Slade and Levine (1991a), Sochava et af. (1991). Pissis er al. (1992), Shah and Ludescher (1992), Sochava and Smirnova (1992)

Other enzymes: Poole and Finney (1984), Morozov and Gevorkian (1985), Levine and Slade (1988a). Slade er al. (1989), Slade and Levine (1991a), Sochava et af. (1991)

Cellulose, hemicelluloae, lignin: Salmen and Back (1977), Kelley et af. (1987), Ostberg et al. (1990), Pissis et al. (1992), Newman (1992)

Poly(hydroxybutyrate): Harrison ef al. (1992)


gelationhetwork formation behavior of major food polymers such as starch and gluten (Table VIIA). Table VIII lists specific references to publications dealing with the glassy state phenomenon in starch or gluten, and Table IX lists those applying a polymer science approach to various aspects of studies of starch gelatinization or retrogradation. In recent years, many other workers have also supported the perspective based on the fundamental behavioral similarities between synthetic polymer-plasticizer and food molecule-water systems (Table VIIB), first popularized by Slade and Levine.

A central theme of the food polymer science approach focuses on the effect of water as a plasticizer on the glass transition and resulting diffusion- limited behavior of water-compatible or water-sensitive amorphous materials or amorphous regions of partially crystalline materials (Levine and Slade, 1988a, 1989b, 1992b; Slade et al., 1989; Slade and Levine, 1991a,b). Water-compatible food polymers such as starch, gluten, and gelatin, for which water is an efficient plasticizer but not necessarily a good solvent, exhibit essentially the same physicochemical responses to plasticization by water as do many water-compatible synthetic polymers (Ellis, 1988; Ahl- neck and Zografi, 1990; Oksanen and Zografi, 1990, 1993; Buera et al., 1992; Harrison et al., 1992; Kalichevsky and Blanshard, 1992d; Hancock and Zografi, 1993) and many readily soluble monomeric and oligomeric carbohydrates (Table XE). This fact illustrates two underlying precepts of


TABLE VII REFERENCES TO STUDIES THAT HAVE EMPLOYED A SYNTHETIC POLYMERS OR

MATERIALS SCIENCE APPROACH TO CHARACTERIZATION OF FOOD

MOLECULE-WATER SYSTEMSa

(A) Studies of Starch or Gluten (Researcher: References) Slade and Levine: Slade (1984), Slade and Levine (1984,1985, 1987b, 1988a-d, 1989,1990,

1991a,b, 1992a-c, 1993a-c), Slade et al. (1987, 1989, 1993), Levine and Slade (1986, 1988a,b,c, 1989a,b,c, 1990, 1991, 1992b, 1993), Levine et nl. (1991, 1992)

Lelievre: Lelievre (1976, 1992a,b), Liu and Lelievre (1991a,b, 1992a,b), Liu et al. (1991) Tomka: (1991), Wittwer and Tomka (1984), Sala and Tomka (1992a,b), Willenbucher et al.

Biliaderis and Maurice: Biliaderis et al. (1985, 1986a,b,c), Maurice et al. (1985), Biliaderis and Galloway (1989), Biliaderis and Zawistowski (1990), Biliaderis and Seneviratne (1990a,b), Biliaderis (1990, 1991a,b, 1992a,b), Biliaderis and Tonogai (1991)

Blanshard: Blanshard (1986, 1987, 1988), Edwards et al. (1987), Marsh and Blanshard (1988), Blanshard and Mitchell (1992), Kalichevsky et al. (1992a-d, 1993a), Kalichevsky and Blanshard (1992b,c,d), Tian and Blanshard (1992b), Arvanitoyannis et al. (1993)

Lillford: (1988), Ablett et al. (1986, 1988, 1993). Edwards et al. (1987), Attenburrow et al. (1989, 1990, 1992), Ablett and Lillford (1991). Davies et al. (1991), Lillford et al. (1992), Attenburrow and Davies (1993)

Hoseney: Hoseney (1991, 1992), Hoseney et al. (1986,1992), Yost and Hoseney (1986), Doescher et al. (1987), Zeleznak and Hoseney (1987a,b), Hoseney and Rogers (1990, 1993), Faubion and Hoseney (1989), Dong (1992)

(1992)

Russell: (1987a). Russell and Oliver (1989) Lund: (1989), Chungcharoen and Lund (1987) Morris: (1990), Ring et al. (1987), I’Anson et al. (1990), Cairns et al. (1991b) Ring: Ring et al. (1987), Orford et al. (1989), Noel et al. (1990), Whittam et al. (1990,

Zobel: Zobel (1988, 1992a,b), Zobel et al. (1988) Lineback: Lineback and Rasper (1988), Saunders et al. (1992) Ghiasi: Skarra et al. (1988), Ghiasi and Skarra (1990) Seib: Shi and Seib (1989, 1992), Rajagopalan and Seib (1992) Smith: Smith (1990, 1992a,b,c), Hutchinson et al. (1989), Orford et al. (1989), Warburton et

1991), Ring and Whittam (1991), Botham et al. (1992), Noel and Ring (1992)

al. (1990, 1992), Ollett et al. (1991, 1993a,b), Parker and Smith (1992), Donald et al. (1993), Kirby et al. (1993)

Gudmundsson and Eliasson (1992), Gudmundsson (1992). Eliasson and Larsson (1993) Eliasson: Eliasson (1989, 1992), Larsson and Eliasson (1991), Larsson (1991),

Colonnn: Colonna et al. (1989), Colonna and Buleon (1992) Doublier: Doublier (1989) Fujio: Fujio and Lim (1989) Masi: Masi (1989) Mita: Mita (1990, 1992) Le Meste and Huang: Le Meste et al. (1991a, 1992), Aynie et al. (1992a,b), Huang et al.

(1992) Kokini: Kokini (1992), Cocero and Kokini (1991, 1993), Lai and Kokini (1991), Cocero et

al. (1992a,b), de Graaf etal. (1992), Kokini et al. (1992a,d), Madeka and Kokini (1992a, 1993)

Stauffer: Stauffer (1990) Shogren: Shogren (1992), Shogren et al. (1992)

(continues)


TABLE VII (Continued)

Lawton: Lawton (1992c), Lawton and Wu (1993) Reid: Taylor et al. (1992) Breslauer: Kaletunc and Breslauer (1992) Sture: Stute (1992) Setser: Kim and Setser (1992) Schroeter: Schroeter and Hobelsberger (1992) Guilbert: Gontard et al. (1992a,b,c) Thompson: Thompson (1992) Rollings: Sullivan et al. (1992) Delcour: Eerlingen and Delcour (1992). Eerlingen et al. (1993) Anon: Jovanovich et al. (1992) Jackson: Zhang and Jackson (1992) Haynes: Haynes et al. (1993) Seow: Seow and Teo (1993), Seow and Thevamalar (1993) Schiraldi: Zanoni et al. (1993)

(B) Other Studies Karel (1985, 1986, 1989, 1990, 1991a,b,c, 1992), Campanella et al. (1987), Gosline (1987).

Le Meste and Duckworth (1988). Simatos and Karel (1988), Cairault et al. (1989), Finegold et al. (1989). Franks (1989, 1991c, 1992a, 1993a,b), Simatos et al. (1989), Franks and Grigera (1990), Johnson et al. (1990). Katsuta and Kinsella (1990), Le Meste and Simatos (1990), Le Meste et al. (1990, 1991b), Lillie and Gosline (1990, 1993), Noel et al. (1990, 1991, 1992, 1993), Ollett and Parker (1990), Orford et al. (1990), Roos and Karel (1990, 1991a-f,h, 1992, 1993), Roozen and Hemminga (1990, 1991), Schiraldi (1990), Van Scoik and Carstensen (1990), Yoshida et al. (1990, 1992), Belton (1991), Blond and Simatos (1991), Bruni and Leopold (1991), Cameron and Donald (1991), Chang and Baust (1991a,b,c), Davis (1991), Franks et al. (1991), Fujio et al. (1991), Given (1991), Izzard et al. (1991), Karathanos et al. (1991, 1992), Karel and Saguy (1991), Kohyama and Nishinari (1991), Le Meste and Huang (1991), Lim and Reid (1991, 1992), Lin et al. (1991), Matthiesen et al. (1991), O’Brien (1991), Parak and Nienhaus (1991), Raemy and Lambelet (1991), Roozen et al. (1991), Scandola et al. (1991), Shukla (1991), Simatos and Blond (1991, 1993), Sochava et al. (1991), Tolstoguzov (1991), van den Berg (1991, 1992). Young and Scholl (1991), Aguilera et al. (1992), Amemiya and Menjivar (1992), Angel1 et al. (1992), Anglea et al. (1992a), Bagley (1992), Berail et al. (1992), Best (1992), Bruin (1992), Buera and Karel (1992), Buera et al. (1992), Chinachoti (1992a), Chuy and Labuza (1992), Franks and Hatley (1992, 1993), Goff (1992), Harrison et al. (1992), Huang (1992, 1993a,b), Jouppila and Roos (1992), Karmas et al. (1992), Lawton (1992a), Levi and Karel (1992b, 1993b), Milczarek et al. (1992), Nelson and Labuza (1992a-c, 1993), Nelson et al. (1992), Newman (1992), Peleg (1992, 1993), Pissis et al. (1992), Reid (1992b), Reid and Hsu (1992), Roos (1992a-d), Roy et al. (1992), Sahagian and Goff (1992, 1993), Sapru and Labuza (1992a,b), Sochava and Belopolskaya (1992), Sochava and Smirnova (1992), te Booy et al. (1992), Tian and Blanshard (1992a), Zasypkin et al. (1992), Allen (1993), Anonymous (1993a), Appelqvist et al. (1993), Arvanitoyannis and Blanshard (1993a,b), Blond (1993), De Bry (1993), Gidley et al. (1993), Hancock and Zografi (1993), Hartel (1993), Haynes and Locke (1993), Haynes et al. (1993), Hegenbart (1993), Hemminga et al. (1993), Huang et al. (1993), Kalichevsky et al. (1993b). Karel et al. (1993a,b), Kerr et al. (1993). MacInnes (1993), Madeka and Kokini (1993). Nelson (1993), O’Donnell (1993), Oksanen and Zografi (1993), Pika1 (1993), Ramanujam et al. (1993), Suzuki and Franks (1993), Williams et al. (1993), Yano (1993), Simatos (1994)



TABLE VIII

IN STARCH OR GLUTEN"

REFERENCES TO PUBLICATIONS DEALING WITH THE GLASSY STATE PHENOMENON

(A) References for Methods Used to Determine TB of Starch, Amylose, and/or Amylopectin

Calculated from theory: Blanshard (1986,1988), Marsh and Blanshard (1988) DSC: Slade (1984), Slade and Levine (1984, 1987b, 198&), Wittwer and Tomka (1984),

Maurice et al. (1985), Yost and Hoseney (1986), Zeleznak and Hoseney (1987a), Zobel et al. (1988). Orford et al. (1989), Stepto and Dobler (1989), Biliaderis (1990, 1991b), Noel et al. (1990), Roper and Koch (1990), Warburton et al. (1990, 1992), Liu and Lelievre (1991b, 1992b), Roos and Karel (1991d,e), Tomka (1991), Whittam et al. (1991), Bizot et 01. (1992), Eerlingen and Delcour (1992), Kaletunc and Breslauer (1992), Kalichevsky and Blanshard (1992d), Lelievre (1992a), Noel and Ring (1992), Zobel (1992a), Donald et al. (1993), Eliasson and Larsson (1993), Gidley et al. (1993), Seow and Teo (1993), van den Berg et al. (1993)

TMA: Le Meste et al. (1991a, 1992), Dong (1992), Hoseney et al. (1992) DSC and TMA: Biliaderis et al. (1986a), Biliaderis (1991a) DSC and NMR: Given (1992), Shogren et al. (1992) D M T A and NMR: Lillford et al. (1992) DSC, DMTA, and NMR: Kalichevsky et al. (1992a,b), Kalichevsky and Blanshard (1992c),

DSC, DMTA, and TDEA: Scandola et al. (1991) DSC and mechanical spectroscopy: Sala and Tomka (1992a) TMA, TSC, and dielectric spectroscopy: Huang et al. (1992) Mechanically (as T at which Characteristic orders-of-magnitude decrease in modulus is

Shogren (1992)

observed): Noel et 01. (1990), Smith (1990, 1992a), Ollett et al. (1991), Kalichevsky et al. (1992a,b), Kalichevsky and Blanshard (1992b,c)

Instron (acoustics): Attenburrow et al. (1992) Other methods: van den Berg (1981,1986), Schroeter and Hobelsberger (1992)

(B) References for Methods Used to Determine T, of Gluten, Glutenin, and/or Gliadin

(1988a), Fujio and Lim (1989), Slade et al. (1989), Hoseney and Rogers (1990), Noel et al. (1990), Hoseney (1991), Cocero et al. (1992a,b), Cherian and Chinachoti (1993), Lawton and Wu (1993)

DSC: Slade (1984), Hoseney et al. (1986), Doescher et al. (1987), Levine and Slade

TMA: Le Meste et al. (1991a, 1992), Dong (1992), Hoseney et al. (1992) DMA: Mita (1990), Summers (1992) Mechanically (as T at which characteristic orders-of-magnitude decrease in modulus is

observed): Attenburrow et al. (1990), Davies et al. (1991), Kalichevsky et al. (1992c), Kalichevsky and Blanshard (1992~)

Madeka and Kokini (1993) DSC and mechanical spectrometry: Cocero and Kokini (1991, 1993), de Graaf et al. (1992),

D M T A and NMR: Lillford et al. (1992) DSC and DMTA: Gontard et al. (1992a,b) D M T A and Instron (acoustics): Attenburrow et al. (1992) Instron (acoustics): Attenburrow and Davies (1993) DSC, DMTA, and NMR: Kalichevsky et a/. (1992~). Kalichevsky and Blanshard (1992~) TMA, TSC, and dielectric spectroscopy: Huang et al. (1992)

(continues)


TABLE VIII (Continued)

(C) References Dealiig with Effects of Tg for Starch andlor Gluten on Processing and Product Properties

Slade (1984), Slade and Levine (1984, 1987b, 1990, 1991a, 1992c, 1993a,b), Maurice et al. (1985), Blanshard (1986), Hoseney et al. (1986), Yost and Hoseney (1986), Doescher et al. (1987), Zeleznak and Hoseney (1987a), Levine and Slade (1988a, 1989b, 1992b), Zobel (1988, 1992a,b), Colonna et al. (1989), Faubion and Hoseney (1989), Shi and Seib (1989, 1992), Slade et al. (1989, 1993), Biliaderis (1990, 1991a,b, 1992b), Hoseney and Rogers (1990, 1993), Noel er al. (1990), Schiraldi (1990), Stauffer (1990), Warburton et al. (1990, 1992), Cairns er al. (1991b), Cocero and Kokini (1991), Davies et al. (1991), Given (1991), Hoseney (1991, 1992), Karathanos et al. (1991), Larsson (1991), Larsson and Eliasson (1991), Le Meste et al. (1991a, 1992), O’Brien (1991), Ollett et al. (1991, 1993a,b), Shukla (1991), Spratt et al. (1991), Tanner ef al. (1991), Amemiya and Menjivar (1992), Attenburrow et al. (1992), Aynie et al. (1992a,b), Best (1992). Bizot et al. (1992). Blanshard and Mitchell (1992), Cocero et al. (1992b), Colonna and Buleon (1992), Dong (1992), Gontard et al. (1992a,c), Gudmundsson (1992), Gudmundsson and Eliasson (1992), Hallberg and Chinachoti (1992), Huang et al. (1992), Jovanovich er al. (1992), Kaletunc and Breslauer (1992), Kalichevsky et al. (1992a,b,c), Karathanos and Saravacos (1992), Kokini er al. (1992c,d), Lawton (1992b), Livings et al. (1992), Parker and Smith (1992), Rajagopalan and Seib (1992), Rao er al. (1992), Roos (1992d), Sala and Tomka (1992a), Saunders et al. (1992), Shogren (1992), Shogren et al. (1992), Smith (1992b,c), Stute (1992), Sullivan et al. (1992), Tian and Blanshard (1992b), Tolstoguzov (1992), van den Berg (1992), Zasypkin et al. (1992), Zhang and Jackson (1992), Anonymous (1993b), Attenburrow and Davies (1993), BeMiller (1993), Cherian and Chinachoti (1993), Donald et al. (1993), Eerlingen et al. (1993), Eliasson and Larsson (1993). Kirby et al. (1993). Lawton and Wu (1993), Mestres et al. (1993), Rohde et al. (1993), Seow and Teo (1993), Seow and Thevamalar (1993), Zanoni et al. (1993)


the food polymer science approach: (1) synthetic amorphous polymers and glass-forming aqueous food materials are fundamentally similar in behavior; and (2) food ingredients can be viewed generically as members of homologous families of completely amorphous or partially crystalline polymers, oligomers, and monomers, soluble in and/or plasticized by water (Levine and Slade, 1988a).

E. “WATER DYNAMICS” AND “GLASS DYNAMICS”

On a theoretical basis of structure-property relationships for synthetic polymers (Table 11), functional properties of food materials during processing and product storage can be successfully explained and often predicted (Sladekevine references in Table I). The food polymer science approach unifies structural aspects of foods, viewed as completely amorphous or partially crystalline polymer systems [the latter based on the classical


TABLE IX REFERENCES TO PUBLICATIONS DESCRIBING A POLYMER SCIENCE APPROACH TO

STUDIES OF STARCH GELATINIZATION AND/OR RETROGRADATIONa

(A)Retrogradation as a Nonequilibrium Polymer Crystallization Process Slade (1984), Slade and Levine (1984, 1987b, 1989d, 1989), Miles et al. (1985a,b), Ring

(1985a,b), Ablett et al. (1986), Blanshard (1986), Ring and Orford (1986), Ring et al. (1987), Russell (1987b), I’Anson et al. (1988), Mestres et al. (1988, 1993), Skarra et al. (1988), Zobel (1988), Clark et al. (1989), Colonna et al. (1989), Gidley (1989), Gidley and Bulpin (1989), Levine and Slade (1989b), Lund (1989), Russell and Oliver (1989), Biliaderis (l990,199la,b, 1992b), Morris (1990), Schiraldi (1990), Biliaderis and Tonogai (1991), Cairns et al. (1991a,b), Cameron and Donald (1991), Chang and Liu (1991), O’Brien (1991), Patil (1991), Shukla (1991), Tanner et al. (1991), Colonna and Buleon (1992), German et al. (1992), Gudmundsson (1992), Jurgens et al. (1992), Michniewicz et al. (1992), Mita (1992), Rao er al. (1992), Roos (1992d), Shi and Seib (1992), Shogren (1992), Shogren et al. (1992), Zhang and Jackson (1992), Eerlingen et al. (1993)

(B) Gelatinization as a Nonequilibrium Polymer Melting Process Slade (1984), Slade and Levine (1984, 1987b, 1988b,c,d, 1989), Kuge and Kitamura (1985),

Maurice et al. (1985), Reid and Charoenrein (1985), Biliaderis et al. (1986a), Blanshard (1986, 1987, 1988), Yost and Hoseney (1986), Burros et al. (1987), Chungcharoen and Lund (1987), Paton (1987), Russell (1987a), Zobel (1988, 1992a), Zobel et al. (1988), Levine and Slade (1989b), Lund (1989), Russell and Oliver (1989), Shi and Seib (1989, 1992), Biliaderis (1990, 1991a,b, 1992b), Knutson (1990), Morris (1990), Schiraldi (1990), Whittam et al. (1990), Chinachoti et al. (1991a), Gidley and Cooke (1991), Kohyama and Nishinari (1991). Larsson and Eliasson (1991), Liu and Lelievre (1991a,b, 1992b), Liu et al. (1991), O’Brien (1991), Patil (1991), Shukla (1991), Colonna and Buleon (1992), Eerlingen and Delcour (1992), Gudmundsson (1992), Gudmundsson and Eliasson (1992), Kalichevsky et al. (1992a), Karathanos and Saravacos (1992), Lelievre (1992a), Noel and Ring (1992), Rajagopalan and Seib (1992), Roos (1992d), BeMiller (1993), Eliasson and Larsson (1993), Seow and Teo (1993), Seow and Thevamalar (1993). Zanoni et al. (1993)

(C) Application of the “Fringed Micelle” Structural Model to Starch Guilbot and Godon (1984), Slade (1984), Slade and Levine (1984, 1987a,b), Zobel (1988,

1992a), Lund (1989), Chinachoti et al. (1991b), Kohyama and Nishinari (1991), Lai and Kokini (1991), Liu et al. (1991), Colonna and Buleon (1992), Eerlingen and Delcour (1992), German et al. (1992), Kokini er al. (1992a), Lelievre (1992a), Mita (1992), Stute (1992), Seow and Thevamalar (1993)

(D) Indirect Plasticizing Effect of Water on T, of Partially Crystalline Starch Lelievre (1976, 1992a), Slade (1984), Slade and Levine (1984, 1987b), Maurice et al. (1985),

Biliaderis et al. (1985, 1986a,b,c), Zobel (1988), Zobel et al. (1988), Biliaderis and Galloway (1989), Biliaderis (1990, 1991a,b, 1992b), Biliaderis and Seneviratne (1990a,b), Biliaderis and Zawistowski (1990), Whittam et al. (1990, 1991). Liu and Lelievre (1991b, 1992a), Gudmundsson (1992), Gudmundsson and Eliasson (1992), Kim and Walker (1992), BeMiller (1993), Seow and Thevamalar (1993)

(continues)


TABLE IX (Continued )

(E) Inappropriateness of the Flory-Huggins Equilibrium Thermodynamic Treatment of Gelatinization

Slade (1984), Slade and Levine (1984, 1987b, 1988c,d, 1989), Biliaderis et al. (1986a), Burros er al. (1987), Paton (1987), Russell (1987a), Mestres et al. (1988), Levine and Slade (1989b, 1992b), Lund (1989), Biliaderis (1990, 1991a,b, 1992b), Knutson (1990), Whittam et al. (1990, 1991), Liu and Lelievre (1991b), Ring and Whittam (1991), Colonna and Buleon (1992), Gudmundsson (1992), Lelievre (1992a), Roos (1992d), Eliasson and Larsson (1993)

(F) Antiplasticizing Effect of Sugars (i.e., Sugar Solutions Relative to Water Alone) on Gelatinization or Retrogradation

Slade (1984), Slade and Levine (1984, 1987b, 1988b,d, 1989, 1993b), Blanshard (1987, 1988), Chungcharoen and Lund (1987), Levine and Slade (1989b, 1992b), Lund (1989), Biliaderis (1990, 1992a), Biliaderis and Seneviratne (1990a,b), Hoseney and Rogers (1990, 1993), I’Anson et al. (1990), Johnson et al. (1990), Morris (1990), Sobczynska et al. (1990), Cairns et al. (1991b), Chinachoti er al. (1991a,b), Kohyama and Nishinari (1991), Nishinari et al. (1991), Shukla (1991). Chinachoti (1992a), Eerlingen and Delcour (1992), Eliasson (1992), Gudmundsson (1992), Kim and Setser (1992), Kim and Walker (1992), Lim er al. (1992), Rajagopalan and Seib (1992). Roos (1992d), Eliasson and Larsson (1993)


“fringed micelle” morphological model (Flory, 1953; Wunderlich, 1973; Billmeyer, 1984) shown in Fig. 2 (Slade and Levine, 1987a)], with functional aspects, dependent on mobility and described in terms of the integrated concepts of “water dynamics” and “glass dynamics.” Through this unifica- tion, the appropriate kinetic description of the non-equilibrium thermomechanical behavior of food systems has been illustrated in terms of state diagrams of temperature vs. composition, viewed in the context of a “dynamics map,” shown in Fig. 3 (Slade and Levine, 1988b). The dynamics map has been used (Slade and Levine, 1988b) to describe the behavior of water-compatible food polymer systems that exist in kinetically metastable glassy or rubbery states always subject to conditionally beneficial or detri- mental plasticization by water. The map domains of moisture content and temperature, traditionally described with only limited success using concepts such as “water activity” (A,) and “bound water,” have been treated alternatively in terms of water dynamics (Slade and Levine, 1991a). As the name implies, water dynamics focuses on (1) the mobility (Chinachoti et al., 1991b; Larsson, 1991; Tanner et al., 1991; Goff, 1992) and eventual “availability” (Franks, 1991b; Ablett and Lillford, 1991; Chinachoti, 1992a) of the plasticizing diluent, be it water alone or an aqueous solution; and (2) an interpretive approach to understanding how to control the mobility


TABLE X REFERENCESO

A. Publications in Which Support Is Expressed for the Perspective Provided by the Concepts of Glass Dynamics and Water Dynamics

(1987), Chungcharoen and Lund (1987), Russell (1987a), Gould and Christian (1988), Karel and Langer (1988), Le Meste and Duckworth (1988), Lillford (1988), Lineback and Rasper (1988), Marsh and Blanshard (1988), Quinquenet et al. (1988), Simatos and Karel (1988), Skarra et al. (1988), Takahashi et al. (1988), Zobel (l988,1992a,b), BeMiller (1989), Biliaderis and Galloway (1989), Cairault et al. (1989), Colonna et al. (1989), Franks (1989, 1990, 1991a,b,c, 1992a,b, 1993a,b), Finegold et al. (1989). Karel (1989, 1990, 1991c, 1992), Lund (1989), Orford et al. (1989). Russell and Oliver (1989), Shi and Seib (1989, 1992), Simatos et al. (1989), Ahlneck and Zografi (1990), Biliaderis (1990, 1991a,b, 1992a,b), Biliaderis and Zawistowski (1990), Biliaderis and Seneviratne (1990a), Caldwell et al. (1990, 1992), Franks and Grigera (1990), Franks and Hatley (1990, 1992, 1993), Hosea et al. (1990), Hoseney and Rogers (1990, 1993), I’Anson et al. (1990), Johnson et al. (1990), Johnston-Banks (1990), Kararli and Catalan0 (1990), Kararli et al. (1990), Le Meste and Simatos (1990), Le Meste et al. (1990, 1991a,b, 1992), Morris (1990), Myers-Betts and Baianu (1990), Noel et al. (1990, 1991), Oksanen and Zografi (1990, 1993), Ollett and Parker (1990), Paakkonen and Roos (1990), Pikal (1990a,b, 1993), Pikal and Shah (1990), Pikal et al. (1990, 1991a,b), Reid (1990, 1992a,b), Roos and Karel (1990, 1991a-h, 1992, 1993), Roozen and Hemminga (1990, 1991) Rubin et al. (1990), Schiraldi (1990), Sobczynska et al. (1990), Stauffer (1990), Warburton et al. (1990), Williams et al. (1990, 1993), Ablett and Lillford (1991), Belton (1991), Blond and Colas (1991), Blond and Simatos (1991), Cairns et al. (1991b), Chang and Baust (1991a,c), Chinachoti et al. (1991a,b), Cocero and Kokini (1991), Davis (1991), Franks and van den Berg (1991). Franks et al. (1991), Fujio et al. (1991), Gaines (1991, 1993), Garbow and Schaeffer (1991), Given (1991), Graf and Saguy (1991), Hatley (1991), Hatley and Franks (1991). Hatley et al. (1991), Hegenbart (1991, 1993), Izzard et al. (1991), Karathanos et al. (1991), Karel and Saguy (1991), Kohyama and Nishinari (1991), Lai and Kokini (1991), Larsson (1991), Larsson and Eliasson (1991), Le Meste and Huang (1991), Lim and Reid (1991, 1992). Liu and Lelievre (1991a,b, 1992a,b), Liu et al. (1991), Ma and Hanvalkar (1991), MacDonald and Lanier (1991), Matthiesen et al. (1991), Maurice et al. (1991), Nishinari et al. (1991), O’Brien (1991), Patil (1991), Ring and Whittam (1991), Roozen et al. (1991), Schenz et al. (1991), Shimada et al. (1991), Shukla (1991), Simatos and Blond (1991, 1993), Tanner et al. (1991), Wasylyk and Baust (1991), Ablett et al. (1992a,b,c, 1993). Aguilera et al. (1992), Angel1 et al. (1992). Anglea et al. (1992b), Attenburrow et al. (1992), Berail et al. (1992), Best (1992), Bizot et al. (1992), Blond (1992, 1993), Boskovic et al. (1992), Buera and Karel (1992). Buera et al. (1992), Chang and Randall (1992), Chinachoti (1992a), Chuy and Labuza (1992), Cocero et al. (1992a,b), Colonna and Buleon (1992), de Graaf et al. (1992), Dong (1992), Eerlingen and Delcour (1992), Eliasson (1992), Gaines et al. (1992a-c, 1993a,b), Goff (1992), Gontard et al. (1992~). Gudmundsson (1992), Gudmundsson and Eliasson (1992), Hallberg and Chinachoti (1992), Hanover (1992), Huang (1992, 1993a,b), Huang et al. (1992,1993), Jouppila and Roos (1992), Kaletunc and Breslauer (1992), Kalichevsky and Blanshard (1992a,b), Kalichevsky et al. (1992a-d, 1993a,b), Karathanos and Saravacos (1992), Karmas et al. (1992), Katsuta ef al. (1992), Kim and Setser (1992), Kim and Walker (1992), Kokini (1992). Kokini et al. (1992a,c), Kresin and Rau (1992), Labrousse et al. (1992), Labuza and Baisier (1992), Labuza and Nelson (1992), Lelievre (1992a),

Blanshard (1986, 1987, 1988), van den Berg (1986,1991,1992), Blanshard and Franks

(continues)


TABLE X (Continued)

Levi and Karel (1992a, 1993b), Lim ef al. (1992). MacInnes (1992, 1993), Michniewicz et al. (1992), Mita (1992), Nelson and Labuza (1992a-c), Nesvadba (1992a), Noel and Ring (1992), Peleg (1992, 1993), Rajagopalan and Seib (1992), Rao et al. (1992), Reid and Hsu (1992), Reid et af. (1992, 1993a,b), Roos (1992a-d), Roy et al. (1992), Sahagian and Goff (1992,1993), Sapru and Labuza (1992a), Saunders et al. (1992), Schroeter and Hobelsberger (1992), Shogren (1992), Sullivan et al. (1992), Taylor et al. (1992). te Booy er al. (1992), Thompson (1992). Tian and Blanshard (1992a,b), Tolstoguzov (1992), van den Berg et al. (1992, 1993), Watase et al. (1992), Zasypkin et al. (1992), Zhang and Jackson (1992), Allen (1993), Anonymous (1993a), Appelqvist et al. (1993), Arvanitoyannis and Blanshard (1993b), BeMiller (1993), Bolton et al. (1992), Carpenter et al. (1993), Crowe and Crowe (1993), Crowe et al. (1993), De Bry (1993), Eerlingen et al. (1993), Eliasson and Larsson (1993), George (1993), Gidley et al. (1993), Goff er al. (1993), Hancock and Zografi (1993). Hartel (1993), Hatley and Mant (1993), Haynes and Locke (1993), Haynes et al. (1993), Hemminga et al. (1993), Karel et al. (1993a,b), Kerr et al. (1993), Koster et af. (1993), Lawton and Wu (1993), Mestres er af. (1993), Nelson (1993), O’Donnell (1993), Ohshima et al. (1993), Ollett et al. (1993a,b), Rohde et al. (1993), Roser and Colaco (1993), Seow and Teo (1993). Seow and Thevamalar (1993), Shalaev and Kanev (1993), Suzuki and Franks (1993), Zanoni et af. (1993)

B. Publications in Which the Applicability of WLF Kinetics to Food Systems at Tg < T < T , (in Contrast to Arrhenius Kinetics at T < Tg or T > T,) Is Discussed

Slade (1984), Slade and Levine (1985, 1987a,b, 1988a,b,c, 1989, 1991a,b, 1993c), Levine and Slade (1986, 1988a,b,c, 1989a-d, 1990, 1992b), Campanella et al. (1987), Gosline (1987), Kelley et al. (1987), Simatos and Karel (1988), Franks (1989, 1991c, 1992b, 1993b), Karel (1989, 1990, 1991c, 1992), Simatos et af. (1989). Slade et al. (1989), Ahlneck and Zografi (1990), Biliaderis (1990, 1991a,b, 1992a,b), Franks and Grigera (1990), Katsuta and Kinsella (1990), Le Meste and Simatos (1990), Lillie and Gosline (1990), Morris (1990). Reid (1990, 1992a), Roos and Karel (1990, 1991a,d-h, 1992, 1993), Blond and Colas (1991), Franks and van den Berg (1991), Franks et al. (1991), Given (1991), Karathanos et al. (1991, 1992), Karel and Saguy (1991), Lim and Reid (1991), Patil (1991), Roozen and Hemminga (1991). Roozen et al. (1991). Shimada et al. (1991), Simatos and Blond (1991, 1993), Aguilera er af. (1992), Angel1 et al. (1992), Buera and Karel (1992), Chinachoti (1992a), Cocero and Kokini (1992), Cocero et al. (1992a,b), Colonna and Buleon (1992), Goff (1992), Huang (1992). Karmas er al. (1992, 1993), Kokini (1992), Kokini et al. (1992d), Levi and Karel (1992a, 1993a,b), Nelson and Labuza (1992a,b, 1993), Peleg (1992,1993), Reid et al. (l992,1993b), Roos (1992c,d), Roy et al. (1992), Sapru and Labuza (1992a,b), Taylor et al. (1992), Zhang and Jackson (1992), Arvanitoyannis and Blanshard (1993a), Blond (1993), Franks and Hatley (1993), Haynes and Locke (1993). Huang et al. (1993), Kalichevsky et al. (1993a), Karel et al. (1993a,b), Kerr et al. (1993), MacInnes (1993), Nelson (1993), Pika1 (1993), van den Berg et al. (1993), Williams et al. (1993)

C. Publications in Which Evidence Suggesting That the Glass Curves of Many Polysaccharides and Proteins Show Common Characteristics Is Reported

Slade (1984), Slade and Levine (1985, 1987b, 1991a, 1993b), Hoseney et al. (1986), van den Berg (1986). Atkins (1987), Kelley ef al. (1987), Zeleznak and Hoseney (1987a), Blanshard (1988), Le Meste and Duckworth (1988), Levine and Slade (1988a, 1989b,

(continues)


TABLE X (Continued)

1992b), Takahashi et al. (1988), Orford et al. (1989), Le Meste and Simatos (1990). Lillie and Gosline (1990, 1993), Noel et al. (1990). Yoshida et al. (1990, 1992), Cocero and Kokini (1991), Liu and Lelievre (1991b, 1992b), Ollett et al. (1991), Roos and Karel (1991b,d), Scandola et al. (1991), Sochava et al. (1991), Tanner et al. (1991), Angell et al. (1992), Attenburrow et al. (1992), Chang and Randall (1992), Cocero et al. (1992a,b), de Graaf et al. (1992), Dong (1992), Gontard et al. (1992c), Gudmundsson (1992). Gudmundsson and Eliasson (1992), Hallberg and Chinachoti (1992), Huang et al. (1992), Kaletunc and Breslauer (1992), Kalichevsky and Blanshard (1992a,b), Kalichevsky et al. (1992a,c, 1993a,b), Kokini et al. (1992d), Kresin and Rau (1992), Lawton (1992a,b), Lelievre (1992a), Le Meste et al. (1992), Madeka and Kokini (1992a,b), Magoshi et al. (1992). Milczarek et af. (1992), Newman (1992), Noel and Ring (1992), Roos (1992b), Sala and Tomka (1992a), Shogren (1992), Sochava and Smirnova (1992), Ablett et al. (1993). Appelqvist et al. (1993), Attenburrow and Davies (1993), Blond (1993), Cherian and Chinachoti (1993), Crowe et al. (1993), Gidley et al. (1993), Kirby et al. (1993), Lawton and Wu (1993), Slade et al. (1993), Yano (1993)

D. Publications in Which the General Correlation between Carbohydrate MW and Dry Tg Is Described

Franks (1985, 1989, 1990, 1993b), Slade and Levine (1985, 1988b, 1991a,b, 1993b), Levine and Slade (1986, 1988a,b,c, 1989a, 1992a,b), Finegold et al. (1989), Green and Angell (1989), Orford et al. (1989, 1990), Noel et al. (1990), Williams et al. (1990), Franks and van den Berg (1991), Franks et al. (1991), Levine et al. (1991, 1992), Lim and Reid (1991). Ring and Whittam (1991), Roos and Karel (1991b,d,e), Franks and Hatley (1992), Jouppila and Roos (1992), Roos (1992a,c), te Booy et al. (1992), van den Berg (1992), Karel et al. (1993b), Nelson (1993). O’Donnell (1993), Slade et al. (1993), van den Berg et al. (1993)

E. Publications in Which the Similarities between Water Plasticization of Low-Molecular- Weight Carbohydrates and of Synthetic Polymers Are Pointed Out

Slade and Levine (1985, 1988b, 1991a,b), Levine and Slade (1986, 1988a,b, 1992b), Orford et al. (1989, 1990), Franks and Grigera (1990), Noel et al. (1990), Blond and Simatos (1991), Davies et al. (1991), Franks et al. (1991), Roos and Karel (1990, 1991d-g), Whittam et al. (1991), Nesvadba (1992b), Slade et al. (1993)

F. Publications in Which TB of Food Polymer-Polymer or Food Polymer-Plasticizer (Water) Blends Is Calculated Using One of Several Equations [(l) Gordon-Taylor,

(2) Couchman-Karasz, or (3) Other] (1) Roos and Karel (1991b,d-h, 1993), Aguilera et al. (1992), Buera et al. (1992), Cocero

et al. (1992a,b), de Graaf et al. (1992), Jouppila and Roos (1992), Roos (1992a-d), Cocero and Kokini (1993), Huang (1993a), Kalichevsky et al. (1993b). Karel et al. (1993b), Nelson (1993)

(2) Finegold et al. (1989), Orford et al. (1989, 1990), Ring and Whittam (1991), Roozen et al. (1991), Kalichevsky and Blanshard (1992a,b), Kalichevsky et al. (1992a-d), Noel and Ring (1992), Yoshida et al. (1992), Arvanitoyannis and Blanshard (1993a). Blond (1993), Donald et al. (1993), Franks (1993b), Suzuki and Franks (1993)

(3) Pika1 and Shah (1990), Kalichevsky and Blanshard (1992d), Roy et al. (1992), Huang (1993b), Oksanen and Zografi (1993)



FIG. 2. line polymers. From Slade and Levine (1987a) with permission of Van Nostrand Reinhold/AVI.

“Fringed micelle” model of the crystalline-amorphous structure of partially crystal-

of the diluent in glass-forming food systems that would be inherently mobile, unstable, and reactive at temperatures and moisture contents corresponding to the rubbery liquid state at T > Tg (Slade and Levine, 1991a). This concept, along with that of glass dynamics, has provided an innovative perspective on water relationships in foods (Duckworth, 1988; Simatos and Karel, 1988; Franks, 1991b; Ablett and Lillford, 1991; Goff, 1992; Karel, 1992; Nelson and Labuza, 1992a; Reid, 1992b; Roos, 1992b; Hegenbart, 1993; Nelson, 1993) (Table IIIB), including, e.g., moisture management and structural stabilization of IMF systems (Slade and Levine, 1985, 1988a, 1991a) and “cryostabilization” of frozen, freezer-stored, and freeze-dried aqueous glass-forming food materials and products (Cole ef al., 1983, 1984; Levine and Slade, 1986,1988a-d, 1989a-d, 1990). This perspective, the focal point of which is the critical importance of the glassy state phenomenon in foods, has received a great deal of recent support from many workers in the field (Table XA).

Glass dynamics deals with the time and temperature dependence of relationships among composition, structure, thermomechanical properties, and functional behavior (Levine and Slade, 1988a). As its name implies, glass dynamics focuses on (1) the glass-forming solids in an aqueous food system; (2) Tg of the resulting aqueous glass that can be produced by cooling to T < Tg; and (3) the effect of the glass transition and its Tg on processing and process control, via relationships between Tg and the temperatures of individual processing steps (Slade and Levine, 1991a,b). This concept emphasizes the operationally immobile (translationally), stable, and unreac-


A W DEFINED - HERE ONLV

W

3 a

k kli a

$

~

REACTIVE EQUILIBRIUM VAPORPHASE

CRysTAulNE SWD

I

r 1 BlOLOQlCAL STATE

EQUILIBRIUM DILUTE SOLUTION ROOM TEMPERATURE

STABLE

FIG. 3. Four-dimensional “dynamics map,” with axes of temperature, concentration, time (expressed as t h , where T is a relaxation time), and pressure, which can be used to describe mobility transformations in nonequilibrium glassy and rubbery systems. From Slade and Levine (1988b) with permission.

tive situation (actually one of kinetic metastability) that can obtain during product storage (of a practical duration) at temperatures and moisture contents corresponding to the glassy solid state at T < T,. It has been used to describe a unifying concept for interpreting “collapse” phenomena (Levine and Slade, 1986, 1988b; Franks, 1989, 1990; Pikal, 1990a,b; Reid, 1990; Roos and Karel, 1991b, 1993; Shimada et al., 1991; Anglea et al., 1992b; Chang and Randall, 1992; Chuy and Labuza, 1992; Karel, 1992; Levi and Karel, 1992a, 1993a,b; Roos, 1992b,c; te Booy et al., 1992; Peleg, 1993; Sahagian and Goff, 1993; van den Berg et al., 1993), which govern a host of time-dependent, translational diffusion-limited, deterioration processes (physical, chemical, or enzymatic) (Karel et al. references in Table IV) that can occur in amorphous food or pharmaceutical materials and products


during storage in the rubbery liquid state at T > Tg (Levine and Slade, 1986, 1988b; Roos, Karel, and coworkers references in Table IV). This unifying concept for interpreting collapse phenomena has also been used to describe starch gelatinization, retrogradation, and annealing in baked goods and other starch-based food systems (Slade, 1984; Slade and Levine, 1984,1987b; Levine and Slade, 1988a, 1989b; Zobel, l988,1992a,b; Larsson and Eliasson, 1991; Shi and Seib, 1992; Seow and Teo, 1993; Seow and Thevamalar, 1993) (Table IX).

A physicochemical mechanism for collapse, derived from Williams- Landel-Ferry (WLF) free-volume theory for (synthetic) amorphous polymers (Williams et al., 1955; Ferry, 1980), has been described (Levine and Slade, 1986). The non-Arrhenius kinetics of collapse in the rubbery liquid state are governed by the mobility of the water-plasticized food polymer matrix. These so-called WLF kinetics (Table XB) depend on the magnitude of AT = T - Tg (Slade, 1984; Slade and Levine, 1985; Levine and Slade, 1986), as defined by WLF theory. Glass dynamics has proved a useful concept for elucidating the physicochemical mechanisms and kinetics of structuraYmechanica1 changes involved in various melting, annealing, and (re)crystallization processes (Levine and Slade, 1988a). Such phenomena are observed in many partially crystalline food polymers and processing/ storage situations, including, e.g., the gelatinization, retrogradation, and annealing of starches (Slade, 1984; Slade and Levine, 1984, 1987b, 1988c; Whittam et al., 1990, Lelievre, 1992a,b; Stute, 1992; Taylor et al., 1992; Zanoni et al., 1993). Glass dynamics has also been used to describe the viscoelastic behavior of amorphous network-forming proteins such as wheat gluten (Slade, 1984; Slade et al., 1989).

F. STATE DIAGRAMS

The fundamental importance of state diagrams (e.g., Fig. 4), and of the glass transition as the critical reference state on such diagrams (which determines technological performance), to understanding and controlling the nonequilibrium behavior of water-containing food systems during processing and product storage, as influenced by the variables of moisture content, temperature, and time, has become widely recognized (Table XIB). The interdependent concepts of water dynamics and glass dynamics, embodied in the state diagram as a dynamics map, have provided insights into the relevance of the glassy reference state to functional aspects of a variety of food systems (Levine and Slade, 1988a; Slade and Levine, 1988b). Process control, product quality, safety, and shelf-life are all dictated by WLF kinetics (Table XB), which are applicable above Tg in the viscoelastic, rubbery liquid state of accelerating mobility and translational diffusion


FIG. 4. Schematic state diagram of temperature versus weight percent (w%) water for an aqueous solution of a hypothetical, glass-forming, small carbohydrate, illustrating how the critical locations of T,’ and W,’ divide the diagram into three distinguishable structure-property domains. Reprinted with permission from Slade, L., and Levine, H. 1991a. Beyond water activity: Recent advances based on an alternative approach to the assessment of food quality and safety. CRC Crir. Rev. Food Sci. Nutr. 30, 115-360. Copyright CRC Press, Inc. Boca Raton, FL.

(Slade and Levine, 1991a, 1993~). In recent years, many workers (Table XIB) have lent their strong support to advocacy of the importance of state diagrams (e.g., Fig. 4), and of the solute-specific, invariant point, T,’ - C,’ (defined below), as the focal point (see Fig. 4) of such dynamics maps, to understanding structure-property relationships of food molecule-water systems. Most recently, a three-dimensional state diagram for a ternary system [two solutes (sucrose and glycine) plus water] has been reported by two groups (Suzuki and Franks, 1993; Shalaev and Kanev, 1993). As pointed out in both reports, this complex state diagram contains a metastable, three-dimensional glass transition surface (as opposed to the typical two-dimensional glass transition curve for a binary solute-water system), which Suzuki and Franks have proposed as a novel concept important to, e.g., the formulation of multicomponent mixtures to be freeze- dried.

T,’ represents the solute-specific subzero T, of the maximally freeze- concentrated, amorphous solutehnfrozen water (UFW) matrix surrounding the ice crystals in a frozen solution (Franks et al., 1977; Schenz et al., 1984; Levine and Slade, 1986) (Table XIA); C,’ is the solute concentration of the Tgr glass, while W,’ is its concentration of UFW. As discussed further later, Tgf and Cgf values, especially of various low-molecular-weight (MW) sugars widely used in foods, have become a topic of so much current interest, as well as considerable controversy and debate (Slade and Levine, 1991a; Levine and Slade, 1992b), that many recent publications have been devoted


TABLE XI REFERENCE^

A. Publications Dealing with T,’ Franks et al. (1977, 1991), Franks (1982, 1983b, 1985,1986a,b, 1989,1990, 1991a,c, 1992b,

1993a,b), Schenz et al. (1984, 1991, 1992), Slade (1984), Slade and Levine (1984, 1985, 1987a,b, 1988a-d, 1989, 1990, 1991a,b, 1992~. 1993b,c), Reid (1985, 1990, 1992a), Levine and Slade (1986, 1988a-d, 1989a-d, 1990, 1991, 1992a,b, 1993), van den Berg (1986, 1992), Blanshard and Franks (1987), Lillford (1988), Marsh and Blanshard (1988), Takahashi et al. (1988), Hatley et al. (1989, 1991), Karel (1989, 1991c), Russell and Oliver (1989), Slade et al. (1989, 1993). Biliaderis (1990, 1991b), Caldwell er al. (1990), Franks and Grigera (1990), Le Meste and Simatos (1990), Noel et al. (1990), Paakkonen and Roos (1990), Pika1 (1990a,b), Williams and Carnahan (1990). Blanshard et al. (1991), Blond and Colas (1991), Blond and Simatos (1991), Franks and van den Berg (1991), Given (1991), Hatley (1991), Hatley and Franks (1991), Izzard et al. (1991), Karel and Saguy (1991), Le Meste and Huang (1991). Levine et al. (1991, 1992), Lim and Reid (1991,1992), Ma and Harwalkar (1991), MacDonald and Lanier (1991), Matthiesen et al. (1991), Maurice et al. (1991). Roos and Karel (1991~-h, 1993), Roozen and Hemminga (1991), Roozen et al. (1991), Wasylyk and Baust (1991), Ablett et al. (1992a,b,c, 1993), Anglea et al. (1992a), Attwool et al. (1992), Berail et al. (1992), Best (1992), Blond (1992, 1993), Chang and Randall (1992). Chinachoti (1992a), Cocero et al. (1992a,b), Dong (1992), Eerlingen and Delcour (1992), Goff (1992), Hanover (1992), Hsu and Reid (1992), Huang (1992), Huang et al. (1992, 1993), Jouppila and Roos (1992), Kokini (1992), Kresin and Rau (1992), Le Meste et al. (1992), Leung et al. (1992), Nelson and Labuza (1992a), Reid and Hsu (1992), Reid et al. (1992, 1993a,b), Roos (1992a-d), Sahagian and Goff (1992), Taylor et al. (1992), van den Berg er al. (1992, 1993), Crowe and Crowe (1993), Crowe et al. (1993), Ford and Dawson (1993), Franks and Hatley (1993), Goff et al. (1993), Hartel (1993), Hatley and Mant (1993), Hegenbart (1993), Hemminga et al. (1993), Karel et al. (1993a,b), Kerr er al. (1993), MacInnes (1993), McCurdy et al. (1993), Nelson (1993), O’Donnell (1993), Ohshima et al. (1993), Shalaev and Kanev (1993), Simatos and Blond (1993), Suzuki and Franks (1993)

B. Publications Dealing with the Importance of State Diagrams and of the T,’-C,‘ Point Schenz et al. (1984, 1991, 1992), Slade (1984), Slade and Levine (1984, 1985, 1987a,b,

1988a-d, 1989, 1990, 1991a,b, 1992c, 1993b,c), Franks (l985,1986a, 1989,1990, 1991c, 1992b, 1993a,b), Burke (1986), Levine and Slade (1986, 1988a-d, 1989a-d, 1990,1991, 1992a,b, 1993), van den Berg (1986,1992). Blanshard and Franks (1987). Lillford (1988), Simatos and Karel (1988), Blond (1989, 1993), Green and Angel1 (1989), Karel (1989, 1992). Simatos et al. (1989), Slade et al. (1989, 1993), Williams and Leopold (1989), Ahlneck and Zografi (1990), Biliaderis (1990, 1991b), Caldwell e ta / . (1990), Franks and Grigera (1990), Le Meste and Simatos (1990), Noel et al. (1990), Williams and Carnahan (1990), Ablett and Lillford (1991), Blanshard et al. (1991), Blond and Colas (1991). Bruni and Leopold (1991), Chang and Baust (1991b), Franks and van den Berg (1991), Franks et al. (1991). Given (1991), Hatley (1991), Hatley and Franks (1991), Hatley et al. (1991), Izzard et al. (1991). Le Meste and Huang (1991), Lim and Reid (1991), Ma and Harwalkar (1991), MacDonald and Lanier (1991), Matthiesen et al. (1991), Roos and Karel (1991d,e-h, 1993), Simatos and Blond (1991, 1993), Ablet et al. (1992a,b,c, 1993), Berail et al. (1992), Best (1992), Chang and Randall (1992), Chinachoti (1992a), Cocero et al. (1992a,b, 1993), Colonna and Buleon (1992), de Graaf et al. (1992), Eerlingen and Delcour (1992), Goff (1992), Huang et al. (1992). Jouppila and Roos (1992), Kawai e ta / . (1992), Kokini (1992), Kokini et al. (1992b,d), Kresin and Rau (1992), Labuza and Nelson (1992), Leopold et al.

(continues)


TABLE XI (Continued)

(1992), Madeka and Kokini (1992a,b, 1993), Nelson and Labuza (1992b, 1993), Reid (1992b). Reid and Hsu (1992), Roos (1992a-d), van den Berg et al. (1992,1993), Cocero and Kokini (1993), Franks and Hatley (1993), Hartel (1993), Hegenbart (1993). Karel et al. (1993a,b), MacInnes (1993), Nelson (1993), Ohshima et al. (1993), Reid et al. (1993a,b), Shalaev and Kanev (1993), Suzuki and Franks (1993)


to discussions of different aspects of this subject (Table XII). Table XI11 presents a survey of recent literature values of T,‘ and C,’ for various low- MW carbohydrates.

Ill. KEY ELEMENTS AND APPLICATIONS OF THE “FOOD POLYMER SCIENCE” APPROACH

A. EFFECT OF MOLECULAR WEIGHT ON T,

For pure synthetic polymers, in the absence of diluent, T, varies with MW in a characteristic and theoretically predicted fashion, which has a significant impact on resulting mechanical and rheological properties (Ferry, 1980; Levine and Slade, 1989b). As illustrated in Fig. 5 , for a homologous series of amorphous linear polymers, Tg increases with increasing number-average MW (z,,) [due to decreasing free volume (Ferry, 1980)], up to a plateau limit for the region of entan@ement coupling in rubberlike viscoelastic random networks [typically at M,, = 1.25 X lo3 to 105 (Graes- sley, 1984)], then levels off with further increases in a,, (Ferry, 1980; Bill- meyer, 1984). Below the entanglement zn limit, there is a theoretical linear relationship between increasing T, and decreasing inverse %,, (Sperling, 1986). The difference in three-dimensional morphology and resultant mechanical and rheological properties between a collection of nonentangling, low-MW polymer chains and a network of entangling, high-MW, randomly coiled polymer chains can be viewed as analogous to the difference between masses of elbow macaroni and spaghetti (Levine and Slade, 1989b). For synthetic polymers, z,, at the boundary of the entanglement plateau often corresponds to about 600 backbone-chain atoms (Sperling, 1986). Since there are typically about 20 to 50 backbone-chain atoms in each polymer segmental unit involved in the cooperative, long-range, translational motions that can occur only at or above T, (Brydson, 1972), entangling high polymers are those with at least about 12 to 30 such segmental units per chain (Levine and Slade, 1989b).


TABLE XI1 REFERENCES TO RECENT PUBLICATIONS CONTAINING DISCUSSIONS OF VALUES OF Tg' AND c,' ANDIOR METHODS OF DETERMINING THOSE VALUES FOR LOW-MOLECULAR-

WEIGHT CARBOHYDRATES"

(A) Methods of Determining C,' Values Levine and Slade (l986,1988b, 1992a,b), Slade and Levine (1988b, 1991a), Blond (1989), Hatley et al. (1989, 1991), Noel et al. (1990), Orford et al. (1990). Schiraldi (1990), Vertucci (1990), Blond and Colas (1991), Blond and Simatos (1991). Hatley (1991), Hatley and Franks (1991), Izzard et al. (1991), Ma and Harwalkar (1991), Raemy and Lambelet (1991), Roos and Karel (199lc,d,f,g,h), Schenz et al. (1991, 1992), Simatos and Blond (1991, 1993), Ablett er al. (1992a,b,c, 1993), Franks (1992b, 1993b), Jouppila and Roos (1992), Kawai et al. (1992), Kresin and Rau (1992), Reid and Hsu (1992), Roos (1992a,c,d), Hatley and Mant (1993), MacInnes (1993), Reid et al. (1993a,b)

(B) T,' Value of -32°C for Sucrose Schenz et al. (1984, 1991), Franks (1985, 1986,a,b, l989,1990,1991c, 1992b, 1993a,b), Slade and Levine (1985, 1988b). Blanshard and Franks (1987), Levine and Slade (1988b, 1992b), Caldwell et al. (1990), Blanshard et al. (1991), Izzard et al. (1991), Franks et al. (1991), Hatley (1991), Hatley and Franks (1991), Hatley etal . (1991), Lim and Reid (1991, 1992), Maurice ef al. (1991), Roos and Karel (1991f), Roozen and Hemminga (1991), Chang and Randall (1992), Chinachoti (1992a), Goff (1992), Leung et al. (1992), Reid and Hsu (1992), Roos (1992d), Sahagian and Goff (1992). te Booy et al. (1992), van den Berg (1992), Goff et al. (1993), Hartel (1993), Hatley and Mant (1993), Hemminga et al. (1993), Kerr et al. (1993), McCurdy et al. (1993), Reid et al. (1993a), Suzuki and Franks (1993), van den Berg et al. (1993)

(C) Different C,' Values for Sucrose MacKenzie (1977), Franks (1983b, 1985, 1986a,b, 1989, 1990, 1991c, 1992b. 1993a,b), Slade and Levine (1985, 1988b, 1991a), Blanshard and Franks (1987), Levine and Slade (1988b, 1992b, 1993), Williams and Carnahan (1989, 1990), Franks and Grigera (1990), Williams et al. (1990). Blanshard et al. (1991), Dereuddre et al. (1991). Franks et al. (1991), Hatley (1991), Hatley and Franks (1991), Hatley et al. (1991), Izzard et al. (1991). Le Meste and Huang (1991), Roos and Karel (1991~-h, 1993), Ablett et al. (1992a,b, 1993), Chinachoti (1992a, 1993), Goff (1992), Hanover (1992), Jouppila and Roos (1992), Kawai et al. (1992), Nesvadba (1992a), Reid and Hsu (1992), Roos (1992a,c,d), van den Berg (1992), Blond (1993), Franks and Hatley (1993). Hartel (1993), Reid ef al. (1993a), Suzuki and Franks (1993), van den Berg et 01. (1993)

(D) T,' as the Temperature at the Midpoint, Rather Than the Onset, of the Transition Franks et al. (1977, 1991), Franks (1982, 1985, 1986a, 1989, 1990, 1992b), Schenz et al. (1984, 1991,1992), Reid (1985, 1990), Levine and Slade (1986, 1988a-d, 1992b), Blanshard and Franks (1987), Hatley et al. (1989, 1991), Caldwell et al. (1990). Franks and Grigera (1990), Franks and Hatley (1990), Le Meste and Simatos (1990), Williams and Carnahan (1990), Blanshard et al. (1991), Blond and Simatos (1991), Hatley (1991), Hatley and Franks (1991), h a r d et al. (1991), Lim and Reid (1991, 1992), Maurice et al. (1991), Wasylyk and Baust (1991), Ablett et al. (1992a,b,c), Berail et al. (1992), Chang and Randall (1992). Reid and Hsu (1992), Blond (1993), Goff et al. (1993), Hatley and Mant (1993), Reid er al. (1993b), Suzuki and Franks (1993), van den Berg et al. (1993)


TABLE XI11 Tg' AND C,' VALUES FOR LOW-MW CARBOKYDRATES: SURVEY OF RECENT LITERATURE DATA"

T8( ("C) Other T i values C,' (w%) Other C l values Compound MW [refs. a-d] [refs.] [refs. b-el [refs.]

Ethylene glycol Ropylene glycol 1,3-Butanediol Glycerol Erythrose Threose Erythritol Thyminose (deoxyribose) Ribulose Xylose Arabmose

Ribose Arabitol Ribitol Xylitol Methyl riboside Methyl xyloside Quinovose (deoxyglucose) F u m e (deoxygalactose) Rhamnose (deoxymannose) Talose Idose Psicose Altrose Gulose Glucose

LYxW

62.1 76.1 90.1 92.1

120.1 120.1 122.1 134.1 150.1 150.1 150.1 150.1 150.1 152.1 152.1 152.1 164.2 164.2 164.2 164.2 164.2 180.2 180.2 180.2 180.2 180.2 180.2

- 85 -67.5 -63.5 - 65 -50 -45.5 -53.5 -52 -50 -48 -47.5 -47.5 - 47 - 47 -47 -46.5 - 53 -49 -43.5 - 43 - 43 -44 -44 -44 -43.5 -42.5 -43

-65 [f.h.iJ,k,aa,ddd]. -95 [q], -100 [eee]

-48 [j,k,l,aa,ddd,eee], -47 [h,i], -60 [z] -48 [h,i], -48.5 [I], -61 [z] -47.5 [aa] -47 [h,i,j,k,aa.ddd,eee]. -49 b], -62 [z] -48 PI

-46.5 [j,k,aa,ddd], -47 [eee], -67 [z]

-57 [z] -55 [z] -44 [aa]

-43.5 [aa]

-43 [f,h,i,j,k,l,r,s,aa,zz,ddd,eee], -43.5 [ccc]. -36.5 [m], -52 [ss], -53 [z], -57 [tt,uu], -42.4 [w], -50 [WW]

34.5 43.9 41.5 54.1 41.8

43.1

69.0 44.8

67.1 52.9 54.9 57.1 51.0 49.8 47.4 47.4 52.6

70.9

54 [j,k,ddd], 54.1 [f,h,i], 80 [ql

69 [h,i,j,k,l,ddd], 78.9 [z] 44.8 [h,i,l], 79.3 [z]

67 [j,k,ddd], 67.1 [h,i]. 81.4 [z] 52.9 [I]

57.1 [j,k,ddd], 80.2 [z]

78.4 [z] 82.8 [z]

70.9 [fj,k,l], 67.1 [xx]. 71.4 [h,i], 74.7 [w], 79.2 [ss,tt], 80.0 [z,ww], 82 [xx], 83 [zzddd,eee]

Fructose

Fructose : glucose (1 : 1 blend) Galactose

Allose Sorbose Mannose

Tagatose Inositol Mannitol Galactitol Sorbitol

Glucoheptose Mannoheptulose Glucoheptulose Perseitol (mannoheptitol) Isomaltulose (palatinose) Nigerose Cellobiulose Isomaltose Sucrose

Gentiobiose Larninaribiose Turanose Mannobiose

180.2

180.2

180.2 180.2 180.2

180.2 180.2 182.2 182.2 182.2

210.2 210.2 210.2 212.2 342.3 342.3 342.3 342.3 342.3

342.3 342.3 342.3 342.3

- 42

-42.5 -41.5

-41.5 -41 -41

-40.5 -35.5 -40 - 39 -43.5

-37.5 -36.5 -36.5 -32.5 -35.5 -35.5 -32.5 -32.5 -32

-31.5 -31.5 -31 -30.5

-42 [f,h,i,j,k,l,s,t,aa,u,ccc,ddd,eee], -48 [v].

-42.5 [j,k,ddd] -41.5 [f,aa], -42 [h,i], -40.5 [j,k,ddd],

-53 [z,ss], -58 [ t t , ~ ~ ]

-4l[eee], -43 [p]. -50 [ww], -51 [z]

-52 [z] -41 [j,k,aa,ddd,eee], -53 [z]

-36 [h,i] -35 [o]

-43.5 [f,j,k,aa&ddd], -43 [ail, -44 [I,ccc,eee], -57 [z]

-32 [I] -32 [f,h,i,j,k,l.n,o,r,t,x,aa,cc,dd,ee,ff,aaa,

ddd,eee,hhh], -33 [gg,hh,zz,ggg], -33.5 [ m l . -34 [ g ~ ~ l , -35 [ibbb], -37 [ww]. -40 [q,ss]. -41 [z,mm,nn,rr], -41.5 [ql, -46 [aa,kkpn jj,tt,uu]

-31 [aa] -30.5 [aa]

51.0

52.0 56.5

64.1 69.0 74.1

42.9 76.9

81.3

56.5

51 [f,j,k,l], 51.3 [h,i], 79 [v], 82.5 [z], 78.6

52 [j,k,ddd] 55 [j,k,ddd], 56.5 [fl. 57.1 [h,i], 71 [p.xx]. 79

[ss,tt], 85 [zz,ddd,eee]

[ww]. 80.5 [z], 84 [xx,eee]

81.0 [z]

[eeel 74.1 [j,k,ddd], 67.6 [xx], 80.1 [z], 83.3 [xx], 84

76.3 [hj]

81.3 [f,h,i,j,k,l,ddd], 70.4 [xx], 81.7 [z], 84.7 [=I, 88 [zz,eeel

58.8 58.8 [I] 64.1 64.1 [f,b,ij,k,l], 66 [qd , 73 [kk], 75.0 [bb], 77

[bbb], 78 [hh], 79.5 [kk,ll,nn,pp,ss,tt], 80 [aa,ii,jj,mm], 81 [ww]. 81.2 [q], 81.7 [z]. 83.0 [ff,zz,eee], 83.3 [hhh], 84 [eee]

79.4

61.0 52.4

(continues)

c e

CL

P N

TABLE XI11 (Continued)

TB) (“(3 Other Tb( values Compound Mw [refs. a-d] [refs.]

G’ (w%) Other Cb( values [refs. b-el [refs.]

Melibiose Lactulose Maltose

Maltulose Trehalose

Cellobiose Lactose

Maltitol Isomaltotriose Panose RafIinose Maltotriose

Nystose Stachyose Maltotetraose Maltopentaose Maltohexaose Maltoheptaose

342.3 342.3 342.3

342.3 342.3

342.3 342.3

344.3 504.5 504.5 504.5 504.5

666.6 666.6 666.6 828.9 990.9

1153.0

-30.5 -30 -29.5

-29.5 -29.5

- 29 -28

-34.5 -30.5 -28 -26.5 -23.5

-26.5 -23.5 -19.5 - 16.5 - 14.5 - 13.5

-37 [z]

-29.5 [fJ,k,aa,zz,ddd], -29.7 [vv], -30 [h,i,l,eee), -31.6 [ccc], -41 [pp,tt], -37 [z], -36 [ss]

-27 [o], -29.5 [j,k,aa,ddd], -30 [h,i,l,eee], -35 [z]. -32 [fffJ

-28 [b,i,o,aa,zz], -35 [ss], -36 [z], -40 [HI. -41 [uu], -45 [OO]

-42 [z]

-26 [h,i] -23.5 [f,j,k,aa~z,ddd], -23 [h,i,w], -23.1

[w], -24 [I.eee,ggg], -29 [u]

-19.3 [w] -15.8 [w]. -18 [w] - 14.5 [zz], - 14.7 [w] -13.2 [w]. -14 [w], -18 [u]

58.1 80.0

833

59.2

62.9 66.7 62.9 58.8 69.0

47.2 64.5 68.0 66.7 78.7

81.7 [z]

80 [f,h,i,j,k,l,pp,ddd], 81.6 [z], 74.5 [bb], 76.7 [w], 81.5 [ss,tt,zz,eee]

83.3 [j,k,l,ddd], 81.6 [z]. 82.6 [h,i], 80 [eee], 70.5 [bb]

58.8 [h,i], 80 [oo], 81.3 [z], 82.5 [ss]

82.9 [z]

58.5 [b,i] 69 [f,h,ij,k,l,ddd], 81 [u]. 78.5 [bb], 81.6 [w].

’5 [=I

75.5 [bb], 82.1 [w] 74.0 [bb], 76.5 [w] 72.5 [bb], 75 [zz], 76.4 [w] 77.4 [w], 78 [u], 79.0 [bb]

References [a] Levine and Slade (1986) [b] Levine and Slade (1988h) [c) Levine and Slade (1988~) [d] Levine and Slade (1988d) [el Slade and Levine (1988b) [f] Goff (1992) [g] Goff er al. (1993) [h] Franks (1985) [i] Franks (1986a) [j] Franks (1990) [k] Franks and Grigera (1990) [I] Franks (1989) [m] Franks (1982) [n] Franks (1986h) [o] Chang and Randall (1992) b ] Blond (1989) [q] Ahlett et al. (1992a) [r] Maurice ef al. (1991) [s] Ruhin et al. (1990) [t] Schenz er al. (1991) [u] Ahlett ef al. (1993) [v] Ablett er al. (1992~) [w] Roozen ef al. (1991) [XI Roozen and Hemminga (1990) [y] Wasylyk and Baust (1991) [z] Roos (1992a) [aaa] Suzuki and Franks (1993) [ccc] Ken er al. (1993) [eee] Franks (1993a) [ggg] Berail ef al. (1992)

[aa] Roos (1992d) [hh] Kawai er al. (1992) [ a ] Blanshard and Franks (1987) [dd] Caldwell et al. (1990) [eel Schenz et al. (1984) [ffl Hatley er al. (1991) [gg] Lim and Reid (1991) [hh] Reid and Hsu (1992) [ii] Izzard er al. (1991) [jj] [kk] Roos and Karel (1991~) [Ill Reid ef al. (1993b) [mm] Williams and Camahan (1990) [nn] Roos and Karel (1991g) [oo] Roos and Karel (1991e) [pp] Roos and Karel (1991d) [qd Nesvadba (1992a) [rr] MacInnes (1993) [ss] Roos and Karel (1991f) [tt] Roos and Karel (1991h) [uu] Roos and Karel (1993) [w] Schenz er al. (1992) [ww] Simatos and Blond (1993) [xx] Simatos and Blond (1991) [yy] te Booy er al. (1992) [zz] van den Berg (1992) [bbh] Shalaev and Kanev (1993) [ddd] Franks (1991~) [a Crowe et al. (1993) [hhh] Hatley and Mant (1993)

Le Meste and H u n g (1991)

a Modified from Levine and Slade (1992h).


400 - 360 -

1 10 100 1000 DP

FIG. 5. Plot of Tg as a function of log DP (degree of polymerization, a measure of molecular weight), for poly (m-methylstyrene) (open circles); poly(methylmethacry1ate) (open triangles); poly(viny1 chloride) (solid circles); isotactic polypropylene (solid triangles); atactic polypropylene (circles, top half solid); and poly(dimethy1siloxane) (circles, bottom half solid). From Sperling, L. H. 1986. “Introduction to Physical Polymer Science.” Copyright 0 1986 Wiley (Interscience), New York. Reprinted by permission of John Wiley & Sons, Inc.

Within a homologous family of food polymers [e.g., from the glucose monomer through maltose, maltotriose, and higher maltooligosaccharides (e.g., maltodextrins) to the amylose and amylopectin high polymers of starch], both Tg (Slade and Levine, 1988b; Levine and Slade, 1989a; Orford el al., 1989,1990; Ring and Whittam, 1991; Roos and Karel, 1991b; Jouppila and Roos, 1992; Karel et al., 1993b) and Tg’ (Levine and Slade, 1986, 1988a-d; Lim and Reid, 1991; Roos and Karel, 1991e; Jouppila and Roos, 1992; Karel et al., 1993b) increase in the same characteristic fashion with increasing solute MW, as demonstrated in Figs. 6 (Levine and Slade, 1986), 7 (Levine and Slade, 1988a), and 8 (Levine and Slade, 1989a). This finding has been shown to be in full accord with the established variation of Tg with MW for homologous families of pure synthetic amorphous polymers (Ferry, 1980; Billmeyer, 1984; Sperling, 1986), described above. The insights resulting from this finding have proved pivotal to the characterization of structure-function relationships in many different types of completely amorphous and partially crystalline food polymer systems (Slade and Lev- ine, 1991a).

The general correlation between increasing saccharide MW and increasing Tg (Levine and Slade, 1988a,b, 1989a; Roos, 1992a) was also confirmed


Gelation, Encapsulation, Thermomechanical Stabilization, Cryostabilizatio:. Facilitation of Drying

.1

0- . 8 ' I d

- 10-

0 -

o -20- r=-0.98

0, -30-

-40-

- 5 0 . . . , . . . . 1 0 5 1 0 1 5 M 2 5 3 0 3 5 4 0

1 Kin (104) r 6 1 DE

- 50- 1 I 0

Gelation, Encapsulation, Thermomechanical Stabilization, Cryostabilizatio:. Facilitation of Drying

0-

. I . . - 10-

o -20-

Q + .E 'E 0, -30

P I

- 4

- Mn = 180181DE

FIG. 6. Variation of the glass transition temperature, Tg'. for maximally frozen 20 w% solutions against M, [expressed as a function of dextrose equivalent (DE)] for a homologous family of commercial starch hydrolysis products (SHPs). DE values are indicated by numbers marked above the x axis. Areas of specific functional attributes, czresponding to the three regions of the diagram, are labeled. Inset: Plot of Tg' versus 1/M, ( X 10,000) for SHPs with a, values below the entanglement limit, illustrating the theoretically predicted linear dependence. From Levine and Slade (1986) with permission.

in recent studies of the T, behavior of amorphous mixtures of low-MW carbohydrates reported by Finegold ef al. (1989) and Orford ef al. (1990). Results of their studies demonstrated that measured Tg values of various dry binary mixtures of mono- and di- or oligosaccharides can be roughly approximated from a mole fraction-weighted average of the individual dry T, values of each saccharide in a given amorphous blend. However, while the general correlation between carbohydrate MW and dry T, has now become well established and widely accepted (Table XD), it should be carefully noted, as illustrated by the Tg data in Table XIV, that T, can vary substantially, even within a series of compounds of the same MW and only the most subtle differences in chemical structure. The reasons for this are not known.

1. Tg of Polymer-Polymer and Polymer-Plasticizer Blends

As alluded to above and reviewed elsewhere (Levine and Slade, 1988a), the correlation between polymer MW and T, can be extended to cover

146

-20 0

L- e

-30

-40

LOUISE SLADE AND HARRY LEVINE

- -

-

r - -0.99

Ti (K) 250

240 230 220

-

- 0 1 2 3 4 5 6

I I 1 1 I 1 1 1 I I i/Mw(x 103)

-10 O[

-20 0

L- e

-30

-40

-

-

- ,

i/Mw(x 103) I I 1 1 I 1 1 1 I I

- -

r - -0.99

Ti (K) 250

240 230 220

0 1 2 3 4 5 6

FIG. 7. Variation of the glass transition temperature, Tg', for maximally frozen 20 w% solutions against MW for a homologous series of maltooligosaccharides from glucose through maltoheptaose. Inset: Plot of Tg' versus l/MW ( X 1000) of solute, illustrating the theoretically predicted linear dependence. Reprinted from Levine and Slade (1988a) with permission of Cambridge University Press. Copyright 1988, Cambridge University Press.

homogeneous amorphous blends of two or more miscible neat polymers, a polymer plus a plasticizer, or two or more polymers plus a plasticizer. To a first approximation, Tg of a homogeneous, multicomponent blend can be estimated from the weight-average composition of a mixture of components with individual Tg values of Tgl, Tg2, Tg3, etc. (Sperling, 1986). This simple approach has been shown to work well for various glass-forming food systems (Levine and Slade, 1988c, 1989a; Berail et al., 1992), in addition to those studied by Finegold et al. (1989) and Orford et af. (1990). For example, the DSC-measured Tg' value of -37.5 iz 1.O"C for various samples of orange juice is well approximated by a calculated value of -37.25"C for a 2 : 1 : 1 blend of sucrose ( Tg' = -32"C), fructose (Tg' = -42"C), and glucose ( Tg' = -43"C), which constitutes the predominant water-soluble solids of typical orange juice (Levine and Slade, 1989~). Similarly, the homogeneous glass formed by a molten 1:l mixture of pure crystalline glucose (dry Tg = 31°C) and fructose (lower dry Tg = 11°C) has a DSC- measured Tg value of 20°C (Slade and Levine, 1988b; Finegold et al., 1989), which is within 1°C of the calculated value.

Beyond such a simple empirical approach to estimating Tg of a polymer- polymer or polymer(s)-plasticizer blend (which, in any event, was used,


Molecular Weight

B

0 0 200 400 600 800 1000 1:

D O J

0 20 40 I

10000/MW

10

FIG. 8. Variation of the glass transition temperature, T, (measured for dry powders), against (A) MW and (B) 10,000/MW, for a homologous series of maltooligosaccharides from glucose through maltoheptaose. From Levine and Slade (1989a) with permission.

in the cases described above, to corroborate measured data, rather than to replace experimental measurements), efforts to predict Tg of such blends, based on theoretical approaches using equations derived empirically or from classical thermodynamics, represent largely an academic exercise. Franks (1993b) has recently stated that “the prediction of glass transition temperatures of mixtures from the properties of the pure components and the mixture composition is uncertain, because the very nature of the glass transition itself is not yet fully understood.” If one needs to know, with confidence, an accurate value for Tg of a blend, in order, e.g., to design a process [e.g., freeze-drying without “melt-back” (Levine and Slade, 1986,


TABLE XIV Tg VALUES FOR LOW-MW CARBOHYDRATES: SURVEY OF RECENT LITERATURE DATA‘

Compound T* (“C)

MW [Refs. a-c] Other Tg values [Refs.]

Xylose Arabinose Lyxose Ribose Xylitol

Fucose (deoxygalactose) Rhamnose (deoxymannose) Talose Tagatose Altrose Glucose

Fructose

Fructose : glucose (1 : 1 blend) Galactose Sorbose Mannose Sorbitol

Isomaltose Sucrose

Turanose Mannobiose Melibiose Maltose

Trehalose Cellobiose Lactose

Maltitol Maltotriose

150.1 9.5 150.1 150.1 8 150.1 -10 152.1 -18.5

164.2 164.2 180.2 11.5 180.2 180.2 10.5 180.2 31

180.2 11, 100

20 180.2 =30, 110 180.2 180.2 30 182.2 -2

342.3 342.3 52

342.3 52 342.3 90 342.3 342.3 43

342.3 79 342.3 77 342.3

344.3 504.5 76

-10 [d,e,z,ll], -11 [g], -13 [i] -18.5 [z], -19 [g], -23 [i],

31 [i]

11.5 [z] 40.5 [Ill 10.5 [z] 21 [f], 29 [j,k,hh], 30 [I], 31

-39 [d,e,ll]

0 [il, 27 [sl

[e,m,z,bb], 33 [n], 36 [i,aa,mm], 37 [I], 38 [g,h], 38.5 [dd], 39 [d,e,o,y,llI

5 [bbl, 7 [g,z,ggl, 10 [iff], 11 [aal, 13 [d,e,o,p], 16 [dd,ll], 17 [f], 100 [el

13 [gl, 20 [d,e,lll, 21 [ol 32 [g,h], 38 [i], 110 [e,zJ] 27 [i] 30 [d,e,z,lll, 31 [il, 36 [gl -2 [q,z,ddl, -3 [d,e,fl, -4 [il,

78 [gl 0 [0,111

52 [e,f,r], 57 [d,e,o,s,t,u], 55 [nn,oo], 56 [v.ii], 56.6 [jj], 61 [mm], 62 [z,bb], 66 [n]. 67 [ij,aa,cc,kk], 69.5 [dd], 70 [g,y,e,lll

52 [ZI 90 [.I 91 [il, 95 [gl 43 [d,e,fl, 70 [nl, 87 [w,z,bbl, 91 [il,

92 [h,i,aal. 94 [.I, 95 [g,y,ddI, 100 [Ill

79 [n,z], 77 [eJ], 107 [i] 77 [d,e,ll] 100 [dd], 101 [i,s,u,x,bb,ii], 103 [z],

106 [aa] 44 [i] 76 [d,e,f,z,ll], 130 [y,dd], 131 [v].

134 [g], 135 [i]

(continues)


TABLE XIV (Continued)

Tg (“C) Compound MW [Refs. a-c] Other Tg values [Refs.]

Nystose Maltotetraose Maltopentaose Maltohexaose Maltoheptaose

666.6 77 666.6 111.5 147 [I] 828.9 125 165 [I] 990.9 134 173 [vl, 175 [g,l,y,ddI

1153.0 138.5

References [a] Slade and Levine (1988bl [b] Slade and Levine (1991a) [c] Levine and Slade (1989a) [d] Franks (1990) [el Franks and Grigera (1990) [f] Franks (1989) [g] Orford ef al. (1990) [h] Noel ef al. (1990) [i] Roos (1992a) 01 Soesanto and Williams (1981) [k] Chan et a/. (1986) [I] Orford ef al. (1989) [m] Ollett and Parker (1990) [aa] Roos and Karel (1991f) [cc] te Booy et al. (1992) [eel Suzuki and Franks (1993) [gg] Kalichevsky and Blanshard (1992b) [ii] Arvanitoyannis and Blanshard (1993a) [kk] Blond (1993) [mm] Murthy ef al. (1993) [OO] Hatley and Mant (1993)

[n] Green and Angel1 (1989) [o] Finegold et al. (1989) [p] Wasylyk and Baust (1991) [q] Quinquenet et al. (1988) [r] Blanshard and Franks (1987) [s] Roos and Karel (1991a) [t] Roos and Karel (1991~) [u] Roos and Karel (1990) [v] Franks ef al. (1991) [w] Roos and Karel (1991b) [XI Shimada et al. (1991) [y] van den Berg et al. (1993) [z] Roos (1992d) [bb] Roos and Karel (1991h) [dd] van den Berg (1992) [ff] Ablett et al. (1992~) [hh] Tian and Blanshard (1992a) bj] Levi and Karel (1992a) [Ill Franks (1993a) [nn] Franks and Hatley (1992)


1988b,c, 1989c; Franks, 1989, 1990, 1991c, 1992b; Pikal, 1990a,b, 1993)] or specify a temperature for product storage (Levine and Slade, l986,1988b,c, 1989a,c), one must measure it, rather than rely on an uncertain, predicted value. Nevertheless, a great deal of recent research effort, especially in the food area (Table XF), has gone into studies aimed at predicting Tg of binary or ternary systems, using equations commonly employed in the synthetic polymers field for different types of binary mixtures, including polymer-polymer and polymer-plasticizer systems (Levine and Slade, 1988a). The oldest and simplest of these equations is that of Gordon and Taylor (1952),


where Tgl, Tg2, and Tg are the T, values (in OK) of components 1 and 2 and of the blend, respectively, w1 and w2 are the weight fractions of components 1 and 2 in the blend, and k is an empirical constant. (Note that if k = 1, this equation reduces to the straightforward weight-averaging approach described above for orange juice.) This Gordon-Taylor equation was subsequently refined by Couchman and Karasz (1978), who defined k as ACpz/ACpl, where ACpl and ACp2 are the changes in heat capacity at the Tg of components 1 and 2, to yield the Couchman-Karasz equation. In either of these equations, extra terms can be added for mixtures with three or more components. As reviewed elsewhere (Levine and Slade, 1988a), both of these equations have been used, successfully only in certain instances, to predict Tg of blends of synthetic amorphous polymers and of synthetic polymers and plasticizers (Lee, water or organic diluents). As summarized in Table XF, both of these equations, as well as others, have also been used in recent years by various workers to predict (1) extents of depression of polymer T, by plasticizing water and/or (2) Tg of homogeneous blends of glass-forming food materials, including binary and ternary mixtures of various small sugars (i.e., saccharide monomers and oligomers), or sugar alcohols, and/or food polymers (e.g., starch, amylose, amylopectin, maltodextrins, dextran, xanthan gum, hyaluronic acid, gluten, glutenin, gliadin, casein, sodium caseinate, and fish protein), in the presence or absence of water. As in the case of synthetic polymers, these studies of food polymer systems have generally met with only mixed success, i.e., the agreement between results for measured vs. predicted T, has sometimes been good and sometimes not (Huang, 1993b).

B. PLASTICIZATION BY WATER

As mentioned earlier, plasticization, and its modulating effect on the temperature location of the glass transition, is a key technological aspect of synthetic polymer science (Sears and Darby, 1982). In that field, the classical definition of a plasticizer is “a material incorporated in a polymer to increase the polymer’s workability, flexibility, or extensibility” (Sears and Darby, 1982). Characteristically, the T, of an undiluted polymer is much higher than that of a typical low-MW, glass-forming diluent. As the diluent concentration of a solution increases, Tg decreases monotonically, because the average MW of the homogeneous polymer-plasticizer mixture decreases, and its free volume, (the volume of the mixture that is not occupied by molecules) increases (Ferry, 1980). Plasticization, on a molecu-


lar level, leads to increased intermolecular space or free volume, decreased local viscosity, and concomitant increased mobility (Ferry, 1980; Lillie and Gosline, 1990; Harrison et al., 1992). Plasticization implies intimate mixing and molecular compatibility, such that a plasticizer is homogeneously blended in a polymer, or a polymer in a plasticizer (Sears and Darby, 1982; Kalichevsky er aL, 1993b). Note that a true solvent, capable of cooperative dissolution of ordered crystals and having high thermodynamic compatibility and miscibility at all proportions, is always also a plasticizer, but a plasticizer is not always a solvent (Sears and Darby, 1982).

It is well known that water, acting as a plasticizer, affects the Tg of completely amorphous polymers and both the Tg and T, of partially crystalline polymers (Rowland, 1980; Levine and Slade, 1988a, 1989b, 1992b; Slade er al., 1989; Slade and Levine, 1991a,b). Water is a “mobility enhancer”; its low MW leads to a large increase in mobility, due to increased free volume and decreased local viscosity (Ferry, 1980), as moisture content is increased from that of a dry solute to a solution (Slade and Levine, 1988a,b). The direct plasticizing effect of increasing moisture content at constant temperature is equivalent to the effect of increasing temperature at constant moisture and leads to increased segmental mobility of chains in amorphous regions of glassy and partially crystalline polymers, allowing, in turn, a primary structural relaxation transition at decreased Tg (Rowland, 1980; Sears and Darby, 1982). Lillie and Gosline (1990) explained plasticization this way: “The plasticizing action of water or heat on an amorphous polymer can be explained in terms of free volume, a loosely defined parameter quantifying the space available for movement of the polymer chain segments. Increasing free volume through the introduction of a low MW solvent, such as water, or through thermal expansion reduces the barriers to rotational and translational motion. This reduction increases molecular mobility and shifts the viscoelastic properties away from the glass point as predicted by the WLF equation.”

Sears and Darby (1982) stated unequivocally that “water is the most ubiquitous plasticizer in our world.” Karel (1985) noted that “water is the most important . . . plasticizer for hydrophilic food components.” Atkins (1987) succinctly stated the important observation that “water acts as a plasticizer, dropping the Tg of most biological materials from about 200°C [for anhydrous polymers, e.g., starch, gluten, gelatin (Levine and Slade, 1988a)l to about -10°C or so [under physiological conditions of water content], without which they would be glassy” [in their native, in vivo state]. The latter Tg of about -10°C [corresponding to Tg’ (Levine and Slade, 1988a)l is characteristic of high-MW biopolymers at or above moisture contents, near 30%, corresponding to physiological conditions. This fact has been reported for many polysaccharides and proteins, including starch,


gluten, and gelatin (Levine and Slade, 1988a), hemicelluloses (Kelley et af., 1987), and elastin (Atkins, 1987). Recently, there has been increasing evidence (Table XC) to suggest that most, if not all, high-MW biopolymers appear to share a common glass curve (or, at least, very similar ones) (Slade and Levine, 1991a, 1993b; Levine and Slade, 1992b). With the weight of this evidence has come recognition (Slade and Levine, 1992a,b, 1993a-c) that the following rule of thumb appears to be widely applicable to many glass-forming polysaccharides and proteins (including starch, amylopectin, amylose, maltooligosaccharides and maltodextrins, hydroxyethyl starch, cellulose, hemicellulose, carboxymethyl celluose, dextran, pullulan, xanthan, hyaluronic acid, gluten, glutenin, gliadin, zein, collagen, gelatin, elastin, keratin, albumins, globulins, casein, poly-L-asparagine, lysozyme, and ribonuclease), and fully describes the "practical" portion of the glass curve [i.e., from dry T, to Tgr (Slade and Levine, 1991a)], which is relevant to many food applications of such high-MW biopolymers: (1) dry T, = 200 2 50"C, (2) T, decreases by =10 t 5"C/w% water at low moisture contents (5 = 10 w% water), (3) T, = room temperature at W, = 20 ? 5 w% water, and (4) T,' = -10 ? 5°C and Wgr = 25-30 w% water.

State diagrams illustrating the extent of the T,-depressing effect of water, as well as the above rule of thumb, have been reported for a wide variety of natural, as well as synthetic, water-compatible, glassy and partially crystalline polymers (Rowland, 1980; Levine and Slade, 1988a, 1989b, 1992b; Slade et al., 1989; Slade and Levine, 1991a,b; Nelson, 1993). In such diagrams, exemplified by the schematic one in Fig. 4, the smooth "glass curve" of T, versus w% composition demonstrates the dramatic effect of water on T,, especially at low moisture contents, where T, typically decreases by =10 t 5"C/w% water (Levine and Slade, 1988a), from the neighborhood of 200°C for an anhydrous polymer (Atkins, 1987). An example of such a state diagram (Fig. 9) depicts the amylopectin of freshly gelatinized starch as a typical water-compatible, completely amorphous polymer that exhibits a T, curve from >125"C for pure anhydrous starch to about -135"C, the T, of pure amorphous solid water (Mayer, 1991), passing through T,' at about -5°C (and W,' = 27 w% water) (Slade, 1984; Slade and Levine, 1984). Figure 9 shows T, of starch decreasing by =6"C/w% water for the first 10 w% moisture, in good agreement with another glass curve for starch, calculated from free volume theory (Blanshard, 1988; Marsh and Blanshard, 1988). Another recently published glass curve for completely amorphous amylopectin (waxy maize starch) samples with moisture contents in the 10-25 w% range, measured by DSC (Kalichevsky et af., 1992a), showed T, continuing to decrease by --7"C/w% water in this range, in agreement with another DSC-measured glass curve, this one for native wheat starch, reported by Zeleznak and Hoseney (1987a). Like the glass curve


I . . . . . . . . . 0 a2 o.lr 0.6 a0

0, (Weight fraction of water)

0

-100

-134

I rg. w

FIG. 9. State diagram, showing the approximate Tg values as a function of mass fraction, for a gelatinized starch-water system. From van den Berg (1986) with permission.

in Fig. 9, the experimental curve reported by Kalichevsky et al. could be smoothly extrapolated to the Tg'-Wg' point for starch (Levine and Slade, 1992b).

Similarly, the glass curve for water-compatible, amorphous wheat gluten in Fig. 10 (Hoseney et al., 1986) shows a decrease in T, from >160"C at <1 w% water to -15°C at -16 w% water, a depression of -1O"C/w% water in this moisture range (Fujio and Lim, 1989; Kalichevsky et al., 1992~). The plasticizing effect of water on gluten continues at higher moisture contents, until T, falls to T,' = -7.5"C and W, reaches W,' = 26 w% water (Slade et al., 1989). As in the glass curve for starch in Fig. 9, a smooth curve of T, versus w% composition results for gluten, when one connects the DSC- measured data points for (1) T, of gluten samples at <2 w% moisture (Fujio and Lim, 1989); (2) T, of commercial vital wheat gluten powder at 6 w% "as is" moisture [from Slade's (1984) original report of a glass transition in gluten]; (3) Tg'-W,' (Slade et al., 1989); and (4) T, of water. Another experimentally measured glass curve, this one for the glutenin component of gluten, likewise shows a monotonic decrease in T, from -110-130°C at 4 w% water to =20°C at 14 w% water, a depression of =9-11 C/w% water in this moisture range (Cocero and Kokini, 1991, 1993). These results for glutenin were recently confirmed by Kalichevsky et al. (1992c), who also


0 4 0 12 16

WATER %

FIG. 10. Change in T, as a function of moisture for a hand-washed and lyophilized wheat gluten. From Hoseney et al. (1986) with permission.

measured Tg for the gliadin component of gluten [as did Cocero et al. (1992a), de Graaf et al. (1992), and Madeka and Kokini (1992a,b, 1993)] and reported that “the Tgs of both glutenin and gliadin are very similar to the Tg of whole gluten.”

Since the first reports of a glass transition in starch (van den Berg, 1981, 1986; Slade, 1984; Slade and Levine, 1984, 1987b, 1988c; Maurice et al., 1985) and in gluten (Slade, 1984; Levine and Slade, 1988a; Slade et al., 1989), and recognition by these workers of the importance of the glass transition temperature, and its dependence on water content, to the structure-function relationships of starch and gluten in baked products and other wheat flour-based foods, the aqueous glass curves of these two major food polymers, and the effect of plasticization by water on their Tgs, have been actively investigated by many other workers (references in Tables VIIIA and VIIIB). Since starch and gluten are the major storage polymers of wheat endosperm, and they are synthesized, stored in the mature seed, and hydrolyzed by enzymes of germination in the same temperature and moisture environment, it appears biochemically and physiologically logical that their aqueous glass curves should be similar, if not identical (Levine and Slade, 1992b). That is, since conditions of synthesis, storage, and use


are identical for the two storage polymers, equivalent performance should reflect equivalent glass curves. In fact, the results and conclusions of many of the studies cited in Table VIII seem to support this suggestion. Parts A and B of that table list references to studies of the Tgs of starch, gluten, and/or their component polymers and to methods used to determine them. Taken together, these Tg results appear to fall in the same region of temperature and moisture content, i.e., a Tg around “room temperature” (=15-25”C) at a moisture content around 15 to 20 w% (Levine and Slade, 1992b). The technological implications of such a Tg and moisture content to the processing and product properties of starch- and/or gluten-based food products (e.g., doughs and baked goods) are obviously profound, as discussed in the references listed in Table VIIIC. Particularly noteworthy in this regard are recent results from Le Meste et al. (1992) on the glass transition of white pan bread, measured (in terms of onset temperature for softening) by TMA. In this first study of the thermomechanical glass transition in an actual, complex food product, they reported a glass curve of Tg versus moisture content for bread, which begins at -165°C at 0 w% water, decreases by =lO°C/w% water from 0 to 10 w% moisture and by =5”C/w% water from 10 to 20 w% moisture, thereby passing through Tg = 20°C at e16.5 w% water, and then levels off at Tg = -12°C (i.e., T g f ) for moisture contents above =25 to 27 w% (i.e., Wgf). Such a glass curve is just what one might expect for a water-plasticized, amorphous, mixed matrix of developed gluten and gelatinized starch, which comprises the continuous, three-dimensional, structural network of fresh-baked bread crumb (Levine and Slade, 1989b; Aynie et al., 1992a,b; Dong, 1992; Le Meste et al., 1992; Hallberg and Chinachoti, 1992; Huang et al., 1992; Kim and Taub, 1993).

In a related vein, also noteworthy are the recent interest and research activity concerning the glass transition and the effect of water plasticization on Tg and resulting rheological properties of corn zein protein and zein- containing doughs (Lawton, 1992a,b; Cocero et al., 1992a,b; Kokini et al., 1992d; Madeka and Kokini, 1992a,b, 1993; Magoshi et al., 1992). Particular attention is due to the meticulous DSC studies by Sochava and co-workers on the glass transition and the effect of water plasticization on Tg in lysozyme and various other globular proteins (Sochava et al., 1991; Sochava and Belopolskaya, 1992; Sochava and Smirnova, 1992).

Returning to the subject of water plasticization of starch, we note the importance of the fact that plasticization of the amorphous regions (e.g., backbone segments and branch points of amylopectin molecules) of native granular starches by sorbed water is neither instantaneous nor simultaneous with the initial swelling caused by water uptake (Slade and Levine, 1991a). It was demonstrated by Aguerre et al. (1989) that “the uptake of water


takes place between the concentric layers” of the granule, leading to “interlamellar expansion of the starch granule structure.” This sorbed water must subsequently diffuse from interlamellar spaces to the amorphous regions of the granule before plasticization of polymer molecules or chain segments in those amorphous regions can begin (Yost and Hoseney, 1986; Slade and Levine, 1991a). The salient point is that while water is unquestionably an effective plasticizer of starch and all other water-compatible food polymers, the mere presence of water does not attest that plasticization has already occurred (Slade and Levine, 1991a). Thus, when one reads the following quote from Hoseney (1986),

When starch is placed in water, the granule is freely penetrated by water, or for that matter, by most small molecules. The starch can hold about 30% of its dry weight as moisture. The granule swells slightly; its increase is generally considered to be about 5%. The volume change and water absorption are reversible, and heating the system to just below its gelatinization temperature will not bring about any other changes,

one must be careful not to equate rapid penetration (e.g., through preexist- ing channels and voids) with rapid plasticization by water, as some have done (Liu and Lelievre, 1991b; Lelievre, 1992a). In fact, plasticization at T below the initial T, of a material is very slow, but it occurs much more rapidly at Tabove the initial T, (Slade and Levine, 1991a). This was graphically demonstrated by some remarkable swelling studies of poly(viny1 chloride) (PVC) described by Sears and Darby (1982). It was found that several “common PVC plasticizers. . . would not swell an unplasticized PVC sheet at room temperature in two years’ time . . . . Yet, all of the plasticizers swelled the rigid PVC sheet at 76”C, the approximate T, of the resin. When the same PVC was hot-compounded with a plasticizer, cooled, and then immersed in various other plasticizers, it would imbibe more plasticizer.”

A new diagnostic experiment (a “thought” experiment, i.e., designed but not yet conducted), outlined in Table XV (Slade and Levine, 1993b), was conceived to reinforce the critical distinction between “presence of water” and “plasticization by water.”

I . Plasticization and Antiplasticization by Sugars

An interesting phenomenon to discuss in this section on plasticization concerns what we have always thought of as the “schizophrenic” behavior of sucrose (or other typical low-MW sugars such as glucose or fructose) in a starch-sucrose-water system (Slade and Levine, 1987b, 1993b; Levine and Slade, 1988c, 1989~). By schizophrenic, we mean showing either of two very different (essentially opposite) personalities, depending on which of two different situations exists; where the two different behaviors can each be


TABLE XV DIAGNOSTIC EXPERIMENT TO DISTINGUISH “PRESENCE OF WATER“ FROM

“PLASTICIZATION BY WATER””

A. Native commercial wheat starch at room temperature and “as is” moisture content: 1. Observe low relative vapor pressure (RVP) at 25°C; this implies that room

temperature, T, is below initial Tg and = 6 w% moisture, W, is below Wg’. 2. So, water is present and acts as a plasticizer, but it is insufficient to depress initial Tg

to below 25°C. B. Add water at room temperature to achieve W such that W8’ > W > 20 w%:

1. Observe high R W at 25°C; this implies that T is below initial T8 and

2. So, more water is present, but it does not act as an instantaneous plasticizer, because T is below initial Tg, even though total (but nonplasticizing) water is sufficient to depress Tg to below T after long relaxation time.

Wg’ > w > 20 w%.

C. Add water at a different temperature, T, such that initial Tg < T < initial T,,, (i.e., gelatinization peak temperature, from DSC), to achieve W such that Wg‘ > W > 20 w% (same Was in part B, above), then cool to room temperature to complete treatment: 1. Both T and W are the same as in part B when RVP is measured at room

temperature, but the path was different, because water was added below (part B) versus above (part C) initial Tg.

2. So, observe low RVP, such that part A RVP < part C RVP < part B RVP.

temperature, from DSC), to achieve higher W such that W > Wg‘, then cool to room temperature: 1. Amylopectin is still partially crystalline, but it is completely plasticized. 2. So, observe high RVP, such that part A RVP < part C RVP C part D RVP C part B

D. Add water at T such that initial T, C T < initial T,,, (i.e., gelatinization peak

RVP. E. Add water at higher T such that T = amylose Tg > initial T,,, (amylopectin), to achieve

W such that W > Wg’ (same W as in part D), then cool to 25°C 1. Amylopectin is completely amorphous and plasticized, and its short branches depress M, of amylopectin, but amorphous, high-MW amylose is also plasticized and contributes to overall Tg (also to network Tg?).

RVP < part B RVP. 2. So, observe high RVP, perhaps part A RVP < part C RVP C part E RVP < part D

a Modified from Slade and Levine (1993b).

described and explained in terms of established concepts, but the underlying reason for the manifestation of two different personalities is not completely understood. As indicated in Table IXF, the “antiplasticizing” effect of sucrose and other sugars (i.e., of a sugar-water solution relative to water alone) on the gelatinization of native starches [normal or waxy (all amylopectin)] (Slade and Levine, 1987b) is well established and widely accepted. In the presence of native starch and excess water (i.e., W > Wg’), sugar behaves as a plasticizing cosolvent with water, such that the sugar-water co-plasticizer, of higher average MW [and lower free volume, so higher Tg


(Ferry, 1980)] than water alone, plasticizes [depresses the temperature of the glass transition of the amorphous regions, which immediately precedes the gelatinization of native, partially crystalline (Slade and Levine, 1988c)l starch less than does water alone, so that the gelatinization temperature (as well as the T, that precedes it) in the presence of sugar (whether initially added to the mixture as a crystalline solid or predissolved in water) is elevated (hence, “anti”) relative to the gelatinization temperature of starch in water alone. Moreover, with increasing concentration of sugar in the three-component mixture (thus, a sugar-water co-plasticizer with increasing average MW, decreasing free volume, and increasing T,, relative to water alone), the magnitude of the antiplasticizing effect increases, and so do T, and the gelatinization temperature (Slade and Levine, 1987b).

If we then take the very same sample and immediately freeze it, we see that T,’ of the homogeneous, completely amorphous, three-component matrix (gelatinized starch or amylopectin + sugar + unfrozen water) surrounding the crystals of pure ice is lower than the corresponding T,’ of gelatinized starch in water alone, falling somewhere between the T,’ values of the sugar-water alone (e.g., -32°C for sucrose, -43°C for glucose) and gelatinized starch-water alone (= -5”C), as determined by the weight ratio of starchmgar in the sample (Levine and Slade, 1988c, 1989~). Thus, in the freeze-concentrated matrix, sugar has changed its behavior by assuming the role of a low-MW co-solute with starch, which lowers the average MW and increases the free volume of the starch-sugar mixture, thus depressing T,’ relative to the T,’ of gelatinized starch-water i:i the absence of sugar. In other words, in this situation, sugar is acting as a plasticizer of starch, which is the behavior one would ordinarily expect of a low-MW oligomer in a homogeneous blend with a high-MW polymer with which it is thermody- namically compatible (Ferry, 1980; Wunderlich, 1980; Sears and Darby, 1982). Analogous behavior, i.e., various sugars (e.g., sucrose, glucose, fructose) and polyols (e.g., glycerol) acting as plasticizers and depressing T, of amorphous starch or amylopectin in three-component aqueous mixtures at low moisture contents (W < Wg’), has been described in several reports (Ollett et al., 1991; Roozen et al., 1991; Kalichevsky et al., 1992b; Kalichevsky and Blanshard, 1992b,c; Shogren et al., 1992; Arvanitoyannis et al., 1993; Attenburrow and Davies, 1993; Barrett et al., 1993; Kirby et al., 1993; Sala and Tomka, 1992a).

Let us summarize the apparent symptoms of sugar’s schizophrenia. In the first situation (gelatinization), sugar acts as a co-solvent with water, whereas starch acts as the sole solute, and its Tg is elevated in the presence of sugar. In contrast, in the second situation (frozen or low-moisture systems), sugar acts as a co-solute with starch, whereas water acts as the sole solvent, and starch’s Tg is depressed in the presence of sugar. Such schizophrenic


behavior of sugars is evidently not limited only to starch-containing systems, but may be generalizable to other polysaccharide- (Levine and Slade, 1989c) and protein-containing systems (Slade et al., 1989) as well. For example, the effect of sucrose and other sugars to elevate the denaturation temperature of various enzymes (e.g., lysozyme) and globular proteins (e.g., egg white, soy) in aqueous solution is well known (Franks, 1988; Harwalkar and Ma, 1990; Ma and Harwalkar, 1991), although the mechanism of this action of sugars to raise the heat-set-gelation temperature of proteins had not previously been ascribed to “ antiplasticization” (Slade and Levine, 1993b), even though such native proteins are known to have amorphous regions that undergo glass transitions (Slade et al., 1989; Sochava et al., 1991; Sochava and Smirnova, 1992). In contrast to this stabilizing effect of sugars, the behavior of various sugars (e.g., sucrose, glucose, fructose) acting as plasticizers and depressing Tg of amorphous gluten or caseinate in three-component aqueous mixtures at low moisture contents (W < Wgr) has been described in several reports (Kalichevsky et al., 1992c, 1993b; Kalichevsky and Blan- shard, 1992~). The analogous effect of sucrose to depress Tg’ of a frozen gluten-sugar-water mixture (with W > Wgr) , relative to the corresponding Tgr of gluten-water in the absence of added sugar, had been reported earlier (Slade et al., 1989).

C. THE DYNAMICS MAP

The key to our perspective on concentrated, water-plasticized food polymer systems relates to recognition of the fundamental importance of the dynamics map shown earlier in Fig. 3. The major area of the map (i.e., the area surrounding the reference state in two dimensions and projecting into the third, time, dimension) represents a nonequilibrium situation corresponding to the temperature-composition region of most far-reaching technological consequence to aqueous food systems (Slade and Levine, 1988b). The critical feature in the use of this map is identification of the glass transition as the reference state (Slade and Levine, 1988a,b; Franks and Grigera, 1990), a conclusion (Levine and Slade, 1988a; Ollett and Parker, 1990) based on WLF theory for glass-forming polymers. This line of demar- cation (representing the glass curve of Tg versus composition) (Levine and Slade, 1986; Simatos and Karel, 1988; Franks, 1989, 1990; Karel, 1992; Karel et al., 1993b) provides a basis for description of the nonequilibrium thermomechanical behavior of water-compatible polymeric materials in glassy and rubbery states (Levine and Slade, 1988a), in response to changes in the key variables of moisture content, temperature, and time (Tsourouflis et al., 1976). Mobility is the transcendent principle underlying definition of the glass transition as the appropriate reference state (Slade and Levine,


1988a), because mobility is key to all transformations in time (or frequency), temperature, and composition between different relaxation states for a technologically practical food (Slade and Levine, 1988b) or biological (Lillie and Gosline, 1990) system (Karel, 1992; Karel et af., 1993b).

As mentioned earlier, the interdependent concepts of water dynamics and glass dynamics embodied in the dynamics map have provided insights into the relevance of the glassy reference state to functional aspects of a variety of food systems (Levine and Slade, 1988a; Slade and Levine, 1988b). For example, the kinetics of all constrained relaxation processes, such as translational or large-scale rotational diffusion (and any chemical or enzymatic reaction dependent thereon), which are governed by the mobility of a water-plasticized polymer matrix in glass-forming systems, vary (from Arrhenius- to WLF-type) between distinct temperature/structure domains, which are divided by this glass transition (Table XB). Thus, while Arrhenius kinetics are applicable below Tg in the glassy solid state of very low mobility and very slow diffusion (the domain of glass dynamics, labeled STABLE in Fig. 3), WLF kinetics (Ferry, 1980) are applicable above Tg in the viscoelastic, rubbery liquid state of accelerating mobility and diffusion (the domain of water dynamics, labeled REACTIVE in Fig. 3) (Slade and Levine, 1988b). The WLF equation (Williams et af., 1955; Ferry, 1980) defines the kinetics of molecular-level relaxation processes (such as the collapse phenomena referred to earlier), which will occur in practical time frames only in the rubbery state above Tg, in terms of an exponential, but non-Arrhenius, function of AT above this boundary condition (Levine and Slade, 1986, 1988b, 1989d, 1992b; Slade and Levine, 1988b, 1991a, 1993c; Franks and Grigera, 1990; Roos and Karel, 1990, 1991a,d-h, 1992, 1993; Franks et af., 1991; Lim and Reid, 1991,1992; Shimada et al., 1991; Buera and Karel, 1992; Karathanos et al., 1992; Karel, 1992; Karmas et af., 1992, 1993; Labrousse et al., 1992; Levi and Karel, 1992a, 1993a,b; Nelson and Labuza, 1992a, 1993; Reid ef af., 1992, 1993b; Karel et af., 1993a,b; Kerr et af., 1993; Nelson, 1993). The highest mobility and most rapid diffusion occur in the region above a second set of reference lines, the equilibrium liquidus and solidus curves, which demark the upper boundary of the WLF region, where Arrhenius kinetics again apply (Levine and Slade, 1989~). Within the WLF region, kinetics accelerate according to the WLF equation, from the extremely steep temperature dependence of WLF kinetics just above Tg to the moderate temperature dependence of Arrhenius kinetics on approaching T, (Slade and Levine, 1988b, 1993~). The WLF equation describes the profound range of kinetics between Tg and T,, with correspond- ingly profound implications for process control, product quality, safety, and shelf life (Levine and Slade, 1989d; Slade and Levine, 1991a). It should be noticed in Fig. 3 that A, is correctly defined only in the region of the map


corresponding to a dilute solution at equilibrium at room temperature (Franks, 1991a,b). In contrast, the actual measured relative vapor pressure (RVP) of an (nonequilibrium) IMF system would approach zero in the limit of the glassy reference state at temperatures below T, and moisture contents below W,, but would increase toward 1.0 with increasing temperature above T, and increasing moisture content above W, (Slade and Lev- ine, 1991a,b).

One particular location among the continuum of T, values along the reference glass curve in Fig. 3 results from the behavior of water as a crystallizing plasticizer and corresponds to an operationally invariant point [defined earlier as T,’ (Table XIA)] on a state diagram for any particular solute. A compilation of recently reported literature values of T,’ for many different low-MW carbohydrates, along with corresponding C,’ values, is shown in Table XIII. As illustrated in the schematic state diagram shown earlier in Fig. 4, T,‘ corresponds to, and is determined by, the point of intersection of the kinetically determined glass curve for homogeneous solute-water mixtures and the nonequilibrium extension of the equilibrium liquidus curve for T,,, of ice (Levine and Slade, 1986, 1988a-d, 1989a-d, 1990). This solute-specific location defines the composition of the glass that contains the maximum practical amount of plasticizing moisture [W,’, expressed as g UFW/g solute or w% water, or alternatively defined in terms of C,’, expressed as w% solute (Levine and Slade, 1986,1988a)l and represents the transition from concentrated fluid to kinetically metastable, dynamically constrained solid that occurs on cooling to T < T,’ (Slade and Levine, 1988b, 1991a). In this homogeneous, freeze-concentrated, solute-- water glass, the water represented by W,’ is not “bound” energetically, but rather rendered unfreezable in a practical time frame due to the immobility imposed by the extremely high local viscosity of =lo1* Pa s at Tg’ (Franks, 1982,1983a, l985,1986a,b, 1989,1990,1991a; Levine and Slade, 1986,1988a, 1989b Slade et al., 1989; Franks and Grigera, 1990; Oksanen and Zografi, 1990, 1993; Belton, 1990, 1991; Slade and Levine, 1991a,b; Tanner et al., 1991; Goff, 1992).

Marsh and Blanshard (1988) documented the technological importance of freeze-concentration in such situations and the practical implication of water being a readily crystallizable plasticizer, characterized by a high value of Tm/Tg ratio = 2 (Soesanto and Williams, 1981; Slade and Levine, 1988b; Wunderlich, 1990). A theoretical calculation (Marsh and Blanshard, 1988) of Tg of a 50% wheat starch gel fell well below the measured value of -5” to -7°C for T,’ (Slade, 1984; Slade and Levine, 1984, 1987b; Liu and Lelievre, 1991b, 1992b), because the calculation, based on free volume theory, did not account for ice formation and resulting freeze-concentration that occur below about -3°C. Recognition of the practical limitation of


water as a plasticizer of water-compatible solutes, due to phase separation of ice (Slade and Levine, 1988b, 1991a; Hancock and Zografi, 1993; Lillie and Gosline, 1993), reconciled the difference between theoretical and measured values of T, (Marsh and Blanshard, 1988). Moreover, the theoretical calculations supported the measured value of -27 w% water (Slade, 1984; Slade and Levine, 1984, 1987b) for W,’, the maximum practical water content of a homogeneous wheat starch-water glass. The calculated water content of a wheat starch glass with T, of about -7°C is -28 w% (Marsh and Blanshard, 1988), and 26 w% for a measured T, [i.e., T,’] of -5°C (Roos and Karel, 1991d).

In the context of the present discussion of the practical importance of T,‘, as well as the earlier discussion of collapse phenomena, a report by Chang and Randall (1992) is particularly noteworthy. They used a combination of DSC and TMA to determine optimum freeze-drying conditions for various multicomponent aqueous solutions of proteins with sugars and/or salts. Their DSC results showed measured values of T,’ [taken as the midpoint, rather than the onset, of the transition (Table XIID), as obtained from derivative heat-flow curves (Levine and Slade, 1986)] for various maximally freeze-concentrated matrices, as well as corresponding lower- temperature values of Tg (arising from only partial freeze-concentration), which were in excellent agreement with previously published results (Levine and Slade, 1988c, 1989c; Slade et al., 1989). More importantly, their TMA results for the same samples (measured in terms of probe penetration, as a consequence of mechanical softening of the ice-containing matrix, at T > T,’), which agreed with their DSC data, verified that structural collapse (“melt-back”) of the amorphous matrix during the ice sublimation stage of their freeze-drying experiments was only observed at freeze-drying temperatures above T,’, but not at T, < T < Tg’. They concluded that T,’ represents “the true collapse temperature,” while the lower-temperature T, “is not related to collapse” (Chang and Randall, 1992). Their conclusion was in full agreement with a concept for interpreting collapse phenomena in frozen food systems, which had been advocated in earlier reports dealing with the theory and practice of “cryostabilization technology” (Levine and Slade, l986,1988b,c, 1989c; Slade and Levine, 1991a), but aspects of which have subsequently been questioned by other workers (Simatos and Karel, 1988; Blond, 1989, 1993; Simatos et al., 1989; Blond and Simatos, 1991; Izzard et al., 1991; Karel, 1991c, 1992; Karel and Saguy, 1991; Le Meste and Huang, 1991; Roos and Karel, 1991c-h, 1993; Simatos and Blond, 1991, 1993; Ablett et al., 1992a,b,c, 1993; Roos, 1992a,c,d; Kerr et al., 1993; MacInnes, 1993; Reid et al., 1993a,b). According to this concept of cryostabilization against collapse, Tg’ (rather than the lower-temperature T g ) is the only glass transition temperature relevant to the stability of ice-containing


foods, against any translational diffusion-limited deterioration process (e.g., structural collapse), during freeze-drying or freezer storage (Levine and Slade, 1989c; Slade and Levine, 1991a). The previous questioning of certain aspects of this concept (discussed further in the next section) has concerned (1) the true identity of the transition (i.e., glass transition or ice melting?) labeled Tgf in typical DSC profiles of frozen sugar solutions [e.g., Fig. 1 in Levine and Slade (1986)], and (2) which of the two DSC-measured transitions, Tg’ or the lower T,, is the predominant one governing stability against collapse processes in frozen foods. The results and conclusions of Chang and Randall (1992) seem to lend strong support to the original interpretation of the cryostabilization concept.

1. Tgr and Ce‘: Current Issues

As mentioned earlier and discussed in previous reviews (Slade and Levine, 1991a; Levine and Slade, 1992b), Tgf and C,’ values, especially of various low-MW sugars widely used in foods, are a topic of much current interest, controversy, and debate (Table XII). The wide variations in recently published T,’ and C,’ values for many low-MW carbohydrates were illustrated earlier in Table XIII. The present discussion focuses on the current status of several key issues.

a. Tgf Values. The following issues around published T,‘ values, which relate mainly to different interpretations of the DSC-measured transitions, were summarized in introductory notes (prepared by Unilever Colworth House and Wageningen University) for an informal discussion conference on “glass transitions in foods,” held at Wageningen University in The Netherlands in May 1990, and attended by about two dozen invited special- ists in this area.

1. “The major transition (the “T,’ glass transition” [measured as the midpoint temperature (Table XIID)]) is the glass transition of the maximally freeze-concentrated solution, and this is assumed to be the only transition of real significance. The other minor transitions [including a lower-temperature T, arising from only partial freeze-concentration during initial freezing (Levine and Slade, 1986, 1988c, 1989c)l tend to be ignored” (Schenz et al., 1984, 1991, 1992; Slade and Levine, 1984, 1985, 1988a,b, 1991a; Franks, l985,1986a,b, l989,1990,1991c, 1992b, 1993a,b; Reid, 1985, 1990; Levine and Slade, 1986,1988a-d, 1989a-d, 1990,1991,1992b; Hatley et al., 1989, 1991; Pikal, 1990a,b, 1993; Roozen and Hemminga, 1990, 1991; Blanshard et al., 1991; Franks et al., 1991; Hatley, 1991; Hatley and Franks, 1991; Lim and Reid, 1991, 1992; Maurice et af., 1991; Roozen et af., 1991;


Attwool et al., 1992; Chang and Randall, 1992; Goff, 1992; Hsu and Reid, 1992; Reid and Hsu, 1992; Reid et al., 1992, 1993b; Sahagian and Goff, 1992, 1993; Taylor et al., 1992; van den Berg, 1992; Goff et al., 1993; Hatley and Mant, 1993; Hemminga et al., 1993; Kerr et al., 1993; McCurdy et al., 1993; van den Berg et al., 1993).

2. “There are two glass transitions, which correspond to either two different conformations or two concentrations of the sugar present, i.e. due to inhomogeneity within the sample” (Simatos and Karel, 1988; Blond, 1989, 1993; Simatos et al., 1989; Blond and Simatos, 1991).

3. “There is only a single glass transition, and the complex nature of the DSC transition is due to an additional stress relaxation process which occurs at the glass transition” (Blond, 1989; Blond and Simatos, 1991; Simatos and Blond, 1991, 1993; Hatley and Mant, 1993).

4. “There is only a single glass transition, and the second major transition in the thermograph, the so-called ‘Tgr transition’, is not a glass transition at all, but is due to the onset of ice melting, which starts to occur soon after the sample goes through the glass transition, and represents the temperature at which the sample changes from kinetic to thermodynamic control” (Le Meste and Simatos, 1990; Noel et al., 1990; Izzard et al., 1991; Le Meste and Huang, 1991; Roos and Karel, 1991c-h, 1993; Ablett et al., 1992a-c, 1993; Jouppila and Roos, 1992; Karel, 1992; Roos, 1992a,c; Karel et al., 1993a,b; MacInnes, 1993).

The above issues were still current in 1993, as demonstrated by the dates of the above-cited references, and by summary notes prepared by M. Karel following (a) the April ’92 Easter School on The Glassy State in Foods and (b) the November 1992 ISOPOW-V Meeting (mentioned earlier with regard to Tables IIIE and IIIG, respectively). After (a), Karel remarked: “With respect to Tgr and Cgr, a consensus is forming, largely due to studies on carefully annealed systems using both mechanical spectrometry and calorimetry. It is evident that the transition phenomenon initiated at Tgr merges with ice melting.” It should be noted that the latter point has never been at issue; the original description of Tgr, according to (1) above and consistent with the schematic state diagram in Fig. 4 (the validity of which is unquestioned), made it clear that the DSC-measured Tgr value coincides, in the time frame of the DSC experiment, with the temperature at which ice formation ceases during cooling and ice melting begins during subsequent warming (Levine and Slade, 1986, 1988b,c, 1989~). After (b), Karel noted that “frozen food stability could be described with reference to a “critical temperature” (T,,), but the data do not allow a confirmation of the correspondence of T,, to Tg or Tgr.” Despite this statement, the fact remains that many workers currently support the view that T,, = Tg’ [based on (l)]; e.g., Tgr = -32°C for sucrose (Table XIIB).


Although not necessarily part of the consensus referred to above by Karel, the following points about interpretations (1) to (4) are worthy of mention. The proponents of (3), who had earlier been proponents of (2), have consistently reported that annealing experiments, performed with galactose solutions to confirm the hypothesis, were not completely successful (Blond, 1989; Blond and Simatos, 1991; Simatos and Blond, 1991,1993). One of the major arguments against (l), by some proponents of (4), has been the claim that the “T,‘ glass transition” cannot be a glass transition at all, but rather must be an ice-melting onset transition, because ACp at Tg’ (e.g., for sucrose) is too large (Izzard et af., 1991; Ablett et al., 1992a,c, 1993), when compared, e.g., to results from theoretical modeling simulations of Cp-T data for sucrose-water systems (Ablett et af., 1992b). However, results of other concurrent analyses of ACp data [for galactose-water (Blond and Simatos, 1991), maltooligosaccharide-water (Schenz et al., 1992), and sucrose-water and stachyose-water (Hatley and Mant, 1993) systems] have shown this argument to be questionable. Moreover, according to the criteria for “fragile” glass-forming liquids described by Angel1 et af. (1992) (discussed further later), one would expect that ACp at T,’ would increase with increasing “fragility” of aqueous solutions of a series of sugars. Franks (1993b) has concluded that “the origin of the relationship between the heat capacity change accompanying the glass transition and the degree of polymerization of the substrate is obscure.” Most recently, Ablett et al. (1993) have conceded that their “explanation for the origin of the single transition [- T,’, based on (l)] for biopolymers [as opposed to the two transitions they claim to observe for low-MW sugars] does mean that it would be valid to use this transition to characterize the T,’ temperature in the case of biopolymers, even though the observed transition probably arises from a different phenomenon.” Similarly, Roos and Karel (1991d,e) have conceded that the two transitions they advocate [as proponents of (4)], T,’ and T,‘ (as onset temperatures), merge to a single transition for high-MW carbohydrates such as starch and low-DE maltodextrins, with a midpoint temperature identical to the corresponding T,‘ values [based on (l)] previously reported for starch and the same low-DE maltodextrins (Slade and Levine, 1984, 1987b, 1988c; Levine and Slade, 1986, 1988b).

As mentioned earlier, a large body of evidence has been built [e.g., references in (1) above, including the freeze-drying study by Chang and Randall (1992), described earlier, and the studies by Lim and Reid (1991, 1992) and Kerr et af. (1993; Reid et al., 1992) of reaction kinetics in relation to Tg’ in frozen model systems, described later], on a foundation of the “cryostabilization technology” concept, which supports the practical importance of T,’ [based on (l)] as the temperature [= Karel’s T,, (Hsu and Reid, 1992; van den Berg, 1992; van den Berg et af., 1993)] critical to frozen


food stability and successful freeze-drying, because below T,’, translational diffusion-limited processes will not occur, on a practical time scale, in ice-containing frozen food systems. The current situation is perhaps best summed up by the pragmatic view espoused by Reid in the following remarks: “Whilst, as scientists, we are deeply committed to determining the true nature of the events which lead to the “humps” and “bumps” in the calorimeter traces, for practical purposes it may not be so important. If product stability can be correlated with any particular bump, then this is of practical value, and should not be overlooked” (Reid et al., 1993a), and “While the discussion of the true nature of these transitions is very important, it does not invalidate the observation that in many ways reaction kinetics are affected by the temperature of the event labelled Tgf by Levine and Slade” (Reid et al., 1993b).

b. C,’ Values. The introductory notes from the above-mentioned Wa- geningen conference had this to say about the issue of Cgf values:

There is concern about how the concentration of the freeze-concentrated glass can be measured. The most widely publicized method [e.g., Levine and Slade, 1986, 1988bI is to determine the total amount of heat required to melt the ice in a known weight of sample [i.e., a single 20 w% solution]. The amount of ice present is then calculated from the known latent heat of ice. However, there is now evidence [references for “other Cg’ values” in Table XIII] to show that this method consistently underestimates the amount of ice present, and consequently underestimates the freeze-concentrated concentration (Cg’). Therefore there is some concern about the validity of published C,’ data on numerous food compounds [i.e., Levine and Slade, 1988b,c].

Two years later, this statement was followed up by Karel’s earlier-quoted remark, at the close of the Easter School at Nottingham, about a consensus forming with regard to Cgf values. He was referring to a widely held belief that C,’ values for many low-MW sugars and polyols all appear to fall in a narrow range around 80 k 5 w% solute (references in Table XIII). This point of view has been supported by painstaking analyses [as exemplified by that of Hatley et al. (1991)l of actural state diagrams, such as the most popular one for sucrose-water, and of the Cgf concentration [by definition, the solute concentration of a solution with no “freezable” water (Levine and Slade, 1986)] corresponding to the point at which measured Tgf coincides with the glass curve constructed from measured values of Tg as a function of solute concentration for highly concentrated [in many instances, supersaturated, and therefore unstable with respect to solute crystallization during instrumental cooling (e.g., MacInnes, 1993)], completely vitrified solutions with values of C, surrounding C,’ (Blond, 1989, 1993; Simatos et al., 1989; Williams and Carnahan, 1990; Blanshard et al., 1991; Franks,


1991c, 1992b, 1993a,b; Franks and van den Berg, 1991; Franks et al., 1991; Hatley et al., 1991; Izzard et al., 1991; Le Meste and Huang, 1991; Roos and Karel, 1991d-h, 1993; Simatos and Blond, 1991, 1993; Ablett et al., 1992a-c, 1993; Leung et al., 1992; Roos, 1992a,c; van den Berg, 1992; Hatley and Mant, 1993; van den Berg et al., 1993).

Today, it seems safe to say that different values of C,’ can result, depending on the method used to determine it for a given low-MW carbohydrate (Tables XIIA, XIIC, and XIII). For example, for sucrose, the “state diagram” method has produced a “rigorous” C,’ value of =80 (24) w% (Tabe XIII), whereas analysis of uncorrected DSC data for a single freeze- concentrated 20 w% solution (not annealed) produced a much lower “apparent” C,’ value of -64 w% (Levine and Slade, 1988b), which has subsequently been confirmed by several other workers (Hatley et aL, 1991; Izzard et al., 1991; Ablett et al., 1992a,b; Nesvadba, 1992a; Reid and Hsu, 1992; Reid et al., 1993a). As shown by the data compiled in Table XIII, similarly large differences between reported C,‘ values determined by these two methods result for many other sugars. It should also be noted, as has been widely reported for sucrose (as well as various other sugars and polyols), that it is generally (1) very difficult to force ice to form, in supersaturated solutions of =65 to 70 w% concentration, without extensive annealing, temperature cycling, or in the absence of stress cracking (which serves to heterogeneously nucleate ice formation) of an otherwise totally vitrified glass (Williams and Carnahan, 1989, 1990; Hosea et al., 1990; Williams et al., 1990, 1993), and (2) essentially impossible, in a realistic experimental time frame, to force ice to form in solutions of > =70 w% concentration, even with extensive annealing (MacKenzie, 1977; Franks, 1982; Levine and Slade, 1988a; Blond, 1989,1993; Simatos et al., 1989; Williams and Carnahan, 1989,1990; Hosea etal., 1990; Roozen and Hemminga, 1990,1991; Williams et al., 1990 Blond and Simatos, 1991; Chang and Baust, 1991b; Dereuddre et al., 1991; Hatley et al., 1991; Izzard et al., 1991; Le Meste and Huang, 1991; Roos and Karel, 1991c,f-h, 1993; Schenz et al., 1991; Simatos and Blond, 1991, 1993; Ablett et al., 1992a-c, 1993; Nesvadba, 1992a; Roos, 1992a,c,d; Hatley and Mant, 1993; Karel et al., 1993b; MacInnes, 1993). Such findings have inadvertently lent support to the practical relevance of the lower “apparent” C,’ values (obtained using freeze-concentrated 20 w% sugar solutions as models for real frozen food systems), rather than the higher “rigorous” C,’ values, to technological issues concerning the ice content of actual frozen foods (Cole et al., 1983, 1984; Slade and Levine, 1991a; Levine and Slade, 1992b).

Despite these facts, or perhaps because of them, the opinion expressed by Karel, about a consensus over C,’ values, glosses over some real problems yet to be confronted and resolved. One of these is the remaining mystery


about why, for most of the low-MW carbohydrates listed in Table XIII, the difference between values of “rigorous” C,’ (determined from measurements on highly concentrated solutions, produced, in essence, by hydrating the dry solute) and “apparent” C,’ (determined from DSC measurements made of the ice-melting endotherm after freeze-concentrating a fairly dilute solution) is quite large (with “rigorous” S “apparent” C,’ in all such cases) (Hatley and Mant, 1993), while for several others (i.e., glucose, mannose, maltose, trehalose, sorbitol), the difference is much smaller or nonexistent. Trehalose is especially unusual in this regard, in that, depending on the experimental method used to determine C,’, trehalose’s C,’ value can be the highest (Levine and Slade, 1988b) or lowest (Kawai et al., 1992) among those for a series of common sugars. Possibly, the answer to this mystery might lie in the relative kinetics of ice formation in undercooled, concentrated versus dilute aqueous solutions of small sugars (Slade and Levine, 1991a; Levine and Slade, 1992b; Roos, 1992a,d), which, evidently, are usually (Franks, 1982), but not always, quite different. But why not always? Of course, it should be kept in mind that both T,’ and C,’ are kinetic parameters measured under nonequilibrium conditions, for solutions whose behavior is under kinetic rather than thermodynamic control (Levine and Slade, 1986, 1988a,b,c; Franks, 1985, 1993a; Noel et al., 1990). As if to balance Karel’s opinion about consensus, Franks (1993b) has offered the following note of caution: “however, the endotherm profile also contains contributions from the heat of dilution and the heat capacity of fusion, which make the exact determination of unfrozen water an uncertain affair.”

Another part of the mystery is illustrated by the C,’ data for the homologous series of maltooligosaccharides listed in Table XIII, which show that, as a general trend, “apparent” C,’ increases and “rigorous” C,’ decreases with increasing solute MW, until the two values essentially merge for maltoheptaose (Levine and Slade, 1988b; Schenz et al., 1992; Ablett et al., 1993). As alluded to earlier, once MW increases to the range of all the high-MW polysaccharide and protein biopolymers (e.g., starch and gluten) characterized to date, the C,’ range moves to -70 to 75 w% solute (Slade and Levine, 1993a-c), and the difference between “rigorous” and “apparent” C,’ values essentially disappears (Slade and Levine, 1984, 1987b, 1991a; Marsh and Blanshard, 1988; Roos and Karel, 1991d; Ablett et al., 1993). Part of the explanation for this lies in the fact that the DSC data corrections (e.g., temperature dependence of latent heat of fusion of ice) alluded to above, necessary to convert “apparent” into “rigorous” C,’, are known to decrease in significance with increasing solute MW (Hatley et al., 1991; Reid and Hsu, 1992). However, this fact does not explain why one obtains the same C,’ value, whether one starts with a dilute or a highly concentrated biopolymer solution; i.e., why the relative kinetics of ice formation in dilute versus


concentrated solutions are evidently the same, without exception, for high- MW, but not for low-MW, solutes. Why do the kinetics of ice formation appear to differ the way they do in aqueous solutions of mono-, oligo-, and polysaccharides of equivalent weight concentration? The expected differences in bulk viscosity among undercooled, rubbery solutions of different MW solutes (Franks, 1982) would seem not to completely account for the C,’ results obtained. Moreover, why do values of “rigorous” C,‘ appear to decrease with increasing solute MW (e.g., from 430 w% for mono- and disaccharides to -70 to 75 w% for polysaccharides) (Roos and Karel, 1991d; Ablett et al., 1993), when one might intuitively expect the opposite relationship, as is, in fact, evidenced by “apparent” C,’ values (Slade and Levine, 1991a)? Before there can be a true consensus about C,’ values, there is obviously much work still to be done on the question of kinetics of ice formation [as well as on the possibly related question of stress relaxation in the freeze-concentrated glass at T,’ (Hatley and Mant, 1993)] in aqueous food carbohydrate solutions, as functions of solute MW and concentration, not to mention molecular conformation (for solutes of equal MW).

2. Diffusion of Water and Solute in Aqueous Food Glasses at T < Tg

Let us quote once again from the introductory notes for the 1990 Wagen- ingen conference, this time on the subject of mobility of water in glassy materials:

It has been proposed that the residual UFW in a frozen solution is kinetically controlled, and not an equilibrium property. During freezing, the solution viscosity rapidly increases until [the solution] finally becomes an amorphous solid, i.e., a glass. One definition of a glass is that the Stokes viscosity of the sample is in excess of 1014 Pa s, and it is assumed that this is also the local viscosity of the water molecules present. Translated into molecular motion terms, this is equivalent to a diffusion rate for the water molecules of a few microns per [decade] [Franks etal., 19911. It has been proposed that the reason why UFW is detected in frozen solutions is not because it becomes “bound,” but because the mobility of the water in the glass is so slow. This means that on a practical time scale, the water molecules will stop diffusing to the ice interface, with the result that ice crystallization stops [Franks, 1982,19851. The problem with this model is that there is an increasing amount of evidence [from NMR, ESR, and sorption studies] to suggest that the mobility of the UFW is much higher than has been proposed.

It should be noted, of course, that the skepticism expressed above begs the question: If the unfrozen, plasticizing water in a glass at T < T,‘ is so mobile, why can’t it continue to crystallize to ice? The fact that it does not, in real time, continue to freeze is not at issue (Slade and Levine, 1991a), and past rationalizations based on the concept of “bound” water are no


longer widely accepted (Simatos and Karel, 1988; Ahlneck and Zografi, 1990; Belton, 1990, 1991; Ablett and Lillford, 1991; Franks, 1991a; Slade and Levine, 1991a; Chinachoti, 1992a; Gidley et aL, 1993; Hegenbart, 1993). The Wageningen notes went on to say: “All these observations [of much higher water mobility in glasses] may be a result of inhomogeneity within the samples, which could arise due to either the structure being porous, or due to there being different domains with different water mobilities. However, if this is the case, then the effect of sample inhomogeneity needs to be built into the model.” This issue of potential sample heterogeneity is certainly valid, even more so for real, complex food products than for simpler model systems, and it needs to be more widely recognized and stressed (Slade and Levine, 1991a; Nelson, 1993), even though, as discussed below, it may account for some but not all such observations of higher- than-expected mobility of water or solute in glasses.

The above question about the “unfreezability” of the plasticizing water in glasses leads to another obvious question. Since UFW is evidently not mobile enough to continue to crystallize in real time at T < Tg‘, why is it that this UFW can be removed, albeit slowly, by diffusion from a glass, e.g., during the desorption (or secondary drying) stage of freeze-drying? As pointed out by Franks (1993b): “The migration of water within, and its removal from, amorphous matrices containing hydrogen-bonding groups is of particular scientific interest and technological importance.” There has been much recent study of the diffusion of water in glassy matrices, with regard to the freeze-drying of proteins, other biologicals, and pharmaceuti- cals (Franks, 1990, 1992b; Pikal and Shah, 1990; Pikal et al., 1990; Pikal, 1993). The specific question of the mobility and desorbability of UFW from glasses during freeze-drying has been explicitly addressed and investigated in detail by Pikal and co-workers, who measured rates of water removal from glassy systems. Their quantitative data led to the conclusion that “while water mobility is greatly restricted in a glassy system (relative to a system above Tg), sufficient [translational] mobility does exist to allow moderate secondary-drying rates even well below Tg” (Pikal and Shah, 1990; Pikal et al., 1990). However, Franks (1990) noted that “clearly, the geometry (texture) of the freeze-concentrate is an important factor, as is also its structural uniformity and the distribution of the residual water” [after the vacuum sublimation or primary drying stage].

The subject of diffusion in aqueous food polymer glasses and rubbers has also been extensively discussed in many previous reports by Karel and co-workers, with regard, e.g., to aspects of food dehydration technology, encapsulation and controlled release, and quality of dehydrated foods (Karel and Flink, 1983; Karel, 1985, 1986, 1989, 1990, 1991c, 1992; Karel and Langer, 1988; Simatos and Karel, 1988; Le Meste and Simatos, 1990;


Karel and Saguy, 1991; Shimada et al., 1991; Karmas et al., 1992, 1993; Labrousse et al., 1992; Levi and Karel, 1992a, 1993a,b; Karel et al., 1993a,b). Encapsulation of labile food components, such as volatile flavors and aromas, in glassy matrices produced, e.g., by freeze-drying, spray-drying, or extrusion is a well-established and well-founded technology. Its fundamental technological basis, indeed its very reason for being, is the greatly enhanced stability that results from successful entrapment of labile, otherwise highly mobile, small molecules in low-moisture food glasses at T < Tg (Karel and Flink, 1983; Karel, 1985, 1986, 1990; Saleeb and Pickup, 1985, 1989; Karel and Langer, 1988; Levine et al., 1991, 1992; Shimada et al., 1991; Boskovic et al., 1992; Bruin, 1992; Karmas et al., 1992,1993; Labrousse et al., 1992; Levi and Karel, 1992a, 1993a,b; Nelson et al., 1992; Karel et al., 1993a,b; Nelson, 1993; Nelson and Labuza, 1993; Peppas, 1993). In other words, minimized diffusion of small, labile components within and out of aqueous glassy matrices is the established functional basis of glass- encapsulation technology. The same technological basis underlies the use of freeze-drying (or other drying methods) for the stabilization of high- value biological and pharmaceutical materials, including enzymes, proteins, and drugs, in low-moisture glassy matrices (Franks, 1989, 1990, 1991a,c, 1992b, 1993a,b; Levine and Slade, 1989c, 1992a; Franks and Hatley, 1990, 1992, 1993; Pikal, 1990a,b, 1993; Pikal and Shah, 1990; Pikal et al., 1990, 1991a,b; Franks and van den Berg, 1991; Franks et al., 1991; Hatley, 1991; Hatley and Franks, 1991; Roser, 1991b; Roy et al., 1992; te Booy et al., 1992; van den Berg et al., 1992, 1993; Ramanujam et al., 1993; Roser and Colaco, 1993). But these facts just lead to another obvious question that one should ask. Since water can be removed from a glass during the desorption stage of freeze-drying, why is it that other volatile molecules, encapsulated in the same glass, are effectively retained, at least partially? Is this due, at least in part, to the fact that water has a lower MW, smaller molecular size, lower Tg, and, therefore, higher mobility than most molecules that one would attempt to glass-encapsulate? The answer to the last question, although not known with certainty (Bruin, 1992), is probably yes. In the context of such questions with, at best, uncertain answers, the following remark by Schlotter (1990) is especially interesting: “It should be noted that above the Tg, and in the melt, diffusion is well behaved compared to that in glassy materials.”

Two years after the Wageningen conference, Karel, in his summary notes for the Easter School, said the following about mobility, of water and solute, in glasses:

It has been shown by many of the participants, using many different techniques, that T8 represents an important transition point for “mobility” in foods. However, stability


of foods is far from assured below Tg Processes [e.g., chemical reactions that are not translational diffusion-limited, including free-radical reactions such as lipid oxidation (Karel, 1986; Nelson and Labuza, 1992b)l continue below this temperature, some at very respectable rates . , . , Water remains highly mobile below Tg Vapor pressure measurements, NMR spectroscopy, as well as computer modeling indicate water binding to specific sites is very transient (picosecond scale [Brady and Ha, 19911). There is no discontinuity in properties of water, itself, in polymer glasses [i.e., with regard to properties such as, or related to, translational or rotational diffusion rates of water, there appears to be no change in their temperature dependence at T, (Franks, 1993b; Pikal, 1993; Simatos and Blond, 1993), as measured, e.g., by NMR (Pikal etal., 1991a; Kalichev- sky et al., 1992b; Ablett et al., 1993; Cherian and Chinachoti, 1993) and/or sorption studies (Oksanen and Zografi, 1993; Pikal, 1993); this behavior of water is in sharp contrast to that of the same properties of the polymer segmental units in amorphous food polymer-water blends (Tanner et al., 1991; Ablett et al., 1993; Franks, 1993b; Pikal, 1993; Reid et al., 1993b; Simatos and Blond, 1993) or of other molecules [probes (Shah and Ludescher, 1993) or reporters] much larger than water, such as dyes (Wesson et al,, 1982; Huang et al., 1987; Ehlich and Sillescu, 1990; Nelson, 1993) or ESR spin probes (Le Meste and Duckworth, 1988 Le Meste et al., 1991b Roozen and Hemminga, 1990, 1991; Bruni and Leopold, 1991, 1992; Roozen et al., 1991; Hemminga et al., 1993), as discussed further below] . . . . Ordering of polymer segments, or of polymer segment- water structures, below T, is implied by the existence of low temperature endotherms [purportedly due to melting of microcrystalline regions that form during aging] in [many glassy] polysaccharides [Bizot et al., 1992; Kalichevsky et al., 1992a; Livings et al., 1992; Shogren, 1992; Appelqvist et al., 1993; Donald et al., 1993; Gidley et al., 19931.

It should be noted that the latter phenomenon, which has generated much current interest (in his summary of advances reported at ISOPOW- V, Karel referred to it alternately as the “so-called Gidley endotherm” and an “apparition” deserving further study), could be due simply to the kind of sample heterogeneity mentioned earlier. In such a case, even for a polysaccharide-water model system, multiple amorphous domains of different size, with different compositions (e.g., moisture contents) and T, values, could coexist, and at least one of these Tgs (but not the major one observed by DSC) could be below the storage (“aging”) temperature, thus allowing recrystallization to occur in that (rubbery) domain, as one would expect (Slade and Levine, 1991a). Alternatively, the endothermic event in question might be due to slow “enthalpic relaxation” of a glassy component, which is a phenomenon quite widely and generally observed after sub-T, annealing (“physical aging”) of plastics and other synthetic amorphous polymers (Petrie, 1975; Wunderlich, 1981; Roos, 1992d; Williams et al., 1993). Enthalpic relaxation, which occurs as a consequence of sufficient short-range mobility at T < Tg, has recently also been reported for low- moisture glasses of food polymers such as starch (Bizot et al., 1992; Shogren, 1992; Seow and Teo, 1993) and gluten (Lawton and Wu, 1993), as well as for low-MW carbohydrates such as sucrose (Hatley and Franks, 1991), maltose (Noel et al., 1991), glycerol and sorbitol (Chang and Baust, 1991a,c;


Franks, 1993b). Indeed, Appelqvist et al. (1993) have mentioned the possibility of enthalpic relaxation as an explanation for their observations, and even pointed out that certain aspects of their results, e.g., dependence on thermal history and timehemperature dependence of the recovery, were consistent with established criteria for enthalpic relaxation (Wunderlich, 1981). If this alternative interpretation were correct, then one would expect the endothermic transition to change in appearance and temperature location under different DSC cooling and heating protocols, and to disappear after a first heating scan that would erase the sample's thermal history (Wunderlich, 1981). Such an experimental approach, which to date has not been applied to aqueous food polymer glasses that have exhibited the low- temperature endotherm in question, would be the surest way to distinguish between a relaxation process under kinetic control and a first-order melting transition under energetic control and thus unaffected by the kinetics of its measurement (Wunderlich, 1990).

Certain other facts, relating to diffusional mobility below versus above T,, are well established. For example, in the synthetic polymers literature, it has been clearly shown that plots of diffusivity or permeability as a function of temperature, for the diffusion of gases [or of dye molecules (Ehlich and Sillescu, 1990)] in amorphous polymers, exhibit a sharp change in slope at T, [see, e.g., Fig. 5.105 in Sears and Darby (1982)], demonstrating that the translational mobility of small probe molecules in amorphous matrices, which is very low (but not zero) at T < T,, increases dramatically as free volume likewise increases at T > Tr This behavior is so diagnostic that the breakpoint in such plots can actually be used to determine a T, value that agrees precisely with the T, measured by DSC (Compan et al., 1993). Exactly the same type of behavior has recently been reported for (1) the diffusivity and permeability of gases in completely amorphous, water-plasticized films of potato or rice starch (with or without added sugars), for which Tg as a function of moisture and sugar content was measured by DSC and DMTA [see Figs. 2-13 in Arvanitoyannis et al. (1993)l; (2) the temperature-dependent rate of water vapor transmission as a function of RVP for water-plasticized films of wheat gluten [see Fig. 8 in Gontard et al. (1992c)], for which T, as a function of moisture content was measured by DSC and DMTA (Gontard et al., 1992b); and (3) the diffusion coefficient of water during gelatinization of raw rice (Lin, 1993), although, in this case [and others involving the kinetics of starch gelatinization, reviewed by Lai and Kokini (1991)], it was not recognized that such behavior is a direct consequence of a glass transition at 63.5"C during heating of partially crystalline glassy starch in excess moisture (Lee, W > Wg') (Maurice et al., 1985; Slade and Levine, 1987b, 1988~). Similarly, the same kind of extreme increase in diffusional mobility with increasing


free volume and decreasing local viscosity at T > Tg has been demonstrated by (1) ESR data for local rotational diffusion of spin probes in frozen aqueous solutions of sugars and maltooligosaccharides, when compared to the DSC-measured Tg’ values of Levine and Slade (1988b-d) for the same solutes (Roozen and Hemminga, 1990,1991; Roozen et al., 1991; Hemminga et al., 1993); (2) ESR data for rotational diffusion coefficient of a spin probe in water-plasticized samples of soybean seeds [or of caseinate proteins (Le Meste and Duckworth, 1988; Le Meste et al., 1991b)], when compared to Tg data for the same samples measured by TSC (Bruni and Leopold, 1991, 1992); (3) Brillouin scattering data, when compared to DSC-measured Tg data, for aqueous sucrose solutions (Hosea et al., 1990); (4) rate data for a diffusion-limited chemical reaction (nonenzymatic browning) in model aqueous amorphous food matrices above and below Tg, which show a clear change in slope at Tg (signifying a sharply increasing reaction rate at T > Tg), in so-called Arrhenius plots of log browning rate versus 1/T (Kar- mas et al., 1992; Karel etal., 1993b); and ( 5 ) water vapor absorption isotherm data for various glassy food polymers, which show a sharp increase in the extent of water sorption at T > Tg, once the sorbing polymer transforms from glassy solid to rubbery liquid as a consequence of dynamic plasticization by the sorbed water (van den Berg, 1981,1986,1991; Slade and Levine, 1985, 1991a; Levine and Slade, 1988a; Cairault ef al., 1989; Slade ef al., 1989; Oksanen and Zografi, 1990,1993; Franks, 1991b; Gontard et al., 1992c; Reid, 1992b; Roos, 1992b; Hancock and Zografi, 1993; van den Berg ef al., 1993; Yano, 1993).

Most recently, Franks (1993b) has summed up the current state of our understanding about mobility and diffusion in aqueous food glasses this way: “In the glass, the water [i.e., UFW] molecule motions are severely retarded [i.e., retarded enough so that this water cannot crystallize to ice], but hardly to the extent of complete inhibition . . . the available evidence suggests that the water molecules constituting the ‘residual moisture’ in a solid solution are able to diffuse rapidly, compared to the motions of the molecules (or polymer segments) making up the glassy phase. The mobility of water within glassy matrices may contribute to the observed chemical reactions that can occur at measurable rates, even below the measured (DSC) Tr” Others have recently reached quite similar conclusions (Johari, 1991; Ablett et al., 1993; Reid ef al., 1993b; Simatos and Blond, 1993).

Relevant new learning, in a recent paper from the synthetic polymers literature by Chin et ul. (1992), could be important in advancing our understanding of the diffusional mobility of water in aqueous food polymer glasses. Chin et al. used NMR to investigate the rotational diffusional motion of polymer chain segments, as a function of temperature, in a two- component, compatible amorphous blend of two polymers with individual


Tgs different by 110°C. They pointed out that such a large difference in Tgs could also apply to typical polymer-diluent blends. As usual, their compatible blend exhibited only a single Tg by DSC, which was intermediate in value between the Tgs of the two pure polymers. Chin et al. stated that this “large Tg gap indicates a significant difference in the intrinsic mobility of the individual components of the blend.” By analogy, we would expect the same to be true of a compatible amorphous blend of water ( Tg = - 135°C) and a typical food polymer (dry Tg = 200°C) or food monomer (e.g., a monosaccharide of dry Tg = 30°C). Chin et al. asked the following questions: “To what degree, if any, is this difference preserved in the dynamics of the two chains near the DSC Tg of the blend? Is there extensive cooperativity between the motions of the two types of chains in the blend or does each component retain certain aspects of the intrinsic mobility indicated by their respective Tgs as pure polymers?” They found evidence of segmental motions in the blend, beginning at temperatures significantly below the blend’s Tg, under experimental conditions in which a one-component polymer glass would only begin to show such local motions at temperatures above its Tg. They also found that these rotational diffusional motions in the blend showed a much broader distribution of relaxation times than would be shown by a single-component polymer glass under the same conditions (Chin et al., 1992).

Let us suggest an explanation, built on fairly self-evident analogies to the findings of Chin et al. (1992) [as well as earlier results of (1) Ehlich and Sillescu (1990) on tracer diffusion below and above Tg, and (2) Roland and Ngai (1991) on dynamical heterogeneity in miscible polymer blends], for why the diffusion rate of water is always greater than the diffusion rate of the solute(s) in an aqueous glass, as a means to explain why water can be slowly removed from an aqueous glass by freeze-drying at a sample temperature below the instantaneous Tg of the blend, but above that of water itself.

The Tg of a binary blend of a compatible diluent (often a fluid solvent or plasticizer at the experimental temperature) with a monomeric material (often a solid solute at the experimental temperature) is determined by the Mw of the blend (Ferry, 1980); Tg corresponds to the limiting free volume associated with the mw of the blend (Slade and Levine, 1991a). In other words, only the average free volume, not its distribution, governs the glass transition (Roland and Ngai, 1991). Thus, the Tg of the blend is higher than that of the diluent (since the material with the lower Tg is designated as the diluent) and lower than that of the solute. The operational interpretation of this rule (which depends on the introduction of a reporter molecule as a third species, at a concentration that is insufficient to alter the Tg of the original blend) is that only a diagnostically selected reporter molecule,


which has a hydrodynamic volume corresponding to that of a species with the same Zw as the a, of the blend, would experience, at the Tg of the blend, a change in slope of the temperature dependence of its relaxation times, such that diffusion below Tg would be constrained to a rate several orders-of-magnitude smaller than the diffusion rate measured at a temperature about 20°C above Tg [rather than several times smaller, as expected for Arrhenius kinetics (Levine and Slade, 1989d)l. [Such behavior can be inferred from the experimental data reported by Furuta e? al. (1984) for the diffusion coefficient of water in aqueous maltodextrin glasses and rubbers.] That is, at the Tg of the blend, the conditions of free volume and local viscosity required for cooperative motions over 100-81 dimensions are just met for the exogenous reporter molecule, but they are still more than adequate for diluent molecules. A corollary to this rule is that, if the experimental temperature were set equal to the Tg of the blend, then the diffusion rate of the diluent, if the diluent were unwarily considered as an endogenous reporter molecule, would be not only higher than the respective rates for the exogenous reporter molecule or for a solute molecule, but would also be more extremely temperature dependent [according to WLF kinetics, themselves temperature dependent! (Slade and Levine, 1988b)l than those rates. In contrast, if the monomeric solute could be monitored as an endogenous reporter molecule, its diffusion rate not only would be lower than that of the exogenous reporter molecule, but would also be much less temperature dependent (according to Arrhenius kinetics). As a result, a small temperature increment of a few degrees near the instantaneous Tg of the blend could have an order-of-magnitude effect to increase the diffusion rate of diluent molecules, with an almost negligible effect on the diffusion rate of solute molecules. Alternatively, if the experimental temperature were set equal to the Tg of the pure diluent, constrained diffusion, characteristic of the solid state, would be observed for all three molecular species. In the case of a nonentangling oligomer, the rules would be the same, but the temperature window between the Tg of the diluent and that of the solute would be wider [as seen from NMR results (Ablett et al., 1993)], such that the disparity would be greater between the diffusion rate of the diluent and that of the solute at an intermediate temperature equal to the Tg of the blend. In the case of a high-MW, entangling polymer, the Tg of the blend should be calculated from the nw of the diluent with the segmental unit of the polymer, using the Tg of the pure diluent and the molecular or segmental Tg of the polymer, rather than its higher network Tg (Levine and Slade, 1989b). We also note another potential consequence of such situations [inferred from the findings of Roland and Ngai (1991)l: for an aqueous food polymer glass, a very broad glass transition can result


from the very different dynamics of the water and solute components of the miscible blend.

D. WLF THEORY AND WLF KINETICS

With the first sentence of their classic paper, Williams, Landel, and Ferry (1955) introduced their new equation by saying “in an amorphous polymer above its glass transition temperature, a single empirical function can describe the temperature dependence of all mechanical and electrical relaxation processes.” Over 30 years later, Sperling (1986) remarked that “for a generation of [synthetic] polymer scientists and rheologists, the WLF equation has provided a mainstay both in utility and theory.” It has become increasingly recognized (Table XB) that (1) the amorphous polymers referred to by Williams et al. include many food polymer, oligomer, and/or monomer systems; (2) in the last several years, Sperling’s remark has begun to apply to food scientists and technologists; and (3) even after decades of investigations and applications by synthetic polymer scientists, there are still new and useful insights to be gleaned from the WLF equation, which are relevant to the behavior of amorphous food molecules and systems (Slade and Levine, 1993~).

1. The WLF Equation

At temperatures above Tg, plasticization (by either water or heat) affects the viscoelastic, thermomechanical, electrical, guestlhost diffusion, and gas permeability properties of completely amorphous and partially crystalline polymer systems (Levine and Slade, 1988a; Lillie and Gosline, 1990). In the rubbery range, above Tg for completely amorphous polymers or between Tg and T,, for partially crystalline polymers [in either case, typically from Tg to about Tg + 100°C for well-behaved synthetic polymers (Slade and Levine, 1988b)], the dependence of viscoelastic and other properties on temperature (i.e., the effect of increasing temperature on relative relaxation times) is successfully predicted (Cowie, 1973; Gray, 1991) by the WLF equation, an empirical equation whose form was originally derived from the free volume interpretation of the glass transition (Williams et al., 1955; Ferry, 1980). The WLF equation can be written as (Williams et al., 1955; Soesanto and Williams, 1981)

where q is viscosity or other diffusion-limited relaxation process, p is density,


and C1 and Cz are coefficients that describe the temperature dependence of the relaxation process at temperatures above the reference temperature. For an undiluted polymer, when Tg is used as the reference temperature, C1 is proportional to the inverse of the free volume of the system at Tg, while Cz is the ratio of the free volume at Tg over the expansion coefficient of the free volume. The expansion coefficient of the free volume is the constant that describes the linear dependence, on AT above Tg, of the increase in free volume due to thermal expansion above Tg (Lee, the difference between the volumes of the rubbery liquid and glassy solid states). C, and Cz are material-specific constants (i.e., not temperature-dependent) for an undiluted polymer, but, importantly, their numerical values vary with composition in the presence of a diluent (Ferry, 1980), such as water (Slade and Levine, 1988b, 1991a, 1993c; Lillie and Gosline, 1990, Nelson, 1993). C1 and C, take on values of “universal constants” [17.44 and 51.6, respectively, as extracted from experimental data on many synthetic amorphous polymers (Williams et al., 1955)] for well-behaved, diluent-free polymers (Slade and Levine, 1988b).

The WLF equation describes the kinetic nature of the glass transition and has been shown to be applicable to any glass-forming polymer, oligomer, or monomer (Ferry, 1980). The equation, with the same values of the “universal constants” (in most cases), has also been reported, especially recently (Table XB), to apply to experimental data (for viscosity, or other mechanical relaxation processes, as a function of temperature) for an ever-increasing variety of food materials and systems, including (1) molten glucose (Williams et al., 1955); (2) concentrated solutions of mixed sugars (Soesanto and Williams, 1981); (3) ice cream and frozen desserts (Slade and Levine, 1985, Levine and Slade, 1989~); (4) amorphous glucose-water mixtures (Chan er af., 1986); ( 5 ) hydrated elastin rubbers (Gosline, 1987); (6) American process cheese (Campanella et af., 1987; Peleg, 1992); (7) concentrated, thermoset, aqueous gels of whey protein (Katsuta and Kinsella, 1990); (8) undercooled melts of fructose or glucose (Ollett and Parker, 1990); (9) amorphous lactose and sucrose powders at low moisture contents (Roos and Karel, l990,1991a,e, 1992; Shimada et al., 1991; Levi and Karel, 1992a, 1993a,b; Nelson and Labuza, 1992a; Peleg, 1992); (10) freeze-concentrated solutions of fructose or glucose (Roos and Karel, 1991f); (11) supersaturated sucrose solutions (Roos and Karel, 1991g); (12) frozen solutions or low-moisture powders of maltodextrins (Lim and Reid, 1991, 1992; Reid, 1992a; Nelson and Labuza, 1992a, 1993); (13) frozen fruits and juices (Huang, 1992; Huang et af., 1993); (14) dehydrated vegetables (Karmas et af., 1992; Nelson and Labuza, 1992a; Kareletaf., 1993a); (15) gelatinized starch (Tayloretaf., 1992); and (16) bacterial spores (Sapru and Labuza, 1992a,b). Notably, Lillie and Gosline (1990) extended Gosline’s (1987) work on rubbery elastin by showing that, over a


range of water contents, the WLFcoefficients, determined from elasticmodu- lus data, had (nonuniversal) average values of C1 = 9.7 and C2 = 28.8. Simi- larly, Cocero and Kokini (1992) reported that the temperature dependence of shear viscosity data for water-plasticized glutenin above its Tg fits the WLF equation with C1 = 10.8 and C, = 40.2. In a related vein, the issue of nonuni- versa1 values of the WLF coefficients, as determined from experimental data, for various synthetic polymers and food systems has been discussed in several papers (Slade and Levine, 1988b, 1993c; Angel1 etal., 1991,1992; Buera and Karel, 1992; Huang, 1992; Nelson and Labuza, 1992a, 1993; Peleg, 1992; Hu- ang et al., 1993; Karel et al., 1993a; Reid et al., 1993b) and in notable depth, in the context of a detailed comparison of the applicability of WLF and Arr- henius kinetics to food systems, in the thesis by Nelson (1993). It is also notable that the WLF equation, with values of the coefficients dictated by experimental data, has been shown to correctly describe the kinetics of chemical degradation, of amorphous freeze-dried drug-excipient mixtures at low moisture contents, at storage temperatures 5"-30"C above Tg (Roy et al., 1992).

The WLF equation defines mobility in terms of the non-Arrhenius temperature dependence of the rate of any diffusion-limited relaxation process occurring at a temperature T, compared to the rate of the relaxation at the reference temperature Tg, expressed in terms of how log q depends on AT, where AT = T - Tg. The equation is valid in the temperature range of the rubbery or undercooled liquid state, where it is typically used to describe time-/temperature-dependent behavior of polymers (Roberts and White, 1973). The equation is based on the assumptions that polymer free volume increases linearly with increasing temperature above Tg, and that segmental or mobile-unit viscosity, in turn, decreases rapidly with increasing free volume (as illustrated implicitly in Fig. 1) (Ferry, 1980). Thus, the greater the AT, the faster a system is able to move (due to increased free volume and decreased mobile-unit viscosity); so the greater is the mobility, and the shorter is the relaxation time (Slade and Levine, 1988b; Lillie and Gosline, 1990).

In essence, the WLF equation and resulting master curve of log (q/qg) versus T - Tg (Williams etal., 1955; Soesanto and Williams, 1981) represent a mobility transformation, described in terms of a time-temperature superposition (Slade and Levine, 1988b). Such WLF plots typically show a five orders-of-magnitude change in q [or in rates of other relaxation processes, such as diffusion (Ehlich and Sillescu, 1990; Karathanos et al., 1991; Karel and Saguy, 1991; Roozen and Hemminga, 1991; Levi and Karel, 1992a, 1993b) or crystallization (Roos and Karel, 1990, 1991a,e, 1992; Shimada et al., 1991; Levi and Karel, 1993b)l over a 20°C interval immediately above Tg (Franks, 1985; Lillie and Gosline, 1990), which is characteristic of WLF


behavior in the rubbery fluid range (Slade and Levine, 1988b; Levine and Slade, 1989c; Noel et al., 1990; Franks et af., 1991; Roos and Karel, 1991g). For example, as demonstrated by Soesanto and Williams (1981), the effects of temperature and concentration on mobility of fluids above Tg can be combined to create a single master curve, which represents the WLF equation. The viscosity data shown in Fig. 11 were obtained for highly concentrated (> 90 w%) aqueous mixtures of sucrose and fructose. These results showed a five orders-of-magnitude change in q of concentrated sugar solutions, over a 20°C interval near Tg, a finding in excellent accord with the behavior predicted by the quantitative form of the WLF equation, with its “universal constants” of C1 = 17.44 and C, = 51.6. These results constituted the first experimental demonstration that concentrated sugar solutions obey the WLF equation quantitatively as well as do synthetic high polymers. As

10-6 ‘ O - k

b97.63 wtX (x.0.7069 P o 96.36 13.89 l t wtX X ( x (1 = 0 . i 7 3 6 1 0.6078)

093.01 wtX (xmO.4379) 0 91.87 wtX (~90.3982)

I 12.5% fructose - 87.5% sucrose

T -Tg (OK)

FIG. 11. Temperature dependence of viscosity for aqueous solutions of a 12.5 : 87.5 (w/w) fructose : sucrose blend, illustrating the fit of the data to the curve of the WLF equation. Reprinted with permission from Soesanto, T., and Williams, M. C. 1981. Volumetric interpretation of viscosity for concentrated and dilute sugar solutions. J. Phys. Chem. 85, 3338-3341. Copyright 1981 American Chemical Society.


mentioned earlier, it had been shown decades before that viscosity data for a completely amorphous, diluent-free glucose melt fit the WLF equation with the same coefficients, and thus that glucose also behaves like a typical well-behaved synthetic polymer (Williams et al., 1955; Matsuoka et al., 1985). In a related vein, it has been confirmed experimentally that undercooled melts of diluent-free fructose or glucose exhibit a five orders-of- magnitude change in q over a 20°C interval above T, (Noel et al., 1990; Ollett and Parker, 1990), as predicted by the WLF equation. Similarly, as illustrated in Fig. 12, Roos and Karel (1990, 1991a,e, 1992; Shimada et al., 1991) showed that the rates of crystallization of sucrose or lactose in completely amorphous powders at low moisture contents change by five orders of magnitude over a 20°C interval above T,, in accord with the WLF, but not the Arrhenius, equation. Likewise, Levi and Karel (1992a, 1993a,b) showed that (1) the rate of diffusion of propanol in low-moisture sucrose rubbers increases by five orders of magnitude, and (2) the relaxation times for propanol release, crystallization, and collapse in anhydrous sugar matrices decrease by five orders of magnitude, from Tg to Tg + 20”C, and that their experimental data fit a curve of the WLF equation with its “universal” values for C1 and C2. It is also interesting to note that Le Meste and Huang (1991), in a TMA study of frozen sucrose solutions, found that for a 20 w% sucrose sample, softening and concomitant flow during heating were observed to be gradual in the temperature range -32°C (i.e., T,’) to -1O”C, but above -10°C (i.e., 22°C above T,), sample volume decreased much more rapidly as a consequence of compression, as expected, based on WLF kinetics (Slade and Levine, 1988b).

Description of the time-/temperature-dependent behavior of food systems by the WLF equation requires selection of the appropriate reference T, for any particular glass-forming material [of any given MW and extent of plasticization (Williams et al., 1955; Soesanto and Williams, 1981; Chan et al., 1986)], be it T, for a low-moisture system (with W < W,’) or Tgf for a frozen system (with W > W g f ) (Levine and Slade, 1986, 1988b, 1989a; Roos and Karel, 1990, 1991a,d-g, 1992; Biliaderis, 1991b; Franks et al., 1991; Shimada et al., 1991; Slade and Levine, 1991a; Karmas et al., 1992, 1993; Levi and Karel, 1992a, 1993a,b). For a typical, diluent-free polymer, T, of the undercooled liquid is defined in terms of an iso-free volume state of limiting free volume (Ferry, 1980), and also, very approximately, as an iso-viscosity state somewhere in the range lo9 to lOI4 Pa s (Soesanto and Williams, 1981; Franks, 1982, 1989, 1990; Angell, 1988; Levine and Slade, 1988a; Simatos and Karel, 1988; Hofer et al., 1989; Koide et al., 1990; Ollett and Parker, 1990; Angell et al., 1991, 1992; Slade and Levine, 1991a; Kalichevsky ef al., 1992b; Karel, 1992). This iso-viscosity state refers to local, not macroscopic, viscosity (Ferry, 1980), a fact that constitutes a


A

Experimental Sucrose 23% rH - Makower and Dye (1956) 24% rH ----- Arrhenius

0 20 40 I

T-Tg ("C)

0' 0 10 20 30 40 !

T-Tg ("C)

FIG. 12. WLF- versus Anhenius-type temperature dependence of the crystallization time [expressed as log 8, (s)], as a function of AT = T - Tg, for (A) completely amorphous sucrose and lactose [adapted from Roos and Karel (1991a) with permission] and (B) completely amorphous lactose [adapted from Roos and Karel(1992) with permission] at low moisture contents.


critical conceptual distinction (Slade and Levine, 1988b), as explained later with regard to Fig. 14D.

a. Viscoelastic Behavior of Aqueous Food Polymer Systems. In the context of the utility of the WLF equation, the underlying basis of the principle of time-temperature superpositioning is the equivalence between time (or frequency) and temperature as they affect molecular relaxation processes that influence the viscoelastic behavior (i.e., the dual characteristics of viscous liquids and elastic solids) of polymeric materials and glass- forming small molecules (Ferry, 1980; Sichina, 1988; Lillie and Gosline, 1990; Mita, 1990). This principle is illustrated in Fig. 13 (Levine and Slade, 1989b), which shows a master curve of modulus as a function of temperature or frequency for a typical partially crystalline synthetic high polymer (Graessley, 1984). Figure 13 has been widely used to describe the viscoelastic behavior of such materials, as exemplified by molten starch during extrusion (Lai and Kokini, 1991) or by a kinetically metastable gelatin gel in an undercooled liquid state, in the context of WLF theory (Borchard et al., 1980; Slade and Levine, 1987a). At T > Tg, gelatin gels manifest a characteristic rubberlike elasticity (Tomka et al., 1975), due to the existence of a network of entangled, randomly coiled chains (Yannas, 1972). With increasing temperature, a gelatin gel traverses the five regions of viscoelastic

I I

Elastic

Rubbery Flow

Or

I I I I I I I I I I I I I I I I

Tg Tm

TEMPERATURE or log FREQUENCY

FIG. 13. Master curve of modulus as a function of temperature or frequency, illustrating the five regions of viscoelastic behavior characteristic of synthetic partially crystalline polymers. From Levine and Slade (1989b) with permission of Van Nostrand ReinholdlAVI.


behavior characteristic of synthetic partially crystalline polymers (Yannas, 1972), as illustrated in Fig. 13: (1) at T < T,, vitrified glass; (2) at T = T,, glass transition to leathery region, typically manifested as a three orders- of-magnitude decrease in modulus (Sperling, 1986); (3,4) at T, < T < T,, rubbery plateau to rubbery flow; and ( 5 ) at T > T,, viscous liquid flow (Le Meste and Simatos, 1990; Le Meste et al., 1992). It is interesting to note that at T, < T < T,, a gelatin gel is freely permeable to diffusion of dispersed dyes and molecules as large as hemoglobin (Slade et al., 1989); only at T < Tg is such dye diffusion greatly inhibited (Wesson et al., 1982). In light of this fact, it is not at all surprising that the self-diffusion coefficient of water in a gelatin gel at T > Tg has been shown to be close to that of water in bulk liquid water (Mel'nichenko et al., 1993). Thus, the local viscosity inside the pores of a gelatin gel network, as measured by the diffusion of water [which shows Arrhenius kinetics (Mel'nichenko et al., 1993) at an experimental temperature >150°C above water's molecular T,], is essentially the same as that of bulk water.

The typical decrease in modulus (usually by =2-4 orders of magnitude) at the glass transition, illustrated in Fig. 13, has also been shown in similar plots of mechanical modulus data (as functions of either temperature or moisture content) recently reported for (1) completely amorphous, water- plasticized matrices of amylopectin and amylopectin-sugar mixtures (Kali- chevsky et al., 1992a,b), gluten and mixtures of gluten with sugars, lipids, emulsifiers, or salts (Kalichevsky et al., 1992c,d), other proteins such as ovalbumin, gelatin, casein, and sodium caseinate (Kalichevsky et al., 1993a,b), mixtures of casein or caseinate with sugars (Kalichevsky et al., 1993b), and 1:l mixtures of amylopectin, casein, and gluten, as well as other multicomponent biopolymer mixtures (Kalichevsky and Blanshard, 1992a,c); (2) extruded starch and starch-glucose mixtures (Noel et al., 1990; Ollett et al., 1991; Smith, 1990, 1992a); (3) dry amorphous dextran and pullulan (Scandola et al., 1991); (4) water-plasticized films of amorphous carboxymethyl cellulose (Yano, 1993); (5) elastic networks of thermoset (via covalent crosslinking) gluten (Attenburrow et al., 1990, 1992; Davies et al., 1991); and (6) hydrated elastin rubbers (Gosline, 1987; Lillie and Gosline, 1990, 1993).

Le Meste et al. (1991a) used TMA to study flour-based food products and found that their viscoelastic properties, as functions of temperature and moisture content, correspond to those characteristic of the glassy state, the glass transition region, or the rubbery plateau. More recently, they published the first thermomechanical study of the glass transition in a complex food product, fresh white pan bread, and reported (Le Meste et al., 1992) that the rigidity modulus drops by three orders of magnitude at T, [i.e., T,' = -12"C, as measured by TMA, for bread with moisture


content > Wgf = 25-30 w%, in comparison to Tgf = -5”C, as measured by DSC (Levine and Slade, 1989b; Slade and Levine, 1991a)l. Those results on the glass transition in white bread were confirmed in follow-up work using DMA (Aynie et al., 1992a,b), which also illustrated the plasticizing effect of water on Young’s modulus. Colleagues of ours at Nabisco (Amem- iya and Menjivar, 1992) recently reported the first measurements of the glass transition in commercial cookies and crackers. Their plots [see Fig. 5 in Slade and Levine (1993b)l of Young’s modulus as a function of moisture content (fat-free dry solids basis), as measured by an Instron three-point- bend test done at room temperature, showed a room-temperature glass transition occurring at a moisture content coinciding with the point at which the modulus shows a sharp drop from its high value at the glassy plateau (Sperling, 1986), as illustrated in Fig. 13. Their results showed that the moisture content at the glass transition decreased from -10 w% for a lean cracker to < =5 w% for a high-sugar, wire-cut cookie (Slade and Levine, 1993b). Attenburrow and Davies (1993) reported similar results for Young’s modulus (measured at room temperature) versus moisture content for another baked product, ice cream wafers, which showed a glass transition at -10 w% moisture. Also noteworthy is a report by Sauvageot and Blond (1991) of the effect of water plasticization on crispness of breakfast cereals, which included plots of sensory crispness versus moisture content [as well as RVP, which they called A,,,] for two commercial grain-based products (Kellogg’s Corn Flakes and Rice Krispies). These plots bear a striking resemblance to the portion of the curve in the glass-leathery-rubbery plateau region of Fig. 13, as well as to previously published plots of modulus versus moisture content for amorphous starch or starch-sugar mixtures (Noel et al., 1990; Ollett et al., 1991; Smith, 1990, 1992a; Kalichevsky et al., 1992a,b). They appear to indicate an abrupt transition, occurring over a narrow range of moisture contents around lo%, from a glassy product of high crispness to a leathery/rubbery one of low crispness. Nevertheless, Sauvageot and Blond surprisingly did not point out the possible connection between the behavior they observed and an underlying glass transition in the water-plasticized products. In contrast, Kaletunc and Breslauer (1992) have reported a direct correlation between increasing sensory crispness and increasing Tg (measured by DSC) for low-moisture corn-flour extrudates, while Labuza and Nelson (1992), Lillford et al. (1992), Parker and Smith (1992), and Tolstoguzov (1992) have all described how the glass transition governs the texture (i.e., hard, brittle, crispy, and/or crunchy below Tg; soft, leathery, rubbery, and/or soggy above Tg) of various low- moisture, cereal-based materials and products, including extruded products. In a related vein, in an older report by Andrieu and Stamatopoulos (1986) on durum wheat pasta drying, a plot of Young’s modulus (measured at


20°C) versus moisture content shows obvious evidence of a glass transition occurring at -16 w% moisture, which those authors discussed in terms of a change in the pasta’s mechanical behavior from plastic to elastic type. It is also interesting to note the results reported by Ollett et al. (1993a,b) on the powder-compaction behavior measured at room temperature for hydrated native potato starch. These results appear to indicate a clearly observable change from glasslike to rubberlike response, at a moisture content (20 w%) coinciding with that at which Young’s modulus, measured at room temperature, exhibits the characteristic drop indicative of the glass transition of completely amorphous starch- or amylopectin-water mixtures (Ollett et al., 1991; Kalichevsky et al., 1992a).

From modulus results for starch-, gluten-, and other protein- and polysaccharide-based systems (Attenburrow et al., 1990; Lillie and Gosline, 1990, 1993; Noel et al., 1990; Smith, 1990, 1992a; Davies et al., 1991; Ollett et al., 1991; Scandola et al., 1991; Gontard et al., 1992b; Kalichevsky and Blansh- ard, 1992a,b; Kalichevsky et al., 1992a-d, 1993a,b; Attenburrow and Davies, 1993; Yano, 1993), it can be seen that the decrement in modulus from below Tg to above Tg for an aqueous glass of a compatible water-solute blend is not identical to the decrement in modulus for a neat glass of the same solute. No explanation for this observation has previously been given; we offer a possible one here.

When the distribution of water is spatially homogenous in the amorphous regions of a compatible aqueous blend, and the water content is less than W,’, then all of the water functions as plasticizing water, so that the Tg of the blend is intermediate between the lower Tg of pure water and the higher (molecular or segmental) Tg of the solute (Slade and Levine, 1991a). (A binary blend is considered here, to avoid complications due to the order of addition of components to the blend.) Addition of water as a plasticizer has a predictable effect on the kinetics of the primary mechanical relaxations of the blend, compared to those of the neat solute. That is, the characteristic decrement in modulus, as the temperature passes from below to above T, during an experimental temperature sweep, would be observed to occur at lower and lower temperatures for aqueous blends with increasing water content than for the neat solute (Cocero and Kokini, 1991; Scandola et al., 1991; Aynie et al., 1992a; de Graaf et al., 1992; Gontard et al., 1992b; Kalichevsky and Blanshard, 1992a; Kalichevsky et al., 1992c, 1993a; Lillie and Gosline, 1993). In this context, the term “plasticization” refers unambiguously to “depression of the Tg of a compatible blend, relative to the Tg of the undiluted solute,” i.e., a kinetic process (Sears and Darby, 1982). Alternatively, at a constant experimental temperature, an inflection point in the decreasing modulus of the sample would be observed as the sample moisture content is increased (Andrieu and Stamatopoulos, 1986; Gosline,


1987; Attenburrow et al., 1990; Lillie and Gosline, 1990, 1993; Smith, 1990, 1992a,c; Davies et af., 1991; Ollett et al., 1991; Amemiya and Menjivar, 1992; Kalichevsky and Blanshard, 1992b; Kalichevsky et al., 1992a-d, 1993b; Attenburrow and Davies, 1993; Yano, 1993). In this case, the term “plasticization” could refer ambiguously either to the kinetic effect of altering the temperature difference between the experimental temperature and the depressed T, of the blend relative to the undiluted solute (Slade and Levine, 1991a) or to a mechanical softening effect. The ambiguity arises from the fact that the mechanical softening may be solely an indirect reflection of the kinetic effect, or it may represent an additional direct mechanical effect independent of the kinetic effect of the plasticizer to depress the T, of the solute-water blend below the T, of the undiluted solute, and thereby change the relationship between the experimental temperature and the instantaneous Tg of the sample. Indeed, the kinetic effect of a plasticizer to depress T, (the a or primary relaxation of the backbone of a polymer or molecular relaxation of a monomer) may be accompanied by either a plasticizing or antiplasticizing effect on T, (the p or secondary, sub-T,, relaxations) (Levine and Slade, 1988a), as is suggested by the tan 6 curves reported for carboxymethyl cellulose (Yano, 1993), dextran, amylose, and pullulan (Scandola et al., 1991). As a result, even in the glassy state, water may serve to decrease or increase the modulus of the glassy aqueous blend, relative to the modulus of the undiluted glass.

A thought experiment, to help clarify the distinction between the kinetic effect of depression of T, by water as a plasticizer and the mechanical effect on the absolute modulus observed at a particular temperature for a particular composition of an aqueous blend, is as follows. Part I of the experiment would demonstrate the ambiguous dual aspects of plasticization by water. At a constant experimental temperature, typically room temperature for convenience, the modulus of an undiluted solute in the glassy state would be compared to the moduli of a series of aqueous blends, prepared so that the distribution of water would be spatially homogeneous throughout the sample (thought experiments allow such a luxurious criterion). Alternative conventions would define a value of W, corresponding to T, = room temperature as the water content at the onset or at the inflection of the decreasing modulus profile.

Part I1 of the experiment would demonstrate unambiguously the independent mechanical aspect of plasticization by water. The absolute modulus of each sample in the series would be measured at an experimental temperature below the respective T, of each sample and at a series of equivalent temperature increments above the respective T, of each sample. The maximum water content in any sample should be less than W,’, so that the modulus of ice would not be convoluted with that of a glassy matrix at any experimental temperature in a practical time frame. The absolute values


of the glassy modulus below the respective T, and of the rubbery modulus at each equivalent temperature increment above T, of each sample in the series, with water content in the range 0% to W, < W,', would reflect the mechanical aspect of water as a plasticizer or antiplasticizer of the solute. At each equivalent temperature, relative to the respective values of T,, the moduli of the aqueous blends might be the same, or larger, or smaller than that of the undiluted parent sample, independent of the kinetic effect of water as a plasticizer that actually determines those respective values of T8' Although the typical decrement in modulus between the glass below T, and the deformable rubber above T, of the undiluted solute would be =3 orders of magnitude (Sperling, 1986), the corresponding decrement for the aqueous blends could be the same, or larger, or smaller than three orders of magnitude. For starch or gluten blends with water, the decrement is often smaller than 3 log units (Andrieu and Stamatopoulos, 1986, Noel et af., 1990; Smith, 1990, 1992a,c; Cocero and Kokini, 1991; Davies et af., 1991; Ollett et af., 1991; Attenburrow et af., 1992; Aynie et af., 1992a; Gontard et af., 1992b; Kalichevsky and Blanshard, 1992a,b; Kalichevsky et af., 1992a-d, 1993a; Attenburrow and Davies, 1993). This might be due to the role of network crosslinking (Ollett et af., 1991; Kalichevsky et af., 1993a) and resultant network T, (Levine and Slade, 1989b) in these cases, e.g., for thermoset gluten, the amorphous glutenin network formed by covalent crosslinking via intermolecular disulfide bonding (Schofield et af., 1984; Slade et af., 1989; Kalichevsky et af., 1993a), and for gelatinized/ retrograded starch, the partially crystalline amylopectin network formed by a combination of interchain entanglements in the amorphous regions and hydrogen-bonded junction zones comprising the microcrystalline regions (Slade, 1984; Miles et af., 1985b; Slade and Levine, 1987b; Smith, 1990, 1992a; Ollett el af., 1991).

Part I11 of the experiment would demonstrate the unambiguous kinetic aspect of plasticization by water. The interpolated temperature location at the onset or at the inflection of the decrement in modulus, determined from the quantized temperature sweep of part I1 of the experiment (or from a judicious continuous temperature sweep experiment, recognizing that use of a constant heating rate results in a variable, increasing value of t h as the temperature increases above the respective T, of the blend), for each sample of the series, would define the kinetic effect of water as a plasticizer to depress the T, of the undiluted solute to the respective T, values of the aqueous blends.

2. Comparison of WLF and Arrhenius Behavior

The WLF equation is not intended for use much below T, (ie., in the glassy solid state) or in the very low viscosity liquid state [q < 10 Pa s


(Soesanto and Williams, 19Sl)], typically 100°C or more above T,, where Arrhenius kinetics apply (Roberts and White, 1973; Johari, 1976; Nielsen, 1977; Ferry, 1980; Robertson, 1985; Levine and Slade, 1988a; Ehlich and Sillescu, 1990; Chang and Baust, 1991c; Franks et al., 1991; Gray, 1991; Karathanos et al., 1991; Noel et al., 1991; Scandola et al., 1991). For partially crystalline polymers, the breadth of the temperature range of the rubbery domain of WLF behavior corresponds to the temperature interval between T, and T, (Wunderlich, 1976; Ferry, 1980), as illustrated in Fig. 1. Cheng (1989) noted that the size of this temperature interval between T, and T, may be as much as several hundred degrees for synthetic high polymers. We reported an analysis of the variation of the size of this temperature interval with the T,/Tg ratio of representational synthetic polymers and glass-forming, low-MW carbohydrates (Slade and Levine, 1988b). The study compared the WLF behavior of kinetically metastable carbohydrate-water systems to the corresponding knowledge base for synthetic polymers. Ac- cording to the conventional description, a typical well-behaved synthetic high polymer (e.g., a representational elastomer) would manifest its T, around 200 K in the completely amorphous state, and its T, around 300 Kin the completely crystalline state (Wunderlich, 1980), so that the ratio of T,,, for the pure crystalline material to Tg for the completely amorphous material is =1.5 (or T,/T, - 0.67) (Brydson, 1972; Wunderlich, 1990). Such a polymer would also have a local viscosity of =lo” Pa s and a free volume fraction of -2.5% at Tg (Ferry, 1980). For this typical well-behaved polymer, WLF kinetics are considered to be operative in a temperature range about from Tg to Tg +lOO°C (Williams et al., 1955). It can be seen that this operational definition is related to the typical T,/Tg ratio of 1.5, since in such a case, the difference in temperature between T, and T, would be -100°C. Figure 14A (Slade and Levine, 1988b) illustrates this conventional description of the relaxation behavior of a typical well-behaved polymer [e.g., the classic model, Hevea rubber (Brydson, 1972; Ferry, 1980; Sperling, 1986), or polyvinyl acetate (Johnson et al., 1980; Matsuoka et al., 1985)], which would obey the standard form of the WLF equation with coefficients C1 = 17.44, and C, = 51.6. In this plot of log aT versus AT [where aT = WLF shift factor, expressed as relative relaxation times (Ferry, 1980)], relaxation time progresses from WLF behavior very near Tg to Arrhenius behavior at -100°C above Tg. Within this temperature range, where technological process control would be expected, relaxation times for WLF behavior near Tg would change by a factor of 10 for every 3°C change in temperature. In contrast, for Arrhenius behavior with familiar Qlo = 2 kinetics above T,, a factor-of-10 change in relaxation time would require a 33°C change in temperature (Slade and Levine, 1988b).

A second class of amorphous polymers was described as typical but not well behaved (Slade and Levine, 1988b), in the sense that they are readily

190 LOUISE SLADE

WLF 2 -20

- T, 0 A T = Texp

4ND HARRY LEVINE

0

0

a+ CD -

-20 200 - T,

A T = Texp 0

FIG. 14. WLF plots of the time-temperature scaling parameter (WLF shift factor, expressed as relative relaxation times), aT, as a function of the temperature differential above the reference state, Tg, with the limiting regions of low and high AT defined by the WLF and Arrhenius kinetic equations, respectively. The curves of the WLF equation (with coefficients C1 and Cz as noted) illustrate the temperature dependence of the relaxation time behavior for hypothetical polymers with T,,,ITp ratios of (A) 1.5 (C, = 17.44, C, = 51.6); (B) 2.0 (C, = 26.4, Cz = 154.8); (C) 1.25 (C, = 10.8, C, = 23.3); (D) 2.0, 1.5, and 1.25. From Slade and Levine (1988b) with permission.

crystallizable (Brydson, 1972; Haward, 1973; Ferry, 1980; Wunderlich, 1980) and thus have a decreased glass-forming tendency (Murthy et al., 1993). Highly symmetrical polymers such as poly(viny1idene chloride) and poly(vinyl cyclohexane), which manifest crystalline melting enthalpies of -170 J/ g, fit this class (Wunderlich, 1980). For such polymers, the T,/Tg ratio is frequently B 1.5, so the temperature range between Tg and T, is S 100°C. Different WLF coefficients would be required to describe their relaxation profile, as illustrated by the plot in Fig. 14B drawn for a hypothetical case where C1 = 26.4 and C, = 154.8. For a representational case of Tg =200 K (with -qg > lo** Pa s and/or free volume fraction < 2.5%) and Tm/Tg = 2 (Tg/Tm = OS), T, would be -400 K. Thus, there would be a region of -200°C in which relaxation times would change from WLF behavior near Tg (in this schematic case, by a factor of 10 for every 6°C) to Arrhenius behavior near T,,, (by a factor of 10 for every 33°C) (Slade and Levine, 1988b). A notable example of a material with T,,,/Tg = 2 is water (Soesanto


and Williams, 1981; Wunderlich, 1990). It is also interesting to note a recent report by Lomellini (1992) that both polystyrene and polycarbonate exhibit WLF behavior, with regard to their melt viscoelastic data, over a surprisingly large temperature range from Tg to about Tg + 185°C. Like synthetic polymers in this class, low-MW carbohydrates with high T,/Tg ratios are reported to have high values of crystalline melting enthalpy and to crystallize readily (Roos, 1992a).

A third class of polymers, often characterized by highly unsymmetrical structures, was described as atypical and poorly behaved (Slade and Levine, 1988b), in that Tg is near T, (Brydson, 1972; Wunderlich, 1980). Murthy et af. (1993) have reported a correlation between decreasing T,/Tg and increasing glass-forming tendency. For such polymers, with T,/Tg -e 1.5 (e.g., ~ 1 . 2 5 , or Tg/T, =0.8), a quantitatively different form of the WLF equation would be required to describe their relaxation profile. For a representational polymer in this class, as illustrated in Fig. 14C, using a hypothetical case where C1 = 10.8 and C, = 23.3, Tg ~ 2 0 0 K (with qg Q 10l2 Pa s and/ or free volume fraction %- 2.5%) and T, ~ 2 5 0 K. Thus, the temperature range in which WLF kinetics would be operative is much smaller than usual. Relaxation times would change from WLF behavior near Tg (in this case, by a factor of 10 for every 2°C) to Arrhenius behavior above T, (by a factor of 10 for every 33°C) over a region of only ~50°C (Slade and Levine, 1988b). The synthetic polymer cited as the classic example of this behavior, which was attributed to anomalously large free volume and high mobility at Tg, is bisphenol polycarbonate, with Tm/Tg =1.18 (Brydson, 1972; Ehlich and Sillescu, 1990). Another important example of this case, with Tm/Tg ~ 1 . 2 5 , is poly(dimethy1 siloxane), which was singled out (Ferry, 1980) for its anomalously high mobility, manifested as anomalously low local effective viscosity (Slade and Levine, 1988b). This category of behavior was also reported (Levine and Slade, 1988a, 1989a) to be exemplified by food materials such as gelatin and native starch (due to nonuniform distribution of moisture in amorphous and crystalline regions of these high polymers at low moisture), as well as the sugars fructose and galactose (due to an anomalously large requirement for rotational free volume of these anhydrous monosaccharides and concomitantly large availability of translational free volume at a given extent of rotational mobility) (Slade and Levine, 1988b).

The three types of behavior exemplified in Figs. 14A-C, in which the T,/Tg ratio is either the typical value of 1.5, or much greater, or much less, were compared, in order to examine how the respective relaxation profiles change in the temperature interval between T, and Tg for representational, diluent-free polymers with a common value of Tg (Slade and Levine, 1988b). As illustrated in Fig. 14D, this analysis revealed the critical significance of


the T,/T, ratio for any given glass-forming polymer. For sufficiently similar values of T,, either different values of the ratio, T,/Tg, or the difference, T, - T,, for different polymers (e.g., food carbohydrates) can be used as an inherent normalizing parameter to compare relative mobilities at Tg and at T s- Tr The greater the value of T,/Tg or T, - T,, the greater is the inherent rigidity of the material at its T,, and the more constrained is its mobility at any given AT above its Tr The case of polymers with similar values of T, is illustrated by a comparison of the local translational mobility of Hevea rubber and polyisobutylene (Ferry, 1980) at an experimental temperature 100 K above their respective values of T, (T, = 200 K for Hevea rubber, 202 K for polyisobutylene). The local viscosity in Hevea rubber was more than two orders of magnitude lower than that in polyisobutylene (Ferry, 1980), and either the smaller ratio of T,/Tg (1.49 for Hevea rubber, 1.57 for polyisobutylene) or the smaller difference of T, - Tg (98 K for Hevea rubber, 115 K for polyisobutylene) served to predict the greater mobility in Hevea rubber (Slade and Levine, 1988b). However, for dissimilar values of Tg, the ratio T,/T,, is still a predictive normalizing parameter for relative mobilities, but the difference, T, - Tg, may no longer be predictive. This case of polymers with dissimilar values of T, is illustrated (Ferry, 1980) by a comparison of the local translational mobility of polystyrene and polyisobutylene at an experimental temperature 100 K above their respective values of T, ( Tg = 373 K for polystyrene). The local viscosity in polystyrene was even lower than that in Hevea rubber, which could be successfully predicted by the even smaller value of T,/Tg (1.37 for polystyrene), but not by the value, even greater than that of polyisobutylene, of T, -T, (138 K for polystyrene) (Slade and Levine, 1988b). In Fig. 14D, the behavior of log aT versus A T is compared for different values of T,/ T, (i.e., -2,1.5, and 1-25), to determine how mobility varies in the kinetically constrained regions of this mobility transformation map. At T P T,, the overall free volume for different polymers may be similar (Ferry, 1980), yet individual free volume requirements for equivalent mobility may differ significantly, as reflected in the T,/Tg ratio. Anisotropy in either rotational mobility [which depends primarily on free volume (Ferry, 1980)] or translational mobility [which depends primarily on local viscosity, as well as free volume (Ferry, 1980)] may be the key determinant of a particular polymer’s relaxation behavior.

In the context of discussing the comparison between WLF and Arrhenius behavior, it is interesting to note the connection between the analysis illustrated in Fig. 14 and Angell’s framework for classifying the behavior of glass-forming liquids (including synthetic amorphous polymers and organic and inorganic liquids) as either “strong” or “fragile” (Angell, 1988; Murthy, 1989; Angell et al., 1991, 1992). The potential for useful insights


resulting from application of Angell’s classification scheme for glass- forming liquids to the behavior of glass-forming food materials (e.g., sugars and polyols), with regard to stability against translational diffusion-limited relaxation processes such as crystallization, was first mentioned by Noel et al. (1990, 1991) and subsequently discussed in depth by Angell, himself (Angell et al., 1992). According to Angell’s terminology, “strong” liquids (exemplified by the network glass former, SO2) are those with (1) a three- dimensional structural network of bonds; (2) resistance of the short-range order of the glassy structure to thermal disruption on heating to T > T,; (3) Arrhenius or almost-Arrhenius relaxation behavior (e.g., linear dependence of log q on 1/T) at T > T, (as well as T < T,); (4) approximately constant q, = 10” Pa s at T, [i.e., the glass transition is an iso-viscous phenomenon only for “strong” liquids (Angell et al., 1991)l; and (5) generally, a small or even undetectable increase in C, at T,. In contrast, “fragile” liquids [exemplified by molecular (e.g., terphenyl) and ionic (e.g., aqueous solutions of LiCl) glass formers] are those with (1) no three-dimensional structural network of bonds; (2) loss of the short-range order of the glassy structure, accompanying the thermal transition at T,; (3) an increasingly exaggerated departure from Arrhenius relaxation behavior at T > T,; (4) variable q, at T, (q, decreases from =lo” Pa s for “strong” liquids to as low as =lo9 Pa s for the most “fragile” liquids) [i.e., the glass transition is not an iso-viscous phenomenon for “fragile” liquids (Angell et al., 1991)l; and ( 5 ) generally, a large increase in C, (as much as a doubling of the glassy-state C, in some cases) at Tg Angell has illustrated the contrasting relaxation behavior of “strong” and “fragile” glass-forming liquids by means of so-called “T,-scaled Arrhenius plots” of log q or log(segmenta1 relaxation time, T ) versus T,/T, referred to as “fragility plots” or “‘strong- to-fragile’ patterns” (Angell, 1988; Angell et al., 1991, 1992). Others have used similar “T,-scaled Arrhenius plots” of log q versus log(T - T,) (Mur- thy, 1989) or log 7 versus T,/T (Ngai and Roland, 1993) or log aT versus T,/T (Roland and Ngai, 1992) as alternative representations of Angell’s “fragility plots.” Roland and Ngai (1992) prefer to call theirs “cooperativity plots,” since the synthetic amorphous polymers they have discussed do not undergo any modification of structure at the glass transition.

There is a revealing analogy between Angell’s “fragility plot” and the WLF plot in Fig. 14D, both of which are normalized to T, as the reference state. The upper curve in Fig. 14D (taken from Fig. 14B) corresponds to that of a “strong” liquid, while the lower curve in Fig. 14D (taken from Fig. 14C) corresponds to that of a “fragile” liquid. Moreover, just as the curves in Fig. 14D intersect the left-hand ordinate (at T = Tg) at different values of log aT, so too do Angell’s “fragility curves’’ not all pass through the same point (in terms of log q or log 7) at T,/T = 1 (Angell et al., 1991,


1992), thus demonstrating that the glass transition is rigorously neither an iso-viscous (Murthy, 1989) nor an iso-relaxation time phenomenon for all glass-forming liquids. Let’s explore the basis of the analogy, in an effort to glean further insights. Angell has described the “strong”-to-“fragile” pattern of increasing deviations from Arrhenius-type behavior (with its mild dependence of q on temperature), using the Vogel-Tammann-Ful- cher (VTF) form of the dependence of q on temperature,

with D as the only variable (Angell et al., 1992). “Strong” liquids are those with large values of D, while “fragile” liquids, with small values of D, are those that show a much greater dependence of q on temperature at T > Tr Angell noted that (1) the VTF and WLF equations have identical mathematical forms and so are interconvertible; (2) for the entire spectrum of glass- forming liquids, including synthetic polymers, existence of a range of D values in the VTF equation implies a range of values for the WLF constant, C2, rather than one universal value; (3) the values of D in the VTF equation and C1 in the WLF equation are numerically proportional; and (4) the WLF constant, C2, is numerically identical to Tg - To, where Tg is the WLF reference temperature and To is a constant from the VTF equation (Angell et al., 1991, 1992). Therefore, “strong” liquids, with large values of D, are those with large values of C1 (i.e., larger than the “universal” value of 17.44, e.g., the polymer represented in Fig. 14B, with C1 = 26.4), while “fragile” liquids, with small values of D, are those with small values of C1 (i.e., smaller than the “universal” value of 17.44, e.g., the polymer represented in Fig. 14C, with C1 = 10.8). For glass-forming liquids that are also crystallizable, such as the representational polymers described with regard to Fig. 14, it follows that Angell’s classification of “strong” corresponds to those liquids with values of Tm/Tg ratios B 1.5 (see, e.g., the top curve in Fig. 14D), his description of “fragile” corresponds to those liquids with Tm/Tg 4 1.5 (see, e.g., the bottom curve in Fig. 14D), while liquids with intermediate “strong-fragile” behavior should be those with Tm/Tg = 1.5 (see, e.g., the middle curve in Fig. 14D).

The synthetic polymers mentioned by Angell as being known to be most “strong” [polyisobutylene (PIB)] and most “fragile” [bisphenol polycarbonate (BPPC), poly(dimethy1 siloxane) (PDMS)] in their relaxation behavior (Angell et al., 1991, 1992) are the same ones that were noted for their extraordinarily high or low Tm/Tg values (PIB, 1.57; BPPC, 1.18; PDMS, 1.25), in the context of Fig. 14. Hevea rubber and polyvinyl acetate are synthetic polymers with Tm/Tg = 1.50 (Ferry, 1980), which would be expected to show intermediate “strong-fragile” behavior. With regard to


glass-forming food materials (e.g., low-MW carbohydrates), glycerol has an extraordinarily high value of Tm/Tg = 1.62 (Slade and Levine, 1991a), and Angell noted that it has a large D, commensurate with “strong” behavior of this hydrogen-bonded liquid (Angell et al., 1992). Angell also remarked that the behavior of glassy water near its Tg (-137°C) appears to be that of a very “strong” liquid (Angell et al., 1992). This is particularly noteworthy, since water has the highest known Tm/Tg value of 2.0 (Wunderlich, 1990), as mentioned earlier with regard to Fig. 14B. In contrast to such “strong” hydrogen-bonded liquids as glycerol and water, with T,,,/Tg values 9 1.5, other hydrogen-bonded liquids, including sorbitol and sugars such as glucose, fructose, maltose, and sucrose, all of which have Tm/Tg values 5 1.43 (Slade and Levine, 1991a) and values of qg at Tg significantly less than lo1* Pa s (Angell el al., 1992), have been classified as somewhat “fragile” (Noel et al., 1990; Angell et al., 1992). As mentioned earlier and noted elsewhere in various contexts by many workers (Williams et al., 1955; Soesanto and Williams, 1981; Slade and Levine, 1988b; Noel et al., 1990; Ollett and Parker, 1990 Roos and Karel, 1991g), one manifestation of the relaxation behavior ( WLF-type) of liquids as “fragile” as glucose, fructose, and sucrose (in the form of either anhydrous melts or highly concentrated aqueous solutions) is a drop in viscosity of five orders of magnitude over the 20°C interval immediately above Tg. Consequently, as Tg is approached from above in this temperature region, the rate of translational diffusion, and thus of diffusion-limited processes such as crystallization that cause time-dependent changes in the quality of foods, will be greatly reduced (Slade and Levine, 1988b; Noel et al., 1990). Noel et al. (1990) have stated that “for ‘fragile’ liquids, small differences in storage temperature in the vicinity of Tg will have dramatic effects on stability and will determine whether a material is stable for ten days or for hundreds of days. . . . The classification of ‘strong’ and ‘fragile’ behavior on the basis of the dependence of liquid viscosity on temperature can give useful insights into stability with respect to crystallization.” Further insight results from the recognition that, for glass-forming liquids that can crystallize, “fragility” evidently increases with decreasing value of Tm/Tg, whereas “strength” appears to increase with increasing value of Tm/Tg.

The relaxation phenomena manifested at Tg represent thermomechanical or electrical properties controlled by the local small-molecule or segmental, rather than macroscopic, environment of a polymer, because the molecular glass transition is a cooperative transition (Roberts and White, 1973; Wund- erlich, 1981; Angell, 1988; Cheng, 1989) resulting from concerted constraints on both translational and rotational mobility. On cooling a viscous fluid of relatively symmetrical mobile units with relatively isotropicmobility, translational motions would be expected to be “locked in” at a higher temperature


before rotational motions, because of the slower structural relaxations associated with larger-scale translational diffusion (Slade and Levine, 1988b). In this case, cooperative constraints of local viscosity and free volume on translational diffusion determine the temperature at which the glass transition is manifested, as a dramatic increase in relaxation times compared to the experimental time frame. However, in the case of motional anisotropy, molecular asymmetry has a much greater effect on rotational than translational diffusion, so rotational motions could be “locked in” before translational motions as the temperature is lowered (Slade and Levine, 1988b).

As revealed by Fig. 14D, a very small T,/Tg ratio (i.e., close to 1.0) is accounted for by an anomalously large free volume requirement for rotational diffusion (Brydson, 1972). When the free volume requirement is so large, a glass transition (i.e., vitrification of the viscous liquid) on cooling can actually occur even when the local viscosity of the system is relatively low (Slade and Levine, 1988b). Thus, instead of the typical “firmness” for a glass Pa s), such a glass (e.g., of bisphenol polycarbonate, or anhydrous fructose or galactose, or any other “fragile” liquid) may manifest a qg 4 10l2 Pa s (Soesanto and Williams, 1981; Angell, 1988; Levine and Slade, 1988a; Slade and Levine 1988a; Angell et al., 1992). In such a glass, the time constant for translational diffusion may be anomalously small, indicative of high translational mobility. In contrast, in the glass of a typical well-behaved polymer, the time constant for translational diffusion would be greater than that for rotational diffusion, so an increase in local viscosity would be concomitant with a decrease in free volume (Slade and Levine, 1988b). The above analysis pointed out the critical significance of anomalously low values of T,/Tg (for the dry solute) close to 1.0 on the mobility, resultant relaxation behavior, and consequent technological process control for nonequilibrium food polymer systems (in the presence of water) in their supraglassy fluid state well above Tg (Slade and Levine, 1988b), in terms of the WLF kinetics of various translational diffusion-limited, mechanical/ structural relaxation processes, such as gelatinization, annealing, and recrystallization of starch (Slade and Levine, 1988c; Biliaderis, 1991b).

Interestingly, Ollett and Parker (1990) have reported a measured value of q = 4.5 Pa s [reported as lo4.’ Pa s by Noel et al. (1990)l for a fructose melt at 100°C. This temperature was selected for discussion, because it represents the higher of the two values of Tg measured by DSC for undiluted, amorphous fructose as 11” and 100°C (Slade and Levine, 1988b) and as 13” and 100°C (Finegold et al., 1989), not 30” and 100°C as stated by Ollett and Parker (1990) and Noel et al. (1990), who took the low value of viscosity at 100°C to mean that the DSC transition at 100°C could not be a glass transition. The dismissal of the higher-temperature DSC transition as a glass transition appeared to be supported by the fact that similar


values of 10” and 13°C for Tg of fructose were obtained by analysis of the temperature dependence of their q data according to the VTF and WLF equations, respectively. However, Ollett and Parker measured shear viscosity (i.e., related to translational mobility), and we hypothesized that the constraint at 100°C relates to rotational restraint to reorientation, in a way in which fructose and galactose have been suggested to be anomalous (Slade and Levine, 1988b). It would be expected, when cooperative relaxations are overwhelmingly predominated by translational constraints, so that a measurement of shear viscosity or translational diffusion can be used to characterize the plasticizing effect of temperature, time, or diluent on mobility (Slade and Levine, 1991a), that the WLF equation, which describes the influence of both translational and rotational constraints on relaxations, would be equivalent to the VTF equation that describes the temperature dependence of viscosity (Ferry, 1980; Angell el al., 1992). In fact, as mentioned earlier, the VTF parameters, D and To, can then be redefined directly in terms of the WLF coefficients, C1 and C,. In contrast, when relaxations are predominated by rotational constraints or depend on both rotational (primarily limited by free volume) and translational (primarily limited by local viscosity) mobility, then analysis of measurements of shear viscosity by the VTF equation, as a means to estimate the value of Tg, would not be expected to match values of Tg obtained from analysis of measurements of C, (by DSC) or dielectric loss according to the WLF equation (Ferry, 1980). The change in C, at the glass transition reflects the change in the expansion coefficient of free volume from that of the glass to that of the undercooled liquid (Wunderlich, 1981, 1990), such that an anomalous requirement of free volume for cooperative, asymmetrical rota- tions could result in the observation by DSC of an anomalously elevated value of Tg, at a temperature where the measured value of local viscosity [e.g., shear viscosity of sufficiently small molecules, to avoid the complica- tion that would be caused by the presence of entanglement networks (Ferry, 1980)] would be anomalously depressed. Similarly, observation of dielectric relaxations, which reflect the combined extent of cooperative rotational constraints and the effects of local viscosity on reorientation of dipoles (Sears and Darby, 1982), would not necessarily provide the same estimation of Tg as a measurement of shear viscosity alone.

Angell has recently suggested an alternative explanation (Angell et al., 1992) for the high-temperature glasslike transition exhibited by a melt of &-fructose, which he too has observed at 100°C by DSC (personal communication, 1993). This explanation is based on his principle of “superstructuring” effects, which involve dynamic phenomena, e.g., slow (much longer than the relaxation time associated with shear viscosity) structural relaxations, that can occur in glass-forming liquids at T S Tg and then


become “locked in” before vitrification of the liquid at Tg, as exemplified by the lambda transition in liquid sulfur (Angell et al., 1987). As a consequence of such “superstructuring” effects, “the properties of a subsequently formed glass can be influenced by heat treatments that are administered at temperatures far above the glass transformation range . . . the structures trapped by quenching through the ‘slow’ superstructuring region are stable against relaxation to the original lower-Tg structures, even when [the glass is] heated well above Tg where thermal history effects are usually removed” (Angell et al., 1987). In other words, such a “superstructuring” effect on the properties of the glass would not be susceptible to the normal effects of annealing near Tg (Angell et af., 1992). The critical consequence of “superstructuring” is that diffusion-limited reactions in low-moisture glass- forming food materials could be arrested at temperatures well above the glass transition. Such diffusional arrest would be “an important deciding factor in the shelf life of both dried and [frozen] foods” (Angell et af., 1992). Angell believes that the high-temperature glasslike transition exhibited by anhydrous melts of fructose (at 100°C) and galactose (at 1lOOC) (Slade and Levine, 1988b) represents such a “superstructuring” effect, i.e., a slow a + /3 conformational relaxation (personal communication, 1993). Obvi- ously, this intriguing phenomenon in fructose and galactose, which is not observed in other monosaccharides such as glucose and mannose (Slade and Levine, 1988b), is worthy of further study that might help us better understand the dramatic, and technologically important, differences between fructose and glucose as moisture-management agents with respect to microbiological activities (e.g., mold-spore germination) in foods (Slade and Levine, 1985, 1988a,b, 1991a).

The temperature location of Tg is determined by free volume, which is proportional to z,, of a material blend (Ferry, 1980). Variations in z,,, of a material of a single chemical type with an intrinsic distribution of MWs or of a materials blend with a compatible diluent, result in variations in the number of free chain ends and small mobile units, which in turn give rise to variations in free volume. The local viscosity of a series of materials or materials blends, evaluated while maintaining constant values of %, and temperature, is proportional to weight-average MW (%,) (Ferry, 1980). Local viscosity is inversely related to translational diffusion and sufficiently asymmetric rotational diffusion, and equivalent to bulk viscosity for nonentangling materials, but dramatically lower than bulk viscosity when entanglement occurs (Ferry, 1980; Slade and Levine, 1988b). The onset of glassy behavior has been related to constraints in free volume for high-MW materials and to constraints in local viscosity for low-MW materials or highly plasticized blends with low-MW diluents (Ferry, 1980).


The location of a particular value of Tg on the state diagram in Fig. 4 for a diluent-plasticized material requires identification of the explanations for both the temperature location and weight-composition location of Tr For a given composition, the temperature location of Tg for both the diluent- - free material and the diluent-plasticized blend depends on free volume and M , (Ferry, 1980). For a given temperature, explanation of the composition location of Tg relates to local viscosity and z,, as previously demonstrated (Slade and Levine, 1988b) by the relationships between Tg’, inverse M, (Fig. 6), and inverse %, for a diverse series of low-MW sugars and polyhydric alcohols and, more clearly, for a homologous series of nonentangling glucose oligomers (Fig. 7). This explanation leads to the expectations, for comparison of materials, that anomalies in the temperature location of Tg arise from anomalies in the free volume requirement for mobility, while anomalies in the composition location of Tg arise from anomalies in the local viscosity requirement for mobility. In turn, these expectations provide intuition for comparison of the effects of materials on particular relaxation processes that become limited primarily by constraints in either rotational or translational mobility (Slade and Levine, 1988b). Examples of relaxation processes that manifest primary rotational constraints include (1) absorption of microwave energy, and (2) creation of critical nuclei during the prerequisite nucleation step of crystallization in the absence of seeding. Examples of relaxation processes that become translationally limited include (1) mositure migration/sorption, which is measured as apparent RVP but often called A,, and (2) growth of crystals during the subsequent propagation step of crystallization (Slade and Levine, 1991a).

Experimental conditions that are located above the reference glass curve result in free volumes greater than those in the glassy state. Typically, free volumes that are sufficient for a relaxations (cooperative translational motions and cooperative, anisotropic rotational motions) are much more than sufficient for /3 relaxations (noncooperative, isotropic rotational motions). This typical behavior is reflected in (1) WLF kinetics of a relaxations above Tg versus Arrhenius kinetics of /3 relaxations below or above Tg, and (2) a dramatic extension of time scales for a! relaxations below Tg, resulting from a discontinuity in the temperature dependence of time scales at Tg versus continuation of local isotropic rotational motions below Tg, without a discontinuity in time scale (Ferry, 1980). Moreover, free volumes that are sufficient for a relaxations are typically coincident with local viscosities that are sufficient for a relaxations, and therefore, also sufficient for /3 relaxations. Accordingly, both the temperature and composition locations of Tg would typically coincide for mobile units of the same molecular volume. However, if free volumes that are sufficient for a relaxations are coincident with local viscosities that are not sufficient for Q relaxations,


then it is possible that even /3 relaxations would become constrained and probable that small extents of anisotropy could become more important. Thus, it is this departure from the normal coincident behavior that gives interesting clues for comparison of materials that show anomalies in either the temperature location or composition location of points on a reference glass curve, and of Tg'-Wg' in particular (Slade and Levine, 1991a).

3. Interesting Analogy between the Michaelis-Menten and WLF Equations

As mentioned earlier, the coefficients of the WLF equation, C1 and C,, have previously been compared to constants that appear in other empirical, descriptive equations that relate viscosity to temperature, such as the VTF equation (e.g., Gray, 1991; Angel1 et al., 1992). The WLF constants have also previously been related to free volume in the glassy state and the expansion coefficient of free volume above Tg (Ferry, 1980). However, despite the practical importance of the WLF equation (Sperling, 1986), some simple, intuitive implications of the WLF coefficients'were previously overlooked in the 38 years since the classic WLF paper appeared. Recogni- tion that the WLF equation has the form of a rectangular hyperbola has revealed such useful, intuitive implications about the WLF coefficients (Slade and Levine, 1993c), in much the same way that recognition of the Michaelis-Menten (MM) equation as a section from a rectangluar hyperbola has made it one of the most useful tools available to the biochemist to study enzyme reaction kinetics and other ligand-binding phenomena (Dixon and Webb, 1964). Perhaps the practical utility of the WLF equation will become even more conceptually appealing to food scientists and other biochemists by consideration of the analogy between the MM and WLF equations, as illustrated by the side-by-side comparison in Fig. 15 (Slade and Levine, 1993~).

A preliminary consideration of the elegance of the generic rectangular hyperbola shown in Fig. 15 will serve to introduce the descriptive analogy. Hyperbolas are distinguished from parabolas by the fact that hyperbolas have asymptotes, and rectangular hyperbolas are further distinguished by the fact that the angle between the asymptotes is 90" (Schwartz, 1960; Strang, 1991). The special case of the rectangular hyperbola of practical interest, familiarly exemplified by Boyle's gas law (I' * V = constant), is represented by the equation X * Y = constant, whose asymptotes coincide with the Cartesian coordinate axes (Schwartz, 1960). When experimental data are observed to have the empirical form of a section from a rectangular hyperbola with the Cartesian coordinate axes as asymptotes, but the origin ( x = 0, y = 0) of the experimental data is not


WLF

Rectangular hyperbola X * Y = constant Translated axes x = A2 + X A1 * X

A 2 + x y = Ai . Constant = AI*A2 Y = -

A2' y = A1 / 2 when x = A2

A2

X ,............ ............. c,

... ' 31

Ci * A T

C2 + AT Equation l o g 5 =-

.--. c2s log aT - C1/2 when A T = C2 .............................................

( WLF region above Ta , Asrhenius region below Tg

undiluted a 4 m * polww with increasing

dilution

V,.S v= - Michaelis-Menten

- Kin

Equation K,+S

v = V,,/2 whenS = K,

1 1 : of- : o s

FIG. 15. Comparative diagrams illustrating the analogy between Michaelis-Menten and WLF kinetics. From Slade and Levine (1993~) with permission. See text for explanation of symbols.

coincident with the origin ( X = 0, Y = 0) of the coordinate axes, the coordinates of the experimental values can be translated, so that they relate to the origin of the coordinate axes (Schwartz, 1960). If Al is the distance along the Y axis between the origin of the experimental data and the origin of the coordinate axes that represent the asymptotes of the experimental data, and A2 is the distance along the X axis between the origin of the experimental data and the origin of the asymptotic coordinate axes, then the equation describing the rectangular hyperbolic form of the translated experimental data may be written,

X * Y = constant

and rearranged as

(where X = A2 + x, Y = Al -y , and constant = A l * A2), (4)


For the experimentalist, the great practical utility of these variations of the equation for the rectangular hyperbola derives from the fact that a simple visual inspection of a graphical display of experimental data that conform to them yields both an initial estimate for the values of the coefficients Al andA2 and immediate intuition about their physical interpretation. The visual inspection is facilitated by the generic attributes of the rectangular hyperbola, as illustrated in Fig. 15: (1) the experimental data x,y represent a section from the rectangular hyperbola defined by X * Y = Al * A2, with the coordinate axes X, Y as asymptotes; (2) the origin of the experimental axes is translated from the origin of the coordinate axes, such that X = A2 + x and Y = Al - y; (3) in the limit as x approaches 00, the slope approaches 0, as y approaches Al, the horizontal asymptote; (4) in the limit as y approaches -00, the slope approaches 1, as x approaches -A2, the vertical asymptote, which is outside of the experimental region of immediate interest, i.e. at values of x < 0; (5) for any value of 0 > x > 0, in the limit as x approaches 0, the slope approaches its maximum value, which is equivalent to the ratio A1/A2; (6) the intermediate value of 00 > x > 0, at which y attains a value of A1/2 (half of the asymptotic value of y), is equivalent toA2; and (7) the inherent property of scaling allows the selection of any alternative experimental origin (x’ = 0’ > 0, y’ = 0’ > 0), so that X = A2’ + x ’ , Y = Al’ - y’, A2’ > A2, Al’ < Al, and A, * A2 = Al’ * A2’.

It is this translated form of the equation for a rectangular hyperbola, whose asymptotes coincide with the Cartesian coordinate axes, that is most familiar to food scientists and other chemists, because measurements that relate the concentration of bound ligand to the free ligand concentration for any type of ligand-binding process that involves a simple dissociation of ligand from independent and equivalent sites, such as Langmuir adsorption (Dixon and Webb, 1964), would be described by this equation. In particular, in the absence of allostery, in nearly all cases where the initial rate of an enzymatic reaction is plotted against the free substrate concentration, the data are observed to fall on a section from a rectangular hyperbola (Dixon and Webb, 1964) described by an equation of this form, the well-known MM equation,

as shown in Fig. 15, where y is the initial reaction velocity, v; x is the substrate concentration, S; Al is the asymptotic rate, V,,, attained as the ultimate initial reaction velocity at infinite substrate concentration, when all binding sites are occupied by substrate (Le., at saturation); and A2 is


the Michaelis constant, K,, defined as the substrate concentration that results in a half-maximal reaction rate. As seen in Fig. 15, in analogy with the generic attributes of the rectangular hyperbola, V,,, can be estimated from the horizontal asymptote, and K,,, can be estimated from the value of the substrate concentration that results in an intial velocity of V,,,,/2 or from the initial slope (V,,,,.JK,) of the experimental curve. The classic theory of enzyme kinetics proposed by Michaelis and Menten in 1913 is based on the suggestion that a dissociation constant K , = K,,, of an enzyme- substrate complex accounts for the rectangular hyperbolic form of the kinetic data (Dixon and Webb, 1964). Even though the proof for any particular theoretical interpretation cannot be inferred from the fact that experimental data exhibit the empirical form of a rectangular hyperbola, the MM equation (and its linearized or logarithmic variations) has maintained its preeminence with physical biochemists (Wyman and Gill, 1990), and every introductory course in food science extends its popularity.

Analogously, the WLF equation ( 2 ) also exhibits the form of a section from a rectangular hyperbola, as shown in Fig. 15, where y is the log of the relaxation shift factor expressed as relative relaxation rates, log aT; x is the differential of the experimental temperature above Tg as the reference temperature, AT; Al is the asymptotic value of log aT, C1, corresponding in number to the ultimate orders-of-magnitude change in relaxation rates from the limit of WLF temperature dependence of kinetics (just above the sub-T, Arrhenius region), through the region where Arrhenius behavior is again achieved (typically above T,,,), and finally to the limiting region of elevated temperatures, such that kinetics are no longer temperature dependent; and A2 is Cz, the value of AT at which log aT attains a value of C1/ 2, half of the ultimate orders-of-magnitude change in relaxation rates above Tg, and also the difference in temperature of the reference temperature, T,, above the asymptotic temperature (equivalent to To, the VTF parameter). Ferry (1980) stated the “rule of thumb that To is usually 4 0 ° C below the glass transition temperature,” but a visual inspection of Fig. 15 shows that Cz = 50 only pertains to certain well-behaved systems, typically with T,/T, = 1.5. One particularly useful aspect of Fig. 15 can be appreciated by the analogy of the scaling property, shown for the generic rectangular hyperbola, to the description by Ferry (1980) of the consequence of increasing dilution on the values of C1 and C2. Just as the value of Al decreases to At’ and the value of A2 increases to A*’, when the experimental origin is translated from (0,O) to (O’,O’), a depression of the T, of the undiluted polymer with increasing dilution results in an effective translation along the AT axis, so that with increasing dilution, “C2 increases while C1 decreases” (Ferry, 1980). As mentioned earlier, the numerical value of C1 is “uni- versally” (Sperling, 1986) 17.44, meaning that there would be a


17+ orders-of-magnitude change in relaxation rates between the sub-T, and super-T, Arrhenius regions for a model “universal” polymer. The “universal” value of C2 is 51.6, which would correspond to an 8+ orders- of-magnitude increase in relaxation rates in the temperature interval between Tg and 50°C above Tr Near Tg, the limiting slope corresponds to C1/C2, so that an order-of-magnitude increase in relaxation rates is achieved in the 3°C interval above Tg for the “universal” polymer. For a typical well- behaved polymer with Tm/Tg = 1.5, the slope of the rectangular hyperbola achieves a value near 0.03 (such that Arrhenius Qlo = 2 kinetics are observed) at a A T = 100°C above Tg, coincident with the value of T, = 300 K for the classic model polymer with Tg = 200K. Thus, this intuition about the physical basis of the WLF constants, revealed by the analogy between MM and WLF kinetics illustrated in Fig. 15, has recently been related to previous empirical observations about implications of the value of the T,/ Tg ratio on mobility (Slade and Levine, 1988b), described earlier in the context of Fig. 14. In essence, the T,/Tg ratio serves as a normalizing factor for intrinsic rigidity, i.e. as a normalizing factor for inherent constraints on relaxation within the WLF rubbery range (Slade and Levine, 1993~).

4. Contrast between WLF and Arrhenius Kinetics

WLF kinetics differ from Arrhenius kinetics in several important respects (Slade and Levine, 1988b; Levine and Slade, 1989c, 1990), and we have emphasized the contrast, because we believe that the qualitative differences are as influential as the quantitative differences in their impact on both experimental approach and technological significance (Levine and Slade, 1989d, 1992b; Slade and Levine, 1991a). A comparison between WLF and Arrhenius kinetics begins with recognition that the temperature dependence of microscopic relaxation parameters (including self-diffusion coefficient, viscosity, rotational and translational relaxation rates or times, and macroscopic processes that rely on them, such as crystallization) changes monotonically from a steep dependence of log relaxation rate on temperature just above Tg to a shallow dependence above T,, i.e., over a material- specific temperature range from Tg to far above Tg (Franks, 1982; Angell, 1988; Slade and Levine, 1988b; Ehlich and Sillescu, 1990; Pika1 and Shah, 1990; Franks et al., 1991; Karathanos et al., 1991; Karel and Saguy, 1991; Karmas et al., 1992; Levi and Karel, 1992a). This realization reveals two underlying diagnostic characteristics that distinguish WLF from Arrhenius kinetics (Levine and Slade, 1989d).

First, the coefficient of the temperature dependence (so-called “activation energy”) is defined as a constant in the expression for Arrhenius kinetics, and a plot of log relaxation rate versus 1/T is a straight line (Le


Meste and Simatos, 1990; Chang and Baust, 1991c; Noel et al., 1991; Scan- dola et al., 1991). But the coefficient itself is temperature dependent in the WLF expression; a plot of log relaxation rate versus 1/T is characteristically curvilinear in the material-specific, rubbery temperature range above Tg (up to T,) (Campanella et al., 1987; Karel, 1989; Ehlich and Sillescu, 1990; Koide et al., 1990; Pika1 and Shah, 1990; Karel and Saguy, 1991; Nelson and Labuza, 1992a; Nelson, 1993), approaching linearity only below Tg or above T, (Levine and Slade, 1989~). Examples of food materialdsystems exhibiting such curvilinear plots were recently reported by (1) Ollett and Parker (1990), log q versus 1/T for undercooled melts of fructose and glucose; (2) Noel etal. (1991), log q versus 1/Tfor low-moisture maltose-water mixtures; (3) Lillie and Gosline (1990), log aT versus T for hydrated elastin; (4) Karmas et al. (1992; Buera and Karel, 1992; Karel et al., 1993b), log rate of nonenzymatic browning versus 1/T for amorphous model systems; (5) Levi and Karel(1992a), log diffusion factor versus 1/Tfor diffusion of volatiles in amorphous sucrose-water matrices; and (6) Nelson (1993), log rate of ascorbic acid oxidation versus 1/T for amorphous maltodextrin- water matrices. The absolute value of the derivative of log relaxation parameter versus 1/T increases as T approaches Tg from above, and decreases abruptly to an approximately constant value as Tfalls below Tg, or decreases gently to an approximately constant value as T is elevated above T, and far above Tg, where the constant value corresponds to the Arrhenius coefficient (activation energy) that characterizes a given system and relaxation process (Levine and Slade, 1989d). The shape of the derivative profile and the temperature range over which the derivative varies are material-specific properties: typically, a range of >lOO°C for materials with T,/Tg > 1.5, or a range of <lOO°C for materials with T,/Tg < 1.5 (Slade and Levine, 1988b). Clearly, it is the fact that the derivative varies in a material-specific and temperature-dependent fashion, rather than the particular magnitude of the derivative (Simatos et al., 1989; Le Meste and Simatos, 1990; Simatos and Blond, 1991,1993), that constitutes the salient feature of WLF kinetics (Levine and Slade, 1989d). It is also important to note, as has been stressed by various workers (Huang, 1992; Huang et al., 1993; Peleg, 1992; Nelson and Labuza, 1992a; Nelson, 19?3), that an apparent absence of curvature, at T > Tg, in a plot of log reaction rate versus 1/T should not be assumed to indicate that the kinetics in question cannot be described by the WLF equation, and so Arrhenius kinetics must apply. Rather, such a linear plot should be taken to indicate that the temperature range over which the experimental reaction rate data were examined was probably too narrow to reveal the curvature expected of WLF-type behavior, illustrated in Fig. 16.

It was previously noted (Slade and Levine, 1991a) that such curvilinear plots of log rate versus 1/T (still referred to as so-called “Arrhenius plots”


A- I / Arrhenius 1uq Y _ _ _ _ _ _ _ _ _ _ - ------ ------

I

AT = T -Tg (“C)

FIG. 16. Variation of the rate of a diffusion-limited relaxation process against AT = T - Tg, as defined by the WLF equation with its “universal” numerical constants of C, = 17.44 and Cz = 51.6 (solid line), or by the Arrhenius equation (Qlo = 2 form) (dashed line). Adapted from Levine and Slade (1989~) by permission of Gordon and Breach Science Publishers, Inc. and Franks et al. (1991) with permission.

by their authors) have also been reported (e.g., LeBlanc et al., 1988) from studies of storage stability (e.g., against different enzymatic activities) of various frozen food systems stored at different freezer temperatures (Duden and Scholz, 1982; LeBlanc et al., 1988; Lee et al., 1988; Baardseth and Naesset, 1989), all above the relevant Tg‘ reference state of the foods in question (Levine and Slade, 1989d). Results of the cited studies demonstrated that enzymatic reaction rates and corresponding rates of quality loss manifest a WLF-type, rather than Arrhenius-type, temperature dependence on AT of the freezer temperature above Tg‘ (Levine and Slade, 1989d; Slade and Levine, 1991a). In a related vein, Lim and Reid (1991, 1992) and Kerr et al. (1993; Reid et al., 1992) have reported an excellent correlation, predicted from WLF theory and “cryostabilization technology” concepts (Levine and Slade, 1988c; Reid, 1990), between increasing rates of diffusion-limited deterioration reactions (e.g., model reaction systems involving enzyme hydrolysis, protein aggregation, or nonenzymatic oxidation of ascorbic acid) and increasing AT above Tg’. This correlation was demonstrated by results of their experimental studies of reaction kinetics in frozen aqueous model systems [formulated using a series of sugars and commercial maltodextrins with a range of previously measured Tg‘ values (Levine and Slade, 1986, 1988b)l stored at a range of different subzero freezer temperatures. Their results also confirmed the expectation (Levine and Slade, 1988c; Reid, 1990) [based on earlier experimental results for a frozen enzyme-substrate model system (Levine and Slade, 1986) and more


recent results for an amorphous enzyme-substrate model system at low moisture (Leopold et af., 1992)] of an essentially zero rate of reaction for each storage situation in which the freezer temperature was below Tg’ of the particular sugar-maltodextrin-based model system (Lim and Reid, 1991,1992; Reid et af., 1992; Kerr et af., 1993). Most recently, Nelson (1993; Nelson and Labuza, 1993) has analyzed reaction kinetics data from the literature for a wide variety of rubbery food systems, and has reported results that can be seen to confirm the applicability of WLF-type kinetics and the inappropriateness of Arrhenius kinetics.

The second diagnostic characteristic that distinguishes WLF from Arr- henius kinetics is that there is no explicit reference temperature in the expression for Arrhenius kinetics (Levine and Slade, 1989d), because, in fact, the implicit reference temperature is taken generically to be 0 K (Slade and Levine, 1991a; Nelson and Labuza, 1992a; Franks, 1993a), regardless of the distinctive thermomechanical properties of a system, and even though Arrhenius kinetics are applicable only below Tg and above T, (Ferry, 1980; Franks, 1982; Levine and Slade, 1988a). In contrast, the WLF equation benefits from an explicit material-specific reference temperature, which is Tg of a component or compatible blend (Ferry, 1980). Therefore, it is critical to note that when the rate or time scale of a relaxation process can be shown to depend on a material-specific reference Tg (e.g., Anonymous, 1993b), Arrhenius kinetics are not applicable to describe mobility transformations (time-temperature-moisture superposition) for that process in the rubbery range from Tg to T,, regardless of whether the average slope of a log k versus 1/T curve can be empirically fitted by a Qlo = n rule and regardless of the particular magnitude of n (Levine and Slade, 1989d; Peleg, 1992). Even today, some food engineers seem to be especially prone to overlook or disregard this fundamental distinction between Arrhenius and WLF kinetics; while they acknowledge the benefits to be gained by treating kinetics data using a material-relevant reference temperature, they still insist on incorrectly characterizing their results in terms of Arrhenius kinetics (Nunes et af., 1993; Lin, 1993). Similarly, in various recent studies of the kinetics of starch gelatinization (reviewed by Lai and Kokini, 1991) and protein denaturation (reviewed by Ma and Harwalkar, 1991), so-called “Arrhenius plots” have been presented, each of which exhibits a sharp break at the material-specific reference temperature, i.e., the gelatinization temperature [actually, at T > Tg (Slade and Levine, 1987b)l for starch or the denaturation temperature [actually, at T > Tg (Slade et af., 1989; Sochava et af., 1991; Sochava and Smirnova, 1992; Tolstoguzov, 1991,1992)] for protein.

In the temperature-composition domain sufficiently above Tg, where equilibrium and steady-state thermodynamics apply, the coefficient of the


temperature dependence of log relaxation rate on 1/Tis defined by Arrhen- ius kinetics to be a constant and is observed to approximate a relatively small constant value over a typical experimental range of about 20°C (Levine and Slade, 1989d). In the increasingly nonequilibrium domain of temperature- composition approaching Tg from above, the coefficient of the temperature dependence of log relaxation rate on 1/T is not a constant and increases ever more rapidly over a range of 20°C (Slade and Levine, 1988b). Typically, Arrhenius rates for aqueous systems above T,,, might increase fourfold over a temperature range of 20°C (Levine and Slade, 1989d), while WLF rates near Tg would increase by four or five orders of magnitude (Franks, 1985; Slade and Levine, 1988b; Levine and Slade, 1989~). As an example illustrating the significance of the difference between WLF and Arrhenius kinetics, Chan et af. (1986) noted that the dielectric relaxation behavior of amorphous glucose plasticized by water is “remarkably similar” to that of synthetic amorphous polymers in glassy and rubbery states. They showed that the rates of this mechanical relaxation process, which depends on rotational rather than translational mobility, follow the WLF equation for water- plasticized glucose mixtures in their rubbery state above Tg, but follow the Arrhenius equation for glucose-water glasses below Tg (Chan et af., 1986). Also noteworthy is Angell’s (1983) pertinent observation that the temperature dependence of the transport and relaxation properties of undercooled liquid water is strikingly non-Arrhenius in the temperature range from T, to the homogeneous nucleation temperature at -40°C. This non-Arrhenius temperature dependence also typifies the case for many other viscous liquid systems that undergo restructuring processes that require the “cooperative involvement of other molecular motions” (Angell, 1983, 1988). Included in these other viscous liquid systems that exhibit non-Arrhenius behavior are concentrated aqueous solutions (analogous to frozen foods) at subzero temperatures (Pika1 and Shah, 1990), according to a suggestion by Hofer et al. (1989).

The impact of WLF behavior on the kinetics of diffusion-limited relaxation processes in water-plasticized, rubbery food polymer systems has been conceptually illustrated by the schematic curves shown in Fig. 16 (Levine and Slade, 1989c; Franks et af., 1991). Relative relaxation rates, calculated from the WLF equation with its “universal” numerical constants, demonstrate the nonlinear logarithmic relationship: For AT = 0”, 3”, 7”, ll”, and 21”C, corresponding relative rates would be 1, 10, lo2, lo3, and lo5, respectively. These rates illustrate the five orders-of-magnitude change, over a 20°C interval above Tg, typically shown by WLF plots, as mentioned earlier with regard to Fig. 11. They are dramatically different from the rates defined by the Qlo = 2 rule of Arrhenius kinetics for dilute solutions, as shown for comparison in Fig. 16. As pointed out earlier with regard to


Fig. 14A, for Arrhenius behavior above T,, a factor-of-10 change in relaxation rate would require a 33°C change in temperature, in contrast to a 3°C change for WLF behavior near Tg of a partially crystalline polymer of T,,,/Tg = 1.5 (Slade and Levine, 1988b). An illustration of the effect of such WLF kinetics, relating to the striking contrast between the relative time frames involved in slow caking and spontaneous agglomeration in amorphous food powders (Levine and Slade, 1989a), is represented by the schematic state diagram in Fig. 17. Caking and agglomeration are identical, translational diffusion-limited collapse processes that can occur in low- moisture, amorphous powders sensitive to plasticization by water and/or heat (Levine and Slade, 1986, 1988b; Peleg, 1993). As implied in Fig. 17, caking and agglomeration are distinguishable only by a WLF-type time- temperature transformation (Slade and Levine, 1988b). It has been shown (Downton et al., 1982; Tardos et al., 1984; Wallack and King, 1988) that spontaneous agglomeration of solid powder particles occurs, at a so-called “sticky point” temperature (Tsp), when q of the liquid phase at the surface of a particle drops to =lo7 Pa s. This q is =lo5 lower than qg. From the WLF equation (which treats viscosity as a diffusion-limited relaxation process), as defined above in Fig. 16, this five orders-of-magnitude difference between qg and qsp corresponds to a AT of =21”C between the Tg that governs caking and the Tsp that governs spontaneous agglomeration of a given powder at a given moisture content. Thus, as indicated on the state diagram in Fig. 17, WLF theory would predict that the Tg and Tsp

g!

E 3 c

AT = 21°C

12

I Water

FIG. 17. Schematic state diagram of temperature versus water content for an amorphous food powder, illustrating the WLF-governed temperature differential (at a given moisture content) between the iso-viscosity lines representing the Tg curve, just above which slow caking can occur, and the “sticky point” temperature ( Tsp) curve, above which spontaneous agglomeration can be effected.


curves should represent parallel iso-viscosity lines. The Tsp curve for fast agglomeration (in seconds at T > Tsp) during processing would lie above the Tg curve for slow caking (after months at T just above Tg) during storage, and the AT of 21°C would reflect the very different time scales for these two collapse processes (Levine and Slade, 1986, 1988b, 1989a; Slade and Levine, 1988b). Recently reported experimental results for amorphous powders of mixed sugars and of maltodextrins (Roos and Karel, 1991a,b) have confirmed the quantitative relationship between Tg and Tsp predicted by Fig. 17. Peleg (1993), too, has recently reassessed and endorsed this relationship.

Another example of WLF-governed relaxation behavior involves the kinetics of (re)crystallization (Levine and Slade, 1988a, 1989b; Slade et al., 1989). Crystallization kinetics, with regard to such common sugars as glucose, sucrose, and lactose, has been a subject with a long history of study by food scientists and technologists (Makower and Dye, 1956; White and Cakebread, 1966; Cakebread, 1969), which has seen much renewed interest in recent years (Slade and Levine, 1988b, 1991a; Levine and Slade, 1989a, 1992b; Roos and Karel, 1990, 1991a,e, 1992, 1993; Van Scoik and Cars- tensen, 1990; Blanshard et al., 1991; Noel et al., 1991; Shimada et al., 1991; Karel, 1992; Levi and Karel, 1992a, 1993b; Roos, 1992a,c; Arvanitoyannis and Blanshard, 1993a,b; Hartel, 1993; Karel et al., 1993a,b). Crystallization is a diffusion-limited process (Baro et al., 1977) that, on a time scale of technological significance, can only occur within the WLF rubbery domain bounded by the temperature limits of Tg and T, (Wunderlich, 1976). As illustrated in Fig. 18 (Levine and Slade, 1989b), the propagation step in the crystallization mechanism approaches a zero rate at T < Tg for an amorphous but crystallizable solute [either polymeric (Wunderlich, 1976) or monomeric (Levine and Slade, 1988a)], initially quenched from the melt or liquid solution state to a kinetically metastable solid state. Due to immobility in the glass, migratory diffusion of either large main- chain segments or small molecules, required for growth of crystals from metastable nuclei, would be inhibited over realistic times (Levine and Slade, 1988a; Chang and Baust, 1991c; Noel et al., 1991). However, propagation rate increases exponentially with increasing AT above Tg (up to T,) (Marsh and Blanshard, 1988; Morris, 1990; Biliaderis, 1991b, 1992b; Arvanitoyannis and Blanshard, 1993a), due to the mobility allowed in the rubbery liquid state. Thus, a recrystallization transition from unstable (i.e., undercooled) amorphous liquid to (partially) crystalline solid may occur at T > Tg (White and Cakebread, 1966; Karel, 1986; Phillips et al., 1986; Levine and Slade, 1988b, 1989a; Noel et al., 1991; Ring and Whittam, 1991; Tian and Blanshard, 1992a; Arvanitoyannis et al., 1993; Eerlingen et al., 1993), with a rate defined by the WLF

GLASS TRANSITIONS 21 1

RATE

NUCLEATION * . PROPAGATION

' . . OVERALL

'CRYSTWUTION . ' * * *

U * . . I

I Tg

GLASS TRANSITION

I Tm

MELTINQ POINT

TEMPERATURE

FIG. 18. Crystallization kinetics of partially crystalline polymers, expressed in terms of crystallization rate as a function of temperature. From Levine and Slade (1989b) with permission of Van Nostrand Reinhold/AVI.

equation (Levine and Slade, 1986). Brydson (1972) illustrated the extreme temperature dependence of, and the remarkably contrasting time frames that resulted for, the crystallization kinetics for a water-plasticized, well- behaved (i.e., T,,,/Tg = lS) , synthetic amorphous polymer, nylon-66. When that polymer's water-plasticized Tg was = room temperature, slow crystallization occurred during room-temperature storage for a period of up to 2 years; however, when the polymer was annealed at a higher temperature, T, = ( Tg + Tm)/2 [i.e., the temperature corresponding to the peak maximum of the overall crystallization rate curve in Fig. 18, also corresponding to the maximum rate of annealing at T, = 0.83Tm (Brydson, 1972; Slade and Levine, 1988d)], a comparable extent of crystallization was achieved in only =30 minutes.

The facts that time-dependent recrystallization can only occur at temperatures above Tg, with a rate increasing with increasing AT above Tg, and manifests kinetics defined by the quantitative WLF (rather than Arrhenius) equation (Levine and Slade, 1988b) have been confirmed in recent experimental studies of the recrystallization of amorphous, freeze-dried sugars (sucrose and lactose) by Karel, Roos, and co-workers (Roos and Karel, l990,1991a,e, 1992,1993; Shimada et al., 1991; Karel, 1992; Levi and Karel, 1992a, 1993a,b; Roos, 1992a,c; Karel et al., 1993a,b). As shown earlier in


Fig. 12A, which combined data for sucrose and lactose from Roos and Karel with a critical datum point (the x highlighted in the oval) from the classic study of sucrose crystallization by Makower and Dye (1956), and in Fig. 12B, experimental curves of the temperature dependence (in terms of AT above T,) of crystallization times for completely amorphous sucrose and lactose at low moisture contents (in the water-plasticized, rubbery liquid state at various AT values above T,) fit the theoretical curve of the WLF equation (with C1 = 17.44 and C2 = 51.6), but not that of the Arrhen- ius equation. Other specific examples of such a recrystallization process [i.e., a collapse phenomenon (Levine and Slade, 1988b)l include ice (Roos and Karel, 1991e-g) and solute [e.g., lactose in dairy products (Blanshard and Franks, 1987)] recrystallization in frozen aqueous food systems at T > T, = T,‘ (Levine and Slade, 1986).

A critical message to be distilled from this section is that the structure- property relationships of water-compatible food polymer systems are dictated by a moisture-temperature-time superposition (Starkweather, 1980; Flink, 1983; Levine and Slade, 1986; Lillie and Gosline, 1990). Referring to the schematic state diagram in Fig. 4 as a conceptual mobility map (representing an extension of the dynamics map in Fig. 3), one sees that the T, curve represents a boundary between nonequilibrium glassy solid and rubbery liquid physical states in which various diffusion-limited processes [e.g., collapse phenomena involving mechanical and structural relaxations (Levine and Slade, 1988b)l either can (at T > T, and W > W,’, the high-moisture portion of the water dynamics domain corresponding to the upper left part of Fig. 4, or T > T, and W < W,’, the low-moisture portion of the water dynamics domain corresponding to the upper right part of Fig. 4) or cannot occur (at T < T,, in the domain of glass dynamics corresponding to the bottom part of Fig. 4) over realistic times (Levine and Slade, 1986,1988a; Franks, 1989,1990; Karel, 1992; Karel et af., 1993b). The WLF equation defines the kinetics of molecular-level relaxation processes [e.g., the diffusion-limited chemical and enzymatic reactions studied by Lim and Reid (1991) and Kerr et al. (1993), mentioned earlier], which will occur in practical time frames only in the rubbery state above Tg, in terms of an exponential, but non-Arrhenius, function of AT above this boundary condition (Levine and Slade, 1986).

E. THERMOSETTING OF AMORPHOUS POLYMERS

The classical thermosetting process for synthetic amorphous polymers, e.g., in the production of rigid epoxy resins and flexible natural or synthetic rubbers, is typically described by a time-temperature transformation (TIT)


reaction diagram (Prime, 1981). As illustrated in Fig. 19 (Sperling, 1986), this TT diagram has been “used to provide an intellectual framework for understanding and comparing the cure and glass transition properties of thermosetting systems” (Sperling, 1986). As reviewed by Prime (1981), thermosetting polymers become infusible (i.e., unmeltable) and insoluble due to chemical crosslinking reactions during curing, which produce a three- dimensional network with essentially infinite MW. Curing of amorphous polymers at a curing temperature (T,,,,), somewhat above the initial molecular Tg of the polymer reactant (Tgo), allows very slow conversion with concomitant elevation of molecular Tg, until molecular Tg approaches TCure, and the system vitrifies. The molecular Tg depends only on the extent of conversion and not on T,,, (Bair, 1985). At a sufficiently higher T,,,, gelation occurs before vitrification, at a characteristic extent of conversion. At this system-specific degree of cure, the partially cured polymer thermosets by undergoing a sudden and thermally irreversible transformation from a viscous liquid (T,,, % Tm of the polymer reactant) to a nonflowable,

Log time

FIG. 19. Thermosetting process for amorphous polymers, as illustrated by a time-temperature transformation reaction diagram. See text for explanation of symbols. From Sperling, L. H. 1986. “Introduction to Physical Polymer Science.” Copyright 0 1986 Wiley (Intersci- ence), New York. Reprinted by permission of John Wiley & Sons, Inc.


elastic network (T,,, > the gel point, initial gel-T,), which rapidly approaches infinite macroscopic viscosity as gel- T, approaches T,,,,, and the system vitrifies. Gelation, per se, does not retard the rate of the curing reaction (which is chemically, rather than diffusion, controlled as long as T,,,, is well above the effective instantaneous network T,), and is detectable by the inability of large gas bubbles to rise in a thermosetting mass (Prime, 1981). In Fig. 19, the S-shaped curing curve between T, and T,, (the network T, of the fully cured system) results, because the reaction rate increases with increasing temperature (Sperling, 1986). Thus, the gel point, initial gel-T,, is the critical minimum value of gel-T, that represents the temperature at which gelation and vitrification (of the elastic gel to a glass) occur simultaneously. As described above, at temperatures between gel-T, and T,,, the thermosetting mass first reacts chemically to form a crosslinked rubbery network, then vitrifies as the continuously increasing network T, becomes coincident with T,,,,, essentially quenching the curing reaction (now diffusion-controlled) before completion, but giving the illusion that the product is completely cured (Sperling, 1986). Once the network T, exceeds T,,,,, the diffusion of reactants becomes so restricted that the reaction rate decreases by more than three orders of magnitude (Bair, 1985). The resulting incompletely cured network would have a T, denoted as ult-T,, where gel-T, < ult-T, < T,, (Prime, 1981). However, since vitrification is thermally reversible, curing can be extended by a postcure heat treatment of the rubbery thermoset at T > ult-T,, or rapidly completed by a postcure heat treatment at T > Tg,.

Importantly, the TTT diagram in Fig. 19 refers to the curing and vitrification kinetics of a neat polymer system, in the absence of plasticizer or solvent. Sorption of, or dilution by, plasticizer has a profound effect on the time frame of the TIT diagram, due to depression of all operative T, values and concomitant acceleration of the curing reactions at a given T,,,,. The role of water as a plasticizer is perhaps even more critical for the storage stability of both the original reaction mixture and the final cured product. Moisture sorption or loss has been shown to be a major factor in the aging behavior of synthetic thermosets (Prime, 1981). Plasticization of the thermoset network depresses network T, and modulus, whereas loss of plasticizer leads to toughness or brittleness.

The TTT diagram in Fig. 19 illustrates the four material states encoun- tered during curing: liquid, rubber (elastomer), ungelled glass, and gelled glass. The time axis signifies measured times to gelation and to vitrification, as a function of cure temperature (Prime, 1981; Sperling, 1986), and under- lines the importance of the kinetics of curing reactions in thermosettable polymers. For example, in a DSC experiment, if a polymer sample (of T, = T,) is heated to T > gel-T, in an initial scan, but then immediately


cooled to T < Tm and rescanned, the extent of curing achieved during the first scan may be undetectably small (Slade et al., 1989). Consequently, Tm in the rescan may appear unchanged, and the sample may behave as a thermoplastic polymer. In contrast, if the same sample is heated to T 5 gel- Tg, or to T > gel-T, and then held for a longer time before cooling and rescanning, the curing reaction can proceed. In this case, the rescan would not show Tm, but rather a new network T, > Tm, and the sample would appear to be thermoset (Slade et al., 1989). An actual example of this contrasting behavior has been reported for curing of an epoxy-amine system in a DSC (Prime, 1981). On holding at T,,, a few degrees above gel- Tg, about an hour was required to achieve the characteristic extent of cure to thermoset. When T,,,, was more than 80°C above gel-T,, the same extent of cure was achieved in about 3 minutes, so that thermosetting could occur within the time frame of a dynamic DSC experiment. Indeed, in the latter case, because T,,,, was more than 20°C above Tg,, complete curing was achieved within 10 minutes.

The apparent parallels between the established behavioral characteristics of thermosetting synthetic amorphous polymers, described above, and the empirically known structure-funtion relationships of wheat gluten protein in various baking applications (and postbaking heat treatments, e.g., microwave “refreshening”) have been described (Slade et al., 1989) and are reviewed below.

1. Wheat Gluten as a Viscoelastic Polymer System

Unique among cereal grain storage proteins, native (“vital”) wheat gluten is a major functional food ingredient, especially as a flour component in baked goods (Bloksma, 1978; Kasarda et al., 1978; Pomeranz, 1978; Magnu- son, 1985). The structure-property relationships of gluten have been described in terms of a highly amorphous, multipolymer system, water- plasticizable but not water-soluble, and capable of forming continuous, multidimensional, viscoelastic films and networks (Schofield et al., 1984; Slade, 1984; Hoseney et al., 1986; Levine and Slade, 1988a; Doescher et al., 1987; Edwards et al., 1987). Gluten also appears to function as either a thermoplastic or thermosetting amorphous polymer in response to the heat-moisture treatment constituted by baking (Slade et al., 1989, 1993; Slade and Levine, 1992c; Levine and Slade, 1993). The thermosetting behavior, when it occurs at practical heating rates and sample concentrations, is considered to be a typical consequence of irreversible protein denaturation and “heat-set gelation” above a critical temperature (Schofield et al., 1984; Magnuson, 1985; Davies, 1986; Schofield, 1986; Ablett et al., 1988; Blan- shard, 1988; Masi, 1989). The resulting dramatic changes in mechanical and


rheological properties (i.e., large increases in viscosity and elastic modulus at temperatures between about 50” and 100°C) of bread dough during baking are responsible in part (along with starch gelatinization) for the transformation from a predominantly viscous dough to a predominantly elastic baked crumb (Schofield et al., 1984; Schofield, 1986; Bloksma, 1986; Blanshard, 1988; Bloksma and Bushuk, 1988; Dreese et al., 1988b; Masi, 1989). The mechanism of such thermo-irreversible “heat-set’’ gelation involves formation of permanent networks, due to rheologically effective, intermolecular crosslinks composed of covalent disulfide bonds, resulting in a polymerization of the gluten proteins (Bloksma, 1978,1986; Pomeranz, 1978; Cantor and Schimmel, 1980; Schofield et al., 1984; Ablett et al., 1988; Dreese et al., 1988a; Shewry et al., 1988; Attenburrow et al., 1990; Hoseney and Rogers, 1990; Davies et al., 1991; Hoseney, 1991). Such completely amorphous, thermoset gel networks (which may also contain entangled polypeptide chains) represent a separate category from two other common types of food polymer gels, i.e., partially crystalline, thermoreversible gels (e.g., starch-water and gelatin-water) and completely amorphous, topolog- ically reversible “entanglement” gels (e.g., sodium caseinate-water, not heat-treated) that can be dispersed by dilution (Slade and Levine, 1987a,b; Burchard, 1988; Slade et al., 1989; Levine and Slade 1988a, 1989b).

Wheat gluten’s unique combination of cohesive, viscoelastic, film- forming, thermoplastic or thermosetting, and water-absorbing properties (Schofield et al., 1984; Magnuson, 1985) is derived from the structure of the native proteins (Payne, 1987). Gluten is a complex of interacting proteins that have been categorized traditionally into two major groups, gliadins and glutenins, based on their solubility and insolubility, respectively, in 70 to 90% aqueous ethanol. An alternative classification has been proposed, which reflects the chemical and genetic relationships of the component polypeptides of gluten (Tatham el al., 1985; Payne, 1987). The groups are (1) sulfur-poor prolamins (o-gliadins), (2) sulfur-rich prolamins (a-, j3-, and y-gliadins; low-MW glutenin subunits), and (3) high-MW prolamins (high- MW glutenin subunits). The broad correspondence between the two classification schemes relates to identification of proteins present as monomers (single polypeptide chains) associated by hydrogen bonding and hydrophobic interactions as gliadins, and proteins present as polymers (multisubunit aggregates) linked by interchain disulfide bonds as glutenins (Payne et al., 1985; Schofield, 1986; Payne 1987). In this context, gliadins are lower-MW proteins (< lo’) that can adopt globular conformations with intramolecular disulfide bonds, a-helical segments, and/or j3-turns, and act as the viscous, extensible, plasticizing component of gluten (Schofield, 1986; Payne, 1987; Blanshard, 1988; Slade ef al., 1989). Glutenins are higher-MW (> 10’) with broader MW distribution, highly elastic, multisubunit proteins [which con-


tain a significant amount of secondary structure in the form of @turns (Tao et al., 1989)] crosslinked by intersubunit disulfide bonds (Schofield, 1986). Glutenins act as the elastic component (Lasztity, 1986; du Cros, 1987; Edwards et al., 1987; Payne, 1987; Blanshard, 1988) of the viscoelastic network that forms in a flour-water dough via crosslinking by thiol-disulfide interchange reactions catalyzed by rheologically active thiols (Kasarda et al., 1978; Pomeranz, 1978; Bloksma, 1978; Schofield, 1986; Ewart, 1988). In this functionality, glutenins are similar to elastin, in that these are the only two protein systems known to exhibit high segmental mobility at high MW and to form hydrated, three-dimensional, open-structured, rubberlike elastic networks via disulfide bonding (Schofield, 1986; Edwards et al., 1987; Ablett et af., 1988; Shewry et al., 1988; Slade et al., 1989).

It is important to stress the distinction between local intermolecular disulfide crosslinking of gliadins (Schofield et al., 1984) and long-range intermolecular disulfide network formation of high-MW glutenins (Weegels et al., 1988; Shewry et al., 1988). “Gluten” defies the usual biological rule of thumb for genetic design of stable proteins (Cantor and Schimmel, 1980): single polypeptide chains should contain either free thiols (cysteine) or disulfides (cystine), but not both in a single gene product. Coexistence of both -SH and -S-S- results in dynamic catalysis of disulfide exchange and consequent lack of definition of highly probable, predominant subunit conformations in the gluten population (Cantor and Schimmel, 1980). Disul- fide exchange, crosslinking, and thermosetting reactions during heat-moisture treatments, such as baking, result in (1) progressive depletion of the least stable crosslinks and catalytically effective thiols and (2) eventual establishment of permanent local crosslinks in gliadins, a permanent long- range (cooperative) network by high-MW glutenins, and residual catalytically ineffective thiols. Extensibility and viscous flow of film-forming gliadins in doughs with moisture contents > W,’ are diminished by disulfide crosslinking during baking (Schofield et al., 1984), resulting in decreased extensibility and increased hardness, but not increased elasticity. Elasticity and rigidity of network-forming high-MW glutenins in such doughs are enhanced by disulfide thermosetting during heat treatment (Weegels et al., 1988; Shewry et al., 1988), which results in an increased number of permanent disulfide junction zones, such that loss of extensibility is accompanied by increased elasticity. Thus, increased firmness results from an increase in loss modulus due to gliadin crosslinking (Schofield et al., 1984) and an increase in elastic modulus due to high-MW glutenin (Weegels et al., 1988; Shewry et al., 1988) network thermosetting. Clearly, simple determination of neither residual thiol content nor the amount of insolubilized protein (Rogers et af., 1990; Huang et al., 1991) is sufficient for interpretation of the mechanism of firming or development of rubbery texture during


heat-moisture treatment of gluten doughs, such as microwave baking or reheating of bread (Slade and Levine, 1992~). The causative insolubilized proteins should be identified by demonstration of gel electrophoresis bands, high-pressure liquid chromatography (HPLC) peaks (Lookhart et al., 1987; Pomeranz et al., 1987; Menkovska et al., 1987; Ng and Bushuk, 1987), or antigens to monoclonal antibodies (Skerritt et al., 1988; Hill et al., 1988), which are present in doughs but missing in baked products. Missing gliadin subunits could account for increased firmness and decreased extensibility without increased rubberiness, while missing glutenin subunits could account for a dramatic increase in rubberiness with only moderate increase in firmness (Levine and Slade, 1989b). These structure-property relationships of gluten are critical to the processing, product quality, and storage stability of various types of flour-based doughs and finished baked goods (Schofield et al., 1984; Davies, 1986; Schofield, 1986; Blanshard, 1988; Slade et al., 1989; Levine and Slade, 1989b; Hoseney and Rogers, 1990; Hoseney, 1991, 1992; Slade and Levine, 1992~).

2. Thermosetting Behavior of Amorphous Gluten

Slade (1984) reported DSC results of Tg measurements from a study of commercial vital wheat gluten, approached as an amorphous, water- and lipid-compatible polymer. Isolated wheat gluten at low moisture (i.e., 6 w% “as is” water content) manifested a Tg of 66°C (Slade, 1984; Levine and Slade, 1988a). Once heated through this glass transition, gluten has sufficient mobility, due to thermal and water plasticization, to form a thermoset network, via disulfide crosslinking. This thermally irreversible thermosetting reaction was suggested (Levine and Slade, 1988a) to be analogous to chemical “curing” of epoxy resin and “vulcanization” of rubber. As described earlier, such classical thermosetting reactions of synthetic amorphous polymers are also only possible at T > Tg (Prime, 1981; Sperling, 1986). In contrast to its response to plasticization by water alone, gluten remained thermoplastic (i.e., its glass transition was reversible) in 1 : 1 co- melts with nonaqueous plasticizers such as triacetin and B12K, the latter a blend of lipids commonly used as a bread-shortening system. However, higher Tg values of 73” and 79°C demonstrated the more dramatic plasticizing effect on gluten of only 6 w% water, compared to 50 w% triacetin or B12K, respectively (Slade, 1984; Levine and Slade, 1988a). This finding, that water is far superior to lipids in its ability to plasticize gluten, was recently corroborated by results of Kalichevsky et al. (1992c,d).

In the context of effects on bread-baking performance, Slade (1984) deduced that the interaction of aqueous gluten with lipids represents another example of relative “antiplasticization” of a polymer by a higher-


MW plasticizer [analogous to the effect of different sugar-water co-solvents on the gelatinization of starch (Slade and Levine, 1987b)], relative to plasticization by water alone. Slade’s (1984) conclusion, that the thermomechanical and rheological properties of wheat gluten can be explained based on recognition of gluten as a water- and lipid-plasticizable, highly amorphous polymer system, was verified, in part, by Tg measurements from a subsequent DSC study by Hoseney et al. (1986). As shown earlier in Fig. 10, their glass curve of Tg versus w% water is a smooth curve from >16OoC for the glassy polymer at 41 w% moisture to 15°C for the rubbery polymer at 16 w% moisture. As discussed earlier, these results exemplified the typical extent of plasticizations by water (i.e., depression of Tg by =lO”C/w% water), previously illustrated for a variety of other water-compatible, amorphous food polymers (Levine and Slade, 1988a, 1989b; Slade et al., 1989). However, Hoseney et al. (1986) did not observe the thermosetting behavior of aqueous gluten, reported by Slade (1984); we suggested that this could have been due to differences in the DSC measurement techniques employed (Slade et al., 1989). More recently in Hoseney’s laboratory, Dong (1992) conducted a TMA study of gluten at low moisture contents and reported that gluten’s Tg was observed to increase to higher temperatures with increasing time of heat treatment at T slightly above Tm, as a consequence of thermosetting via intermolecular disulfide crosslinking. Dong explained the contrast between her results and the earlier ones of Hoseney et al. (1986) by pointing out that TMA showed greater sensitivity than did DSC to changes caused by heat treatment of gluten, because TMA is known to be more sensitive to rheological changes occurring at and below the gel point of a thermosetting polymer (Prime, 1981). Slade’s (1984) original conclusion, regarding the basis of its structure-property relationships in the amorphous polymeric nature of wheat gluten, was also verified by another DSC study (Fujio and Lim, 1989) of the effect of moisture content on Tg and the resulting properties of gluten. Fujio and Lim’s results showed that heat treatment of native gluten at low moisture produced an amorphous polymeric material that, once softened to a rubbery liquid (of increased compressibility under pressure and resulting higher density) due to plasticization by heat and moisture, underwent a diagnostic color change (as a consequence of the pressure-induced increase in density) above a critical temperature (which decreased with increasing moisture) corresponding to Tg. Additionally, a recent DSC study by Lawton and Wu (1993) on the thermal behavior of annealed wheat gluten produced new experimental results that confirmed the findings of earlier workers, with regard to gluten’s behavior as an amorphous, water-plasticizable polymer, which is analogous to the behavior of many synthetic amorphous polymers.


It has been suggested (Slade et al., 1989) that the enhanced ability of amorphous gluten in hard wheat flour-water doughs heated to T > Tg to thermoset (i.e., undergo an irreversible structural transformation from viscous liquid polymer system with transient crosslinks to elastic, permanently crosslinked gel with a higher network Tg) represents a critical aspect of the bakingholume-determining mechanism for white pan bread, and is partly responsible (along with starch gelatinization) for the mechanical and rheological differences between dough and baked bread. Recent results that lend direct support to this suggestion have come from a study by Van Vliet et al. (1992) of the mechanism of gas retention in bread doughs from hard wheat flours. Their results showed that in biaxially extended doughs, hydrated gluten is oriented and birefringent [as also reported previously by Slade et al. (1989)], and that gluten in this condition can thermoset during baking, thus contributing to the setting of the bread crumb and its transformation from foam to sponge. In a related vein, the dramatic increases in viscosity and elastic modulus of starch-free gluten-water doughs heated to temperatures between 4 0 ” and 100”C, leading to gelation via network formation by protein-protein aggregation at T > 80°C (Schofield et al., 1984; Davies, 1986; Ablett et al., 1988; Blanshard, 1988; Dreese et al., 1988a; Shewry et al., 1988; Weegels et al., 1988; Masi, 1989), are strongly suggestive of a molecular mechanism involving a classical thermosetting process via covalent crosslinking in an amorphous polymer system (Levine and Slade, 1989b). It is well known that the effect of shortening (mixed lipids) on flour-water bread doughs produces breads of increased baked loaf volume, by allowing greater oven spring during baking (Moore and Hoseney, 1986). This effect has been interpreted (Slade et al., 1989) as arising from lipid plasticization of gluten to extend its period of thermoplas- ticity early in the baking process, prior to thermosetting of the water- plasticized gluten network at the end of baking, in addition to the better known role of lipid to retard starch pasting. It has also been suggested (Slade et al., 1989) that the characteristically rubberyAeathery texture produced by microwave reheating of bread and other baked products results in large part from a postcure heat treatment of the thermoset gluten polymer, which effects a further increase in crosslinking density with a concomitant further elevation of network Tg. In a related vein, we have described a new hypothesis [reviewed elsewhere (Slade et al., 1989)] for the role of gluten in the “spreading”/baking mechanism of sugar-snap cookies. This hypothesis dif- ferentiates, based on flour functional properties, between certain glutens that thermoset during baking (as in bread) of inferior cookies made from poor-cookie-quality hard wheat flours and other glutens that remain thermoplastic until structural collapse occurs during baking of superior cookies made from excellent-quality soft wheat flours (Stauffer, 1990, 1992).


In the context of the earlier description of the gelation mechanism and viscoelastic properties of thermosetting amorphous polymers, the parallels seem obvious (Slade et al., 1989) among the established behavioral characteristics of thermosetting synthetic amorphous polymers, the equally well- documented behavior of thermosetting partially crystalline and paracrystal- line keratin biopolymer systems (Whewell, 1977; Milczarek et al., 1992), and the empirically known structure-function relationships of hard wheat glutens in the baking of excellent breads and inferior sugar-snap cookies (Bloksma, 1978; Kasarda et al., 1978; Pomeranz, 1978). Moreover, the general response of fully baked products to subsequent microwave heating is consistent with a postcure reaction stage in a continuous, thermoset gluten network. Such a partially thermoset network, incompletely cured during baking, with incident T, = ult-T, (as a result of removal from the oven, rather than intervening vitrification), could have its curing process extended or even completed during microwaving, where superheated steam could provide internal temperatures greater than those during oven baking. Con- sequently, the network T, would increase to a higher ult-T, or even to T,,. The resulting texture of a product after cooling to room temperature would be characterized by increased rubberinessAeatheriness, due to increased macroscopic viscosity, owing to the change in AT between network Tg and room temperature (Slade et al., 1989). The maximum elevation of T, effected by curing, from molecular Tgo to network T,,, varies significantly depending on the system, e.g., from <30°C for an unidentified DuPont elastomer (Maurer, 1981) to >170"C for a bisphenol-A epoxy (Prime, 1981). Considering the approximate value of Tm for developed gluten in unbaked bread dough [i.e., gluten T,' of about -7.5"C (Slade et al., 1989)] and the comparative textural attributes at room temperature of the dough, the fresh oven-baked loaf, and the product microwaved after baking, we have speculated (Slade et al., 1989) on the values of ult-T, after baking or after microwaving. For AT = 33°C (room T - Tm), dough firmness is characteristically low. Fresh-baked and cooled bread is clearly elastic (diagnostic of a critical extent of network formation), but also quite deformable, so that AT must exceed 0. Thus, baked ult-T, is between -7.5" and 25°C. The leathery texture of microwave-reheated bread is diagnostic of AT Q 20°C. In the worst case, microwaved ult-T, may reach 25°C or above! As observed for the aging of commercial synthetic polymer thermosets, moisture will play a critical role in aging of the baked product. Thus, ult-T, and product modulus could be elevated even further by moisture redistribution or drying (as an additional consequence of microwave heating) or depressed by hu- midification (Slade et al., 1989).

Such apparent analogies imply a relatively high potential for displacement of the distribution of free thiol groups and intramolecular disulfide bonds


toward intermolecular disulfide crosslinking and network thermosetting during baking in glutens from hard wheat flours, without underestimating the contribution of leached-amylose networks [especially from chlorinated flours, as emphasized by Blanshard (1986)l or pentosan networks (Slade et al., 1989; Slade and Levine, 1992~). Blanshard (1986) described the progressive firming of cakes during cooling after baking in terms of Tg of amylose and the plasticizing effect of moisture and short amylopectin branches. Because the cake crumb structure depends on leached-amylose networks [so that replacement of wheat starch by waxy corn starch results in cake collapse (Blanshard, 1986)], the operative Tg is network Tg. In glutens from excellent cookie-quality soft wheat flours used in optimized cookie formulations, the converse argument, suggested by cookie-baking results reviewed elsewhere (Slade et al., 1989,1993; Slade and Levine, 1992c; Levine and Slade, 1993), implies a lower potential for such a displacement of the distribution toward disulfide crosslinking during baking. These specula- tive conclusions were offered (Slade et al., 1989) to stress the importance of network Tg, rather than molecular or segmental Tg, to the rheological behavior of baked products (Levine and Slade, 1989b; Slade and Levine, 1992c), in the hope that they would stimulate future research.

Subsequent research in several laboratories (Attenburrow et al., 1990, 1992; Rogers et al., 1990; Davies et al., 1991; Huang et al., 1991; Dong, 1992; Tian and Blanshard, 1992b) has focused on the suggestion that thermosetting of water-plasticized gluten, through high-temperature treatment by microwave or conventional heating, results from disulfide-crosslinked network formation and concomitant elevation of gluten’s Tg, which could account for the textural toughening of reheated bread (Slade et al., 1989). The previously mentioned findings of Dong (1992) appear to lend strong support to this concept. In contrast, workers at Unilever (Davies et al., 1991; Attenburrow et al., 1992) reported “anomalous behavior” that could call the concept into question. They made DMTA measurements of storage modulus for hydrated glutens (21% moisture) and found, unexpectedly, a lower Tg for gluten heat-set at 150°C than for gluten heat-set at 90°C. This finding appeared to contradict results of their earlier study (Attenburrow et al., 1990), which were interpreted as showing that increasing temperature treatments (up to -130°C) of rubbery gluten dough result in an increased amount of intermolecular crosslinking. A possible explanation for the “anomalous behavior” they observed [which they speculated could be due to a reduction in polymer chain length (Attenburrow et al., 1992)] is that it resulted from a thermosetting reaction (“curing”) carried out in air, leading not only to crosslinking but also to oxidative degradation at 150°C (but not at 90°C) and, thus, to a net reduction in Tg; if the reaction had been carried out in a nitrogen atmosphere, the gluten heat-set at 150°C


would have shown a higher Tg than that for gluten heat-set at 90°C. This possibility is suggested by a recent finding for synthetic polyimide thermosets; i.e., postcuring a thermoset polyimide in air can oxidatively degrade the polymer, but postcuring it in nitrogen will raise its Tg (Studt, 1992). Further support for this explanation could come from the fact that the measured Tg for gluten heat-set at 150°C was around -10°C (Davies et al., 1991; Attenburrow et al., 1992), a value more suggestive of a molecular Tg (i.e., close to Tg' = -7.5"C) for somewhat degraded gluten than of a network Tg (expected to be well above 0°C) for extensively crosslinked gluten (Slade et al., 1989). In a related study by Tian and Blanshard (1992b), who used NMR to investigate the toughening of bread by microwave reheating, their data could be interpreted as showing that microwave heating results in a thermoset gluten network with a rheologically unique distribution of disulfide crosslinks, rather than an increase in the number of crosslinks (Schofield et al., 1983). Such an interpretation would contrast with the questionable conclusion reached by Rogers er al. (1990) that disulfide crosslinking of gluten cannot explain the toughening of bread heated or baked in a microwave. Obviously, still more research is warranted, to clarify the relationship among gluten thermosetting, disulfide crosslinking, network Tgr and the practical problem of textural toughening of microwaved or conventionally reheated baked goods.

3. Thermoplastic and Thermoset Behavior of Aqueous Gluten Systems

As mentioned earlier, gluten-water blends are capable of exhibiting thermoplastic or thermoset behavior. With respect to gluten model systems, it has been shown that heat damage in gluten-water blends depends on both temperature and moisture content, and that thiol-disulfide exchange reactions are a critical feature of the changes in both mechanical and thermodynamic properties. These effects of heat-moisture treatment were measured by changes in interfacial behavior, distribution, and reactivity of thiol and disulfide groups; rheology; solubility; and gel electrophoresis to identify the missing bands in the solubilized protein (Schofield et al., 1983; Weegels er al., 1988,1991; Weegels, 1991; Weegels and Hamer, 1991; Elias- son and Silverio, 1991). At 65% moisture, progressive disulfide exchange leading to enhanced thermostability of crosslinking of high-MW (HMW) glutenins begins below 70"C, and unfolding and disulfide crosslinking of relatively heat-labile, sulfur-rich (a) gliadins occur above 70°C, while relatively heat-stable, sulfur-poor (0) gliadins are not polymerized even at 100°C (Schofield et al., 1983). At 25% moisture and 8O"C, HMW glutenins become crosslinked by disulfides, leading to a more rigid glutenin matrix, but lower-MW glutenins are less affected, and gliadins are essentially un-


changed (Weegels et al., 1988; Ablett et al., 1988). At moisture contents below 21%, even the HMW glutenins are progressively less affected (Weegels et al., 1988).

Similar experiments have not been conducted in gluten model systems containing sucrose, which would be much more appropriate for understanding gluten functionality in rich cracker and cookie products (Slade and Levine, 1992~). Actual baking experiments, followed by protein solubiliza- tion and gel electrophoresis, give clues to the expected behavior for ternary gluten-sucrose-water model systems. During baking of lean white pan bread (high moisture, low sugar), both HMW glutenins and gliadins (sulfur- rich a-, p-, and y gliadins most affected, sulfur-poor o-gliadins least affected) become crosslinked and insoluble in the absence of p-mercaptoethanol (Lookhart et al., 1987; Menkovska et al., 1987, 1988). During baking of sugar-snap cookies with high-quality soft wheat flours, HMW glutenins are unaffected, and gliadins are essentially unaffected (Pomeranz et al., 1987, 1989). The absence of extensive polymerization of glutenins and S-rich gliadins in cookie systems, compared to the presence of disulfide crosslinking of glutenins and S-rich gliadins in bread crumb (even greater in bread crust), was said to be “unexpected and unexplainable” (Pomeranz et al., 1987), since the internal temperature of the cookie exceeded that of the bread crumb during baking. However, the high sucrose concentration in the AACC sugar-snap cookie dough retarded thermosetting of glutenins and S-rich gliadins, just as it retards the gelatinization of starch, compared to the extent of starch gelatinization in bread crumb (Slade and Levine, 1992~).

It is important to stress that the thermosetting of gluten proteins by disulfide crosslinks occurs with essentially no change in the measurable content of total free -SH groups; rather, there is a progressive shift in the location of free -SH groups from a sodium dodecyl sulfate (SDS)- extractable to an SDS-inextractable form (Schofield et al., 1983). Thus, it is the progressive disulfide exchange, from the relatively thermolabile distribution characteristic of native gluten proteins in aqueous doughs toward an ultimate thermostable distribution characteristic of gluten networks with maximum rigidity and network Tg, that accounts for the mechanical changes from a thermoplastic gluten system in dough to a thermoset gluten in baked products (Schofield et al., 1984; Levine and Slade, 1989b). As mentioned earlier, this mechanism of thermosetting is entirely analogous to the thermosetting of synthetic polymers (Slade et al., 1989); with increasing reaction temperature, the instantaneous value of macroscopic Tg increases from an initial minimum value for the thermoplastic polymer chain (T,), through an intermediate higher value for the partially thermoset network (gel-T, or ult-T,), to an ultimate maximum T, (Tg,) of the maximally thermoset network (Sperling, 1986). For a homopolymer system, progres-


sive thermosetting and increase in network Tg occur with a progressive increase in crosslink density and network rigidity (Prime, 1981). But, the very key to gluten functionality is that gluten is not a homopolymer system! The intra- and intermolecular locations and distribution of thermolabile disulfide bonds among the different native gliadin and glutenin proteins of gluten differ from the progressively more thermostable population of disulfide bonds that results from disulfide exchange among gluten proteins that have undergone conformational (secondary and tertiary) and associa- tive (quaternary) changes during heat-moisture treatments (Schofield et af., 1984). These differences in the spatial and energetic distribution of covalent disulfide linkages, from the thermoplastic distribution associated with the dominance of molecular Tg in the dough to the thermoset distribution associated with the dominance of network Tg in the baked product, and the decreased tendency for ongoing disulfide exchange to depart from the thermoset distribution (Weegels, 1991), compared to the relatively facile disulfide exchange in the thermoplastic distribution, account for the characteristic difference in mechanical properties of a dough at room temperature from a baked product that has been cooled to room temperature (Schofield, 1986; Levine and Slade, 1989b).

With respect to the mechanical behavior and sensory texture of heat- moisture-treated gluten-based systems, it may be instructive to distinguish possible pathways that lead to different “ultimate thermosets.” A critical feature of nonequilibrium systems, such as thermosets, whose behavior depends on plasticization by time, temperature, and diluents of molecular and network glass transitions, is that the previous history of the system may be even more important than its final steady state (Slade and Levine, 1991a). We could compare the ultimate thermosets that would result from a two-step heating process with that from a single-step process. The ultimate distribution of disulfide linkages that would result from a two-step process of slow heating through the 55”-75”C temperature region where glutenins crosslink (Schofield et af., 1983) to a resilient network noted for its elastic recovery, followed by generation of a tough, leathery, elastic network resulting from postcuring of the glutenin network and participation of gliadins (Schofield et af., 1983), on continued heating past 130°C (Attenburrow et af., 1990; Cocero et af., 1992b), would not be the same as the distribution that would result, on instantaneously rapid heating to, and spontaneous crosslinking of gliadins above, 130”C, in a hard, brittle, short network via a one-step process. The generation of fewer crosslinks with greater rheological effectiveness, in terms of elastic recovery and chewy sensory texture, by the two-step process might well be associated with greater protein extractability, compared to the generation of greater insolubilization of protein and short, brittle sensory texture by the one-step process. The two-


step process might be identified with conventional bread baking, followed by microwave reheating, or with atmospheric steam cooking, followed by microwave superheated-steam treatment, whereas the one-step process could represent initial microwave baking. The protein extractability, rheological behavior, and sensory attributes of the final steady states resulting from such alternative pathways would not be expected to be, and have in fact been shown not to be (Rogers et al., 1990; Huang et al., 1991), the same. In contrast, the direct contribution of apolar interactions to the mechanical behavior of heat-moisture-treated gluten systems is not expected to be a dominant feature. Although an increasing extent of hydrophobic interactions among gluten proteins with increase in temperature during heat-moisture treatments is expected to be indirectly important to gluten functionality, through their modulation of disulfide exchange reactions during baking (Weegels et al., 1991), the relative importance of hydrophobic interactions is diminished on cooling (Privalov and Gill, 1988), when the baked product is returned to room temperature.

IV. RESEARCH NEEDS: OUTSTANDING PROBLEMS, ISSUES, AND UNANSWERED QUESTIONS

In the context of this concluding section on research needs and unsolved problems, it is apropos to quote the following remark by Franks (1992a), contained in a recent editorial on “the importance of being glassy:”

It is clear that amorphous solids and glass transitions impinge directly on several aspects of low temperature science. What is not clear to some of the practitioners who rush to apply these concepts is that the actual physical nature of the glass transition and its proper description are still subject to some very basic mysteries which continue to baffle physicists and chemists alike. We thus possess several applicable technologies and phenomenologies but lack a proper scientific understanding. Perusal of the relevant literature reveals much point scoring, overt and covert, rather than genuine attempts to solve the major outstanding problems.

In Section 11, C, we discussed the wide variety of experimental methods that have been used recently to determine Tg in aqueous food polymer systems (Table V). As a practical matter, there is a need for further research with regard to the issue of the agreement, or lack of agreement, among Tg values measured for the same sample by different methods, or even by the same method (e.g., DSC) using different experimental conditions. Since the glass transition is a kinetically controlled change in state, but not an equilibrium thermodynamic change of phase (Ferry, 1980), it is not at all surprising that the measured value of Tg (be it dry T,, Tg at 0 < W, < W,’,


or Tg') depends on the temperature-time (or -frequency) conditions of its measurement (Wunderlich, 1981, 1990). For example, for Tg measured by DSC, an order-of-magnitude decrease in heating rate (e.g., from 10" to l"C/min) would generally result in a decrease in T, of -3°C (Wunderlich, 1981; Noel et al., 1991) [as predictable from the WLF equation (Slade and Levine, 1991a)l. Similarly, for Tg measured by DMA, a decade drop in measurement frequency would result in a decrease in T, of about 3"-6°C (MacInnes, 1993). Consequently, a review of earlier reported T, values for pure amorphous glucose (Slade and Levine, 1991a), measured [mainly by DSC or differential thermal analysis (DTA)] over the past 60+ years, showed a spread of lO"C, while the more recent Tg values for glucose listed in Table XIV show an even bigger spread of nearly 20°C. In contrast, as shown by Schenz et al. (1984; Levine and Slade, 1990), when the same sample is measured by different thermal analysis methods, using essentially the same time-temperature conditions of measurement, it is possible to obtain the same Tg value, at least for a simple model system. For a 20 w% sucrose-water solution, measured by DSC (transition midpoint temperature), DMA (loss modulus peak temperature), and TMA (midpoint temperature of dimension change), Schenz et al. observed the same T,' value of -32"C, in agreement with other measurements subsequently made by many other workers (Table XIIB). As shown by the detailed studies by Kalichev- sky et al. (1992a-c), when the samples to be measured are model systems much more complex than a sucrose solution (e.g., aqueous amylopectin or gluten, with or without added sugars), and the techniques employed to determine Tg are as diverse in their measurement aspects as are DSC, DMTA, NMR, and Instron, less than perfect agreement among measured Tg values is only to be expected. Of special note, the results of Kalichevsky et al. (1992a-c) demonstrated that Tg, as determined by NMR [which can measure T, for a sample in terms of the onset of mobility (above a so- called "rigid lattice limit") on a size scale of 410 A], was generally observed at a significantly lower temperature (for a given sample composition and moisture content) than was the Tg determined by DSC or DMTA, which are considered to be able to detect the onset of longer-range, cooperative mobility in a glassy domain of >lOO-A dimension (Wunderlich, 1981,1990; Slade and Levine, 1991a). Also noteworthy are the recent results of a careful thermal analysis study by Cassel and Twombly (1991), who compared the T, values measured by DSC, DMA, and TMA for a well-characterized sample of epoxy resin. They reported that "the best agreement in Tg values between the various measurement methods was obtained between: DSC heat flow midpoint; TMA onset of expansion, flexure or stress relief; and the onset of the storage modulus for DMA data taken at 0.1 Hz. These were all within one degree . . . . The best correlation with the DSC and


TMA data to be obtained from a 1-Hz DMA run was that of the onset of the loss modulus peak." As reviewed elsewhere (Levine and Slade, 1990; Schenz et al., 1991), thermomechanical methods such as DMA or TMA are generally considered to have greater inherent sensitivity, for the detection of lower-intensity transitions such as glass transitions (in comparison to crystalline melting transitions), than do thermal methods such as DSC or DTA (Kalichevsky et al., 1993a). Thus, for T, measurements on real, multicomponent, complex food products, in which moisture and water- compatible solids can be heterogeneously distributed, which can result in the existence of multiple amorphous domains (large and small) with different T, values (Slade and Levine, 1991a), it is generally recommended that DMA or TMA be used, in addition to or in place of DSC (e.g., Dong, 1992; Le Meste et al., 1992; Hoseney et al., 1992), when one has a choice among these techniques. In a related vein, it has been noted that, from a food technology/product development point of view, it would be desirable to have alternative experimental methods that would be easier to use, faster to perform, and less expensive than DSC, DMA, or TMA, for the accurate measurement of Tg of real foods (Hegenbart, 1993). Such methods do not currently exist.

There is a widely agreed need for further research with regard to several questions, of more academic interest, about T, values. One unanswered question concerns the fundamental relationship (presumably, cause-and- effect) between the molecular structure of a given glass former and its Tg (i.e., dry Tg, Tg', or low-moisture Tg). This question was alluded to in Section 111, A, with reference to the dry T, values compiled in Table XIV, which showed that Tg can vary substantially, even within a series of carbohydrates of the same MW and only the most subtle differences in molecular structure.

Another question, discussed in detail in Section III,C,l,a, relates specifically to the significant differences among reported T,' values (as illustrated in Table XIII) and the various diverse interpretations of experimental DSC heat-flow curves, which have given rise to such differences. Although much of the recent debate and controversy over Tg' values has centered around the Tg' for sucrose, which has been by far the most studied sugar, the T,' value for glycerol, as reported by only a few groups (Levine and Slade, 1988b; Franks, 1985,1993a; Ablett et al., 1992a; Roos, 1992d), has nevertheless also become a bone of contention. We have long considered glycerol a unique (at least, in our experience) and perplexing case, in that its T,' value, as we have measured it [-65"C (Levine and Slade, 1988b)], cannot be made to fit on a smooth glass curve drawn between the T, for water (-135°C) and that for neat glycerol [-78" (Ablett et al., 1992a) or -93°C (Hatley and Franks, 1991; Roos, 1992d; Franks, 1993a)], because, as right-


fully pointed out by Ablett et al. (1992a), our T,’ value is higher than glycerol’s dry Tg. In contrast, the Tg’ values of -95” or -100°C are lower than the corresponding dry T, values reported by Ablett et al. (1992a) and Franks (1993a), respectively, and so can be made to fit a smooth glass curve for the glycerol-water system, as would ordinarily be required (Slade and Levine, 1991a). In the absence of any evidence to support the possibility of glycerol-water hydrate formation as an explanation for our anomalously high T,’ value, glycerol stands alone as the only low-MW carbohydrate, among hundreds we have analyzed, for which our measured T,’ value cannot be made to fit a smooth glass curve anchored by the measured dry T, value for the solute (Slade and Levine, 1988b). It would certainly be desirable to understand the reason for, and possible functional significance of, this anomalous behavior of glycerol.

As discussed in detail in Section III,C,l,b, another part of the recent Tg’-Cg’ controversy has involved the issue of the disparity between reported C,’ values (Table XIII) for many low-MW carbohydrates. There is an obvious polarization between the “rigorous” C,’ values, all of which fall around 80 w% solute, and the “apparent” C,’ values, which appear to vary with T,’ and solute MW [i.e., show a general trend of increasing C,‘ with increasing T,’ and MW (Levine and Slade, 1986, 1988b,c)] in a way that one might intuitively expect and view as consistent with practical, albeit empirical, experience. That experience begs the following questions: 1) Doesn’t every food technologist “know” that, to fabricate a frozen product (e.g., ice cream) with minimum ice content and maximum textural softness, or an IMF with minimum “A,” (RVP) [if one accepts the premise that C,’ and RVP both depend in a related fashion on plasticizer mobility (Slade and Levine, 1991a)], one would formulate with low-MW sugars (e.g.. glucose) rather than polymeric carbohydrates (e.g., maltodextrin)? (2) Doesn’t everyone accept the rule of thumb that ice content in a frozen product, or RVP of an IMF, would generally increase (and so would C,’) with increasing MW for a homologous series of solutes (Slade and Levine, 1991a)? Such deliberately prejudicial questions are simply intended to em- phasize the need for a deeper understanding [rather than the point scoring referred to by Franks (1992a)l of the reasons underlying the polarization between “rigorous” and “apparent” values of C,’.

Also in need of a deeper understanding is the relationship between “A,” and T,, where these properties of a food system represent the traditional (“water activity” concept) versus the modern approach (“food polymer science,” sometimes referred to as “glass transition theory”) to assessment of food quality, safety, and stability (Slade and Levine, 1991a). As alluded to in Section II,E, the relationship between these properties has been dealt with in considerable depth by various groups in recent years (Slade and


Levine, 1985, 1988a, 1991a; Simatos and Karel, 1988; Karel, 1989; Roos and Karel, 1990, 1991a,b, 1993; Franks, 1991a,b; Chuy and Labuza, 1992; Nelson and Labuza, 1992a,b, 1993; Nelson et al., 1992; Reid, 1992b; Roos, 1992b; Sapru and Labuza, 1992a,b; Karel et al., 1993a,b; Nelson, 1993). Worthy of special note in this regard is Roos’ novel and original treatment of experimental data for water adsorption isotherm at 25”C, water content at Tg = 25”C, and corresponding “A,,,”, all plotted together on a single graph to highlight “the critical water content and ‘‘A,.,” values . . . that depress Tg to ambient temperature” for various food materials (Roos, 1992b). Despite such useful and informative approaches, there is still a need for greater understanding that would be more directly applicable to practical food technology/product development problems (Hegenbart, 1993).

Another much-debated issue, in need of resolution through further research, concerns the breadth of applicability of WLF-type kinetics (in place of Arrhenius kinetics) to predict rates of deterioration processes in food products at T > Tg (Levine and Slade, l986,1988b,c, 1989d; Simatos et af., 1989; Slade and Levine, 1991a; Buera and Karel, 1992; Nelson and Labuza, 1992a; Peleg, 1992; Nelson, 1993). As discussed in Section III,D,l, with reference to Table XB, the recent controversy about the appropriateness of the WLF equation to describe kinetics in rubbery food systems has focused primarily on two aspects: (1) the specific values (i-e., “universal” versus nonuniversal) of the WLF coefficients, C1 and C2; and (2) the correct Tg to define as the WLF reference state. With regard to the question about the WLF coefficients, experimental results from Reid’s group (Lim and Reid, 1991, 1992; Reid et al., 1992, 1993b; Kerr et al., 1993) and Labuza’s group (Nelson and Labuza, 1992a; Sapru and Labuza, 1992a,b; Nelson, 1993), when viewed together with those from Karel and co-workers (references in Table XB), have clearly indicated that whatever the most suitable values of C1 and C2 may be, the temperature dependence of reaction rates on T - Tg, as defined by the WLF equation, evidently applies to a wide range of diffusion-limited processes in food systems. Nevertheless, a better understanding of the reasons behind case-by-case differences in C1 and C2 values would certainly be desirable.

The question about the correct Tg reference state for use in the WLF equation, when this equation is applied to reaction kinetics data for translational diffusion-limited deterioration processes in partially melted, rubbery frozen food systems at Tg’ < T < T,, has become one of the most controversial and passion-filled issues of the day. We have argued that the appropriate reference state for an ice-containing, rubbery product should be Tg‘, rather than a lower Tg corresponding to a less freeze-concentrated solute-UFW mixture (Levine and Slade, 1988c, 1989c,d; Slade and Levine,


1991a). Our reasoning [based on the conceptual framework of WLF theory (Ferry, 1980)] is that the correct glassy reference state for an ice-containing system should represent the composition (C,’) that a maximally freeze- concentrated sample would have come from on warming (from T 5 T,’ to T,’ < T < T,), and that it would, in practice, go back to on recooling (from T,’ < T < T, to T 5 T,’) at a realistic rate, as the ice-containing, partially melted (at T,’ < T < T,) sample would slowly track along the liquidus curve between the temperature limits of T,’ and maximum T,,, (e.g., see Fig. 4) during fluctuations in freezer-storage temperature. On the other side of this heated argument, Karel (in his summary notes from both the Easter School and ISOPOW-V conferences) and others (Kerr et al., 1993; Reid et al., 1993b; Simatos and Blond, 1993) have claimed that the proper reference state should be a lower T, than T,’, which would represent the more dilute solute-UFW glass (of C = C, < C,’) that could result if the partially melted, rubbery sample at T,,, > T > T,’ were quench-cooled [at an unrealistically high rate (Levine and Slade, 1989c)l to a temperature below the glass curve (i.e., T < T, at C,), but they have not presented any experimental evidence that would unequivocally support this claim. [Results reported by Kerr et al. (1993) and Reid et al. (1993b) were inconclusive on this issue, showing essentially the same excellent correlations of reaction rates with T - T,’ and with T - T, (r2 = 0.98 and 0.99, respectively).] On the contrary, as discussed in Section III,C, Chang and Randall (1992) showed convincingly that T,’, not the lower T,, is the relevant T, that governs collapse during freeze-drying. Moreover, Reid et al. (1993b) have conceded that “the empirical observation remains that T - T,’ is linearly correlated to reaction rates, whatever the event signified by Tg’.” Obviously, this controversial question needs to be addressed dispassionately and resolved by further research.

Some other unresolved issues that have been highlighted in this review include the following:

1. The usefulness of theoretical equations for predicting T, of mixtures (Section 111,AJ)

2. The functional significance of the similarities versus differences in the aqueous glass curves for high-MW food polymers (Section II1,B)

3. The underlying reason for the “schizophrenic” behavior of small sugars as plasticizers or antiplasticizers of food polymers (Section III,B,l) 4. The extent of the diffusional mobility of water (and other low-MW,

volatile or labile components) and matrix solutes in aqueous food glasses at T < T,, and of the instability that could result from diffusion-controlled deterioration reactions, were they to occur at appreciable rates in the glassy solid (Section III,C,2)


5. The reason for the difference in the decrement in modulus at Tg for aqueous versus anhydrous food polymer glasses (Section III,D,l,a)

6. The relationship and underlying significance of “stronglfragile” behavior and T,/Tg ratio of glass-forming liquids (Section 111,D,2)

7. The origin and functional significance of the higher-temperature, glasslike transition in pure amorphous fructose and galactose (Section III,D,2)

8. The role of wheat gluten in the mechanism of toughening of high- moisture baked goods during postbake heat treatments-classical thermosetting behavior of amorphous polymers? (Section 111,E).

In view of the many unanswered questions and unresolved issues highlighted in this section, it is not surprising that the subject of this review has represented a powerful force capable of drawing scientists from several different disciplines (e.g., food, pharmaceutical, biotechnology) together in recent years to confer and debate. Two previous conferences devoted to the subject of glasses and glass transitions in foods, one in 1990 at Wageningen University and the other the 1992 Easter School at Nottingham University, have already been discussed. Another such meeting, an extraordinary 2- day discussion symposium on the “Chemistry and Application Technology of Amorphous Carbohydrates,’’ conceived by Professor Felix Franks, took place in the Spring of 1995. Professor Franks described as follows the situation that prompted him to propose this symposium:

The physical properties of amorphous carbohydrates in the anhydrous state or at a low moisture content play an important role in the processing and product quality of cereal- based and various other foods and the stabilization of pharmaceutical and biotechnologi- cal products (e.g., as excipients in freeze-drying). There is an increasing awareness that, in all such applications, thermomechanical properties partly determine the choice of suitable formulations. Despite their increasing importance in food and pharmaceutical process technology, the chemistry of such amorphous sugars, perhaps containing a small amount of moisture, is substantially unexplored. Formulations and recipes are usually arrived at on a hit-or-miss basis with little basic understanding of the reasons for success or failure.

About 30 invited experts were brought together

to discuss and define the relevant problems, rank them in order of importance, and suggest effective experimental, theoretical, and computational methods for their study. The following properties and attributes are among those that require an improved understanding:

Relationships between molecular structure and glass transition [discussed above and

Chemical reactivity of solid sugars [discussed in Section III,C,2]; Chemical and biochemical reactions and kinetics in supersaturated carbohydrate mix-

in Section III,A];

tures [discussed above and in Section III,D];


Physical processes (e.g., crystallization) of, and within, amorphous carbohydrates; effects of residual moisture [discussed in Section III,D,4];

Solid solutions involving carbohydrates [Franks, 1993bl; Dynamics of small molecules (e.g., water) within amorphous carbohydrates [discussed

Suitable model systems and analytical methods [discussed above and in Section II,C]; Role of carbohydrates as bioprotectants against desiccation [Green and Angell, 1989;

Franks etnl., 1991; Roser, 1991,a,b; Colaco etal., 1992; Levine and Slade, 1992a; Carpenter et al., 1993; Crowe and Crowe, 1993; Crowe et al., 1993; Franks, 1993a,b; Roser and Colaco, 1993; Table IIID].

in Section III,C,2];

A summary report of the proceedings and recommendations of the symposium will be prepared and issued by the conveners (Professor Franks and H. Levine).

Also worthy of note here was the commencement in late 1993 of phase I1 of the ACTIF research program at Nottingham University. Following up on the work of ACTIF I, ACTIF I1 proposed “to investigate the significance of the ‘rubbery state’ in foods in relation to processing, quality and storage life” ( John Blanshard et al., personal communication, 1993).

V. CONCLUSIONS AND FUTURE PROSPECTS

The field of food science and technology has recently enjoyed a seemingly exponential growth of interest in concepts and studies centered around the glassy state phenomenon and glass transition in foods, the applications of state diagrams, WLF kinetics, and the plasticizing effect of water on Tg. This spurt of interest stems from a growing realization of the importance of glassy and rubbery states to the quality (including textural, structural, and rheological attributes), safety, and stability of foods. This state of affairs is evidenced by the fact that more than 500 of the references cited in this review deal with the glassy state phenomenon and glass transitions in food materials and systems, and more than 400 of those have been published since 1990. Results of such studies have demonstrated the major opportunity offered by the food polymer science approach to expand not only our quantitative knowledge but also, of broader practical value in the context of industrial applications, our qualitative understanding of food products and processes, with respect to (1) structure-functional property and water relationships and (2) the influences of nonequilibrium glassy and rubbery states on time-dependent mechanical and rheological properties related to quality and storage stability. In the rest of this decade, we expect to witness even greater growth and interest in this exciting area, because it offers so many challenging questions still to be answered, while promising so many opportunities for technological advancement.


ACKNOWLEDGMENTS

We thank all of the following scientists, whose preprint manuscripts, cited in this paper, were made available to us prior to their publication: S. Ablett, G. Allen, J. Amemiya, A. Angell, G. Attenburrow, J. Baust, D. Best, C. Biliaderis, H. Bizot, J. Blanshard, G. Blond, J. Brady, K. Breslauer, J. Carpenter, P. Colonna, L. De Bry, A. Donald, W. Dong, A. Eliasson, F. Franks, C. Gaines, M. Gidley, P. Given, D. Goff, N. Gontard, M. Gudmundsson, R. Hartel, L. Haynes, S. Hegenbart, M. Hemminga, C. Hoseney, V. Huang, M. Izzard, M. Kalichevsky, T. Kararli, M. Karel, J. Kokini, K. Koster, T. Labuza, J. Lawton, J. Lelievre, M. Le Meste, P. Lillford, M. Lillie, R. Ludescher, W. MacInnes, T. Maurice, P. Nesvadba, K. Nishinari, C. O’Donnell, M. Ollivon, M. Peleg, M. Pikal, D. Reid, S. Ring, Y. Roos, G. Saravacos, T. Schenz, A. Schiraldi, P. Seib, C. Seow, E. Shalaev, R. Shogren, T. Shukla, D. Simatos, A. Smith, I. Sochava, V. Tolstoguzov, I. Tomka, C. van den Berg, R. Williams, B. Wunderlich, and G. Zografi. [8/22/93]

REFERENCES

Ablett, S., and Lillford, P. J. 1991. Water in foods. Chem. Br. 27, 1024-1026. Ablett, S., Attenburrow, G. E., and Lillford, P. J. 1986. The significance of water in the baking

process. In “Chemistry and Physics of Baking” (J. M. V. Blanshard, P. J. Frazier, and T. Galliard, eds.), pp. 30-41. Royal Society of Chemistry, London.

Ablett, S., Barnes, D. J., Davies, A. P., Ingman, S. J., and Patient, D. W. 1988. C13 and pulse NMR spectroscopy of wheat proteins. J. Cereal Sci. 7, 11-20.

Ablett, S., Izzard, M. J., and Lillford, P. J. 1992a. DSC study of frozen sucrose and glycerol solutions. J. Chem. SOC., Faraday Trans. 88,789-794.

Ablett, S., Clark, A. H., Izzard, M. J., and Lillford, P. J. 1992b. Modeling of heat capacity-temperature data for sucrose-water systems. J. Chem. SOC., Faraday Trans. 88,795-802.

Ablett, S. , Izzard, M. J., Lillford, P. J., Arvanitoyannis, I., and Blanshard, J. M. V. 1992c. Calorimetric study of the glass transition occurring in fructose solutions. Carbohydr. Res.

Ablett, S., Darke, A. H., Izzard, M. J., and Lillford, P. J. 1993. Studies of the glass transition in maltooligomers. In “The Glassy State in Foods” (J . M. V. Blanshard and P. J. Lillford, eds.), pp. 189-206. Nottingham Univ. Press, Loughborough.

Aguerre, R. J., Suarez, C., and Viollaz, P. E. 1989. Swelling and pore structure in starchy materials. J. Food Eng. 9, 71-80.

Aguilera, J. M., Levi, G., and Karel, M. 1992. Effect of water content on the glass transition and caking of fish protein hydrolyzates. Biorechnol. Prog. 9,651-654 (1993).

Ahlneck, C., and Zografi, G. 1990. The molecular basis of moisture effects on the physical and chemical stability of drugs in the solid state. Int. J. Pharm. 62, 87-95.

Alfonso, G. C., and Russell, T. P. 1986. Kinetics of crystallization in semicrystalline/amorphous polymer mixtures. Macromolecules 19,1143-1152.

Allen, G. 1993. A history of the glassy state. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 1-12. Nottingham Univ. Press, Loughborough.

Amemiya, J., and Menjivar, J. A. 1992. Mechanical properties of cereal-based food cellular systems. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 207.

Andrieu, J., and Stamatopoulos, A. A. 1986. Durum wheat pasta drying kinetics. Lebensm.- Wiss. Technol. 19,448-456.

Angell, C. A. 1983. Supercooled water. Annu. Rev. Phys. Chem. 34,593-630.

246313-22 (1993).


Angell, C. A. 1988. Perspective on the glass transition. J. Phys. Chem. Solids 49, 863-871. Angell, C. A., Green, J., List, D., Qing, Z., and Senapati, H. 1987. New modes of glass

formation using negative pressure quenching and superstructuring principles. Diffus. Defect Data W54, 77-92.

Angell, C. A., Monnerie, L., and Torell, L. M. 1991. Strong and fragile behavior in liquid polymers. In “Structure, Relaxation, and Physical Aging of Glassy Polymers” (R. J. Roe and J. M. O’Reilly, eds.), pp. 3-9. Materials Research Society, Pittsburgh.

Angell, C. A., Bressel, R. D., Green, J. L., Kanno, H., Oguni, M., and Sare, E. J. 1992. Liquid fragility and the glass transition in water and aqueous solutions. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Abstr. 3.2; J. Food Eng. 22, 115-142 (1994).

Anglea, S. A., Karathanos, V., and Karel, M. 1992a. Low temperature transitions in fresh and osmotically dehydrated plant materials. Biotechnol. Prog. 9,204-209 (1993).

Anglea, S. A,, Wang, J., and Karel, M. 1992b. Structural collapse during pulsed drying of model systems and plant tissues. CoFE ’92-AIChE 1992 Summer Nat. Meet., Minneapolis, Abstr. 6d.

Anonymous, 1993a. R&D horizons. Prepared Foods 162( l), 45-46. Anonymous, 1993b. News and views-What happens to starch in the extruder? Carbohydr.

Polym. 21, 61-62. Appelqvist, I. A. M., Cooke, D., Gidley, M. J., and Lane, S. J. 1993. Thermal properties of

polysaccharides at low moisture. 1. An endothermic melting process and water-carbohydrate interactions. Carbohydr. Polym. 20,291-299.

Arvanitoyannis, I., and Blanshard, J. M. V. 1993a. Rates of crystallization of dried lactose-sucrose mixtures. J. Food Sci. 59, 197-205 (1994).

Arvanitoyannis, I., and Blanshard, J. M. V. 1993b. Physico-chemical studies on a wide composition range of low moisture glucose-fructose mixtures. I. Rates of crystallization. Int. J . Food Sci. Technol., submitted.

Arvanitoyannis, I., Blanshard, J. M. V., Ablett, S., Izzard, M. J., and Lillford, P. J. 1992. Calorimetric study of the glass transition occurring in aqueous glucose : fructose solutions. J . Sci. Food Agric. 63, 177-188 (1993).

Arvanitoyannis, I., Kalichevsky, M., Blanshard, J. M. V., and Psomiadou, E. 1993. Study of diffusion and permeation of gases in undrawn and uniaxially drawn films made from potato and rice starch conditioned at different relative humidities. Carbohydr. Polym.

Atkins, A. G. 1987. Basic principles of mechanical failure in biological systems. In “Food Structure and Behaviour” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 149-176. Academic Press, London.

Attenburrow, G. E., and Davies, A. P. 1993. The mechanical properties of cereal-based foods in and around the glassy state. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 317-331. Nottingham Univ. Press, Loughborough.

Attenburrow, G. E., Goodband, R. M., Taylor, L. J., and Lillford, P. J. 1989. Structure, mechanics, and texture of a food sponge. J . Cereal Sci. 9,61-70.

Attenburrow, G. E., Barnes, D. J., Davies, A. P., and Ingman, S. J. 1990. Rheological properties of wheat gluten. J. Cereal Sci. 12, 1-14.

Attenburrow, G. E., Davies, A. P., Goodband, R. M., and Ingman, S. J. 1992. The fracture behavior of starch and gluten in the glassy state. J. Cereal Sci. 16, 1-12.

Attwool, P. T., Leung, H. K., and Spurlock, L. L. 1992. Characterization of frozen desserts using DMA. Inst. Food Technol. Annu. Meet, New Orleans, LA, Abstr. 692.

Aubin, M., and Prud’homme, R. E. 1988. Analysis of the glass transition temperature of miscible polymer blends. Macromolecules 21,2945-2949.

24, 1-15 (1994).


Aynie, S., Le Meste, M., and Isnard, M. 1992a. Use of DMA for the study of the glass transition (Tg) in white bread. Easter Sch., Sci. Technol. Glassy State Foods, 53rd, Nottingham University, UK, Abstr. P2.

Aynie, S. , Le Meste, M., Martin, P., and Bekaert, D. 1992b. Relationship betweenviscoelasticity and textural stability of cereal foods. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Poster 23.

Baardseth, P., and Naesset, E. 1989. The effect of lipid-degrading enzyme activities on quality of blanched and unblanched frozen stored cauliflower estimated by sensory and instrumental analysis. Food Chem. 32, 39-46.

Bagley, E. B. 1992. Mechanistic basis of rheological behavior of foods. In “Physical Chemistry of Foods” (H. G. Schwartzberg and R. W. Hartel, eds.), pp. 573-593. Dekker, New York.

Bair, H. E. 1981. Thermal analysis of additives in polymers. In “Thermal Characterization of Polymeric Materials” (E. A. Turi, ed.), pp. 845-909. Academic Press, Orlando, FL.

Bair, H. E. 1985. Curing behavior of epoxy resin above and below Ts Polym. Prepr. Am. Chem. SOC., Dir. Polym. Chem. 26, 10.

Baro, M. D., Clavaguera, N., Bordas, S., Clavaguera-Mora, M. T., and Casa-Vazquez, J. 1977. Evaluation of crystallization kinetics by DTA. J. Therm. Anal. 11, 271-276.

Barrett, A. M., Kaletunc, G., Rosenberg, S. , and Breslauer, K. J. 1993. Effect of sucrose and moisture on the mechanical, thermal, and structural properties of corn extrudates. lnsr. Food Technol. Ann. Meet., Chicago, Abstr. 623.

Batzer, H., and Kreibich, U. 1981. Influence of water on thermal transitions in natural polymers and synthetic polyamides. Polym. Bull. 5, 585-590.

Belton, P. S. 1990. Can NMR give useful information about the state of water in foodstuffs? Comments Agric. Food Chem. 2,179-209.

Belton, P. S. 1991. The physical state of water in foods. In “Food Freezing: Today and Tomorrow” (W. B. Bald, ed.), pp. 1-13. Springer-Verlag, London.

BeMiller, J. N. 1989. As reported by M. A. Hill. Carbohydr. Polym. 10, 64. BeMiller, J. N. 1993. Starch-based gums. In “Industrial Gums: Polysaccharides and Their

Derivatives” (R. L. Whistler and J. N. BeMiller, eds.), pp. 579-600. Academic Press, San Diego, CA.

Berail, J., Caron, A., Mifsud, H., and Miguel, L. 1992. Study of the behavior of aqueous solutions at low temperature by differential calorimetric analysis. Calorim. Therm. Anal.

Best, D. 1992. New perspectives on water’s role in formulation. Prepared Foods 161(9), 59-66. Biliaderis, C. G. 1990. Thermal analysis of food carbohydrates. In “Thermal Analysis of Foods”

(V. R. Harwalkar and C. Y. Ma, eds.), pp. 168-220. Elsevier Applied Science, London. Biliaderis, C. G. 1991a. Non-equilibrium phase transitions of aqueous starch systems.

In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 251-273. Plenum, New York.

Biliaderis, C. G. 1991b. The structure and interactions of starch with food constituents. Can. J . Physiol. Pharmacol. 69, 60-78.

Biliaderis, C. G. 1992a. Characterization of starch networks by small strain dynamic rheometry. In “Developments in Carbohydrate Chemistry” (R. J. Alexander and H. F. Zobel, eds.), pp. 87-135. Am. Assoc. Cereal Chem., St. Paul, MN.

Biliaderis, C. G. 1992b. Structures and phase transitions of starch in food systems. Food Technol. 46(6), 98-109.

Biliaderis, C. G., and Galloway, G. 1989. Crystallization behavior of amylose-V complexes: Structure-property relationships. Carbohydr. Res. 189, 31-48.

Biliaderis, C. G., and Seneviratne, H. D. 1990a. On the supermolecular structure and metastability of glycerol monostearate-amylose complex. Carbohydr. Polym. 13, 185-206.

23,455-462.


Biliaderis, C. G., and Seneviratne, H. D. 1990b. Solute effects on the thermal stability of glycerol monostearate-amylose complex superstructures. Curbohydr. Res. 208, 199-213.

Biliaderis, C. G., and Tonogai, J. R. 1991. Influence of lipids on the thermal and mechanical properties of concentrated starch gels. J . Agric. Food Chem. 39,833-840.

Biliaderis, C. G.. and Zawistowski, J. 1990. Viscoelastic behavior of aging starch gels: Effects of concentration, temperature, and starch hydrolysates on network properties. Cereal Chem. 67,240-246.

Biliaderis, C. G., Page, C. M., Slade, L., and Sirett, R. R. 1985. Thermal behavior of amylose- lipid complexes. Curbohydr. Polym. 5,367-389.

Biliaderis, C. G., Page, C. M., Maurice, T. J., and Juliano, B. 0.1986a. Thermal characterization of rice starches: A polymeric approach to phase transitions of granular starch. J. Agric. Food Chem. 34,6-14.

Biliaderis, C. G., Page, C. M., and Maurice, T. J. 1986b. Non-equilibrium melting of amylose-V complexes. Curbohydr. Polym. 6,269-288.

Biliaderis, C. G., Page, C. M., and Maurice, T. J. 1986c. Multiple melting transitions of starch/ monoglyceride systems. Food Chem. 22, 279-295.

Billmeyer, F. W. 1984. “Textbook of Polymer Science,” 3rd ed. Wiley (Interscience), New York.

Bizot, H., Leroux, B., and Whittam, M. 1992. Hydrated starch fractions and their glass transition properties: Agreements and questions. In “Workshop on Cereal Polysaccharides.” INRA- Nantes, Le Croisic, France.

Blanshard, J. M. V. 1986. The significance of the structure and function of the starch granule in baked products. In “Chemistry and Physics of Baking” (J. M. V. Blanshard, P. J. Frazier, and T. Galliard, eds.), pp. 1-13. Royal Society of Chemistry, London.

Blanshard, J. M. V. 1987. Starch granule structure and function: Physico-chemical approach. In “Starch: Properties and Potential” (T. Galliard, ed.), pp. 16-54. Wiley, New York.

Blanshard, J. M. V. 1988. Elements of cereal product structure. In “Food Structure-Its Creation and Evaluation” (J. M. V. Blanshard and J. R. Mitchell, eds.), pp. 313-330. Butterworths, London.

Blanshard, J. M. V., and Franks, F. 1987. Ice crystallization and its control in frozen food systems. In “Food Structure and Behaviour” (J. M. V. Blanshard and P. Lillford, eds.), pp. 51-65. Academic Press, London.

Blanshard, J. M. V., and Mitchell, J. R. 1992. Starch gels. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Abstr. 2.3.

Blanshard, J. M. V., Muhr, A. H., and Gough, A. 1991. Crystallization from concentrated sucrose solutions. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 639-655. Plenum, New York.

Bloksma, A. H. 1978. Rheology and chemistry of dough. In “Wheat Chemistry and Technol- ogy” (Y. Pomeranz, ed.), 2nd ed., pp. 523-584. Am. Assoc. Cereal Chem., St. Paul, MN.

Bloksma, A. H. 1986. Rheological aspects of structural changes during baking. In “Chemistry and Physics of Baking” (J. M. V. Blanshard, P. J. Frazier, and T. Galliard, eds.), pp. 170-178. Royal Society of Chemistry, London.

Bloksma, A. H., and Bushuk, W. 1988. Rheology and chemistry of dough. In “Wheat Science and Technology,” (Y. Pomeranz, ed.), Vol. 2, 3rd ed., pp. 131-217. Am. Assoc. Cereal Chem., St. Paul, MN.

Blond, G. 1989. Water-galactose system: Supplemented state diagram and unfrozen water. Cryo-Lett. 10, 299-308.

Blond, G. 1992. The glassy state in frozen foods and its influence on mechanical properties. In “ISOPO W-V-Water in Foods: Fundamental Aspects and Their Significance in Rela- tion to Processing of Foods, Valencia, Spain,” Abstr. 5.3.


Blond, G. 1993. Mechanical properties of frozen model solutions. J. Food Eng. 22,253-269 (1994).

Blond, G., and Colas, B. 1991. Freezing in polymer-water systems. In “Food Freezing: Today and Tomorrow” (W. G. Bald, ed.), pp. 27-43. Springer-Verlag, London.

Blond, G., and Simatos, D. 1991. Glass transition of the amorphous phase in frozen aqueous systems. Thermochim. Acta 175, 239-247.

Blum, F. D., and Nagara, B. 1986. Solvent mobility in gels of atactic polystyrene. Polym. Prepr., Am. Chem. Soc., Div. Polym. Chem. 27(1), 211-212.

Bolton, K. J., Dodd, C. E. R., Gould, G. W., and Waites, W. M. 1992. The microbiology of glassy states. Easter Sch. Sci. Technol. Glassy State Foods, 53rd, Nottingham University, UK.

Bone, S., and Pethig, R. 1982. Dielectric studies of the binding of water to lysozyme. J. Mol. Biol. 157,571-575.

Bone, S. , and Pethig, R. 1985. Dielectric studies of protein hydration and hydration-induced flexibility. J. Mol. Biol. 181, 323-326.

Borchard, W., Bremer, W., and Keese, A. 1980. State diagram of the water-gelatin system. Colloid Polym. Sci. 258,516-526.

Boskovic, M. A., Vidal, S. M., and Saleeb, F. Z. 1992. Spray-dried fixed flavorants in a carbohydrate substrate and process. U.S. Pat. 5,124,162.

Botham, R. L., Noel, T. R., Ring, S. G., and Whittam, M. A. 1992. Structural relaxations of cereal biopolymers. Am. Assoc. Cereal Chem., 77rh Ann. Meet., Minneapolis, Abstr. 62.

Boyer, R. F., Baer, E., and Hiltner, A. 1985. Concerning gelation effects in atactic polystyrene solutions. Macromolecules 18,427-434.

Brady, J. W., and Ha, S. N. 1991. Molecular dynamics simulation of the aqueous solvation of sugars. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 739-751. Plenum, New York.

Bruin, S. 1992. Flavor retention in dehydration processes. In “Advances in Food Engineering” (R. P. Singh and M. A. Wirakartakusumah, eds.), pp. 15-32. CRC Press, Boca Raton, FL.

Bruni, F., and Leopold, A. C. 1991. Glass transitions in soybean seed. Plant Physiol. 96,

Bruni, F., and Leopold, A. C. 1992. Pools of water in anhydrobiotic organisms. Biophys. J .

Brydson, J. A. 1972. The glass transition, melting point and structure. In “Polymer Science” (A. D. Jenkins, ed.), pp. 194-249. North-Holland Publ., Amsterdam.

Buchanan, D. R., and Walters, J. P. 1977. Glass-transition temperatures of polyamide textile fibers. I. Effect of water. Text. Res. J. 47, 398-406.

Buera, M. P., and Karel, M. 1992. Application of the WLF equation to describe the combined effects of moisture and temperature on non-enzymatic browning rates in food systems. J. Food Process. Preserv. 17,31-45 (1993).

Buera, M. P., Levi, G., and Karel, M. 1992. Glass transition in poly(vinylpyrro1idone): Effect of molecular weight and diluents. Biotechnol. Prog. 8, 144-148.

Burchard, W. 1988. Entanglement and reversible gelation for polymers of different architec- tures. Prog. Colloid Polym. Sci. 78,63-67.

Burke, M. J. 1986. The glassy state and survival of anhydrous biological systems. In “Mem- branes, Metabolism, and Dry Organisms” (A. C. Leopold, ed.), pp. 358-363. Cornell Univ. Press, Ithaca, NY.

Burros, B. C., Young, L. A., and Carroad, P. A. 1987. Kinetics of corn meal gelatinization at high temperature and low moisture. J. Food Sci. 52, 1372-1380.

Cairault, A., Bizot, H., Durning, C. J., and Roques, M. 1989. Approaching viscoelastic diffusion in polyacrylamide gels. In “Drying ’89” (A. S. Mujumdar, ed.). Hemisphere, New York.

Cairns, P., I’Anson, K. J., and Morns, V. J. 1991a. The effect of added sugars on the retrogradation of wheat starch gels by X-ray diffraction. Food Hydrocolloids 5, 151-153.

660-663.

63,663-672.


Cairns, P., Miles, M. J., and Morris, V. J. 1991b. Studies of the effect of the sugars ribose, xylose, and fructose on the retrogradation of wheat starch gels by X-ray diffraction. Carbohydr. Polym. 16,355-365.

Cakebread, S. H. 1969. Factors affecting the shelf life of high boilings. Manuf Confect.

Caldwell, K. B., Goff, H. D., and Maurice, T. J. 1990. The use of thermal mechanical analysis to determine the influence of carbohydrates on stability of frozen dairy products. J. Dairy Sci. 73, Suppl. 1, 95.

Caldwell, K. B., Goff, H. D., and Stanley, D. W. 1992. A low-temperature SEM study of ice cream. 11. Influence of selected ingredients and processes. Food Struct. 11, 11-23.

Cameron, R. E., and Donald, A. M. 1991. Small-angle X-ray scattering and DSC from starch and retrograded starch. In “Food Polymers, Gels and Colloids” (E. Dickinson, ed.), pp. 301-309. Royal Society Chemistry, Cambridge.

Campanella, 0. H., Popplewell, L. M., Rosenau, J. R., and Peleg, M. 1987. Elongational viscosity measurements of melting american process cheese. J. Food Sci. 52,1249-1251.

Cantor, C. R., and Schimmel, P. R. 1980. “Biophysical Chemistry: Part I. The Conformation of Biological Macromolecules.” Freeman, San Francisco.

Carpenter, J. F., Prestrelski, S. J., Anchordoguy, T. J., and Arakawa, T. 1993. Interactions of stabilizers with proteins during freezing and drying. In “Protein Formulations and Deliv- ery” (J. L. Cleland and R. S. Langer, eds.), ACS Symp. Ser. ACS Books, Washington, DC.

Cassel, B., and Twombly, B. 1991. Glass transition determination by TMA, a DMA, and a DSC. In “Materials Characterization by Thermomechanical Analysis” (A. T. Riga and C. M. Neag, eds.), ASTM STP 1136, pp. 108-119. ASTM, Philadelphia.

Chan, R. K., Pathmanathan, K., and Johari, G. P. 1986. Dielectric relaxations in the liquid and glassy states of glucose and its water mixtures. J . Phys. Chem. 90, 6358-6362.

Chang, B. S., and Randall, C. S. 1992. Use of subambient thermal analysis to optimize protein lyophilization. Cryobiology 29, 632-656.

Chang, S. M., and Liu, L. C. 1991. Retrogradation of rice starches studied by DSC and influence of sugars, NaC1, and lipids. 1. Food Sci. 56,564-566.

Chang, 2. H., and Baust, J. G. 1991a. Physical aging of glassy state: DSC study of vitrified glycerol systems. Cryobiology 28,87-95.

Chang, 2. H., and Baust, J. G. 1991b. Further inquiry into the cryobehavior of aqueous solutions of glycerol. Cryobiology 28,268-278.

Chang, Z . H., and Baust, J. G. 1991c. Physical aging of the glassy state: Sub-T, ice nucleation in aqueous sorbitol systems. J. Non-Cryst. Solids. 130, 198-203.

Cheng, S. Z. D. 1989. Thermal characterization of macromolecules. J. Appl. Polym. Sci.: Appl. Polym. Symp. 43,315-371.

Cheng, S. Z. D., and Wunderlich, B. 1987. Glass transition and melting behavior of poly(oxy- 2,6-dimethyl-1,4-phenylene). Macromolecules 20, 1630-1637.

Cherian, G., and Chinachoti, P. 1993. *H and I7O NMR of water mobility in glassy and rubbery gluten. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 358.

Chin, Y. H., Zhang, C., Wang, P., Inglefield, P. T., Jones, A. A., Kambour, R. P., Bendler, J. T., and White, D. M. 1992. Glass transition dynamics in a compatible blend by two- dimensional solid-state NMR. Macromolecules 25, 3031-3038.

Chinachoti, P. 1992a. Water mobility and its relation to functionality of sucrose-containing food systems. Inst. Food Technol. Annu. Meet., New Orleans, Abstr. 499; Food Technol.

Chinachoti, P. 1992b. Bread staling study on mechanical properties, starch recrystallization, 17-0 NMR water mobility, and thermal DMA and DSC analysis. In “ISOPOW-V-

49,41-44.

47(1), 134-140 (1993).


Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods” Valencia, Spain, Poster 9.

Chinachoti, P., White, V. A., Lo, L., and Stengle, T. R. 1991a. Application of high-resolution C13, 017, and Na23 NMR to study the influences of water, sugar, and NaCl on starch gelatinization. Cereal Chem. 68,238-244.

Chinachoti, P., Kim-Shin, M. S., Mari, F., and Lo, L. 1991b. Gelatinization of wheat starch in the presence of sucrose and NaCI: Correlation between gelatinization temperature and water mobility as determined by 017 NMR. Cereal Chem. 68,245-248.

Chungcharoen, A., and Lund, D. B. 1987. Influence of solutes and water on rice starch gelatinization. Cereal Chem. 64,240-243.

Chuy, L., and Labuza, T. P. 1992. Caking and stickiness of dairy-based food powders as related to glass transition. J . Food Sci. 59, 43-46 (1994).

Clark, A. H., Gidley, M. J., Richardson, R. K., and Ross-Murphy, S. B. 1989. Rheological studies of aqueous amylose gels: The effect of chain length and concentration on gel modulus. Macromolecules 22,346-351.

Cocero, A. M., and Kokini, J. L. 1991. The study of the glass transition of glutenin using small amplitude oscillatory rheological measurements and DSC. J. Rheol. 35, 257-270.

Cocero, A. M., and Kokini, J. L. 1992. Prediction of the temperature dependence of the apparent shear viscosity of 40% moisture glutenin using Arrhenius and WLF equations. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 60.

Cocero, A. M., and Kokini, J. L. 1993. Development of a preliminary glutenin state diagram. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 34.

Cocero, A. M., Madeka, H., and Kokini, J. L. 1992a. Simulation of the rheological properties of cereal proteins. In “Theoretical and Applied Rheology” (P. Moldenaers and R. Keunings, eds.), pp. 699-701. Elsevier, Amsterdam.

Cocero, A. M., Madeka, H., and Kokini, J. L. 1992b. The role of the glassy state in developing phase diagrams to predict rheological properties of cereal biopolymers. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 63.

Cocero, A. M., Madeka, H., and Kokini, J. L. 1993. Development of state diagrams for zein, gliadin, and glutenin. Inr. Congr. Eng. Food, 6rh, Chiba, Japan, Abstr. 1.02.

Colaco, C., Sen, S., Thangavelu, M., Pinder, S., and Roser, B. 1992. Extraordinary stability of enzymes dried in trehalose: Simplified molecular biology. Bio/Technology 10,1007-1011.

Cole, B. A., Levine, H., McGuire, M. T., Nelson, K. J., and Slade, L. 1983. Soft, frozen dessert formulation. U.S. Pat. 4,374,154.

Cole, B. A., Levine, H., McGuire, M. T., Nelson, K. J., and Slade, L. 1984. Soft, frozen dessert formulation. U.S. Pat. 4,452,824.

Colonna, P., and Buleon, A. 1992. New insights on starch structure and properties. In “Cereal Chemistry and Technology: A Long Past and a Bright Future,” 9th Int. Cereal Bread Congr., Paris.

Colonna, P., Tayeb, J., and Mercier, C. 1989. Extrusion cooking of starch and starchy products. In “Extrusion Cooking” (C. Mercier, P. Linko, and J. M. Harper, eds.), pp. 247-319. Am. Assoc. Cereal Chem., St. Paul, MN.

Compan, V., Lopez, M. L., Monferrer, L., Garrido, J., Riande, E., and San Roman, J. 1993. Determination of the Tg of poly(cyclohexy1 acrylate) from oxygen permeability measurements. Polymer 34,2971-2914.

Couchman, P. R., and Karasz, F. E. 1978. A classical thermodynamic discussion of the effect of composition on glass-transition temperatures. Macromolecules 11, 117-119.

Cowie, J. M. G . 1973. “Polymers: Chemistry and Physics of Modern Materials.” Intertext, New York.

Crowe, J. H., and Crowe, L. M. 1993. Evidence for direct interaction between disaccharides and dry phospholipids. Cryobiology 30,226-227.


Crowe, J. H., Crowe, L. M., and Carpenter, J. F. 1993. Preserving dry biomaterials: The water replacement hypothesis. BioPharm 6(3), 28-37; 6(4), 40-43.

Davies, A. P. 1986. Protein functionality in bakery products. In “Chemistry and Physics of Baking” (J. M. V. Blanshard, P. J. Frazier, and T. Galliard, eds.), pp. 89-104. Royal Society of Chemistry, London.

Davies, A. P., Ingman, S. J., and Attenburrow, G. 1991. Plasticization and mechanical properties of heat-set wheat gluten. In “Gluten Proteins 1990” (W. Bushuk and R. Tkachuk, eds.), pp. 21-28. Am. Assoc. Cereal Chem., St. Paul, MN.

Davis, E. A. 1991. The “Nobel” science of foods. Cereal Foods World 36, 1030. Davis, E. A. 1992a. Updates on temperature sensing and thermal analysis techniques. Cereal

Davis, E. A. 1992b. Some advances during 1992. Cereal Foods World 37,726. De Bry, L. 1993. Anthropology of the Maillard reaction. In “5th International Symposium

on Maillard Browning” Minneapolis, MN, Aug. 29-Sept. 1. de Graaf, E. M., Madeka, H., Cocero, A. M., and Kokini, J. L. 1992. Determination of the

effect of moisture on gliadin glass transition using mechanical spectrometry and DSC. Biotechnol. Prog. 9,210-213 (1993).

Dereuddre, J., Hassen, N., Blandin, S., and Kaminski, M. 1991. Resistance of alginate-coated somatic embryos of carrot to desiccation and freezing in liquid nitrogen. 2. Thermal analysis. Cryo-Lett. 12, 135-148.

Foods World 37, 395-396.

Dixon, M., and Webb, E. C., eds. 1964. “Enzymes.” Academic Press, New York. Doescher, L. C., Hoseney, R. C., and Milliken, G. A. 1987. Mechanism for cookie dough

setting. Cereal Chem. 61, 158-163. Domszy, R. C., Alamo, R., Edwards, C. O., and Mandelkern, L. 1986. Thermoreversible

gelation and crystallization of homopolymers and copolymers. Macromolecules 19,

Donald, A. M., Warburton, S. C., and Smith, A. C. 1993. Physical changes consequent on the extrusion of starch. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 375-393. Nottingham Univ. Press, Loughborough.

Dong, W. 1992. Measurement of certain rheological properties and transformation in wheat flour. M.S. Thesis, Kansas State University, Manhattan.

Donhowe, D. P., Hartel, R. W., and Bradley, R. L. 1991. Determination of ice crystal size distributions in frozen desserts. J . Dairy Sci. 74, 3334-3344.

Doublier, J. L. 1989. Rheological properties of cereal carbohydrates. In “Dough Rheology and Baked Product Texture” (H. Faridi and J. M. Faubion, eds.), pp. 111-155. Van Nostrand-Reinhold/AVI, New York.

Downton, G. E., Flores-Luna, J. L., and King, C. J. 1982. Mechanism of stickiness in hygroscopic, amorphous powders. Ind. Eng. Chem. Fundam. 21,447-451.

Dreese, P. C., Faubion, J. M., and Hoseney, R. C. 1988a. Effect of different heating and washing procedures on dynamic rheological properties of wheat gluten. Cereal Foods World 33, 225-228.

Dreese, P. C., Faubion, J. M., and Hoseney, R. C. 1988b. Dynamic rheological properties of flour, gluten, and gluten-starch doughs. Cereal Chem. 65, 348-353.

Duckworth, R. B. 1988. Foreword. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), p. ix. Elsevier Applied Science, London.

du Cros, D. L. 1987. Glutenin proteins and gluten strength in durum wheat.J. Cereal Sci. 5,3-12. Duden, R., and Scholz, A. 1982. Enzymatic decomposition of phospholipids in frozen parsley.

Z . Ernaehrungswiss. 21, 266-271. Edwards, S. F., Lillford, P. J., and Blanshard, J. M. V. 1987. Gels and networks in practice

and theory. In “Food Structure and Behaviour” (J. M. V. Blanshard and P. Lillford, eds.), pp. 1-12. Academic Press, London.

310-325.


Eerlingen, R. C., and Delcour, J. A. 1992. Glastransitie in voedingspolymeersystemen: De zetmeelverstijfseling. Ingenieursblad 61(9), 50-54.

Eerlingen, R. C., Crombez, M., and Delcour, J. A. 1993. Enzyme-resistant starch. I. Quantita- tive and qualitative influence of incubation time and temperature of autoclaved starch on resistant starch fraction. Cereal Chem. 70, 339-344.

Ehlich, D., and Sillescu, H. 1990. Tracer diffusion at the glass transition. Macromolecules

Eisenberg, A. 1984. The glassy state and the glass transition. In “Physical Properties of Polymers” (J. E. Mark, A. Eisenberg, W. W. Graessley, L. Mandelkern, and J. L. Koenig, eds.), pp. 55-95. Am. Chem. SOC., Washington, DC.

Eliasson, A X . 1989. Rheological properties of cereal proteins. In “Dough Rheology and Baked Product Texture” (H. Faridi and J. M. Faubion, eds.), pp. 67-110. Van Nostrand-ReinholdlAVI, New York.

Eliasson, A.-C. 1992. A calorimetric investigation of the influence of sucrose on the gelatinization of starch. Carbohydr. Polym. 18,131-138.

Eliasson, A.-C., and Larsson, K. 1993. “Cereals in Breadmaking.” Dekker, New York. Eliasson, A X . , and Silverio, J. 1991. Interfacial behavior of gluten proteins after heat treat-

ment. In “Gluten Proteins 1990” (W. Bushuk and R. Tkachuk, eds.), pp. 11-19. Am. Assoc. Cereal Chem., St. Paul, MN.

23,1600-1610.

Elliott, S. R. 1990. “Physics of Amorphous Materials,” 2nd ed. Wiley, New York. Ellis, T. S. 1988. Moisture-induced plasticization of amorphous pol yamides and their blends.

J . Appl. Polym. Sci. 36,451-466. Ellis, T. S., Jin, X., and Karasz, F. E. 1984. The water-induced plasticization behavior of semi-

crystalline polyamides. Polym. Prepr. Am. Chem. SOC., Div. Polym. Chem. 25(2), 197-198. Ewart, J. A. D. 1988. Thiols in flour and breadmaking quality. Food Chem. 28,207-218. Faubion, J. M., and Hoseney, R. C. 1989. The viscoelastic properties of wheat flour doughs.

In “Dough Rheology and Baked Product Texture” (H. Faridi and J. M. Faubion, eds.), pp. 29-66. Van Nostrand-Reinhold/AVI, New York.

Ferry, J. D. 1980. “Viscoelastic Properties of Polymers,” 3rd ed. Wiley, New York. Finegold, L., Franks, F., and Hatley, R. H. M. 1989. Glasdrubber transitions and heat capacities

of binary sugar blends. J. Chem. Soc., Faraday Trans. I 85,2945-2951. Finley, J. W., Verduin, P., Arciszewski, H. E., and Biggs, R. H. 1992. Cookies with reduced

sucrose content and doughs for production thereof. US. Pat. 5,080,919. Finney, J. L., and Poole, P. L. 1984. Protein hydration and enzyme activity: Role of hydration-

induced conformation and dynamic changes in activity of lysozyme. Commenfs Mol. Cell. Biophys. 2, 129-151.

Flink, J. M. 1983. Structure and structure transitions in dried carbohydrate materials. In “Physical Properties of Foods” (M. Peleg and E. B. Bagley, eds.), pp. 473-521. AVI, Westport, CT.

Flory, P. J. 1953. “Principles of Polymer Chemistry.” Cornell Univ. Press, Ithaca, NY. Flory, P. J. 1974. Introductory lecture-Gels and gelling processes. Faraday Discuss. Chem.

Ford, A. W., and Dawson, P. J. 1993. The effect of carbohydrate additives in the freeze- drying of alkaline phosphatase. J. Pharm. Pharmacol. 45, 86-93.

Franks, F. 1982. The properties of aqueous solutions at subzero temperatures. In “Water: A Comprehensive Treatise” (F. Franks, ed.), Vol. 7, pp. 215-338. Plenum, New York.

Franks, F. 1983a. Bound water: Fact and fiction. Cryo-Left. 4, 73-74. Franks, F. 1983b. Solute-water interactions: Do polyhydroxy compounds alter the properties

SOC. 57,7-18.

of water? Cryobiology 20,335-345.


Franks, F. 1985. Complex aqueous systems at subzero temperatures. In “Properties of Water in Foods” (D. Simatos and J. L. Multon, eds.), pp. 497-509. Martinus Nijhoff, Dordrecht, The Netherlands.

Franks, F. 1986a. Metastable water at subzero temperatures. J . Microsc. (Oxford) 141,243-249. Franks, F. 1986b. Unfrozen water: Yes; unfreezable water: Hardly; bound water: Certainly

Franks, F. 1988. Protein hydration. In “Characterization of Proteins” (F. Franks, ed.), pp.

Franks, F. 1989. Improved freeze-drying: An analysis of the basic scientific principles. Process

Franks, F. 1990. Freeze drying: From empiricism to predictability. Cryo-Lett. 11, 93-1 10. Franks, F. 1991a. Hydration phenomena: An update and implications for the food processing

industry. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 1-19. Plenum, New York.

Franks, F. 1991b. Water activity: A credible measure of food safety and quality? Trends Food Sci. Technol. 2(3), 68-72.

Franks, F. 1991c. Freeze-drying: From empiricism to predictability. The significance of glass transitions. Dev. Biol. Stand. 74, 9-19.

Franks, F. 1992a. The importance of being glassy. Cryo-Lett. 13, 349-350. Franks, F. 1992b. Freeze-drying: A combination of physics, chemistry, engineering and eco-

nomics. Jpn. J. Freezing Drying 38, 5-16. Franks, F. 1993a. Product stability during drying. Biopreserv. Symp., BioPharm Conc, Cam-

bridge, MA. Franks, F. 1993b. Solid aqueous solutions. IUPAC Int. Con$ Solution Chem., 23rd, Leicester

University, U K Pure Appl. Chem. 65,2527-2537. Franks, F., and Grigera, J. R. 1990. Solution properties of low molecular weight polyhydroxy

compounds. In “Water Science Reviews” (F. Franks, ed.), Vol. 5, pp. 187-289. Cambridge Univ. Press, Cambridge, UK.

Franks, F., and Hatley, R. H. M. 1990. Water-soluble or water-swellable compositions for storage of enzymes and other biological materials. Eur. Pat. Appl. EP 383,569.

Franks, F., and Hatley, R. H. M. 1992. Storage of materials. US . Pat. 5,098,893. Franks, F., and Hatley, R. H. M. 1993. Stable enzymes by water removal. In “Stability and

Stabilization of Enzymes” (W. J. J. van den Tweel, A. Harder, and R. M. Buitelaar, eds.), pp. 45-54. Elsevier, Amsterdam.

Franks, F., and van den Berg, C. 1991. A rational approach to the drying of labile biologicals: The role of salts and stabilizers. In “Topics in Pharmaceutical Sciences” (D. J. A. Crom- melin and K. K. Midha, eds.), pp. 233-240. Medpharm Scientific, Stuttgart.

Franks, F., Asquith, M. H., Hammond, C. C., Skaer, H. B., and Echlin, P. 1977. Polymeric cryoprotectants in the preservation of biological ultrastructure. I. J. Microsc. (Oxford)

Franks, F., Hatley, R. H. M., and Mathias, S. F. 1991. Materials science and the production of shelf-stable biologicals. BioPharm 4(9), 38-55.

Fried, J. R., Lai, S. Y., Kleiner, L. W., and Wheeler, M. E. 1982. Experimental assessment of the thermodynamic theory of the compositional variation of Tg: PVC systems. J. Appl. Polym. Sci. 27,2869-2883.

Fujio, Y., and Lim, J. K. 1989. Correlation between the glass-transition point and color change of heat-treated gluten. Cereal Chem. 66, 268-270.

Fujio, Y., Hayashi, N., and Hayakawa, I. 1991. Effect of moisture content on flow behavior of molten soy-protein isolate under an elevated temperature. Int. J . Food Sci. Technol.

not. Cryo-Lett. 7, 207.

127-154. Humana Press, Clifton, NJ.

Biochem. 24(1), R3-R7.

110,223-238.

26945-51.


Furuta, T., Tsujimoto, S., Makino, H., Okazaki, M., and Toei, R. 1984. Measurement of diffusion coefficients of water and ethanol in aqueous maltodextrin solution. J. Food Eng. 3, 169-186.

Fuzek, J. F. 1980. Glass transition temperature of wet fibers: Its measurement and significance. ACS Symp. Ser. 127,515-530.

Gaeta, S., Apicella, A., and Hopfenberg, H. B. 1982. Kinetics and equilibria associated with the absorption and desorption of water and lithium chloride in an ethylene-vinyl alcohol copolymer. J . Membr. Sci. 12, 195-205.

Gaines, C. S. 1991. Instrumental measurement of the hardness of cookies and crackers. Cereal Foods World 36,989-996.

Gaines, C. S. 1993. Collaborative studies on the baking quality of cookie flour by wire-cut type formulations (AACC Methods 10-53 and 10-54). Cereal Foods World 38, 26-30.

Gaines, C. S., Kassuba, A,, and Finney, P. L. 1992a. Instrumental measurement of cookie hardness. I. Assessment of methods. Cereal Chem. 69, 115-119.

Gaines, C. S., Kassuba, A., Finney, P. L., and Donelson, J. R. 1992b. Instrumental measurement of cookie hardness. 11. Application to product quality variables. Cereal Chem. 69,120-125.

Gaines, C. S., Kassuba, A., and Finney, P. L. 1992c. Impact of baking company and flour quality on commercial cookie and cracker hardness. Cereal Chem., submitted.

Gaines, C. S., Kassuba, A., Finney, P. L., and Donelson, J. R. 1993a. Instrumental measurement of cookie hardness. 111. Comparison of wire-cut and sugar-snap cookie formulations. Cereal Chem., submitted.

Gaines, C. S., Kassuba, A., and Finney, P. L. 1993b. Instrumental measurement of cookie hardness. IV. Miag Multomat mill stream selection. Cereal Chem., submitted.

Garbow, J. R., and Schaefer, J. 1991. Magic-angle 13C NMR study of wheat flours and doughs. J. Agric. Food Chem. 39,877-880.

George, R. M. 1993. Freezing processes used in the food industry. Trends Food Sci. Technol.

German, M. L., Blumenfeld, A. L., Guenin, Y. V., Yuryev, V. P., and Tolstoguzov, V. B. 1992. Structure formation in systems containing amylose, amylopectin, and their mixtures. Carbohydr. Polym. 18,27-34.

Ghiasi, K., and Skarra, L. L. 1990. Taco shell and method of manufacture. U.S. Pat. 4,950,490. Gidley, M. J. 1989. Molecular mechanisms underlying amylose aggregation and gelation.

Macromolecules 22,351-358. Gidley, M. J., and Bulpin, P. V. 1989. Aggregation of amylose in aqueous systems: The effect

of chain length on phase behavior and aggregation kinetics. Macromolecules 22,341-346. Gidley, M. J., and Cooke, D. 1991. Aspects of molecular organization and ultrastructure in

starch granules. Biochem. SOC. Trans. 19, 551-555. Gidley, M. J., Cooke, D., and Ward-Smith, S. 1993. Low moisture polysaccharide systems:

Thermal and spectroscopic aspects. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 303-316. Nottingham Univ. Press, Loughborough.

Given, P. S. 1991. Molecular behavior of water in a flour-water baked model system. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 465-483. Plenum, New York.

Goff, H. D. 1992. Low-temperature stability and the glassy state in frozen foods. Food Res. In?. 25, 317-325.

Goff, H. D., Caldwell, K. B., Stanley, D. W., and Maurice, T. J. 1993. The influence of polysaccharides on the glass transition in frozen sucrose solutions and ice cream. J. Dairy Sci. 76,1268-1277.

Gontard, N., Guilbert, S., and Cuq, J. L. 1992a. Edible wheat gluten films: Influence of the main process and environmental conditions on thermal, mechanical, and barrier proper-

4, 134-138.


ties. 204th Am. Chem. SOC. Natl. Meet., Washington, DC, Div. Agric. Food Chem., Ab- str. 217.

Gontard, N., Guilbert, S., and Cuq, J. L. 1992b. Mechanical and thermal properties of an edible wheat gluten film plasticized with water. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods” Valencia, Spain, Poster 20.

Gontard, N., Guilbert, S., and Cuq, J. L. 1992c. Water and glycerol as plasticizers affect mechanical and water vapor barrier properties of an edible wheat gluten film. J. Food Sci. 58,206-211 (1993).

Gordon, M., and Taylor, J. S. 1952. Ideal copolymers and the second-order transitions of synthetic rubbers. I. Non-crystalline copolymers. J. Appl. Chem. 2, 493-500.

Gosline, J. M. 1987. Structure and mechanical properties of rubber-like proteins in animals. Rubber Chem. Technol. 60,417-438.

Gould, G. W., and Christian, J. H. B. 1988. Characterization of the state of water in foods- Biological aspects. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 43-56. Elsevier Applied Science, London.

Graessley, W. W. 1984. Viscoelasticity and flow in polymer melts and concentrated solutions. In “Physical Properties of Polymers” (J. E. Mark, A. Eisenberg, W. W. Graessley, L. Mandelkern, and J. L. Koenig, eds.), pp. 97-153. Am. Chem. SOC., Washington, DC.

Graf, E., and Saguy, I. S. 1991. R&D process. In “Food Product Development” (E. Graf and I. S. Saguy, eds.), pp. 49-89. Van Nostrand-Reinhold, New York.

Gray, F. M. 1991. “Solid Polymer Electrolytes,” pp. 149-156. VCH, New York. Green, J. L., and Angell, C. A. 1989. Phase relations and vitrification in saccharide-water

solutions and the trehalose anomaly. J. Phys. Chem. 93,2880-2882. Gudmundsson, M. 1992. Cereal starches: Physicochemical properties and retrogradation. Doc-

toral Thesis, Lund University, Sweden. Gudmundsson, M., and Eliasson, A.-C. 1992. Comparison of thermal and viscoelastic proper-

ties of four waxy starches and the effect of added surfactant. Staerke 44, 379-385. Guilbot, A., and Godon, B. 1984. Le pain rassis. Cah. Nutr. Diet 19, 171-181. Hallberg, L. M., and Chinachoti, P. 1992. DMA for glass transitions in long shelf-life bread.

Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 694 J . Food Sci. 57,1201-1204. Hancock, B. C., and Zografi, G. 1993. The use of solution theories for predicting water vapor

absorption by amorphous pharmaceutical solids: A test of the Flory-Huggins and Vrentas models. Pharm. Res. 10,1262-1267.

Hanover, L. M. 1992. Crystalline fructose: Production, properties, and applications. I n “Starch Hydrolysis Products” (F. W. Schenck and R. E. Hebeda, eds.), pp. 201-231. VCH, New York.

Hardman, T. M. 1986. Interaction of water with food components. In “Interactions of Food Components” (G. G. Birch and M. G. Lindley, eds.), pp. 19-30. Elsevier Applied Sci- ence, London.

Hardy, G., Cser, F., and Takacs, E. 1973. Investigation of the binary system of a crystalline monomer and its polymer. J. Polym. Sci., Polym. Symp. 42, 663-669.

Harrison, S. T. L., Chase, H. A., Amor, S. R., Bonthrone, K. M., and Sanders, J. K. M. 1992. Plasticization of poly(hydr0xybutyrate) in vivo. Int. J. Biol. Macromol. 14, 50-56.

Hartel, R. W. 1992. Solid-liquid equilibrium: Crystallization in foods. In “Physical Chemistry of Foods” (H. G. Schwartzberg and R. W. Hartel, eds.), pp. 47-81. Dekker, New York.

Hartel, R. W. 1993. Controlling sugar crystallization in food products. Food Technol. 47(11), 99-107.

Hartel, R. W., and Shastry, A. V. 1991. Sugar crystallization in food products. CRC Crit. Rev. Food Sci. Nutr. 30,49-112.


Hanvalkar, V. R., and Ma, C. Y., eds. 1990. “Thermal Analysis of Foods.” Elsevier Applied Science, London.

Hatley, R. H. M. 1991. The effective use of DSC in the optimization of freeze-drying processes and formulations. Dev. Biol. Stand. 74, 105-122.

Hatley, R. H. M., and Franks, F. 1991. Applications of DSC in the development of improved freeze-drying processes for labile biologicals. J . Therm. Anal. 37, 1905-1914.

Hatley, R. H. M., and Mant, A. 1993. Determination of the unfrozen water content of maximally freeze-concentrated carbohydrate solutions. Int. J. Biol. Macromol. 15,227-232.

Hatley, R. H. M., Franks, F., and Green, M. 1989. A novel data acquisition, retention and examination system (DARES) for DSC. Thermochim. Acta 156,247-257.

Hatley, R. H. M., van den Berg, C., and Franks, F. 1991. The unfrozen water content of maximally freeze-concentrated carbohydrate solutions: Validity of the methods used for its determination. Cryo-Lett. 12, 113-124.

Haward, R. N. 1973. The nature of polymer glasses, their packing density, and mechanical behavior. In “The Physics of Glassy Polymers” (R. N. Haward, ed.), pp. 1-51. Applied Science, London.

Haynes, L. C., and Locke, J. P. 1993. Measurement of dielectric relaxation time to characterize plasticizer content in a food polymer. 28th Annu. Int. Microwave Power Inst. Symp., Montreal.

Haynes, L. C., Locke, J. P., and Spence, L. A. 1993. Dielectric measurements of water in food polymers. Am. Chem. SOC. 27th MARM, Hofstra University, New York, Abstr. 84.

Hegenbart, S. 1991. In control: Taming moisture with gums and starches. Food Prod. Des.

Hegenbart, S . 1993. Exploring dimensions in intermediate moisture foods. Food Prod. Des.

Hemminga, M. A., Roozen, M. J. G. W., and Walstra, P. 1993. Molecular motions and the glassy state. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 157-171. Nottingham Univ. Press, Loughborough.

Herrington, T. M., and Branfield, A. C. 1984. Physico-chemical studies on sugar glasses. J. Food Technol. 19,409-435.

Hill, A. S., Skerritt, J. H., and Watson, A. R. 1988. Simple and rapid test method for gluten in starches, bakery raw materials and processed foods. Am. Assoc. Cereal Chem., 73rd Annu. Meet., San Diego, Abstr. 66.

Hiltner, A., and Baer, E. 1986. Reversible gelation of macromolecular systems. Polym. Prepr., Am. Chem. SOC., Div. Polym. Chem. 27(1), 207.

Hirsh, A. G. 1987. Vitrification in plants as a natural form of cryoprotection. Cryobiology

Hoeve, C. A. J. 1980. The structure of water in polymers. ACS Symp. Ser. 127,135-146. Hoeve, C. A. J., and Hoeve, M. B. J. A. 1978. The glass point of elastin as a function of

diluent concentration. Org. Coat. Plast. Chem. 39,441-443. Hoeve, C. A. J., and Hoeve, M. B. J. A. 1980. The glass point of elastin as a function of

diluent concentration. Polym. Eng. Sci. 20,290-293. Hofer, K., Hallbrucker, A., Mayer, E., and Johari, G. P. 1989. Vitrified dilute aqueous solutions.

3. Plasticization of water’s H-bonded network and the glass transition temperature’s minimum. J. Phys. Chem. 93,4674-4677.

Hosea, T. J. C., Ng, S. C., and Oates, C. G. 1990. A Brillouin scattering study of the glass transition in sucrose. Food Hydrocolloids 4,137-147.

Hoseney, R. C. 1986. “Principles of Cereal Science and Technology.” Am. Assoc. Cereal Chem., St. Paul, MN.

1(7), 27-38.

3(4), 28-54.

24,214-228.


Hoseney, R. C. 1991. Wheat gluten: Rheological and gas retaining properties. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 657-666. Plenum, New York.

Hoseney, R. C. 1992. Physical chemistry of bread dough. In “Physical Chemistry of Foods” (H. G. Schwartzberg and R. W. Hartel, eds.), pp. 443-457. Dekker, New York.

Hoseney, R. C., and Rogers, D. E. 1990. The formation and properties of wheat flour doughs. CRC Crit. Rev. Food Sci. Nutr. 29, 73-93.

Hoseney, R. C., and Rogers, D. E. 1993. Mechanism of sugar functionality in cookies. In “The Science of Cookie and Cracker Production” (H. Faridi, ed.), pp. 203-225. Van Nostrand-Reinhold/AVI, New York (1994).

Hoseney, R. C., Zeleznak, K., and Lai, C. S. 1986. Wheat gluten: A glassy polymer. Cereal Chem. 63,285-286.

Hoseney, R. C., Seib, P. A., and Dong, W. 1992. A cereal chemist’s approach to studying Tr Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 59.

Hsu, J. H., and Reid, D. S. 1992. Determination of Tg’ in a variety of products using DSC and DETA. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 855.

Huang, V. T. 1992. Reaction kinetics in food systems: WLF vs. Arrhenius models. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 818.

Huang, V. T. 1993a. A thermomechanical method for determining collapse temperatures of food powders. Int. Congr. Eng. Food, 6th, Chiba, Japan, Abstr. P1.03.

Huang, V. T. 1993b. Mathematical modelling of the glass transition curves of glassy food polymers. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 911.

Huang, V. T., Hoseney, R. C., Graf, E., Ghiasi, K., Miller, L. C., Weber, J. L., Gaertner, K. C., Matson, K. L., Hunstiger, A. M., Rogers, D. K., and Saguy, I. A. 1991. Starch- based products for microwave cooking or heating. US. Pat. 5,035,904.

Huang, V. T., McIntyre, F., Haynes, L., Levine, H., and Slade, L. 1992. Glass transitions in starch, gluten, and bread as measured by TMA, TSC, and dielectric spectroscopy methods. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 64.

Huang, V. T., Levine, H., and Slade, L. 1993. Reaction kinetics in food systems: WLF vs. Arrhenius models. Food Powder Manufacture, in press (1994).

Huang, W. J., Frick, T. S., Landry, M. R., Lee, J. A., Lodge, T. P., and Tirrell, M. 1987, Tracer diffusion measurement in polymer solutions near the glass transition by forced Rayleigh scattering. AIChE J. 33,573-582.

Hutchinson, R. J., Mantle, S. A., and Smith, A. C. 1989. The effect of moisture content on the mechanical properties of extruded food foams. J . Mater. Sci. 24,3249-3253.

I’Anson, K. J., Miles, M. J., Morris, V. J., Ring, S. G., and Nave, C. 1988. Study of amylose gelation using synchrotron X-ray source. Carbohydr. Polym. 8,45-53.

I’Anson, K. J., Miles, M. J., Morris, V. J., Besford, L. S., Jarvis, D. A., and Marsh, R. A. 1990. The effects of added sugars on the retrogradation of wheat starch gels. J. Cereal Sci. 11, 243-248.

Izzard, M. J., Ablett, S., and Lillford, P. J. 1991. A calorimetric study of the glass transition occurring in sucrose solutions. In “Food Polymers, Gels and Colloids” (E. Dickinson, ed.), pp. 289-300. Royal Society of Chemistry, Cambridge, UK.

Jeffery, M. S. 1992. The key functional properties of sucrose in chocolate and sugar confection- ery. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 498.

Jin, X., Ellis, T. S., and Karasz, F. E. 1984. The effect of crystallinity and crosslinking on the depression of the glass transition temperature in nylon 6 by water. J. Polym. Sci., Polym.

Johari, G. P. 1976. Glass transition and secondary relaxations in molecular liquids and crystals. Phys. Ed. 22,1701-1717.

Ann. N.Y. Acad. Sci. 279, 117-140.


Johari, G. P. 1991. Dielectric behavior of H-bonded liquids and amorphous and crystalline solids. J. Mol. Struct. 250, 351-384.

Johnson, G. E., Bair, H. E., Matsuoka, S., Anderson, E. W., and Scott, J. E. 1980. Water sorption and its effects on a polymer’s dielectric behavior. ACS Symp. Ser. 127,451-468.

Johnson, J. M., Davis, E. A., and Gordon, J. 1990. Interactions of starch and sugar water measured by ESR and DSC. Cereal Chem. 67,286-291.

Johnston-Banks, F. A. 1990. Gelatine. In “Food Gels” (P. Harris, ed.), pp. 233-289. Elsevier Applied Science, London.

Jolley, J. E. 1970. The microstructure of photographic gelatin binders. Photogr. Sci. Eng.

Jouppila, K., and Roos, Y. 1992. Glass transitions of sugars and sugar alcohols. In “ISOPOW- V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Process- ing of Foods” Valencia, Spain, Poster 8.

Jovanovich, G., Zamponi, R. A., Lupano, C. E., and Anon, M. C. 1992. Effect of water content on the formation and dissociation of the amylose-lipid complex in wheat flour. J. Agric. Food Chem. 40,1789-1793.

Jurgens, A., Gijssel, J. V., and van der Kamp, J. W. 1992. The use of thermal analysis and rheology to assess cereal polysaccharides thermal processing. In “Workshop on Cereal Polysaccharides.” INRA-Nantes, Le Croisic, France.

Kainuma, K., and French, D. 1972. Naegeli amylodextrin and its relationship to starch granule structure. 11. Role of water in crystallization of B-starch. Eiopolymers 11,2241-2250.

Kakivaya, S . R., and Hoeve, C. A. J. 1975. The glass transition of elastin. Proc. Natl. Acad. Sci. U S A . 72,3505-3507.

Kaletunc, G., and Breslauer, K. J. 1992. Glass transitions of extrudates: Relationship with processing-induced fragmentation and end-product attributes. Cereal Chem. 70,

Kalichevsky, M. T., and Blanshard, J. M. V. 1992a. A study of the effect of water on the glass transition of 1 : 1 mixtures of amylopectin, casein, and gluten using DSC and DMTA. Carbohydr. Polym. 19,271-278.

Kalichevsky, M. T., and Blanshard, J. M. V. 1992b. The effect of fructose and water on the glass transition of amylopectin. Carbohydr. Polym. 20,107-113 (1993).

Kalichevsky, M. T., and Blanshard, J. M. V. 1992c. Studies of the glass transition in multicomponent biopolymer systems. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 61.

Kalichevsky, M. T., and Blanshard, J. M. V. 1992d. Theories of polymer plasticization applied to amylopectin. Easter Sch., Sci. Technol. Glassy State Foods, 53rd, Nottingham University, UK, Abstr. P3.

Kalichevsky, M. T., and Blanshard, J. M. V. 1992e. The effect of fructose on the mechanical properties of food biopolymers. Easter Sch., Sci. Technol. Glassy State Foods, 53rd, Nottingham University, UK, Abstr. P4.

Kalichevsky, M. T., Jaroszkiewicz, E. M., Ablett, S., Blanshard, J. M. V., and Lillford, P. J. 1992a. The glass transition of amylopectin measured by DSC, DMTA and NMR. Carbo- hydr. Polym. 18,77-88.

Kalichevsky, M. T., Jaroszkiewicz, E. M., and Blanshard, J. M. V. 1992b. A study of the glass transition of amylopectin-sugar mixtures. Polymer 34,346-358 (1993).

Kalichevsky, M. T., Jaroszkiewicz, E. M., and Blanshard, J. M. V. 1992c. The glass transition of gluten. 1. Gluten and gluten-sugar mixtures. Int. J. Biol. Macromol. 14, 257-266.

Kalichevsky, M. T., Jaroszkiewicz, E. M., and Blanshard, J. M. V. 1992d. The glass transition of gluten. 2. The effect of lipids and emulsifiers. Int. J. Eiol. Macromol. 14,267-273.

14,169-177.

548-552 (1993).


Kalichevsky, M. T., Blanshard, J. M. V., and Marsh, R. D. L. 1993a. Applications of mechanical spectroscopy to the study of glassy biopolymers and related systems. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 133-156. Nottingham Univ. Press, Loughborough.

Kalichevsky, M. T., Blanshard, J. M. V., and Tokarczuk, P. F. 1993b. Effect of water content and sugars on the glass transition of casein and sodium caseinate. Int. J. Food Sci. Technol. 28, 139-151.

Kararli, T. T., and Catalano, T. 1990. Stabilization of misoprostol with hydroxypropyl methyl- cellulose (HPMC) against degradation by water. Pharm. Res. 7, 1186-1189.

Kararli, T. T., Hurlbut, J. B., and Needham, T. E. 1990. Glass-rubber transitions of cellulosic polymers by DMA. J. Pharm. Sci. 79,845-848.

Karathanos, V. T., and Saravacos, G. D. 1992. Water diffusivity in the extrusion cooking of starch materials. In “Food Extrusion Science and Technology” (J. L. Kokini, C. T. Ho, and M. V. Karwe, eds.), pp. 177-187. Dekker, New York.

Karathanos, V. T., Vagenas, G. K., and Saravacos, G. D. 1991. Water diffusivity in starches at high temperatures and pressures. Biotechnol. Prog. 7, 178-184.

Karathanos, V. T., Anglea, S., and Karel, M. 1992. Collapse of structure during drying of vegetables. Drying Technol. 11,1005-1023 (1993).

Karel, M. 1985. Effects of water activity and water content on mobility of food components, and their effects on phase transitions in food systems. In “Properties of Water in Foods” (D. Simatos and J. L. Multon, eds.), pp. 153-169. Martinus Nijhoff, Dordrecht, The Nether- lands.

Karel, M. 1986. Control of lipid oxidation in dried foods. I n “Concentration and Drying of Foods” (D. MacCarthy, ed.), pp. 37-51. Elsevier Applied Science, London.

Karel, M. 1989. Role of water activity. I n “Food Properties and Computer-Aided Engineering of Food Processing Systems” (R. P. Singh and A. G. Medina, eds.), pp. 135-155. Kluwer, Dordrecht, The Netherlands.

Karel, M. 1990. Encapsulation and controlled release of food components. I n “Advances in Bio- and Food Process Engineering” (M. A. Rao and H. Schwartzberg, eds.), pp. 277-294. Dekker, New York.

Karel, M. 1991a. Advances in science and engineering-A challenge to food technology. Food

Karel, M. 1991b. Impact of innovation in food technology on the use of refrigeration. Proc. Int. Congr. Refrig. Idth, 1991 Vol. 1, pp. 35-41.

Karel, M. 1991c. Physical structure and quality of dehydrated foods. In “Drying ’91” (A. S. Mujumdar and I. Filkova, eds.), pp. 26-35. Elsevier, Amsterdam.

Karel, M. 1992. Optimization of quality of dehydrated foods and biomaterials. In “Drying ’92,” Part A, pp. 3-16. Elsevier, Amsterdam.

Karel, M. 1993. Effects of glass transitions on processing and storage of foods. Inf. Congr. Eng. Food, 6th, Chiba, Japan, Abstr. 1.01.

Karel, M., and Flink, J. M. 1983. Some recent developments in food dehydration research. Adv. Drying 2, 103-153.

Karel, M., and Langer, R. 1988. Controlled release of food additives. ACS Symp. Ser. 370, 177-191.

Karel, M., and Saguy, I. 1991. Effects of water on diffusion in food systems. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 157-173. Plenum, New York.

Karel, M., Buera, M. P., and Roos, Y. 1993a. Effects of glass transitions on processing and storage. I n “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 13-39. Nottingham Univ. Press, Loughborough.

Aus~. 43(10), 459-463.


Karel, M., Anglea, S., Buera, P., Karmas, R., Levi, G., and Roos, Y. 1993b. Stability-related transitions of amorphous foods. Thermochim. Acta 246,249-269 (1994).

Karger, N., and Ludemann, H. D. 1991. Temperature dependence of the rotational mobility of the sugar and water molecules in concentrated aqueous trehalose and sucrose solutions. Z. Naturforsch., C: Biosci. 46, 313-317.

Karmas, R., Buera, M. P., and Karel, M. 1992. Effect of glass transition on rates of nonenzymatic browning in food systems. J . Agric. Food Chem. 40,873-879.

Karmas, R., Buera, M. P., and Karel, M. 1993. Effect of temperature, moisture, and the glass transition on Maillard browning in PVP matrices. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 918.

Kasarda, D. D., Nimmo, C. C., and Kohler, G. 0.1978. Proteins and amino acid composition of wheat fractions. In “Wheat Chemistry and Technology” (Y. Pomeranz, ed.), 2nd ed., pp. 227-299. Am. Assoc. Cereal Chem., St. Paul, MN.

Katsuta, K., and Kinsella, J. E. 1990. Effects of temperature on viscoelastic properties and activation energies of whey protein gels. J. Food Sci. 55, 1296-1302.

Katsuta, K., Miura, M., and Nishimura, A. 1992. Kinetic treatment for rheological properties and effects of saccharides on retrogradation of rice starch gels. Food Hydrocolloids

Kauzmann, W. 1948. Nature of the glassy state and behavior of liquids at low temperatures. Chem. Rev. 43,219-256.

Kawai, H., Sakurai, M., Inoue, Y., Chujo, R., and Kobayashi, S. 1992. Hydration of oligosaccharides: Anomalous hydration ability of trehalose. Cryobiology 29,599-606.

Keinath, S. E., and Boyer, R. F. 1981. Thermomechanical analysis of Tg and T > Tg transitions in polystyrene. J. Appl. Polym. Sci. 26, 2077-2085.

Kelley, S. S., Rials, T. G., and Glasser, W. G. 1987. Relaxation behavior of the amorphous components of wood. J. Muter. Sci. 22, 617-624.

Kerr, W. L., Lim, M. H., Reid, D. S., and Chen, H. 1993. Chemical reaction kinetics in relation to glass transition temperatures in frozen food polymer solutions. J. Sci. Food Agric.

Kim, C. S., and Walker, C. E. 1992. Effects of sugars and emulsifiers on starch gelatinization evaluated by DSC. Cereal Chem. 69,212-217.

Kim, S. S., and Setser, C. S. 1992. Wheat starch gelatinization in the presence of polydextrose or hydrolyzed barley P-glucan. Cereal Chem. 69,447-452.

Kim, Y. K., and Taub, I. A. 1993. Compositional influences on glass transition in model bread systems. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 618.

Kirby, A. R., Clark, S. A., Parker, R., and Smith, A. C. 1993. The effects of plasticizers on the deformation and fracture behavior of wheat starch. In “Food Colloids and Polymers: Structure and Dynamics.” Royal Society of Chemistry, London.

Knutson, C. A. 1990. Annealing of maize starches at elevated temperatures. Cereal Chem.

Kohyama, K., and Nishinari, K. 1991. Effect of soluble sugars on gelatinization and retrogradation of sweet potato starch. J. Agric. Food Chern. 39,1406-1410.

Koide, M., Matusita, K., and Komatsu, T. 1990. Viscosity of fluoride glasses at glass transition temperature. J . Non-Cryt. Solids 125, 93-97.

Kokini, J. L. 1992. Measurement and simulation of shear and shear-free (extensional) flows in food rheology. In “Advances in Food Engineering” (R. P. Singh and M. A. Wirakartaku- sumah, eds.), pp. 439-462. CRC Press, Boca Raton, FL.

Kokini, J. L., Lai, L. S., and Chedid, L. L. 1992a. Effect of starch structure on starch rheological properties. Food Technol. 46(6), 124-139.

6,187-198.

61,51-56.

61,376-384.


Kokini,, J. L., Solberg, M., and Henrikson, F. 1992b. A new model for food research: CAFT- Rutgers University. Trends Food Sci. Technol. 3, 128-132.

Kokini, J. L., Chang, C. N., and Lai, L. S. 1992c. The role of rheological properties on extrudate expansion. In “Food Extrusion Science and Technology” (J. L. Kokini, C. T. Ho, and M. V. Karwe, eds.), pp. 631-652. Dekker, New York.

Kokini, J. L., Madeka, H., Cocero, A. M., and de Graaf, E. 1992d. Predicting rheology of corn and wheat proteins. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods” Valencia, Spain, Abstr. 5.2.

Koster, K. L. 1991. Glass formation and desiccation tolerance in seeds. Plant Physiol. %,

Koster, K. L., and Leopold, A. C. 1988. Sugars and desiccation tolerance in seeds. Plant Physiol. 88,829-832.

Koster, K. L., Webb, M. S., Bryant, B., and Lynch, D. V. 1993. Vitrified sugars alter the phase behavior of phosphatidylcholine: Relevance to desiccation tolerance. Biochim. Biophys.

Kresin, M., and Rau, G. 1992. The glass transition of hydroxyethyl starch. Cryo-Lett. 13,

Kuge, T., and Kitamura, S. 1985. Annealing of starch granules-Warm water treatment and heat-moisture treatment. J . Jpn. SOC. Starch Sci. 32, 65-83.

Labrousse, S., Roos, Y.. and Karel, M. 1992. Collapse and crystallization in amorphous matrices with encapsulated compounds. Sci. Aliment. 12,757-769.

Labuza, T. P., and Baisier, W. M. 1992. The kinetics of nonenzymatic browning. In “Physical Chemistry of Foods” (H. G. Schwartzberg and R. W. Hartel, eds.), pp. 595-649. Dekker, New York.

Labuza, T. P., and Nelson, K. 1992. Glass transitions, state diagrams, and texture of cereal products. CoFE ‘92-AIChE 1992 Summer Natl. Meet., Minneapolis, Paper 2e.

Lai, L. S., and Kokini, J. L. 1991. Physicochemical changes and rheological properties of starch during extrusion (a review). Biotechnol. Prog. 7,251-266.

Larsson, I., and Eliasson, A.-C. 1991. Annealing of starch at an intermediate water content. Staerke 43, 221-231.

Larsson, K. 1991. Structure of the starch granule-a curved crystal. Acta Chem. Scand.

Lasztity, R. 1986. Recent results in the investigation of the structure of the gluten complex. Nahrung 30,235-244.

Lawton, J. W. 1992a. The glass transition in corn proteins: Its importance to the viscoelastic properties of corn doughs. 83rd Am. Oil Chem. SOC. Annu. Meet., Toronto, Abstr. 30-3.

Lawton, J. W. 1992b. Viscoelasticity of zein-starch doughs. Cereal Chem. 69, 351-355. Lawton, J. W., and Wu, Y. V. 1993. Thermal behavior of annealed acetic acid-soluble wheat

LeBlanc, E. L., LeBlanc, R. J., and Blum, I. E. 1988. Prediction of quality in frozen cod fillets.

Lee, C. Y., Smith, N. L., and Hawbecker, D. E. 1988. Enzyme activity and quality of frozen

Lelievre, J. 1976. Theory of gelatinization in a starch-water-solute system. Polymer 17,

Lelievre, J. 1992a. Thermal analysis of carbohydrates as illustrated by aqueous starch systems. In “Developments in Carbohydrate Chemistry” (R. J. Alexander and H. F. Zobel, eds.), pp. 137-161. Am. Assoc. Cereal Chem., St. Paul.

Lelievre, J. 1992b. Thermal analysis of carbohydrates as illustrated by aqueous starch systems. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 234.

302-304.

Acta 1193, 143-150 (1994).

371-378,

45,840-843.

gluten. Cereal Chem. 70,471-475.

J. Food Sci. 53,328-340.

green beans as affected by blanching and storage. J . Food Qual. 11,279-287.

854-858.


Le Meste, M., and Duckworth, R. B. 1988. Effect of water content on the mobility of solute molecules and of protein side chains in caseinate preparations. Int. J. Food Sci. Technol.

Le Meste, M., and Huang, V. T. 1991. Thermomechanical properties of frozen sucrose solutions. Inst. Food Technol. Annu. Meet., Dallas, TX, Abstr. 653; J. Food Sci. 57,

Le Meste, M., and Simatos, D. 1990. La transition vitreuse. J. Ind. Aliment. Agric. 107,

Le Meste, M., Viguier, L., Lorient, D., and Simatos, D. 1990. Rotational diffusivity of solutes in concentrated caseinates: Influence of glycosylation. J. Food Sci. 55,724-727.

Le Meste, M., Panama, J., Anderson, G., Lentz, R., and Huang, V. T. 1991a. Thermomechanical analysis of cereal products. Am. Assoc. Cereal Chem., 76th Annu. Meet., Seattle, WA, Abstr. 235.

Le Meste, M., Voilley, A., and Colas, B. 1991b. Influence of water on the mobility of small molecules dispersed in a polymeric system. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 123-138. Plenum, New York.

Le Meste, M., Huang, V. T., Panama, J., Anderson, G., and Lentz, R. 1992. Glass transition of bread. Cereal Foods World 37, 264-267.

Leopold, A. C., Bruni, F., and Williams, R. J. 1992. Water in dry organisms. In “Water and Life” (Somero et al., eds.), pp. 161-169. Springer-Verlag, Berlin.

Leung, H. K., Hong, Y. C., and Attwool, P. T. 1992. Glass transitions of carbohydrate solutions as determined by DMA. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 859; J . Food. Sci., submitted (1993).

Levi, G., and Karel, M. 1992. The effect of crystallization on volatile release from amorphous sugar rubbers in the vicinity of Tg. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 691.

Levi, G., and Karel, M. 1993a. Effect of temperature above Tg and moisture on propanol release in fully collapsed matrix. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 621.

Levi, G., and Karel, M. 1993b. The effect of phase transitions on release of n-propanol entrapped in carbohydrate glasses. J. Food Eng. 24, 1-13 (1994).

Levine, H., and Slade, L. 1986. A polymer physico-chemical approach to the study of commercial starch hydrolysis products (SHPs). Carbohydr. Polym. 6,213-244.

Levine, H., and Slade, L. 1988a. Water as a plasticizer: Physico-chemical aspects of low- moisture polymeric systems. In “Water Science Reviews” (F. Franks, ed.), Vol. 3, pp. 79-185. Cambridge, Univ. Press, Cambridge, UK.

Levine, H., and Slade, L. 1988b. Collapse phenomena-A unifying concept for interpreting the behavior of low-moisture foods. In “Food Structure-Its Creation and Evalua- tion’’ (J. M. V. Blanshard and J. R. Mitchell, eds.), pp. 149-180. Butterworths, London.

Levine, H., and Slade, L. 1988c. Principles of cryostabilization technology from structure/ property relationships of water-soluble food carbohydrates-A review. Cryo-Lett. 9,

Levine, H., and Slade, L. 1988d. Thermomechanical properties of small carbohydrate-water glasses and “rubbers”: Kinetically-metastable systems at subzero temperatures. J. Chem. SOC., Faraday Trans. 1 84,2619-2633.

Levine, H., and Slade, L. 1989a. Interpreting the behavior of low-moisture foods. In “Water and Food Quality” (T. M. Hardman, ed.), pp. 71-134. Elsevier, London.

Levine, H., and Slade, L. 1989b. Influences of the glassy and rubbery states on the thermal, mechanical, and structural properties of doughs and baked products. In “Dough Rheology

23,457-466.

1230-1233 (1992).

5-12.

21-63.


and Baked Product Texture” (H. Faridi and J. M. Faubion, eds.), pp. 157-330. Van Nostrand-Reinhold/AVI, New York (1990).

Levine, H., and Slade, L. 1989~. A food polymer science approach to the practice of cryostabilization technology. Comments Agric. Food Chem. 1,315-396.

Levine, H., and Slade, L. 1989d. Response to the letter by Simatos, Blond, and Le Meste on the relation between glass transition and stability of a frozen product. Cryo-Lett.

Levine, H., and Slade, L. 1990. Cryostabilization technology: Thermoanalytical evaluation of food ingredients and systems. In “Thermal Analysis of Foods” (V. R. Harwalkar and C. Y. Ma, eds.), pp. 221-305. Elsevier Applied Science, London.

Levine, H., and Slade, L. 1991. Polymer physicochemical characterization of oligosaccharides. ACS Symp. Ser. 458,219-260.

Levine, H., and Slade, L. 1992a. An alternative view of trehalose functionality in drying and stabilization of biological materials. BioPharm 5(4), 36-40.

Levine, H., and Slade, L. 1992b. Glass transitions in foods. In “Physical Chemistry of Foods” (H. G. Schwartzberg and R. W. Hartel, eds.), pp. 83-221. Dekker, New York.

Levine, H., and Slade, L. 1993. The glassy state in applications for the food industry, with an emphasis on cookie and cracker production. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 333-373. Nottingham Univ. Press, Loughborough.

Levine, H., Slade, L., Van Lengerich, B., and Pickup, J. G. 1991. Glassy matrices containing volatile and/or labile components, and processes for preparation and use thereof. U.S. Pat. 5,009,900.

Levine, H., Slade, L., Van Lengerich, B., and Pickup, J. G. 1992. Double-encapsulated compositions containing volatile and/or labile components, and processes for preparation and use thereof. U.S. Pat. 5,087,461.

Lillford, P. J. 1988. The polymer/water relationship-Its importance for food structure. In “Food Structure-Its Creation and Evaluation” (J. M. V. Blanshard and J. R. Mitchell, eds.), pp. 75-92. Butterworths, London.

Lillford, P. J., Ablett, S., Davies, A. P., and Appelqvist, I. A. M. 1992. Glass transition in carbohydrates and cereal composites. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, Abstr. 60.

Lillie, M. A,, and Gosline, J. M. 1990. The effect of hydration on the dynamic mechanical properties of elastin. Biopolymers 29, 1147-1160.

Lillie, M. A., and Gosline, J. M. 1993. The effects of swelling solvents on the glass transition in elastin and other proteins. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 281-301. Nottingham Univ. Press, Loughborough.

Lim, H., Setser, C. S., Paukstelis, J. V., and Sobczynska, D. 1992. 017 NMR studies on wheat starch-sucrose-water interactions with increasing temperature. Cereal Chem. 69,382-386.

Lim, M. H., and Reid, D. S. 1991. Studies of reaction kinetics in relation to the Tg‘ of polymers in frozen model systems. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 103-122. Plenum, New York.

Lim, M. H., and Reid, D. S. 1992. The effect of thermal behavior of solutes on reaction kinetics in frozen model systems. CoFE ’92-AIChE I992 Summer Natl. Meet., Minneapolis, Paper 16f.

Lin, L. S., Yuen, H. K., and Varner, J. E. 1991. DSC of plant cell walls. Proc. Natl. Acad. Sci. U.S.A. 88,2241-2243.

Lin, S. H. 1993. Water uptake and gelatinization of white rice. Lebensm.-Win. Technol.

Lineback, D. R., and Rasper, V. F. 1988. Wheat carbohydrates. In “Wheat Chemistry and Technology” (Y. Pomeranz, ed.), 3rd ed., Vol. 1, pp. 277-372. Am. Assoc. Cereal Chem., St. Paul, MN.

10,347-370.

26,276-278.


Liu, H., and Lelievre, J. 1991a. Effects of heating rate and sample size on DSC traces of starch gelatinized at intermediate water levels. Stuerke 43,225-227.

Liu, H., and Lelievre, J. 1991b. A differential scanning calorimetry study of glass and melting transitions in starch suspensions and gels. Curbohydr. Res. 219,23-32.

Liu, H., and Lelievre, J. 1992a. A DSC study of melting transitions in aqueous suspensions containing blends of wheat and rice starch. Carbohydr. Polym. 17,145-149.

Liu, H., and Lelievre, J. 1992b. Transitions in frozen gelatinized-starch systems studied by DSC. Curbohydr. Polym. 19, 179-183.

Liu, H., Lelievre, J., and Ayoung-Chee, W. 1991. A study of starch gelatinization using DSC, X-ray, and birefringence measurements. Curbohydr. Res. 210,7947.

Livings, S. J., Donald, A. M., and Smith, A. C. 1993. Ageing in confectionary wafers. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 507-511. Nottingham Univ. Press, Loughborough.

Lomellini, P. 1992. WLF vs. Arrhenius behavior: Polystyrene melt viscoelasticity revised. Polymer 33,4983-4989.

Lookhart, G. L., Menkovska, M., and Pomeranz, Y. 1987. Changes in the gliadin fraction(s) during breadmaking: Isolation and characterization by HPLC and PAGE. Am. Assoc. Cereal Chem., 72nd Annu. Meet., Nashville, Abstr. 24.

Lund, D. B. 1989. Starch gelatinization. In “Food Properties and Computer-Aided Engineering of Food Processing Systems” (R. P. Singh and A. G. Medina, eds.), pp. 299-311. Kluwer, Dordrecht, The Netherlands.

Luyet, B. 1960. On various phase transitions occurring in aqueous solutions at low temperatures. Ann. N.Y. Acud. Sci. 85,549-569.

Ma, C. Y., and Harwalkar, V. R. 1991. Thermal analysis of food proteins. Adv. Food Nutr. Res. 35, 317-366.

MacDonald, G. A., and Lanier, T. 1991. Carbohydrates as cryoprotectants for meats and surimi. Food Technol. 45(3), 150-159.

MacInnes, W. M. 1992. DMTA of sucrose solutions and other model food systems. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Rela- tion to Processing of Foods” Valencia, Spain, Poster 3.

MacInnes, W. M. 1993. Dynamic mechanical thermal analysis of sucrose solutions during freezing and thawing. I n “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 223-248. Nottingham Univ. Press, Loughborough.

MacKenzie, A. P. 1977. Non-equilibrium freezing behavior of aqueous systems. Philos. Trans. R. SOC. London, Ser. B 278,167-189.

Madeka, H., and Kokini, J. L. 1992a. Effect of moisture and temperature on transitions of zein and gliadin. CoFE ’92-AIChE 1992 Summer Natl. Meet., Minneapolis, Paper 16b.

Madeka, H., and Kokini, J. L. 1992b. Effect of moisture on glass transition of zein and gliadin. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 164.

Madeka, H., and Kokini, J. L. 1993. Rheological properties of gliadin and zein as a function of moisture content and state diagrams. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 33.

Magnuson, K. M. 1985. Uses and functionality of vital wheat gluten. Cereal Foods World

Magoshi, J., Nakamura, S., and Murakami, K. 1992. Structure and physical properties of seed proteins. I. Glass transition and crystallization of zein protein from corn. J. Appl. Polym. Sci. 45,2043-2048.

Makower, B., and Dye, W. B. 1956. Equilibrium moisture content and crystallization of amorphous sucrose and glucose. J. Agric. Food Chem. 4,72-77.

Mancini, L. 1993. Technology leadership-Winning the technology stakes. Food Eng.

30, 179-181.

65(1), 93-100.


Mandelkern, L. 1986. Thennoreversible gelation and crystallization from solution. Polym. Prepr., Am. Chem. SOC., Div. Polym. Chem. 27(1), 206.

Marsh, R. D. L., and Blanshard, J. M. V. 1988. The application of polymer crystal growth theory to the kinetics of formation of the B-amylose polymorph in a 50% wheat starch gel. Carbohydr. Polym. 9,301-317.

Marshall, A. S., and Petrie, S. E. B. 1980. Thermal transitions in gelatin and aqueous gelatin solutions. .I. Photogr. Sci. 28, 128-134.

Masi, P. 1989. Study of the influence of temperature on the rheological behavior of gluten by means of dynamic mechanical analysis. I n “Food Properties and Computer-Aided Engineering of Food Processing Systems” (R. P. Singh and A. G. Medina, eds.), pp. 357-362. Kluwer, Dordrecht, The Netherlands.

Matsuoka, S. 1992. “Relaxation Phenomena in Polymers.” Hanser, Munich. Matsuoka, S., Williams, G., Johnson, G. E., Anderson, E. W., and Furukawa,T. 1985. Phenome-

nological relationship between dielectric relaxation and thermodynamic recovery processes near the glass transition. Macromolecules 18,2652-2663.

Matthiesen, A. B., Hanson, C. E., McIntyre, F., Ibar, J. P., Huang, V. T., and Le Meste, M. 1991. Thermal analysis instrument for the determination of T, of water and sucrose solutions. Insr. Food Technol. Annu. Meet., Dallas, TX, Abstr. 717; Food Technol.

Maurer, J. J. 1981. Elastomers. In “Thermal Characterization of Polymeric Materials” (E. A. Turi, ed.), pp. 571-708. Academic Press, Orlando, FL.

Maurice, T. J., Slade, L., Page, C., and Sirett, R. 1985. Polysaccharide-water interactions- Thermal behavior of rice starch. In “Properties of Water in Foods” (D. Simatos and J. L. Multon, eds.), pp. 211-227. Martinus Nijhoff, Dordrecht, The Netherlands.

Maurice, T. J., Asher, Y. J., and Thomson, S. 1991. Thennomechanical analysis of frozen aqueous systems. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 215-223. Plenum, New York.

Mayer, E. 1991. Calorimetric glass transitions in the amorphous forms of water: A comparison. J. Mol. Srrucr. 250,403-411.

McCurdy, R. D., Goff, H. D., and Stanley, D. W. 1993. Unique rheological and stabilizing properties of dextran-Implications as a functional ingredient in food. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 383.

Megret, C., Guantieri, V., Lamure, A., Pieraggi, M. T., Lacabanne, C., and Tamburro, A. M. 1992. Phase transitions and chain dynamics, in the solid state, of a pentapeptide sequence of elastins. Inr. J. Biol. Macromol. 14,45-49.

Mel’nichenko, Y. B., Klepko, V. V., and Shilov, V. V. 1993. Self-diffusion of water in gelatin gels. 1. Macroscopic measurements by tracer technique. Polymer 34,1019-1023.

Menkovska, M., Pomeranz, Y., Lookhart, G. L., and Shogren, M. D. 1987. Gliadin fractions in breadmaking: Differences in flours varying in breadmaking potential. Am. Assoc. Cereal Chem., 72nd Annu. Meet., Nashville, Abstr. 141.

Menkovska, M., Pomeranz, Y., Lookhart, G. L., and Shogren, M. D. 1988. Gliadin in crumb of bread from high-protein wheat flours of varied breakmaking potential. Cereal Chem.

Mestres, C., Colonna, P., and Buleon, A. 1988. Gelation and crystallization of maize starch

Mestres, C., Colonna, P., Alexandre, M. C., and Matencio, F. 1993. Comparison of various

Michniewicz, J., Biliaderis, C. G., and Bushuk, W. 1992. Effect of added pentosans on some

45(7), 106-107.

65, 198-201.

after pasting, drum-drying, or extrusion cooking. J. Cereal Sci. 7, 123-134.

processes for making maize pasta. J. Cereal Sci. 17, 277-290.

properties of wheat bread. Food Chem. 43,251-257.


Milczarek, P., Zielinski, M., and Garcia, M. L. 1992. The mechanism and stability of thermal

Miles, M. J., Morris, V. J., and Ring, S. G. 1985a. Gelation of amylose. Carbohydr. Res.

Miles, M. J., Morris, V. J., Orford, P. D., and Ring, S. G. 1985b. Roles of amylose and amylopectin in gelation and retrogradation of starch. Carbohydr. Res. l35,271-281.

Mita, T. 1990. Effect of aging on the rheological properties of gluten gel. Agric. Biol. Chem.

Mita, T. 1992. Structure of potato starch pastes in the aging process by the measurement of

Moore, W. R., and Hoseney, R. C. 1986. Effects of flour lipids on expansion rate and volume

Morozov, V. N., and Gevorkian, S. G. 1985. Low-temperature glass transition in proteins.

Morris, V. J. 1990. Starch gelation and retrogradation. Trends Food Sci. Technol. 1,2-6. Moy, P., and Karasz, F. E. 1980. The interactions of water with epoxy resins. ACS Symp. Ser.

Murthy, N. S . , Stamm, M., Sibilia, J. P., and Krimm, S. 1989. Structural changes accompanying

Murthy, S. S. N. 1989. Strength and fragility in glass-forming liquids. J. Phys. Chem. 93,

Murthy, S . S. N., Gangasharan, and Nayak, S . K. 1993. Novel DSC studies of supercooled organic liquids. J . Chem. Soc., Faraday Trans. 89, 509-514.

Myers-Betts, P. A., and Baianu, I. C. 1990. Approaches to protein hydration and water activity in food proteins by NMR and vapor pressure equilibration techniques. J. Agric. Food Chem. 38,1171-1177.

Nakamura, K., Hatakeyama, T., and Hatakeyama, H. 1991. Formation of the glassy state and mesophase in the water-sodium alginate system. Polym. J. 23,253-258.

Nelson, K. A. 1993. Reaction kinetics of food stability: Comparison of glass transition and classical models for temperature and moisture dependence. Doctoral thesis, University of Minnesota, St. Paul.

Nelson, K. A., and Labuza, T. P. 1992a. Water activity and food polymer science: Implications of state on Arrhenius and WLF kinetics in predicting shelf life. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Abstr. 6.1; J. Food Eng. 22, 271-289 (1994).

Nelson, K. A., and Labuza, T. P. 1992b. Relationship between water and lipid oxidation rates: Water activity and glass transition theory. In “Lipid Oxidation in Foods” (A. J. St. Angelo, ed.), pp. 93-103. Am. Chem. SOC., Washington, DC.

Nelson, K. A,, and Labuza, T. P. 1992c. Glass transition theory and texture of cereal foods. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 513-517. Nottingham Univ. Press, Loughborough.

Nelson, K. A., and Labuza, T. P. 1993. Application of WLF and Arrhenius models to chemical reactions in glassy and rubbery systems. Inst. Food Technol. Annu. Meet., Chicago, Ab- str. 620.

Nelson, K. A., Labuza, T. P., Reineccius, G. A., and Ma, Y. 1992. The stability of spray-dried microcapsules as a function of Tg. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 858.

Nesvadba, P. 1992a. Glass transitions in aqueous solutions and foodstuffs. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 523-526. Nottingham Univ. Press, Loughborough.

transitions in hair keratin. Colloid Polym. Sci. 270, 1106-1115.

l359257-269.

54,927-935.

their dynamic moduli. Carbohydr. Polym. 17,269-276.

of bread baked in a resistence oven. Cereal Chem. 63,172-174.

Biopolymers 24, 1785-1799.

127,505-513.

hydration in nylon 6. Macromolecules 22,1261-1267.

3347-3351.


Nesvadba, P. 1992b. Freezing of foods studied by DSC. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Poster 4.

Nesvadba, P. 1993. Glass transition temperatures in water-containing foods. Inf. Congr. Eng. Food, 6th, Chiba, Japan, Abstr. P1.06.

Newman, R. H. 1992. Solid-state I3C NMR spectroscopy of multiphase biomaterials. ACS Symp. Ser. 489,311-319.

Ng, P. K. W., and Bushuk, W. 1987. Relationship between high molecular weight subunits of glutenin and breadmaking quality. Am. Assoc. Cereal Chem., 72nd Annu. Meet., Nash- ville, TN, Abstr. 142.

Ngai, K. L. and Roland, C. M. 1993. Intermolecular cooperativity and the temperature dependence of segmental relaxation in semicrystalline polymers. Macromolecules 26,2688-2690.

Nielsen, L. E. 1977. “Polymer Rheology.” Dekker, New York. Nishinari, K., Watase, M., Williams, P. A., and Phillips, G. 0. 1991. Effect of sugars and

polyols on water in agarose gels. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 235-249. Plenum, New York.

Noel, T. R., and Ring, S. G. 1992. A study of the heat capacity of starcldwater mixtures. Carbohydr. Res. 227,203-213.

Noel, T. R., Ring, S. G., and Whittam, M. A. 1990. Glass transitions in low-moisture foods. Trends Food Sci. Technol. 1(3), 62-67.

Noel, T. R., Ring, S. G., and Whittam, M. A. 1991. Kinetic aspects of the glass-transition behavior of maltose-water mixtures. Carbohydr. Res. 212, 109-117.

Noel, T. R., Ring, S. G., and Whittam, M. A. 1992. Dielectric relaxations of small carbohydrate molecules in the liquid and glassy states. J. Phys. Chern. 96, 5662-5667.

Noel, T. R., Ring, S. G., and Whittam, M. A. 1993. Relaxations in supercooled carbohydrate liquids. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 173-187. Nottingham Univ. Press, Loughborough.

Nunes, R. V., Swartzel, K. R., and Ollis, D. F. 1993. Thermal evaluation of food processes: The role of a reference temperature. J . Food Eng. 20, 1-15.

O’Brien, J. 1991. Characterizing food and its ingredients. Trends Food Sci. Technol. 2(4), 79. O’Donnell, C. D. 1993. The formulation challenge: Formulating for the big freeze. Prepared

Ohshima, T., Suzuki, T., and Koizumi, C. 1993. New developments in surimi technology. Trends Food Sci. Technol. 4, 157-163.

Oksanen, C. A,, and Zografi, G. 1990. The relationship between Tg and water vapor absorption by PVP. Pharm. Res. 7,654-657.

Oksanen, C. A., and Zografi, G. 1993. Molecular mobility in mixtures of absorbed water and solid poly(vinylpyrro1idone). Pharm. Res. 10, 791-799.

Ollett, A. L., and Parker, R. 1990. The viscosity of supercooled fructose and its glass transition temperature. J . Texture Stud. 21,355-362.

Ollett, A. L., Parker, R., and Smith, A. C. 1991. Deformation and fracture behavior of wheat starch plasticised with water and glucose. .I. Muter. Sci. 26, 1351-1356.

Ollett, A. L., Kirby, A. R., Clark, S. A., Parker, R., and Smith, A. C. 1993a. The effect of water content on the compaction behavior of potato starch. Sraerke 45,51-55.

Ollett, A. L., Kirby, A. R., Parker, R., and Smith, A. C. 1993b. A comparative study of the effects of water content on the compaction behavior of some food materials. Powder Technol., in press.

Ollivon, M. R. 1991. Calorimetric and thermodielectrical measurements of water interactions with some food materials. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 175-189. Plenum, New York.

Foods 162(6), 55-56.


Orford, P. D., Parker, R., Ring, S. G., and Smith, A. C. 1989. The effect of water as a diluent on the glass transition behavior of malto-oligosaccharides, amylose and amylopectin. Int. J. Biol. Macromol. 11,91-96.

Orford, P. D., Parker, R., and Ring, S. G. 1990. Aspects of the glass transition behavior of mixtures of carbohydrates of low molecular weight. Carbohydr. Res. 196, 11-18.

Ostberg, G., Salmen, L., and Terlecki, J. 1990. Softening temperature of moist wood measured by DSC. Holzforschung 44,223-225.

Paakkonen, K., and Roos, Y. 1990. Effects of drying conditions on water sorption and phase transitions of freeze-dried horseradish roots. J. Food Sci. 55,206-209.

Parak, F., and Nienhaus, G. U. 1991. Glass-like behavior of proteins as seen by Mossbauer spectroscopy. J. Non-Cryst. Solids Wl-133,362-368.

Parker, R., and Smith, A. C. 1992. The mechanical properties of starchy food materials at large strains and their ductile-brittle transitions. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 519-522. Nottingham Univ. Press, Loughborough.

Patil, S. K. 1991. Starches in bakery foods. AIB Res. Dep. Tech. Bull. 13(6), 1-8. Paton, D. 1987. DSC of oat starch pastes. Cereal Chem. 64,394-399. Payne, P. I. 1987. Genetics of wheat storage proteins and the effect of allelic variation on

bread-making quality. Annu. Rev. Plant Physiol. 38, 141-153. Payne, P. I., Holt, L. M., Jarvis, M. G., and Jackson, E. A. 1985. Fractionation of endosperm

proteins of bread wheat. Cereal Chem. 62,319-326. Peleg, M. 1992. On the use of the WLF model in polymers and foods. CRC Crit. Rev. Food

Sci. Nutr. 32, 59-66. Peleg, M. 1993. Glass transitions and the physical stability of food powders. Zn “The Glassy

State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 435-451. Nottingham Univ. Press, Loughborough.

Peppas, N. A. 1993. Glassylrubbery transition in relation to penetrant transport and controlled release of encapsulated food ingredients. 206rh Am. Chem. SOC. Nut. Meet., Chicago, Abstr. 117.

Petrie, S. E. B. 1975. The problem of thermodynamic equilibrium in glassy polymers. In “Polymeric Materials: Relationships between Structure and Mechanical Behavior” (E. Baer and S. V. Radcliffe, eds.), pp. 55-118. Am. SOC. Metals, Metals Park, OH.

Phillips, A. J., Yarwood, R. J., and Collett, J. H. 1986. Thermal analysis of freeze-dried products. Anal. Proc. (London) 23,394-395.

Pikal, M. J. 1990a. Freeze-drying of proteins. I. Process design. BioPharm 3(8), 18-27. Pikal, M. J. 1990b. Freeze-drying of proteins. 11. Formulation selection. BioPharm 3(9), 26-30. Pikal, M. J. 1993. Stability of amorphous solids: Dynamics, reactivity, and the glass transition.

Biopreserv. Symp. BioPharm Con$, Cambridge, MA. Pikal, M. J., and Shah, S. 1990. The collapse temperature in freeze drying: Dependence on

measurement methodology and rate of wate removal from the glassy phase. Znt. J . Pharm.

Pikal, M. J., Shah, S., Roy, M. L., and Putman, R. 1990. The secondary drying stage of freeze drying: drying kinetics as a function of temperature and chamber pressure. Int. J . Pharm.

Pikal, M. J., Dellerman, K., and Roy, M. L. 1991a. Formulation and stability of freeze-dried proteins: Effects of moisture and oxygen on the stability of freeze-dried formulations of human growth hormone. Dev. Biol. Stand. 74,Zl-38.

Pikal, M. J., Dellerman, K. M., Roy, M. L., and Riggin, R. M. 1991b. The effects of formulation variables on the stability of freeze-dried human growth hormone. Pharm. Res. 8,427-436.

Pissis, P., and Apekis, L. 1991. A dielectric study of molecular mobility at glass transition. J . Non-Cryst. Solids 131-133,95-98.

62,165-186.

60,203-217.


Pissis, P., Anagnostopoulou-Konsta, A., Apekis, L., Daoukaki-Diamanti, D., Christodoulides, C., and Sideris, E. G. 1992. Dielectric studies on glass transitions in biological systems. IEEE Trans. Electr. Insul. 27, 820-825.

Pomeranz, Y. 1978. Composition and functionality of wheat-flour components. I n “Wheat Chemistry and Technology” (Y. Pomeranz, ed.), 2nd ed., pp. 585-674. Am. Assoc. Cereal Chem., St. Paul, MN.

Pomeranz, Y., Lookhart, G. L., Rubenthaler, G. L., and Alberts, L. A. 1987. Changes in gliadin proteins during cookie making. Am. Assoc. Cereal Chem., 72nd Annu. Meet., Nashville, Abstr. 140.

Pomeranz, Y., Lookhart, G. L., Rubenthaler, G. L., and Albers, L. 1989. Changes in gliadin proteins during cookie making. Cereal Chem. 66, 532-534.

Poole, P. L., and Finney, J. L. 1983. Sequential hydration of a dry globular protein. Biopolymers

Poole, P. L., and Finney, J. L. 1984. Sequential hydration of dry proteins. Biopolymers 23,

Prime, R. B. 1981. Thermosets. In “Thermal Characterization of Polymeric Materials”

Privalov, P. L., and Gill, S. J. 1988. Stability of protein structure and hydrophobic interaction.

Quinquenet, S., Grabielle-Madelmont, C., Ollivon, M., and Serpelloni, M. 1988. Influence of

Raemy, A., and Lambelet, P. 1991. Thermal behavior of foods. Thermochim. Acta 193,

Rajagopalan, S . , and Seib, P. A. 1992. Granular cold water-soluble starches prepared at atmospheric pressure. J. Cereal Sci. 16, 13-28.

Ramanujam, R., Heaster, J., Huang, C., Jolly, J., Koelbl, J., Lively, C., Ogutu, E., Ting, E., Treml, S., Aldous, B., Hatley, R., Mathias, S. , Franks, F., and Burdick, B. 1993. Ambient- temperature-stable molecular biology reagents. Bio Techniques 14,470-475.

Rao, P. A., Nussinovitch, A., and Chinachoti, P. 1992. Effects of selected surfactants on amylopectin recrystallization and on recoverability of bread crumb during storage. Cereal Chem. 69,613-618.

Reid, D. S. 1985. Correlation of the phase behavior of DMSONaCUwater and glycerol/NaCY water as determined by DSC with their observed behavior on a cryomicroscope. Cryo- Lett. 6, 181-188.

22,255-260.

1647-1666.

(E. A. Turi, ed.), pp. 435-569. Academic Press, Orlando, FL.

Adv. Protein Chem. 39, 191-234.

water on pure sorbitol polymorphism. J. Chem. SOC., Faraday Trans. I 84,2609-2618.

417-439.

Reid, D. S. 1990. Optimizing the quality of frozen foods. Food Technol. 44(7), 78-82. Reid, D. S. 1992a. Glass transition in the freezing process. In “ISOPOW-V-Water in Foods:

Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Abstr. 11.3.

Reid, D. S. 1992b. Water relations of foods, a key to product stability. In “Advances in Food Engineering” (R. P. Sin& and M. A. Wirakartakusumah, eds.), pp. 3-14. CRC Press, Boca Raton, FL.

Reid, D. S., and Charoenrein, S. 1985. DSC studies of starch-water interaction in gelatinization process. In “Proceedings of the 14th Annual NATAS Conference,” pp. 335-340. NATAS, San Francisco.

Reid, D. S., and Hsu, J. 1992. The estimation of the concentration of the maximally freeze- concentrated matrix in a frozen system. J . Food Sci., submitted.

Reid, D. S. , Chen, H., Lim, M. H., and Kerr, W. 1992. The effect of the properties of the glassy state on the kinetics of reactions in frozen systems. Inst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 856.


Reid, D. S., Hsu, J., and Kerr, W. 1993a. Calorimetry. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 123-132. Nottingham Univ. Press, Lough- borough.

Reid, D. S., Kerr, W., and Hsu, J. 1993b. The glass transition in the freezing process. In “Water in Foods: Fundamental Aspects and Their Significance in the Processing of Foods” (D. Simatos, ed.). Elsevier, London; J. Food Eng. 22,483-494 (1994).

Ring, S. G. 1985a. Observations on crystallization of amylopectin from aqueous solution. Znt. J . Biol. Macromol. 7,253-254.

Ring, S. G. 1985b. Studies on starch gelation. Staerke 37, 80-83. Ring, S. G., and Orford, P. D. 1986. Recent observations on retrogradation of amylopectin.

In “Gums and Stabilizers for the Food Industry 3” (G. 0. Phillips, D. J. Wedlock, and P. A. Williams, eds.), pp. 159-165. Elsevier Applied Science, London.

Ring, S. G., and Whittam, M. A. 1991. Linear dextrins-Solid forms and aqueous solution behavior. ACS Symp. Ser. 458,273-293.

Ring, S. G., Colonna, P., l’Anson, K. J., Kalichevsky, M. T., Miles, M. J., Morris, V. J., and Orford, P. D. 1987. Gelation and crystallization of amylopectin. Carbohydr. Res.

Roberts, G. E., and White, E. F. T. 1973. Relaxation processes in amorphous polymers. In “The Physics of Glassy Polymers” (R. N. Haward, ed.), pp. 153-222. Applied Science, London.

Robertson, R. E. 1985. Segmental mobility in the equilibrium liquid below the glass transition. Macromolecules 18, 953-958.

Rogers, D. E., Doescher, L. C., and Hoseney, R. C. 1990. Texture characteristics of reheated bread. Cereal Chem. 67,188-191.

Rohde, F., Normand, M. D., and Peleg, M. 1993. Effect of equilibrium relative humidity on the mechanical signatures of crunchy food materials. Biotechnol. Prog. 9,497-503.

Roland, C. M., and Ngai, K. L. 1991. Dynamical heterogeneity in a miscible polymer blend. Macromolecules 24,2261-2265.

Roland, C. M., and Ngai, K. L. 1992. Normalization of the temperature dependence of segmental relaxation times. Macromolecules 25,5765-5768.

Roos, Y. 1987. Effect of moisture on the thermal behavior of strawberries studied using DSC. J. Food Sci. 52,146-149.

Roos, Y. 1992a. Melting and glass transitions of low molecular weight carbohydrates. Carbo- hydr. Res. 238,39-48 (1993).

Roos, Y. 1992b. Water activity and physical state effects on amorphous food stability. Znst. Food Technol. Annu. Meet., New Orleans, LA, Abstr. 690; J. Food Process. Preserv. 16,

Roos, Y. 1992c. Characterization of food polymers using state diagrams. CoFE ’92-AZChE 1992 Summer Natl. Meet., Minneapolis, MN, Paper 16a; J. Food Eng. 24,339-360 (1995).

Roos, Y. 19926 Phase transitions and transformations in food systems. In “Handbook of Food Engineering” (D. R. Heldman and D. B. Lund, eds.), pp. 145-197. Dekker, New York.

Roos, Y., and Jouppila, K. 1993. Glass transitions and physical state of dehydrated milk products. Int. Congr. Eng. Food, 6th, Chiba, Japan, Abstr. 1.03.

Roos, Y., and Karel, M. 1990. DSC study of phase transitions affecting the quality of dehydrated materials. Biotechnol. Prog. 6,159-163.

Roos, Y., and Karel, M. 1991a. Plasticizing effect of water on thermal behavior and crystallization of amorphous food models. J. Food Sci. 56,38-43.

Roos, Y., and Karel, M. 1991b. Phase transitions of mixtures of amorphous polysaccharides and sugars. Biotechnol. Prog. 7,49-53.

Roos, Y., and Karel, M. 1991c. Phase transitions of amorphous sucrose and frozen sucrose solutions. J. Food Sci. 56, 266-267.

162,277-293.

433-447 (1993).


Roos, Y., and Karel, M. 1991d. Water and molecular weight effects on glass transitions in amorphous carbohydrates and carbohydrate solutions. J . Food Sci. 56, 1676-1681.

Roos, Y., and Karel, M. 1991e. Applying state diagrams to food processing and development. Food Technol. 45(12), 66-71,107.

Roos, Y., and Karel, M. 1991f. Non-equilibrium ice formation in carbohydrate solutions. Cryo-Lett. 12, 367-376.

Roos, Y., and Karel, M. 1991g. Amorphous state and delayed ice formation in sucrose solutions. Int. J . Food Sci. Technol. 26,553-566.

Roos, Y., and Karel, M. 1991h. Thermal history and properties of frozen carbohydrate solutions. Proc. Int Congr. Refrig., 18th, 1991, Vol. 4, pp. 1829-1833.

Roos, Y., and Karel, M. 1992. Crystallization of amorphous lactose. J. Food Sci. 57,775-777. Roos, Y., and Karel, M. 1993. Effects of glass transitions on dynamic phenomena in sugar-

containing food systems. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 207-222. Nottingham Univ. Press, Loughborough.

Roozen, M. J. G. W., and Hemminga, M. A. 1990. Molecular motion in sucrose-water mixtures in the liquid and glassy state as studied by spin probe ESR. J. Phys. Chem. 94,7326-7329.

Roozen, M. J. G. W., and Hemminga, M. A. 1991. Molecular motion in carbohydrate + water mixtures in the liquid and glassy states as studied by spin-probe ESR. In “Food Polymers, Gels and Colloids” (E. Dickinson, ed.), pp. 531-536. Royal Society Chemistry, Cam- bridge, UK.

Roozen, M. J. G. W., Hemminga, M. A., and Walstra, P. 1991. Molecular motion in glassy water-maltooligosaccharide (maltodextrin) mixtures as studied by conventional and saturation-transfer spin-probe ESR spectroscopy. Carbohydr. Rex 215,229-237.

Roper, H., and Koch, H. 1990. The role of starch in biodegradable thermoplastic materials. Staerke 42, 123-130.

Roser, B. 1991a. Trehalose, a new approach to premium dried foods. Trends Food Sci. Technol.

Roser, B. 1991 b. Trehalose drying: A novel replacement for freeze-drying. BioPharm 4(8),

Roser, B., and Colaco, C. 1993. A sweeter way to fresher food. New Sci. 138(1873), 24-28. Rowland, S. P., ed. 1980. “Water in Polymers,” ACS Symp. Ser. 127. Am. Chem. SOC.,

Washington, DC. Roy, M. L., Pikal, M. J., Rickard, E. C., and Maloney, A. M. 1992. The effects of formulation

and moisture on the stability of a freeze-dried monoclonal antibody-vinca conjugate: A test of the WLF glass transition theory. Dev. Biol. Stand. 74, 323-340.

Rubin, C. A., Wasylyk, J. M., and Baust, J. G. 1990. Investigation of vitrification by NMR and DSC in honey: A model carbohydrate system. J. Agric. Food Chem. 38,1824-1827.

Russell, P. L. 1987a. Gelatinization of starches of different amyloselamylopectin content- DSC study. J. Cereal Sci. 6 , 133-145.

Russell, P. L. 1987b. Aging of gels from starches of different amylose/amylopectin content studied by DSC. J. Cereal Sci. 6, 147-158.

Russell, P. L., and Oliver, G. 1989. The effect of pH and NaCl content on starch gel aging. A study by DSC and rheology. J. Cereal Sci. 10, 123-138.

Sahagian, M. E., and Goff, H. D. 1992. Glass transitions of sucrose and polysaccharide solutions and their relevance to food stability. Proc. Can. Chem. Eng. Conf, 42nd, pp. 301-302.

Sahagian, M. E., and Goff, H. D. 1993. Stabilizer-sucrose interactions at subzero temperatures. Can. Inst. Food Sci. Technol. Annu. Meet., Toronto. Thermochim. Acta, in press.

Sala, R., and Tomka, I. 1992a. Water uptake in partially frozen and plasticized starch. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 483-489. Nottingham Univ. Press, Loughborough (1993).

2(7), 166-169.

47-53.


Sala, R., and Tomka, I. 1992b. On the equation of state of thermoplastic starch. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 475-482. Nottingham Univ. Press, Loughborough (1993).

Saleeb, F. Z., and Pickup, J. G. 1985. Fixing volatiles in an amorphous substrate and products therefrom. U.S. Pat. 4,532,145.

Saleeb, F. Z., and Pickup, J. G. 1989. Fivation of volatiles in extruded glass substrates. U.S. Pat. 4,820,534.

Salmen, N. L., and Back, E. L. 1977. Influence of water on Tg of cellulose. Tappi 60,137-140. Sapru, V., and Labuza, T. P. 1992a. Temperature dependence of thermal inactivation rate

constants of Bacillus stearothermophilus spores in a glassy state. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 499-505. Nottingham Univ. Press, Loughborough.

Sapru, V., and Labuza, T. P. 1992b. Glassy state in bacterial spores predicted by polymer glass-transition theory. J . Food Sci. 58, 445-448.

Saunders, S. R., Hamann, D. D., and Lineback, D. R. 1992. A systems approach to food material adhesion. Lebensm- Wiss. Technol. 25,309-315.

Sauvageot, F., and Blond, G. 1991. Effect of water activity on crispness of breakfast cereals. J . Texture Stud. 22, 423-442.

Scandola, M., Ceccorulli, G., and Pizzoli, M. 1991. Molecular motions of polysaccharides in the solid state: Dextran, pullulan, and amylose. Znt. J . Biol. Macromol. 13,254-260.

Schenz, T. W., Rosolen, M. A., Levine, H., and Slade, L. 1984. DMA of frozen aqueous solutions. In “Proceedings of the 13th Annual NATAS Conference” (A. R. McGhie, ed.), pp. 57-62. NATAS, Philadelphia.

Schenz, T. W., Israel, B., and Rosolen, M. A. 1991. Thermal analysis of water-containing systems. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 199-214. Plenum, New York.

Schenz, T. W., Courtney, K., and Israel, B. 1992. Studies on the magnitude of the glass transitions of glucose polymers. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 469-474. Nottingham Univ. Press, Loughborough; Cryo-Lett. 14,

Schiraldi, A. 1990. Calorimetry, thermal analysis and chemical thermodynamics in food science. Thermochim. Acta 162,253-264.

Schlotter, N. E. 1990. Diffusion of small molecules in glassy polymer thin films studied by waveguide Raman techniques. J. Phys. Chem. 94,1692-1699.

Schofield, J. D. 1986. Flour proteins: Structure and functionality in baked products. In “Chemis- try and Physics of Baking” (J. M. V. Blanshard, P. J. Frazier, and T. Galliard, eds.), pp. 14-29. Royal Society of Chemistry, London.

Schofield, J. D., Bottomley, R. C., Timms, M. F., and Booth, M. R. 1983. The effect of heat on wheat gluten and the involvement of sulphydryl-disulphide interchange reactions. J . Cereal Sci. 1,241-253.

Schofield, J. D., Bottomley, R. C., LeGrys, G. A., Timms, M. F., and Booth, M. R. 1984. Effects of heat on wheat gluten. In “Gluten Proteins” (A. Graveland and J. H. E. Moonen, eds.), pp. 81-90. TNO, Wageningen, The Netherlands.

Schroeter, J., and Hobelsberger, M. 1992. On the mechanical properties of native starch granules. Staerke 44,247-252.

Schwartz, A. 1960. “Analytic Geometry and Calculus.” Holt, Rinehart & Winston, New York. Sears, J. K., and Darby, J. R. 1982. “The Technology of Plasticizers.” Wiley Interscience,

Seow, C. C., and Teo, C. H. 1993. Annealing of granular rice starches: Interpretation of the

91-96 (1993).

New York.

effect on phase transitions associated with gelatinization. Staerke 45,345-351.


Seow, C. C., and Thevamalar, K. 1993. Internal plasticization of granular rice starch by hydroxypropylation: Effects on phase transitions associated with gelatinization. Staerke

Shah, N. K., and Ludescher, R. D. 1992. Hydration and the internal dynamics of hen egg white lysozyme. Proc. SPIE-h i . SOC. Opt. Eng. 1640, 174-179.

Shah, N., and Ludescher, R. 1993. Phosphorescence probes of the glass transition. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 622.

Shalaby, S. W. 1981. Thermoplastic polymers. In “Thermal Characterization of Polymeric Materials” (E. A. Turi, ed.), pp. 235-264. Academic Press, Orlando, FL.

Shalaev, E. Y., and Kanev, A. N. 1993. Study of the solid-liquid state diagram of the water-glycine-sucrose system. Cryobiology 31,374-382 (1994).

Sharples, A. 1972. Crystallinity. In “Polymer Science” (A. D. Jenkins, ed.), pp. 251-321. North-Holland Publ., Amsterdam.

Shewry, P. R., Halford, N. G., Field, J. M., and Tatham, A. S. 1988. The structure and functionality of wheat gluten proteins. In “Proceedings of the 38th Australian Cereal Chemistry Conference” (L. Murray, ed.), pp. 113-116. Cereal Chem. Div., Royal Austra- lian Chem. Inst., Sydney.

Shi, Y. C., and Seib, P. A. 1989. Properties of wheat starch compared to normal maize starch. In “Wheat is Unique” (Y. Pomeranz, ed.), pp. 215-234. Am. Assoc. Cereal Chem., St. Paul.

Shi, Y. C., and Seib, P. A. 1992. The structure of four waxy starches related to gelatinization and retrogradation. Carbohydr. Res. 227,131-145.

Shimada, Y., Roos, Y., and Karel, M. 1991. Oxidation of methyl linoleate encapsulated in amorphous lactose-based food model. J. Agric. Food Chem. 39,637-641.

Shogren, R. L. 1992. Effect of moisture content on the melting and subsequent physical aging of corn starch. Carbohydr. Polym. 19,83-90.

Shogren, R. L., Swanson, C. L., and Thompson, A. R. 1992. Extrudates of corn starch with urea and glycols: structure/mechanical property relations. Staerke 44,335-338.

Shukla, T. P. 1991. A new development in carbohydrate research. Cereal Foods World 36,

Sichina, W. J. 1988. Predicting mechanical performance and lifetime of polymeric materials. Am. Lab. 20(1), 42-52.

Simatos, D. 1994. Influence of macromolecules on glass transition in frozen systems. Food Macromol. Colloids, Dijon, France, 1994.

Simatos, D., and Blond, G. 1991. DSC studies and stability of frozen foods. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 139-155. Plenum, New York.

Simatos, D., and Blond, G. 1993. Some aspects of the glass transition in frozen food systems. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 395-415. Nottingham Univ. Press, Loughborough.

Simatos, D., and Karel, M. 1988. Characterizing condition of water in foods: Physico-chemical aspects. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 1-41. Elsevier Applied Science, London.

Simatos, D., Blond, G., and Le Meste, M. 1989. Relation between glass transition and stability of a frozen product. Cryo-Lett. 10,77-84.

Skarra, L. L., Ghiasi, K., Evans, J. R., and Graf, E. 1988. Tortilla and method of manufacture. US. Pat. 4,735,811.

Skemtt, I. H., Temperley, L. G., and Lew, P. Y. 1988. Probing cereal grain protein structure using monoclonal antibodies. Am. Assoc. Cereal Chem., 73rd Annu. Meet., San Diego, Abstr. 22.

Slade, L. 1984. Starch properties in processed foods: Staling of starch-based products. Am. Assoc. Cereal Chem., 69th Annu. Meet., Minneapolis, MN., Abstr. 112.

45,85-88.

251-252.


Slade, L., and Levine, H. 1984. Thermal analysis of starch and gelatin. I n “Proceedings of the 13th Annual NATAS Conference” (A. R. McGhie, ed.), p. 64. NATAS, Philadelphia.

Slade, L., and Levine, H. 1985. Intermediate moisture systems; concentrated and supersaturated solutions; pastes and dispersions; water as plasticizer; the mystique of “bound” water; thermodynamics versus kinetics. In “Faraday Division, Royal Society of Chemistry Discussion Conference-Water Activity: A Credible Measure of Technological Perfor- mance and Physiological Viability?” pp. 24-27. Cambridge University, Cambridge, UK.

Slade, L., and Levine, H. 1987a. Polymer-chemical properties of gelatin in foods. In “Advances in Meat Research” (A. M. Pearson, T. R. Dutson, and A. Bailey, eds.), Vol. 4, pp. 251-266. AVI, Westport, CT.

Slade, L., and Levine, H. 1987b. Recent advances in starch retrogradation. In “Industrial Polysaccharides” (S. S. Stivala, V. Crescenzi, and I. C. M. Dea, eds.), pp. 387-430. Gordon & Breach, New York.

Slade, L., and Levine, H. 1988a. Structural stability of intermediate moisture foods-A new understanding? In “Food Structure: Its Creation and Evaluation” (J. M. V. Blanshard and J. R. Mitchell, eds.), pp. 115-147. Butterworth, London.

Slade, L., and Levine, H. 1988b. Non-equilibrium behavior of small carbohydrate-water systems. Pure Appl. Chem. 60,1841-1864.

Slade, L., and Levine, H. 1988~. Non-equilibrium melting of native granular starch. I. Tempera- ture location of the glass transition associated with gelatinization of A-type cereal starches. Carbohydr. Polym. 8,183-208.

Slade, L., and Levine, H. 1988d. Thermal analysis of starch. I n “1988 CRA Scientific Confer- ence,’’ pp. 169-244. Corn Refiners Association, Washington, DC.

Slade, L., and Levine, H. 1989. A food polymer science approach to selected aspects of starch gelatinization and retrogradation. In “Frontiers in Carbohydrate Research” (R. P. Millane, J. N. BeMiller, and R. Chandrasekaran, eds.), Vol. 1, pp. 215-270. Elsevier Applied Science, London.

Slade, L., and Levine, H. 1990. Structure-function relationships of sucrose and flours in cookie systems. Am. Assoc. Cereal Chem., 75th Annu. Meet., Dallas, Abstr. 28.

Slade, L., and Levine, H. 1991a. Beyond water activity: Recent advances basedon an alternative approach to the assessment of food quality and safety. CRC Crit. Rev. Food Sci. Nutr.

Slade, L., and Levine, H. 1991b. A polymer science approach to structure/property relationships in aqueous food systems: Non-equilibrium behavior of carbohydrate-water systems. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 29-101. Plenum, New York.

Slade, L., and Levine, H. 1992a. A polymer science approach to glass transitions in cereal- based foods. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, MN, Abstr. 58.

Slade, L., and Levine, H. 1992b. Dependence of the glass transition on composition and chemical structure: Special implications for flour functionality in cookie baking. In “ISOPOW-V-Water in Foods: Fundamental Aspects and Their Significance in Relation to Processing of Foods, Valencia, Spain,” Abstr. 3.3.

Slade, L., and Levine, H. 199%. Structure-function relationships of cookie and cracker ingredients. In “The Science of Cookie and Cracker Production” (H. Faridi, ed.), pp. 23-141. Chapman & Hall, New York (1994).

Slade, L., and Levine, H. 1993a. Water relationships in starch transitions. In “Workshop on Cereal Polysaccharides.” INRA-Nantes, LeCroisic, France; Carbohydr. Polym., 21,

Slade, L., and Levine, H. 1993b. Water and the glass transition-Dependence of the glass transition on composition and chemical structure: Special implications for flour functional-

30,115-360.

105-131.


ity in cookie baking. In “Water in Foods: Fundamental Aspects and Their Significance in the Processing of Foods” (P. Fito, A. Mulet, and B. McKenna, eds.), pp. 143-188. Elsevier, London (1994); J. Food Eng. 22,143-188 (1994).

Slade, L., and Levine, H. 1993c. The glassy state phenomenon in food molecules. I n “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 35-101. Nottingham Univ. Press, Loughborough.

Slade, L., Altomare, R., Oltzik, R., and Metcalf, D. G. 1987. Accelerated staling of starch- based food products. U.S. Pat. 4,657,770.

Slade, L., Levine, H., and Finley, J. W. 1989. Protein-water interactions: Water as a plasticizer of gluten and other protein polymers. In “Protein Quality and the Effects of Processing” (R. D. Phillips and J. W. Finley, eds.), pp. 9-124. Dekker, New York.

Slade, L., Levine, H., Ievolella, J., and Wang, M. 1993. The glassy state phenomenon in applications for the food industry. Application of the food polymer science approach to structure-function relationships of sucrose in cookie and cracker systems. J. Sci. Food Agric. 63, 133-176.

Smith, A. C. 1990. Studies on the physical structure of starch-based materials in the extrusion cooking process. Extrusion Rheol. Foods Symp., Rutgers University, New Brunswick, NJ.

Smith, A. C. 1992a. Studies on the physical structure of starch-based materials in the extrusion cooking process. I n “Food Extrusion Science and Technology” (J. L. Kokini, C. T. Ho, and M. V. Karwe, eds.), pp. 573-618. Dekker, New York.

Smith, A. C. 1992b. Mechanical properties of formed starch. Am. Assoc. Cereal Chem., 77th Annu. Meet., Minneapolis, MN, Abstr. 206.

Smith, A. C. 1992c. The structure and rheology of processed starch materials. I n “Advances in Food Engineering” (R. P. Singh and M. A. Wirakartakusumah, eds.), pp. 393-407. CRC Press, Boca Raton, FL.

Sobczynska, D., Setser, C., Lim, H., Hansen, L., and Paukstelis, J. 1990. Use of C13 and 017 NMR to study wheat starch-water-sugar interactions with increasing temperatures. I n “NMR Applications in Biopolymers” (J. W. Finley, S. J. Schmidt, and A. S. Serianni, eds.), pp. 461-480. Plenum, New York.

Sochava, I. V., and Belopolskaya,T. V. 1992. Thermally induced globular protein gels: Peculiar- ities of formation, melting, and restoration of gels of different structure. Food Hydrocol- loids 6, 97-114.

Sochava, I. V., and Smirnova, 0. I. 1992. Heat capacity of hydrated and dehydrated globular proteins. Denaturation increment of heat capacity. Food Hydrocolloids 6,513-524 (1993).

Sochava, I. V., Tsereteli, G. I., and Smirnova, 0. I. 1991. Denaturation increment of heat capacity in globular proteins and its relation to glass transition. Biofizika 36,432-436.

Soesanto, T., and Williams, M. C. 1981. Volumetric interpretation of viscosity for concentrated and dilute sugar solutions. J . Phys. Chem. 85,3338-3341.

Sperling, L. H. 1986. “Introduction to Physical Polymer Science.” Wiley (Interscience), New York.

Spratt, W. A., Paton, D., and Timbers, G. E. 1991. Microwave-puffable half-products of starch- containing material and their production process. U.S. Pat. 4,990,348.

Starkweather, H. W. 1980. Water in nylon. ACS Symp. Ser. 127,433-440. Stauffer, C. E. 1990. “Functional Additives for Bakery Foods.” AVI, New York. Stauffer, C. E. 1992. The sweet dimension. Baking & Snack 14(5), 24-28. Stepto, R. F. T., and Dobler, B. 1989. Method of producing destructurized starch. Eur. Pat.

Strang, G. 1991. “Calculus.” Wellesley-Cambridge Press, Wellesley, MA. Studt, T. 1992. Polyimides: Hot stuff for the ’90s. R&D Mag., August, pp. 30-34.

Appl. EP 0326517Al.


Stute, R. 1992. Hydrothermal modification of starches: The difference between annealing and heathoisture-treatment. Staerke 44,205-214.

Sullivan, C. R., Corona, A., and Rollings, J. E. 1992. Chromatographic technologies for macromolecular starch characterization. In “Developments in Carbohydrate Chemistry” (R. J. Alexander and H. F. Zobel, eds.), pp. 193-238. Am. Assoc. Cereal Chem., St. Paul, MN.

Summers, F. 1992. Dynamic mechanical analysis applications of natural products. 83rd Am. Oil Chem. SOC. Annu. Meet., Toronto, Abstr. 30-5.

Suzuki, T., and Franks, F. 1993. Solid-liquid phase transitions and amorphous states in ternary sucrose-glycine-water systems. J. Chem. Soc., Faraday Trans., in press.

Takahashi, T., Hirsh, A., Erbe, E., and Williams, R. J. 1988. Mechanisms of cryoprotection by extracellular polymeric solutes. Biophys. J . 54,509-518.

Takushi, E., Asato, L., and Nakada, T. 1990. Edible eyeballs from fish. Nature (London)

Tanner, S . F., Ring, S. G., Whittam, M. A., and Belton, P. S. 1987. High resolution solid state C13 NMR study of some alpha (1-4) linked glucans: The influence of water on structure and spectra. Int. J. Biol. Macromol. 9,219-224.

Tanner, S. F., Hills, B. P., and Parker, R. 1991. Interactions of sorbed water with starch studied using proton NMR spectroscopy. J. Chem. Soc., Faraday Trans. 87,2613-2621.

Tao, H. P., Cornell, D. G., and Kasarda, D. D. 1989. Surface and optical properties of wheat glutenin monolayers. J. Cereal Sci. 10, 5-18.

Tardos, G., Mazzone, D., and Pfeffer, R. 1984. Measurement of surface viscosities using a dilatometer. Can. J. Chem. Eng. 62, 884-887.

Tatham, A. S., Miflin, B. J., and Shewry, P. R. 1985. Beta-turn conformation in wheat gluten proteins and gluten elasticity. Cereal Chem. 62,405-412.

Taylor, T. A., Singh, R. P., and Reid, D. S. 1992. Modeling retrogradation: A unified food polymer science approach. Inst. Food Technol., Annu. Meet., New Orleans, LA, Abstr. 696A.

te Booy, M. P. W. M., de Ruiter, R. A., and de Meere, A. L. J. 1992. Evaluation of the physical stability of freeze-dried sucrose-containing formulations by DSC. Pharm. Res.

Thompson, D. B. 1992. Structure and functionality of carbohydrate hydrocolloids in food systems. I n “Developments in Carbohydrate Chemistry” (R. J. Alexander and H. F. Zobel, eds.), pp. 315-342. Am. Assoc. Cereal Chem., St. Paul, MN.

Tian, W., and Blanshard, J. M. V. 1992a. The molecular reorganization of amorphous glucose during the glass-rubber transition-An FTIR spectroscopic study. Appl. Spectrosc., in press.

Tian, W., and Blanshard, J. M. V. 1992b. Toughening of bread by microwave heating-A physicochemical study by ‘H NMR. J. Cereal Sci., submitted.

Tian, W., and Blanshard, J. M. V. 1992c. The molecular reorganization of glassy glucose during the glass-rubber transition. Easter Sch., Sci. Technol. Glassy State Foods, 53rd, Nottingham University, UK, Abstr. P15.

To, E. C., and F l i , J. M. 1978. “Collapse,” a structural transition in freeze-dried carbohydrates. J . Food Technol. 13,551-594.

Tolstoguzov, V. B. 1991. Functional properties of food proteins and role of protein-polysaccharide interaction. Food Hydrocolloids 4,429-468.

Tolstoguzov, V. B. 1992. Thermoplastic extrusion-The mechanism of the formation of extrudate structure and properties. J. Am. Oil Chem. SOC., in press.

Tolstoguzov, V. B., and Nesmeyanov, A. N. 1992. Thermoplastic extrusion-The mechanism of the formation of extrudate structure and properties. 83rd Am. Oil Chem. SOC. Annu. Meet., Toronto, Abstr. 30-4.

345,298-299.

9,109-114.


Tomka, 1.1986. Thermodynamic theory of polar polymer solutions. Polym. Prepr., Am. Chem. SOC., Div. Polym. Chem. 27(2), 129.

Tomka, I. 1991. Thermoplastic starch. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 627-637. Plenum, New York.

Tomka, I., Bohonek, J., Spuhler, A., and Ribeaud, M. 1975. Structure and formation of the gelatin gel. J. Photogr. Sci. 23, 97-103.

Tsourouflis, S., Flink, J. M., and Karel, M. 1976. Loss of structure in freeze-dried carbohydrate solutions. J. Sci. Food Agric. 27, 509-519.

Turi, E. A. 1981. “Thermal Characterization of Polymeric Materials.” Academic Press, Or- lando, FL.

van den Berg, C. 1981. Vapour sorption equilibria and other water-starch interactions: A physico-chemical approach. Doctoral thesis, Agricultural University, Wageningen, The Netherlands.

van den Berg, C. 1986. Water activity. In “Concentration and Drying of Foods” (D. MacCarthy, ed.), pp. 11-36. Elsevier Applied Science, London.

van den Berg, C. 1991. Food-water relations: Progress and integration, comments and thoughts. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 21-28. Plenum, New York.

van den Berg, C. 1992. Glass transitions in carbohydrates. Curbohydr. Neth. 8,23-25. van den Berg, C., Franks, F., and Echlin, P. 1992. Glass transitions and structure of freeze-

dried sugars, in view of stabilization. Stab. Stubil. Enzymes Conf, Maastricht, Netherlands. van den Berg, C., Franks, F., and Echlin, P. 1993. The ultrastructure and stability of amorphous

sugars. In “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 249-267. Nottingham Univ. Press, Loughborough.

Van Scoik, K. G., and Carstensen, J. T. 1990. Nucleation phenomena in amorphous sucrose systems. Inr. J . Pharm. 58, 185-196.

Van Vliet, T., Janssen, A. M., Bloksma, A. H., and Walstra, P. 1992. Strain hardening of dough as a requirement for gas retention. J. Texture Stud. 23,439-460.

Vertucci, C. W. 1990. Calorimetric studies of the state of water in seed tissues. Biophys. J.

Wallack, D. A., and King, C. J. 1988. Sticking and agglomeration of hygrosopic, amorphous carbohydrate and food powders. Biotechnol. Prog. 431-35.

Walton, A. G. 1969. Nucleation in liquids and solutions. In “Nucleation” (A. C. Zettlemoyer, ed.), p. 225. Dekker, New York.

Wang, S., and Chinachoti, P. 1993. Phase transition of fructose, glucose, and sucrose upon water sorption. Inst. Food Technol. Annu. Meet., Chicago, Abstr. 619.

Warburton, S. C., Donald, A. M., and Smith, A. C. 1990. The deformation of brittle starch foams. J. Muter. Sci. 25,4001-4007.

Warburton, S . C., Donald, A. M., and Smith, A. C. 1992. Structure and mechanical properties of brittle starch foams. J. Muter. Sci. 27, 1469-1474.

Wasylyk, J. M., and Baust, J. G. 1991. Vitreous domains in an aqueous ribose solution. In “Water Relationships in Foods” (H. Levine and L. Slade, eds.), pp. 225-234. Plenum, New York.

Watase, M., Nishinari, K., and Hatakeyama, T. 1992. DSC study on properties of water in concentrated agarose gels. Food Hydrocolloids, in press.

Weegels, P. L. 1991. Sensitive measurement of contents and reactivities of thiol and disulphide in soluble and insoluble proteins. In “Gluten Proteins 1990” (W. Bushuk and R. Tkachuk, eds.), pp. 414-419. Am. Assoc. Cereal Chem., St. Paul, MN.

Weegels, P. L., and Hamer, R. J. 1991. Improving the baking quality of wheat gluten. In “Gluten Proteins 1990” (W. Bushuk and R. Tkachuk, eds.), pp. 219-223. Am. Assoc. Cereal Chem., St. Paul, MN.

58, 1463-1471.


Weegels, P. L., Verhoek, A., and Hamer, R. J. 1988. A comprehensive study of the effect of heat on wheat gluten. Am. Assoc. Cereal Chem., 73rd Annu. Meet., San Diego, Abstr. 133.

Weegels, P. L., Marseille, J. P., de Jager, A. M., and Hamer, R. J. 1991. Structure-function relationships of gluten proteins. In “Gluten Proteins 1990” (W. Bushuk and R. Tkachuk, eds.), pp. 98-111. Am. Assoc. Cereal Chem., St. Paul, MN.

Wesson, J. A., Takezoe, H., Yu, H., and Chen, S . P. 1982. Dye diffusion in swollen gels by forced Rayleigh scattering. J. Appl. Phys. 53, 6513-6519.

Whewell, C. S. 1977. The chemistry of wool finishing. I n “Chemistry of Natural Protein Fibers” (R. S. Asquith, ed.), pp. 333-370. Plenum, New York.

White, G. W., and Cakebread, S . H. 1966. The glassy state in certain sugar-containing food products. J. Food Technol. 1,73-82.

Whittam, M. A. 1992. Dynamics of starch- and maltose-water mixtures. In “Workshop on Cereal Polysaccharides.” INRA-Nantes, Le Croisic, France.

Whittam, M. A., Noel, T. R., and Ring, S. G. 1990. Melting behavior of A- and B-type crystalline starch. Int. J. Biol. Macromol. 12, 359-362.

Whittam, M. A., Noel, T. R., and Ring, S . G. 1991. Melting and glasshbber transitions of starch polysaccharides. In “Food Polymers, Gels and Colloids” (E. Dickinson, ed.), pp. 277-288. Royal Society of Chemistry, Cambridge, UK.

Willenbucher, R. W., Tomka, I., and Muller, R. 1993. Thermally induced structural transitions in the starch-water system. I n “The Glassy State in Foods” (J. M. V. Blanshard and P. J. Lillford, eds.), pp. 491-497. Nottingham Univ. Press, Loughborough; in “Carbohy- drates in Industrial Synthesis” (M. A. Clarke, ed.). Verlag, Berlin.

Williams, M. L., Landel, R. F., and Ferry, J. D. 1955. Temperature dependence of relaxation mechanisms in amorphous polymers and other glass-forming liquids. J . Am. Chem. SOC.

Williams, R. J. 1991. Changes in glass transition temperatures during germination of pea

Williams, R. J., and Camahan, D. L. 1989. Association between ice nuclei and fracture

Williams, R. J., and Camahan, D. L. 1990. Fracture faces and other interfaces as ice nucleation

Williams, R. J., and Leopold, A. C. 1989. Glassy states in dormant corn embryos. Thermochim.

Williams, R. J., Mehl, P., and Carnahan, D. L. 1990. Ice nucleation and growth in vitrifiable solutions. In “Cryopreservation and Low Temperature Biology in Blood Transfusion” (C. T. Smit Sibinga, P. C. Das, and H. T. Meryman, eds.), pp. 71-86. Kluwer, Dordrecht, The Netherlands.

Williams, R. J., Hirsh, A. G., Takahashi, T. A., and Meryman, H. T. 1993. What is vitrification and how can it extend life? Jpn. J. Freezing Drying, in press.

Wittwer, F., and Tomka, I. 1984. Polymer composition for injection moulding. Eur. Pat. Appl. 0118240A2.

Wunderlich, B. 1973. “Macromolecular Physics,” Vol. 1. Academic Press, New York. Wunderlich, B. 1976. “Macromolecular Physics,” Vol. 2. Academic Press, New York. Wunderlich, B. 1980. “Macromolecular Physics,” Vol. 3. Academic Press, New York. Wunderlich, B. 1981. The basis of thermal analysis. I n “Thermal Characterization of Polymeric

Wunderlich, B. 1990. “Thermal Analysis.” Academic Press, Boston. Wyman, J., and Gill, S. J. 1990. “Binding and Linkage.” University Science Books, Mill

77,3701-3706.

embryos. Plant Physiol. %S, 62.

interfaces in sucrose : water glasses. Thermochim. Acta 155, 103-107.

sites. Cryobiology 27, 419-482.

AC~U 155, 109-114.

Materials” (E. A. Turi, ed.), pp. 91-234. Academic Press, Orlando, FL.

Valley, CA.


Xenopoulos, A,, Roles, K., and Wunderlich, B. 1992. A possible glass transition for poly(L-

Yannas, I. V. 1972. Collagen and gelatin in the solid state. J. Macromol. Sci., Rev. Macromol.

Yano, S. 1993. Dynamic viscoelastic properties of carboxymethylcellulose during isothermal

Yoshida, H., Hatakeyama, T., and Hatakeyama, H. 1990. Phase transitions of the water-xan-

Yoshida, H., Hatakeyama, T., and Hatakeyama, H. 1992. Effect of water on the main chain

Yost, D. A,, and Hoseney, R. C. 1986. Annealing and glass transition of starch. Staerke

Young, R. D., and Scholl, R. 1991. Proteins as complex systems. J. Non-Cryst. Solids

Zanoni, B., Schiraldi, A., and Simonetta, R. 1993. A naive model of starch gelatinization kinetics. J. Food Eng. 24, 25-33 (1994).

Zasypkin, D. V., Yuryev, V. P., Alexeyev, V. V., and Tolstoguzov, V. B. 1992. Mechanical properties of the products obtained by the thermoplastic extrusion of potato starch-soybean protein mixtures. Curbohydr. Polym. 18, 119-124.

Zeleznak, K. J., and Hoseney, R. C. 1987a. The glass transition in starch. Cereal Chem. 64,121-124.

Zeleznak, K. J., and Hoseney, R. C. 1987b. Characterization of starch from bread aged at different temperatures. Stuerke 39,231-233.

Zhang, W., and Jackson, D. S. 1992. Retrogradation behavior of wheat starch gel with differing molecular profiles. J. Food Sci. 57, 1428-1432.

Zobel, H. F. 1988. Starch crystal transformations and their industrial importance. Stuerke

Zobel, H. F. 1992a. Starch granule structure. In “Developments in Carbohydrate Chemistry” (R. J. Alexander and H. F. Zobel, eds.), pp. 1-36. Am. Assoc. Cereal Chem., St. Paul, MN.

Zobel, H. F. 1992b. Starch: Sources, production, and properties. In “Starch Hydrolysis Prod- ucts” (F. W. Schenck and R. E. Hebeda, eds.), pp. 23-44. VCH, New York.

Zobel, H. F., Young, S. N., and Rocca, L. A. 1988. Starch gelatinization: An X-ray diffraction study. Cereal Chem. 65,443-446.

methionine) and poly(L-serine). Polymer 34,2559-2563 (1993).

Chem. C7,49-104.

water sorption. Polymer 34,2528-2532.

than system. Polymer 31,693-698.

motion of polysaccharide hydrogels. ACS Symp. Ser. 489,217-230.

38,289-292.

131-133,302-309.

40,1-7.

This Page Intentionally Left Blank

ADVANCES IN FOOD AND NUTIUTION RESEARCH, VOL. 38

CORN WET MILLING: SEPARATION CHEMISTRY AND TECHNOLOGY

DAVID S. JACKSON AND DONALD L. SHANDERA, JR.

Department of Food Science and Technology University of Nebraska-Lincoln

Lincoln, Nebraska 68583

I.

11. 111.

IV. V. VI. VII.

VIII.

Introduction A. Basic Process and Products B. Industry Scale Corn: Structure and Types Used Steeping: Process and Equipment A. Cleaning B. Steeping C. Steepwater Absorption D. Role of Lactic Acid E. Role of Sulfur Dioxide F. Kernel Degradation G. Role of Temperature H. Biochemical Effects Milling and Final Processing Laboratory versus Commercial Milling Research to Improve Wet Milling End Products A. Products Derived from Starch B. Nonstarch By-products Summary References

I. INTRODUCTION

A. BASIC PROCESS AND PRODUCTS

The typical corn kernel (Zea mays L) contains approximately 70 to 73% starch, 9 to 10% protein, 4 to 5% fat, 1 to 2% ash, 2% sugars, and 9 to 10%

271

Copyright Q 1995 by Academic Press, lee. All rights of reproduction in any form reserved.

272 DAVID S. JACKSON AND DONALD L. SHANDERA, JR.

crude fiber. The purpose of corn wet milling is to separate the kernel into its constituent chemical components. Wet milling processing begins with steeping whole kernel corn in an aqueous solution of sulfur dioxide and lactic acid (produced by microorganisms) at 50°C for 24 to 48 hours. The corn is then coarsely ground and the lipid-containing germ and fibrous hull portions are separated. After the remaining components are more finely ground, the starch and protein are separated using hydrocyclones, essentially continuous centrifuges; corn starch is slightly more dense than corn protein. Germ is further processed into oil and the protein and fiber components are usually blended and used as animal feeds. The wet starch is either dried, chemically modified to change its functional properties, converted into intermediate-sized glucose polymers, or fully depolymerized into sugars. Starch is also often used as a raw ingredient for adjacent processing facilities that produce ethanol or other alcohols and other industrial chemicals. The wet milling process is depicted in Fig. 1.

Shel led C o r n

Corn Cleaners 5.

Centrifugal Hydroclone Mills + Screens +Separators Starch Washing

Starch Fiber (Zein) Slurry

J. 4 Gluten

3. Separators

Steep + Germ + Grinding Tanks

5.

Steepwater Germ + 5.

Refineries

Sweeteners, Alcohols. F e e d Products

Organic Acids

Starch.Driers Modification Tanks

4 Starch Driers

Roasters

Dextrins w Unmodi f ied Modi f ied Starch S t a r c h

FIG. 1. Wet milling process flow. Based on Anonymous (1994). Used with permission. Artwork modified by David S. Jackson.

CORN WET MILLING 273

B. INDUSTRY SCALE

Approximately 7000 to 9000 million bushels of corn are produced yearly in the United States; 20% is used for domestic food and industrial purposes (including wet milling) and 60% for feed. The remainder is exported or stored. Wet milling is a capital- and energy-intensive process: A typical small plant costs approximately $100 million to build and bring on-line and processes more than 50,000 bushels of corn daily. The wet milling industry purchases approximately $5 billion (1992 US$) of corn, goods, and services each year (Anonymous, 1993).

II. CORN: STRUCTURE AND TYPES USED

Zea mays L . (corn) has been the major source of refined starch in the United States and most Western countries since the mid-1800s (Watson, 1984). Corn is a relatively cheap commodity ($2.00 to $2.50 per Winchester bushel) and found in good supply. High-ash-content wheat flour is the second largest starch source, but its production is small in comparison to corn. Sorghum was once wet milled for starch, but because corn has a higher milling efficiency, sorghum is no longer being wet milled in the United States. A thorough knowledge of the basic structure of the corn kernel is essential to understanding and ultimately improving the wet milling process.

Corn, which is better known as maize outside the United States, is a member of the grass family Gramineae. Domesticated cultivars are subspe- cies of 2. mays. Corn is a member of the cereal grains, which produce dry, one-seeded fruits. Several varieties of corn have been bred for specific uses. Types are currently classified as dent, popcorn, flint, floury, and sweet corn (Watson, 1987). Other mutant varieties include opaque, amylomaize, waxy, and dull waxy. Yellow dent, initially bred for animal feed, is most common and milled in highest quantities by the wet milling industry. Yellow dent is of low cost, is widely available, has a high starch content, and contains large amounts of yellow pigments. Yellow pigments, which concentrate into the gluten fraction during milling, are valued by the poultry industry (Watson, 1984). Waxy, dull waxy, high amylose, and other mutant types are milled for the specialty starches that they contain. White corn is commonly used for specialty food uses, such as alkaline processing. Flint-type corns are most commonly grown in Europe and Turkey.

Corn has an atypical structure compared with other fruits of the Grami- neae family. The corn kernel (Fig. 2) is a caryopsis that is usually shelled

274 DAVID S . JACKSON AND DONALD L. SHANDERA, JR.

FIG. 2. A corn kernel section. Based on Anonymous (1994). Used with permission. Artwork modified by Kristi A. Snell.

from the female inflorescence known as the ear (cob). The major structural components of the caryopsis are the germ, endosperm, pericarp, and tip cap. The composition of each corn type varies. Bebic (1982) compared yellow dent, flint, waxy, and opaque-2 varieties. In comparison to the other corn types, opaque-2 hybrids contained higher levels of albumen (water soluble) and globulin (NaCl soluble) proteins at the expense of prolamin (ethanol soluble) protein content. The waxy mutant had the largest percentage of pericarp, whereas the dent hybrids contained the most starch. Opaque-2 mutant, bred for higher lysine and tryptophan content, was less dense and had the least amount of pericarp. Breeding programs have used modifier genes to increase opaque-2 corn’s endosperm density by increasing g-zein protein composition two to five times (now termed quality protein maize) (Wallace et al., 1990). Commercially viable high-oil corn, used largely for animal feed, has also become available.

The seed fruit is protectively covered by several layers of dead, cellulosic tube cells termed the pericarp (Wolf et al., 1952). The pericarp is the true fruit coat. It is collectively composed of all layers exterior of the seed coat. The outermost layer of the pericarp, the epidermis, is covered by a waxy cutin layer that retards moisture absorption into the caryopsis. The meso- carp cross cell layer contains pits and open areas and provides capillary interconnections between all cells and facilitates water absorption (Watson, 1987). The pericarp extends to the base of the kernel and connects to the tip cap. Inside the tip cap are spongy, star-shaped cells that form an open structure that is continuous with the cross-cell layers of the pericarp (Wolf et al., 1952). This structure is responsible for liquid uptake from the tip cap to pericarp’s open cellular layers in intact kernels (Cox et al., 1944). The pericarp and tip cap account for approximately 5 and 0.8% of the dry kernel weight, respectively.

The outermost layer of the endosperm is enveloped by thick, tough structural cells. A thin, hyaline, and almost invisible membrane, termed


the seed coat, girdles the outermost structure of the true seed (Wolf et af., 1952). It adheres tightly to the outer surface of the aleurone layer and is thought to impart semipermeable properties to the endosperm. The outer layer of the endosperm is a single layer of cells called the aleurone. These cells, granular in appearance, contain protein but little or no starch. The aleurone accounts for approximately 3% of the kernel weight (Watson, 1984). This section of the kernel is collected in the fiber fraction during wet milling.

The endosperm constitutes approximately 82 to 84% of the dry weight of the kernel and contains 86 to 89% of the starch (Watson, 1984). Starch granules of normal dent corn contain approximately 75% amylopectin and 25% amylose. Amylose-to-amylopectin ratios, and hence starches’ functional properties, have been genetically manipulated to form many unique hybirds (such as dull waxy mutants). The endosperm tissue also contains 74% of the kernel’s protein and 16% of the kernel’s lipid. Most of the endosperm’s lipid is bound to cell contents and only 18% is found as triacylglycerides (Watson, 1987).

The endosperm is composed of elongate cells packed with starch granules of 5 to 30 pm embedded in a continuous protein matrix (Watson, 1987). These cells are large with thin walls of cellulosic material (Wolf et af., 1952). The intergranular starch-protein matrix is deposited and contained within the large cellular structures. The binding and structural integrity between the protein and starch make corn kernels quite hard. Kernel fracturing occurs through the starch granules, resulting in a large number of visible broken granules, instead of severing the strongly bound protein matrix (Hoseney, 1986). Duvick (1961) observed by light microscopy that “a trans- parent glue (clear viscous cytoplasm)’’ surrounded the starch granules and protein bodies within cells of mature endosperm. The encompassing protein matrix is composed of an amorphous protein material (Christianson et af., 1969). The fact that water alone will not allow a good separation of protein and starch during wet milling suggests that the protein composition is quite different in corn than it is in wheat (Hoseney, 1986). The bonds holding the matrix proteins together can be loosened by treatments with alkali, reducing agents such as mercaptoethanol (ME), or sulfites used in the wet milling process (Wall and Paulis, 1978).

Protein bodies contained within the matrix were first described by Duvick (1961), who concluded that they were the sites of zein deposition. Zein is an alcohol extractable protein, a prolamin. It is a major component of protein bodies within the kernel. In its native state, it displays high- molecular-mass aggregates (Watson, 1987). Although zein’s amino acid composition has a low sulfhydryl content, zein bodies can be peptidized by ME, dithiothreitol, or other reducing agents (Watson, 1987). The appar-


ent sizes, as shown by sodium dodecyl sulfate-polyacrylamide electrophoresis (SDS-PAGE), suggest that the aggregates are dimers, trimers, and sometimes even larger sets. Because of the amino acid composition of zein, isoelectric points cannot be reliably assigned (Watson, 1987).

Unlike other cereals, the corn kernel endosperm can contain both translucent (vitreous, hard, horny) and opaque (floury) areas within each kernel (Figs. 2, 3). Dent hybrids are usually the result of crossing very vitreous, flint varieties with soft, floury varieties and have an average ratio of about 1 : 2 floury : horny regions (Wolf et al., 1952). The ratio varies greatly among corn types and can be correlated to protein content and composition (Wolf et al., 1952; Paiva et al., 1991). Furthermore, individual kernels from the same field, or even ear, differ significantly in the vitreous-to-floury ratio (Paulsen et al., 1983; Watson, 1987). The translucent or horny section of the endosperm is tightly compacted and contains polygon-shaped starch granules. The dense horny endosperm results in good dry milling characteristics, but requires long steeping periods for efficient protein-starch separation during wet milling (Watson, 1984). The floury or opaque region devel- ops during drying as the endosperm shrinks. The shrinking tears and separates cells, forming voids or air pockets. This region is characterized by larger cells, large round starch granules, and a relatively thin protein matrix (Wolf et al., 1952). The porous texture of the floury region allows for easy recovery of starch (Watson, 1984).

The germ is the embryo or reproducing section of the kernel and is alive and respirating in a typical, healthy kernel. The germ is composed of a plumule, a radicle, and a scutellum. The scutellum functions as a nutritive organ for the plumule and makes up 10 to 12% of the kernel dry weight. The major part of the scutellum is composed of thick-walled isodiametric cells densely packed with cytoplasm. Oil is contained within these cells as droplets (Jacks etal., 1967). The scutellum’s cells contain 84% of the kernel’s lipids; nearly all are free triacylglycerides (Tan and Morrison, 1979). The scutellum’s surface, adjacent to the endosperm, is covered by the secretory epithelium; it contains amylase enzymes that diffuse into the endosperm and digest starch during germination (Dure, 1963). The scutellar epithelium adheres to the endosperm by an insoluble substance composed of pentogly- cans and proteins (Seckinger et al., 1960). This bond, which resists chemical or physical means of separating the germ and endosperm, is responsible

FIG. 3. Scanning electron micrographs of corn kernel endosperm scraped from a typical corn hybrid. (a) Hard (horny, vitreous) endosperm “enrobing” starch granules. (b) Soft (floury) endosperm with individual starch granules clearly visible among some protein material. Bar = 10 pm.


for the prolonged steeping period requirements necessary for its efficient separation (Watson, 1984). The germ contains 22% of the kernel protein, 82% of the ash, and 65% of the sugar (Watson, 1984).

Watson (1987) described the important structural components of the kernel as pectic substances, hemicelluloses, and cellulose. Sandstead et al. (1978) found that corn bran was composed of 70% hemicellulose, 23% cellulose, and 0.1% lignin on a dry weight basis. No detailed structural analysis of maize kernel cell wall polysaccharides has been reported, but a general description can be based on our present understanding of cell walls (Watson, 1987). Endosperm cell walls have only thick primary cell walls, whereas the cells of the pericarp and tip cap also have secondary cell walls. Cellulose, a linear homopolymer of 10,000 D-glucose units linked p(1,4), is the basic structural unit of higher plant cell walls (Aspinall, 1982). Several matrix polysaccharides are associated with cellulose in both primary and secondary cell walls. Arabinoxylans, polymers with backbones of /3(1,4)-linked D-xylose with (13)-linked L-arabinose branches, account for 90 to 95% of corn seed hemicellulose (Oomiya and Imazato, 1982). They are the major component of the primary cell walls of monocotyledonous and endosperm cells. Primary cell walls of monocotyledonous cells also contain arabinogalactans, xyloglucans, some pectic polysaccharides, and one or more arabinans.

The structural integrity of the kernel (caryopsis) has dictated the starch recovery (milling) process used. An understanding of kernel structure is of utmost importance for selecting methods of degrading the kernel structure and releasing free starch granules on milling.

Ill. STEEPING: PROCESS AND EQUIPMENT

A. CLEANING

Received shelled corn is first cleaned to remove broken kernels, chaff, pieces of cob, and other undesirable foreign matter. The presence of foreign matter may contaminate milled fractions or interfere with the free flow of steepwater within and between the steeps. Steeping of broken kernels releases starch in the steepwater, which gelatinizes and causes an undesirable viscosity during evaporation of the steepwater into steep liquor (May, 1987).

B. STEEPING

Steeping is the first critical processing step of wet milling. It is the aqueous soaking of whole corn kernels under controlled conditions of temperature,


time, sulfur dioxide concentration, and lactic acid concentration (Watson, 1984). Too much emphasis cannot be placed on the importance of steeping for ensuring a smooth milling operation and a maximum yield of prime- quality starch (Cox et al., 1944). The chief purposes of the steeping process are to (1) to soften the corn kernel for milling, (2) to reduce or inhibit the activity of undesirable microorganisms, and (3) to assist in pure starch recovery (Bartling, 1940). Proper steeping at 48 to 50°C for 35 to 50 hours allows a milling recovery of 90% of the kernel’s starch with a residual protein content of 0.40% or less (Watson, 1984). Altering or “rushing” the steeping process results in inferior yields and impure fractions.

Typical industrial steeping methods have changed little over the past 100 years. Commercial steeping systems are a battery of 6 to 20 (usually 10-12) conical-bottomed tanks of 50- to 330-metric tons (t) (2000-13,000 bu) capacity connected in series by a system of piping and pumps (Blanchard, 1992). Older systems moved steepwater from steep to steep in batches. In newer systems, steepwater is continuously circulated countercurrent to the age of the steeping corn. Fresh steepwater, to which 0.10 to 0.20% sulfur dioxide has been added, is continuously applied to the steep tank containing the longest (oldest) steeping corn. Sulfur dioxide is obtained in liquid form or from burning of elemental sulfur on site (Blanchard, 1992). Water for steeping is usually recycled from other processing areas of the plant, typically starch washwater, to conserve water. Approximately 1.2 to 1.4 m3 of water is used to steep each metric ton (8-9 galhu) of corn (May, 1987). The steepwater is circulated through the battery by continuously drawing from an orifice in the bottom of each steep containing older corn and adding to the top side of the next steep in the battery series containing slightly fresher corn. As a steep containing the oldest, fully steeped corn is drained and removed from the battery series for milling, a new tank (steep) of fresh corn is added in the series from the opposing end, and each tank in the series shifts in a countercurrent direction relative to the flowing steepwater.

The chemical concentration of the steepwater continuously changes as it circulates from older to newer corn. Carbohydrates and denatured sulfoproteins leach from the kernels and increase in concentration within the steepwater as it circulates. Sulfur dioxide concentration exponentially diminishes from the initial concentration of 0.10 to 0.20% to approximately 0.02% (Wagoner, 1948). A s the sulfur dioxide concentration diminishes, Lactobacillus species ferment leached reducing sugars into lactic acid from about midway in the battery series to the newest corn (Oomiya and Imazato, 1982). These strains of Lactobacillus, which closely resemble Lactobacillus bulgaricus, have a higher tolerance of acids, sulfur dioxide, and elevated temperature than the putrefactive microorganisms that would otherwise predominate in aqueous corn steeps (Liggett and Koffler, 1948). The steep-

280 DAVID S. JACKSON AND DONALD L. SHANDERA. JR.

ing conditions provide a selective medium for colonization by Lactobacillus. The biota found in each plant may differ based on natural selection or adaption to the steephouse environment. In older steeps, the pH is balanced between the acidic sulfur dioxide and buffering compounds leached from the corn. The combined weak acidity of the lactic acid and buffering action of leached solubles maintain a pH of 3.8 to 4.1 as sulfur dioxide is removed by the corn kernels. The desired pH is regulated in the steephouse by balancing the number of steeps in the system, concentration of sulfur dioxide, steepwater flow, and timing of corn addition and removal for grinding (Watson, 1984).

The steepwater that has passed through the steep battery is diverted to an evaporator. Approximately 0.7 to 0.8 m3/t (4.5-5.5 galhu) is drawn off as light steepwater containing about 6% solids at 4" BaumC (Blanchard, 1992). An evaporator concentrates the collected inorganic matter, phytin, proteins, and carbohydrates to a 50% solids by-product usually termed steep liquor. The condensate has a pH of about 3 (Blanchard, 1992).

C. STEEPWATER ABSORPTION

New corn generally contains 13 to 15% moisture, but high-moisture corn, usually available during and shortly after the harvest season, generally has better milling properties (Wagoner, 1948). The initial moisture content of corn influences its end-use properties (Dorsey-Redding et al., 1990). Mois- ture within equilibrated kernels is not uniformly distributed (Song et al., 1992).

Cleaned grain entering the steeping battery is first immersed in light steepwater containing water, lactic acid, leached solubles, and a very low concentration of sulfur dioxide (Roushdi et al., 1979). Because the pericarp is impregnated with waxes and lipids, the outer surfaces of the kernel are semipermeable to water and solutes (Wolf et al., 1952). The pericarp has a permeability of 1.8 X g cm/cm2 atm s at 50" under Fick's diffusion model (Tolaba et al., 1990); thus, the outer surfaces of the kernel are not a major route for absorption of steepwater. Steepwater, however, does enter the kernel's tip cap, penetrating the open structure of the spongy cells interconnecting the tip cap to the pericarp, and flows through the loose, open cross-cell and tubular layers between the pericarp and seed coat. The flux of steepwater is from the basal end of the kernel to the crown. The large intercellular spaces in the cross-cell layers form a major moisture transfer route from the tip cap to the endosperm (Watson, 1984). Therefore, the tip cap is the major route for steepwater uptake into the kernel. Song et al. (1992) showed, by magnetic resonance imaging, that the glandular layer of the scutellum (the tissue between the germ and floury


endosperm) is a path for moisture transfer to central areas of the kernel, such as the floury endosperm. This moisture penetration route has less resistance than moisture transfer through the pericarp and vitreous endosperm to the floury endosperm. Because of density, structural, and compositional differences within the kernel, moisture transfer through the kernel is not uniform with respect to time. Moisture penetrates the floury endosperm and other open structures more easily than the vitreous (horny) endosperm. Moisture diffuses through the endosperm and germ about eight times faster than the pericarp (Tolaba et al., 1990). The kernels absorb approximately 0.5 m3/t (3.5 gal/bu) and reach a maximum moisture content of approximately 45% in the first 8 to 10 hours of steeping (May, 1987).

Kernels swell appreciably on steepwater absorption. Aqueous absorption causes a volume expansion of 55 to 65% (Anderson, 1963). The kernels mainly have a linear hygroscopic expansion with a typical cubic expansion of 0.018 to 0.020 m3/m3/% moisture content for dent hybrids, with higher- density varieties exhibiting a larger coefficient of expansion (Muthukumarap- pan and Gunasekaran, 1990).

D. ROLE OF LACTIC ACID

Lactic acid, in conjunction with water, is considered the first major steeping chemical in which kernels are steeped. The resulting reduction in internal pH, along with high steeping temperatures, results in the death of respirating cells (Watson, 1984). Cell membranes become porous and allow leaching of solubles and absorption of sulfur dioxide.

Commercial wet millers cite three reasons for maintaining lactic acid fermentations within the steeping batteries: (1) separation of milled components (especially starch and gluten) is optimized, (2) the effectiveness of sulfur dioxide in preventing growth of undesirable microorganisms is most operative at pH 3.8 to 4.2, and (3) mineral scale deposition on water evaporation equipment is lessened at lower pH (Boundy et al., 1967). Lactic acid also affects the form of the protein found in the steepwater. If lactic acid concentrations are not high enough in the steepwater sent to the evaporators, the proteins are in a form that can be heat-denatured; they coagulate readily to form a “liver paste” inside the evaporator tubes that readily collects and is difficult to remove (Blanchard, 1992).

The importance and role of lactic acid in protein dispersion and starch release are disputed among scientists and those in the wet milling industry. Cox et al. (1944) reported that the protein disintegration effect of steeping kernels in 0.1% lactic acid alone was approximately equivalent to that brought about by water alone, and kernel softening action was roughly equal to that found after a 0.1% sulfur dioxide steep. Steeping with 0.2 or


1.0% lactic acid produced protein disintegration approximately equivalent to a 0.1% sulfur dioxide steep and a softening action equivalent to 0.2% sulfur dioxide. Steep solution containing both lactic acid and sulfur dioxide softened and disintegrated the kernel to a similar extent as with higher sulfur dioxide concentrations. Watson (1967) stated that lactic acid aided in softening the corn, but Watson and Sanders (1961), in work on 10-mm- thick sections of endosperm, concluded that neither lactic acid, nor any other acids, had an effect on starch release below pH 9.0. Roushdi et af. (1979) reported that high lactic acid production resulted in lower yields and quality of starch than steeping with only high levels of sulfur dioxide. Eckoff and Tso (1991a) reported increased starch yield and reduced sulfur dioxide requirements with lactic acid addition to steeps. If the pH of a solution is not low enough to cause formation of sulfur dioxide ionic species, the gaseous molecule complexes with glucose and other carbonyl compounds in proportion to their concentration, thus becoming unavailable (Pointing and Johnson, 1945; Gehman and Osman, 1954). Lactic acid’s pH-lowering effect may enhance the effectiveness of sulfur dioxide while steeping under low concentrations. Recent response surface work in our laboratory has shown that levels of lactic acid play a pivotal role in softening kernel structure, and influence the absorption rate of sulfur dioxide (Shandera et af., 1995).

E. ROLE OF SULFUR DIOXIDE

Sulfur dioxide was initially added as an antimicrobial agent to control putrefactive organisms, but it was found indispensable for obtaining optimum starch yields and purity. Because of the solubility and bonding properties of the endosperm proteins, water alone will not disrupt starch granules from the adhering protein matrix (Cox et af., 1944). Although steeping kernels reach their maximum moisture content in the first 8 to 10 hours, the time required for maximum starch yields is at least 40 to 50 hours (Watson et al., 1951).

As a steep shifts in position within the battery, its sulfur dioxide concentration increases. Because kernels reach their maximum moisture content before immersion in steepwater with high sulfur dioxide concentrations, high concentrations of sulfur dioxide are not directly transported into the kernel by steepwater, as with lactic acid. Fan et af. (1965) found that the uptake of sulfurous acid during steeping of corn could be modeled using Fick’s second law equation with a diffusivity of lo7 cm2 s-’. The pericarp is a diffusive barrier, even to the gaseous sulfur dioxide (Eckoff and Okos, 1990). Sulfur dioxide primarily enters the tip cap at the basal end of a sound kernel and travels through the cross and tube cells of the pericarp surrounding and into the endosperm. The absorption route is similar to


that of steepwater. Sulfur dioxide also diffuses through the germ into the endosperm. Adsorption of gaseous sulfur dioxide within yellow dent kernels increases with decreasing temperature. Eckoff and Okos (1990) found 250% greater adsorption at 3°C than at 25°C [760 mm Hg with 30% (w.b.) kernel moisture] The adsorption of sulfur dioxide was independent of vapor pressure and kernel moisture content (above 20%). The similarity of activation energy for sulfur dioxide transfer in the kernel to that found in water by Chang and Rochelle (1981) suggests that moisture is an important factor in ionic transfer of sulfur dioxide through the kernel.

Increasing steepwater sulfur dioxide concentrations excessively may not increase absorption. Nierle (1972) estimated that only 5.7% of the sulfur dioxide used in steeping is absorbed by the corn. Of this, 45% is absorbed by the endosperm protein, 2% in the starch, and 40% in the germ. Only 12% of the absorbed sulfur dioxide reacts with the matrix protein. Eckoff and Okos (1990) showed that the germ holds three times more, and the bran holds 50% more, sulfur dioxide per unit weight than the endosperm when kernels are gassed with sulfur dioxide. Unbound sulfur dioxide is driven off during milling and processing of the steep kernels.

F. KERNEL DEGRADATION

The effects of steeping agents (5OoC, 24-hour steep) on corn kernel structure were first comprehensively observed using microscopy by Cox and co-workers (1944). Modern scanning electron microscopy can also be used to aid in our understanding of kernel degradation during wet milling (Fig. 4). Water is an essential steeping chemical, in that it hydrates the kernel components and acts as a transportation medium for both lactic acid and sulfur dioxide. When steeping with sulfur dioxide, the endosperm's protein matrix gradually swells, becomes globular, and finally disperses over time. The degree of protein degradation (peptidization) was shown to be directly related to starch yield and ease of milling. Starch purity is usually proportional to sulfur dioxide concentration. Cox et al. (1944) found that protein dispersion increased with increasing sulfur dioxide concentration to 0.4%. Lactic, acetic, and hydrochloric acids were tested as steeping agent replacements for sulfur dioxide. Although lactic acid produced some softening action, other acids had little effect. Sulfur dioxide's mode of action was concluded to be other than acidic.

Wagoner (1948) noted that cross cuts of kernels steeped in commercial countercurrent batteries for 12 hours had an intact protein matrix, starch was still held tenaciously, and cellulose cell walls were very firm. Only slight changes were noted in cross sections of steeping kernels until after 24 hours, when the sulfite content of the steepwater began to rise rapidly. Thereafter,


FIG. 4. Scanning electron micrographs of corn kernels steeped in water or 1% lactic acid and 0.15% sulfur dioxide. As steeping time increased, the lactic acid-sulfur dioxide steep solution was able to disrupt the hard endosperm so that individual starch granules are clearly visible. Small holes are visible in the hard endosperm starch granules of the water-treated samples; these represent broken granules, which are more fragile than the protein matrix. (Also compare with Fig. 3.) Bar = 10 pm. Hard endosperm section is on the left, and soft endosperm section is on the right. (a) Steeped in water 3 hours. (b) Steeped in lactic acid-sulfur dioxide 3 hours. (c) Steeped in water 53 hours. (d) Steeped in lactic acid-sulfur dioxide 53 hours.

protein matrix degradation continued as the ions migrated throughout the interior of the kernel. After approximately 50 hours, a well-steeped kernel showed no protein matrix structure, but had intact cellulose walls and free starch granules. Most of the protein network was dispersed into the steep solution, and the remainder was folded back around the cellulose walls.


FIG. 4 (Continued).

Watson and Sanders (1961) studied sulfur dioxide effects on matrix protein under conditions where diffusion of sulfur dioxide was not limiting. Sections of horny (vitreous) endosperm, 10 mm thick, were steeped under high, constant bisulfite concentrations. Loss of starch granules from the protein matrix, on gentle agitation, was 50 to 70% complete in 1 hour and complete in 4 hours at 52°C in 0.2% bisulfite solution. The rate and extent of starch release from the matrix increased with bisulfite concentration (0.05-0.2%) and with increasing steeping temperature (52-60°C). Further experiments showed that at pH levels below 9, the pH of the medium did not have a significant effect on starch release.

Sulfuric ionic species are responsible for degrading the intercellular structure of kernels. Sulfurous acid does not exist as a distinct chemical species.


In water, sulfite anion, bisulfite anion, and sulfur dioxide (as a dissolved gas) exist in equilibrium as a sequence of acids and bases based on sulfurous acid, H2S03 (King et al., 1981). The sulfuric species present in aqueous solution relates to the pK, of the species and the pH of the solution. The steeping kernel’s internal pH of 3.6 to 4.0, reported by Wahl(1970), favors the bisulfite anion. Swan (1957) demonstrated that sulfite or bisulfite ions react with the sulfhydryl groups of cystine in wool protein, reducing the disulfide linkage between tertiary structures. The result is a protein fraction (P’) having one cysteine (-SH) and a second protein (P”) on which a S- sulfo derivative of the cysteine moiety has been formed:

P’S-SP” + HS03- = P’SH + P”SS03- (native protein) (reduced) (sulfoprotein)

(1)

Boundy et al. (1967) confirmed this reaction (1) within the intercellular protein matrixes of corn. Soluble S-sulfoproteins are produced and thus prevent reformation of disulfide bonds and structure. Proteins become either peptidized and soluble, thus leaching into the steepwater, or are denatured and folded back against the cell walls. Although other alkaline reducing agents give similar results, none are of commercial importance.

Separation of protein and starch is not always complete. Goossens et al. (1988) described grits (inseparables) as dense particles from the horny endosperm. Their structure consists of a three-dimensional honeycomb- like protein matrix filled with small starch granules. Many of the grits are enclosed by the fibrous cell wall. Grits result when steeping and milling fail to recover available starch, thus resulting in yield losses. The glutenin portion of the protein matrix can only be completely dissolved with a combination of reducing agents and hydrogen bond breaking chemicals (NaOH, urea, detergent) (Watson, 1984); thus, typical steeping and milling procedures may not completely degrade kernel structure.

G. ROLE OF TEMPERATURE

Steeping is generally done at 48” to 52°C to control microbial growth (Watson, 1984). Steeping temperatures above 55°C are unfavorable for endemic lactobacilli (May, 1987). Maintaining steeping temperatures of 45” to 55°C favors lactic acid production (Watson, 1967). Putrefaction and production of butyric acid and alcohols by other microorganisms occur during steeping at temperatures less than 45°C (May, 1987).

Higher steeping temperatures should increase steeping efficiency. Chemi- cal reaction rates generally increase with temperature. Diffusion of moisture is generally enhanced by the temperature of the fluid medium and has an exponential relationship with the inverse of the fluid temperature (Steffe


and Singh, 1980; Walton et al., 1988). Adsorption of water vapor into kernels increases with absolute temperature in an Arrhenius-type relationship (Mu- thukumarappan and Gunasekaran, 1990). Cox et al. (1944) reported that higher steeping temperatures (38-55°C) increased endosperm protein peptidization and dispersion. Temperature does not influence volumetric expansion of the kernel on moisture uptake (Muthukumarappan and Guna- sekaran, 1990).

H. BIOCHEMICAL EFFECTS

Steeping effects are both chemical and enzymatic. The added sulfur dioxide and microorganisms producing lactic acid are considered the main effectors of the process. Other research, however, has pointed out the additional importance of catabolic enzymes present in the kernel and those produced during fermentation by Lactobacillus sp. Only about 10% of the nitrogen in corn is in the form of nonprotein nitrogen and, therefore, accounts for only about 25% of the soluble nitrogen obtained during steeping. Only enzymatic degradation of dissolved protein can account for the fact that 85% of the nitrogen in incubated steepwater is dialyzable nonprotein nitrogen (Watson, 1984). Although lactic acid bacteria generally do not show much proteolytic activity, the species found in the steeps do contribute to significant protein degradation (Watson et al., 1955). The bacteria populate the soft tissue region of the tip cap, and they may have a proteolytic effect on the matrix protein. Wall and Paulis (1978) have disputed the microbial-proteolytic hypothesis and have, instead, attributed this proteolytic effect to the kernel’s indigenous proteases. Their theory was based on comparing normal- and high-temperature dried corn steeped in the presence of lactobacilli. Poor starch and protein separation differences were attributed to heat denaturation of catabolic enzymes within the kernel. Eckoff et al. (1991b) have attributed poor starch and protein separation to morphological changes of proteins in the heat-damaged kernel. Franzke and Wahl(l970) suggested that at least some biochemical degradation of kernel structure during steeping was due to hydrolyzing enzymes (i.e., amylases, proteases). Proteolytic activity during steeping is generally due to the activation of endogenous enzymes by sulfur dioxide, and these proteases improve the release of corn starch (Wahl, 1970).

IV. MILLING AND FINAL PROCESSING

After steeping, the softened grain is degerminated in two stages by coarse grinding in an attrition mill (Blanchard, 1992). The degerminating mill has


two counterrotating disks with intermeshing fingers that tear apart, rather than crush, the kernel. The disks are adjusted for freeing the maximum amount of intact germs. Germs are swollen and rubbery due to steeping. Hydroclones are used to separate the germs, which are less dense and contain most of the oil, from the rest of the kernel constituents, which are in a starch slurry of 7" to 8" BaumC (Watson, 1984). Loose starch and gluten are removed from germs by washing with 1.2 to 1.3 m3 of water per metric ton of corn on 50" wedge-wire screens (May, 1987). After washing, the germs are dewatered and further processed to extract corn oil. The operation is optimized to recover a germ fraction containing 45 to 50% oil (May, 1987).

The remaining material consists of a starch slurry, gluten, fiber, and kernel fragments (Blanchard, 1992). On screening the slurry, the overs, containing mostly fiber (pericarp) and chunks of horny endosperm, are reground in fine-grinding mills to liberate most of the remaining starch and gluten. The second grind mills may be either of the impact type known as Entoleter mill or of the combined attrition and impact type such as the Bauer mill (Watson, 1984). The fiber is separated by pumping the slurry with considerable force on 120" wedge-wire screens, then washed, dewatered, and dried.

Gluten is separated from the remaining starch slurry, soluble impurities, and high-protein substances in high-speed nozzle centrifuges. Separation results from density differences between gluten (1.06 specific gravity) and starch (1.6 specific gravity) (May, 1987). Older systems separated starch and gluten by settling in large-diameter tanks. The gluten is thickened to 12% solids in centrifuges, dewatered to 42% solids by vacuum filtration, and dried to 88% solids for sale as a protein concentrate.

The remaining starch slurry still contains 3 to 5% protein and must be further purified. The desired final protein concentration is less than 0.4%, with about 0.01% free protein (May, 1987). The slurry is washed of the remaining gluten with fresh water in a countercurrent fashion using multiple stages of centrifuges. The starch is then dewatered, dried, and/or modified in further processing steps. The removed protein consists primarily consists of starch-protein complexes, termed middlings, which are recycled back to the primary separation step.

V. LABORATORY VERSUS COMMERCIAL MILLING

Most researchers have conducted laboratory steeping using batch methods. Steepwater, to which all chemicals are initially added, is held at the desired temperature in a single tank, with or without recirculation (Cox et al., 1944; Zipf, 1951; Watson et al., 1955; Wahl, 1970; Roushdi et al., 1979;


Krochta et af., 1981; Wehling et af., 1993; Eckoff et af., 1993). Wagoner (1948), Watson et af. (1951), and Steinke and Johnson (1991) have chal- lenged the use of laboratory batch steeping methods because of the lower starch yields, timing of sulfur dioxide addition, and solubles exposure with respect to commercial countercurrent systems. Structural changes within the kernel, over time of steeping, differ between batch and countercurrent steeping systems (Steinke and Johnson, 1991). Watson et af. (1951) and Steinke and Johnson (1991) devised elaborate laboratory countercurrent steeping systems to mimic commercial systems. Nevertheless, high variation in the growth of lactobacilli, lactic acid fermentation, and the need for continuous operations have hindered the application of such systems to the laboratory bench top. Watson et af. (1955) and Watson and Sanders (1961) conducted wet milling research using batch steeping, not the developed countercurrent system. Weller et af. (1988) mimicked a countercurrent system by continuously pumping fresh steep solution through a set series of steeps, but only the last steep in the series was milled. This system was a simplified way of immersing new corn in light steepwater containing collected solubles and lactic acid; but microbial growth was still uncontrolled and large amounts of experimental materials were used to produce this effect. Laboratory batch steeping may be more sensitive and have less inherent variability with respect to comparison studies (Fox et af., 1992). Although commercial systems are countercurrent, most university and industrial laboratory research is conducted using batch steeps.

Lactic acid use in laboratory steeping has also been inconsistent. Ander- son (1963), Krochta et af. (1981), Neryng and Reilly (1984), and Steinke and Johnson (1991) used batch steeping methods without lactic acid addition. Eckhoff et af. (1993), Wehling et af. (1993), Ling and Jackson (1991), Weller et al. (1988), Wahl (1970), Franzke and Wahl (1970), and Watson et al. (1955) steeped with added lactic acid. The inconsistent use of lactic acid in laboratory steeping stems from confusion as to its function as a steeping agent.

During batch steeping, work in our laboratory has suggested that both lactic acid and sulfur dioxide play important roles in the degradation of kernels (Shandera et af., 1995). Most notably, yields of starch are influenced by both lactic acid and sulfur dioxide concentrations (Fig. 5). It should, however, be noted that the impact of these chemicals during commercial steeping (especially regarding concentration effects) would be different; corn kernels are exposed to changing concentrations of lactic acid and sulfur dioxide during the commercial countercurrent steeping process.

Most authors have based, usually with slight modifications, their milling techniques on the previous work of Watson et al. (1951,1955) and Anderson (1963). These methods degerminate the kernels with a modified Waring


FIG. 5. Yield of starch obtained during batch laboratory corn wet milling with various concentrations of lactic acid and sulfur dioxide. The response surface model was significant at P = 0.015.

blender, fitted with a blunt blade and operated at a reduced speed to prevent germ damage. Germs are floated by adjusting specific density of the slurry and skimmed off. The slurry is either finely ground in a Quaker or similar crushing mill, or it is reground in the Waring blender operating at a high speed. Bran is recovered and washed on 55- to 48-pm mesh (No. 200-300 U.S.) standard testing sieves, nylon cloth, or linen cloth. Tabling has been the standard for separating starch and protein fractions. Steinke and Johnson (1991) separated starch by centrifugal techniques instead of tabling and identified an “inseparables” fraction consisting of small aggregates of starch and protein, which may be similar to the “grits” described by Goossens et al. (1988). Eckoff et al. (1993) investigated the accuracy and reproducibility of “traditional” laboratory wet milling techniques. Two critical control points were identified: (1) timing between the end of pumping the mill starch slurry during tabling to start of starch washing step, and (2) completeness of germ skimming and recovery. To increase precision between operators, they proposed controlling process water inputs and the lengths of individual process stages. The fiber fraction was separated using a mechanical vibrating screen, but hand mixing allowed more complete removal of starch and gluten from the fiber. Separating starch by tabling, in comparison to industrial techniques, was deemed to have the limitation of obtaining higher-quality starch in exchange for a lower starch yield and protein concentration in the gluten fraction. A gluten fraction was obtained by filtering the runoff from tabling the starch/protein slurry.

VI. RESEARCH TO IMPROVE WET MILLING

Various mechanical and chemical methods to reduce steeping time, and therefore input costs, have been explored. Superficially scratching the peri-


carp to increase water uptake in kernels had little effect in shortening steeping (Roushdi et al., 1979). Krochta et al. (1981) reduced steeping energy input by reducing the corn-to-steepwater ratio or by degerminating kernels between two steeping periods. Neryng and Reilly (1984) increased the pH of their steep solution to 7 without affecting starch yields, but found difficulties in regulating microbial growth. They also found that preensiling high-moisture corn before milling increased starch yields, but resulted in decreased gluten recovery while increasing the protein content of steepwater solids. Meuser et al. (1985) proposed producing starch without sulfur dioxide using high pressure to disintegrate the morphological and structural components of the corn. Hassanean and Abdel-Wahed (1986) agitated steeping corn within and between steeps for faster moisture penetration and increased leaching of solubles, but current commercial steeping facilities would need to be modified, and energy requirements are higher. Eckhoff and Tso (1991a) found that pretreating corn with gaseous sulfur dioxide before steeping significantly increased starch yields when steeping for only 6, 12, and 24 hours. Mistry and Eckhoff (1992) proposed alkaline instead of sulfuric steeping of corn, but they found that starch obtained through alkaline isolation contained significantly more protein than traditional wet milling methods.

Over the past decade, several authors have studied the feasibility of using enzymes to enhance or replace current methods through direct addition of hydrolytic enzymes to steeping systems. Each has investigated a method based on knowledge of kernel structure, current steeping methods, and biochemical processes that take place.

Based on prior studies done on starch slurries and milling grits treated with various preparations of pepsin, papain, bromelain, and trypsin, Roushdi et al. (1979) increased the purity of starch by adding proteolytic enzymes to steeping corn. Because most proteases are effective only at pH 6 or higher, a significant shift in the sulfur dioxide species equilibrium from HS03- to SO?- could affect disulfide reducing capacities of kernel proteins by the steepwater (King et al., 1981). The penetration of the high-molecular- weight enzymes into the endosperm was limited by the pericarp seed coat membrane. The limitations to this method are the large quantities of enzyme used and the need for lower steeping temperatures. Commercial protease usage on steeping grits from high-temperature dried corn was found to increase starch yield and purity (Eckhoff and Tso, 1991b). When steeping grits (endosperm chunks of corn), enzymehubstrate contact was not hindered.

Several investigators have explored enzymatic degradation of the corn’s fibrous structure to aid steeping chemical penetration and separation of kernel constituents. Caransa et al. (1988) used cell wall-degrading enzymes during steeping to improve the separation of germ and starch from fiber

292 DAVID S. JACKSON AND DONALD L. SHANDERA. JR.

and gluten, to improve the quality of corn steep liquor by preventing the accumulation of phytic acid in corn steep liquor, and to shorten the steeping time. Ling and Jackson (1991) steeped corn with a commercial enzyme preparation of cellulase, P-glucanase, and arabinoxylanase activity, thus degrading the pericarp and cellular walls. Mean starch yields were increased only 1.2 percentage units, but starch yields were always higher, with short- ened steeping times (24 hours) with enzyme usage. Within internal kernel layers, an inhibited enzyme-substrate contact between the starchy endosperm and the cellulosic pericarp layers was theorized to be a limiting factor. Steinke and Johnson (1991) steeped with a commercial enzyme preparation of cellulase, hemicellulase, P-glucanase, pectinase, and bromelain (0.25% v/v, each) using a laboratory countercurrent steeping system. Although levels of enzymes used would be uneconomical, corn milled more easily, even with only a 24-hour steeping period. Degradation of pericarp and cell walls allowed a clean separation of layers from the endosperm aleurone layer and underlying starchy cells on milling. Ironically, water diffusion into the kernel did not increase with enzymatic degradation of cellular components. The protease degrades starch-protein complexes, thus increasing starch purity and decreasing the amount of inseparables obtained.

Enzyme use in commercial steeping is limited due to mobility restrictions within the kernel, high steeping temperatures, sulfur dioxide, low pH, and relatively high cost. Possible negative, residual enzymatic action on final products and processes must also be considered.

VII. END PRODUCTS

A. PRODUCTS DERIVED FROM STARCH

Although starch, fiber, germ, and bran are the primary components from corn wet milling, the commercial refining process further converts all of these components into more readily marketable forms. Starch polymers (amylose and amylopectin) are, by far, the most readily convertible components because they are composed of glucose monomers.

1. Native Starches

After starch is washed, it can be dried to obtain common unmodified corn starch or prepared for modification into dextrins. Native starches can consist of “regular” corn starches that have granules containing 25% amylose polymers and 75% amylopectin polymers. Regular starches, when


cooked and cooled, produce firm, white opaque gels. High-amylose starches, commercially available from corn with either 50 or 70% amylose polymers (50 or 30% amylopectin polymers) can be cooked to produce very firm white to whitish yellow opaque gels. Cooked waxy starch gels, consisting of granules with nearly 100% amylopectin polymers, produce weaker, “slimy or stringy” gray-white translucent gels. Starches from corn with genetic modifications other than high amylose or waxy hybrids are also available. None, however, have approached the commercial success of waxy or high-amylose starches.

2. Modified Starches

Regular corn and its common high-amylose and waxy mutants provide a wide range of unique functionality for paper, food, and other industrial uses; their overall functionality and usefulness, however, can be extended by chemical and/or physical modifications. Modifications, typically to waxy or regular starches, include chemical crosslinking, chemical stabilization, bleaching, oxidizing, acid thinning, and ionic modification.

a. Crosslinking and Stabilization. Crosslinked starches are chemically treated to create stable gels after cooking. Chemical connections are established between adjacent starch polymers that result in less viscosity on cooking, because there is less granular swelling. In addition, however, there is also less viscosity breakdown during cooking and further processing (i.e., pumping), as the starch polymer structure is reinforced against shear- induced cleavage. Also, there is less subsequent retrogradation on cooling, as the reinforced chemical backbone prevents close polymer alignment. Waxy starch is often crosslinked; crosslinking eliminates the characteristic “slimy” gel texture, while retaining the less viscous translucent gel (compared with regular starch). Common crosslinking agents include sodium trimetaphosphate, phosphorus oxychloride, and mixed adipic-acetic anhy- drite (Blanchard, 1992).

Stabilized starches are similar to crosslinked starches in that a chemical group is attached to the starch polymers. Typically, hydroxyl groups are processed to undergo esterification (with acetic anhydride, vinyl acetate, mixed acetic-adipic anhydride, monosodium orthophosphate, or succinic anhydride) or etherification (with propylene oxide or acrolein) (Blanchard, 1992). Unlike crosslinking, adjacent polymers are not physically linked. This results in starches that swell to high viscosities (as high as their native counterparts), but form less firm gels and resist retrogradation; the substi- tuted group prevents close alignment of adjacent polymers.


b. Acid ModiJication. Acid-modified starches result when native starch is exposed to small amounts of acid (usually 0.5-0.25 N sulfuric acid) at temperatures 10" to 20°C below the starch gelatinization temperature. When compared with native starch, acid-modified starches do not attain as high a viscosity; consequently, more concentrated starch slurries can be prepared. Acid modification, however, creates polymer mixtures that appear to contain larger amounts of amylose (as amylose and as amylose- like polymers because of partial amylopectin depolymerization) ( Jackson et al., 1992, Zhang and Jackson, 1992). This polymer molecular size profile results in a starch that attains the same degree of gel firmness as its nonmodi- fied native counterpart. Very firm gels can be created by cooking high concentrations of acid-modified starch; this starch is commonly used to manufacture jelly bean-type candies.

c. Other ModiJications. Native starches can be further whitened and thermophilic spores destroyed by bleaching. Starch is treated with small amounts of hydrogen peroxide, peracetic acid, or sodium hypochlorite (Blanchard, 1992) to remove color-forming impurities and/or thermophiles. This results in starch that is ideally suited for canning and is used when starch is mixed with other bright-white ingredients (i.e., powdered sugar mixtures).

In a related process, starch oxidation is also a modification using sodium hypochlorite. Similar to acid modification, oxidizing results in partially depolymerized polymers. Oxidizing, however, also introduces carboxyl and carbonyl groups that inhibit amylose and amylose-type molecular reassocia- tion after cooking (Blanchard, 1992). This results in a soft clear gel.

3. Products Using Fermentation or Enzymatic Technologies

Numerous products widely sold today, and a host of additional products destined to be made in the future, are based on technology designed to depolymerize starch into glucose or glucose oligomers.

a. Sweeteners and Maltodextrins. Corn sweeteners, depending on the desired end product, are produced using an acid, acid-enzyme, or entirely enzymatic process. Traditional corn sweeteners [with dextrose equivalents (DE)' approximately equal to 421 were manufactured as early as 1866; they typically were produced using hydrochloric acid (Hebeda, 1987). The acid depolymerizes the starch in an essentially random fashion. Because acid

' Dextrose is identical to glucose; the term is usually reserved for glucose obtained from starch. The extent of starch hydrolysis is measured according to a dextrose equivalent (DE) scale. DE is the reducing (group) content of a depolymerized starch expressed as dextrose (dry basis). Starch has a 0 DE value; pure glucose has a DE of 100.


can catalyze reactions other than depolymerization, this process is limited to syrups with 42 DE or less.

Forty-two-DE syrups can be produced using acid, or starch can be lique- fied using a-amylase. Syrups of DE greater than 42 are prepared from 42- DE syrup using a mixture of a-amylase and glucoamylase enzymes. These syrups can be clarified and refined to produce high-glucose syrups or crystalline glucose.

High-glucose syrups can be further converted to products high in fructose. Although dependent on temperature, pH, and other factors, fructose is approximately 1.8 times as sweet as sucrose and 2.0 to 2.4 times as sweet as glucose. Usually, 42-DE corn syrups are treated with immobilized isomerase that converts some of the glucose into fructose. As a fructose-glucose equilibrium is established at approximately 51% fructose at 60°C, a good commercially viable product (with a reasonable isomerization reaction time) is produced with 42% fructose and 58% glucose (Hebeda, 1987; Blanchard, 1992). High-fructose corn syrups (HFCSs) are produced by chromatographically separating the fructose from a glucose-fructose mixture. The purified fructose can be crystallized or reblended to produce fructose syrups of various percentages. Commercially important HFCSs include 42% fructose, 55% fructose, and 90% fructose. Corn sweeteners are used in a wide array of food products because of their low cost, easy availability, and liquid (pumpable) state. In many sweetened products, such as soft drinks for example, HFCS has largely replaced sucrose as the nutritive sweetener of choice.

Maltodextrins are depolymerized starches with of less than 20 DE. They are produced similarly to corn syrups, but the enzymatic or acid depolymerization process is halted before the DE value reaches 20. Maltodextrins are essentially odor and taste free. Maltodextrins make excellent bulking agents and are used in puddings, soups, frozen desserts, and dry mixes (Hebeda, 1987).

b. Fuel Alcohol and Industrial Chemicals. Many wet milling plants not only have further refining capabilities to hydrolyze starch into sweeteners, they may also have (instead of or together with) facilities to produce alcohols or other industrial and food chemicals from glucose. Typically, these refineries rely on enzymes and yeast to convert starch into ethyl alcohol and carbon dioxide. A starch slurry from wet milling is converted to glucose using acid or enzymatic processes described earlier. The glucose syrup is then mixed with a small amount of steepwater (high in proteins and other nutrients required by yeast) and fermented by yeast. Microorgan- isms can also be used to convert glucose into a wide variety of other industrial chemicals, including many widely used organic acids.


B. NONSTARCH BY-PRODUCTS

In an effort to make the wet milling process as efficient and economical as possible, almost all of the by-products from corn wet milling are used. One of the most valuable by-products, on a weight basis, is the corn oil- containing germ fraction. After recovery during wet milling, the water in the germ is mechanically removed using a screw-type press and then dried in a steam dryer. In some cases, the dried germ is not extracted at a wet milling plant, but is shipped to a large facility specifically designed for oil production. The spent germ cake (germ after oil is extracted) is rich in relatively high-quality protein and makes an excellent feed for pigs and poultry. Its typical composition is 20% starch, 25% protein (N X 6.25), 1% fat, 10% crude fiber, and 25% pentosans (Anderson and Watson, 1982).

The gluten (protein) fraction can be processed separately or incorporated into other fractions for use as animal feed. This decision is based largely on the size of the plant and the economics associated with the animal feed market in the geographic region surrounding the production facility. If processed separately, gluten is mechanically dewatered and then dried using a rotary dryer. Gluten meals usually consist of 69% protein (N X 6.25), 19% starch, and 3% oil (Blanchard, 1992). In geographic regions where consumers prefer yellow-pigmented poultry, gluten meals make excellent poultry feeds because they are high in yellow pigments. The yellow xantho- phyll pigments contribute to a bird’s skin and egg yolk color.

The largest feed fraction obtained from wet milling is gluten feed. This fraction is a combination of corn fiber, spent germ cake (when available on site), and steepwater. Germ cake and/or gluten meal are added to obtain a total protein content of approximately 21% (10-12% moisture basis). Gluten feed is usually sold wet to nearby cattle feeders. The feed can also be dried, but dry feed is not as economically attractive to feeders who have other sources of grain and oilseed proteins available.

VIII. SUMMARY

Corn wet milling is a complicated, large-scale, and efficient industrial process designed to separate the chemical components from corn kernels. The success of wet milling, in terms of maximum yields, is largely dependent on the success of the steeping process. Improper steeping, or steeping of corn kernels that have unusual physical or chemical structures, results in lost product and lower profits. Steeping processes, however, are still based largely on an art that was developed more than 100 years ago. In the next 20 years, major seed corn producers expect that the market for corn with


unique starch characteristics and corn bred to produce speciality chemicals or have an altered composition will expand substantially. As the market increases for speciality corn, corn genetically bred with unique starch characteristics or corn with altered chemical composition, a thorough scientific understanding of steeping chemistry and the entire wet milling process will become increasingly important.

REFERENCES

Anderson, R. A. 1963. Wet milling properties of grains: Bench scale study. Cereal Sci. Today

Anderson, R. A., and Watson, S. A. 1982. The corn milling industry. In “CRC Handbook of Processing and Utilization in Agriculture” (I. A. Wolf, ed.), Vol. 2, Part 1, pp. 31-61. CRC Press, Boca Raton, FL.

8,190-195,221.

Anonymous. 1993. “Corn Annual 1993.” Corn Refiners Association, Washington, DC. Anonymous. 1994. “Corn Starch,” 9th ed. pp. 7,9. Corn Refiners Association, Washington, DC. Aspinall, G. 0. 1982. Chemistry of cell wall polysaccharides. In “The Biochemistry of Plants”

Bartling, F. W. 1940. Wet process corn milling. No. 5. The steep house. Am. Miller 68 (August),

Bebib, Z. 1982. Composition and structure of maize (kernel) endosperm on the process of double enzyme hydrolysis Zea mays. In “Maize: Recent Progress in Chemistry and Technology” (G. E. Inglett, ed.), pp. 181-191. Academic Press, New York.

Blanchard, P. H. 1992. “Technology of Corn Wet Milling and Associated Processes.” Elsevier, New York.

Boundy, J. A., Turner, J. E., Wall, J. S., and Dimler, R. J. 1967. Influence of commercial processing on composition and properties of corn zein. Cereal Chem. 44,281-287.

Caransa, A,, Simell, M., Lehmussaari, A., Vaara, M., and Vaara, T. 1988. A novel enzyme application for corn wet milling. SfarcWStaerke 40,409-41 1.

Chang, C. S., and Rochelle, G. T. 1981. Sulfur dioxide adsorption into aqueous solutions.

Christianson, D. D., Nielsen, J. C., Khoo, U., Wolf, M. J., and Wall, J. S . 1969. Isolation and chemical composition of protein bodies and matrix protein in corn endosperm. Cereal Chem. 46, 372-348.

Cox, M. J., Macmasters, M. M., and Hilbert, G. E. 1944. Effects of the sulfurous acid steep in corn wet milling. Cereal Chem. 21,447-465.

Dorsey-Redding, C., Hurburgh, C. R., Jr., Johnson, L. A., and Fox, S. R. 1990. Adjustment of maize quality data for moisture content. Cereal Chem. 67,292-295.

Dure, L. S. 1963. Site of original and extent of activity of amylases in maize germination. Plant Physiol. 35,925-934.

Duvick, D. N. 1961. Protein granules of maize endosperm cells. Cereal Chem. 38,374-385. Eckoff, S . R., and Okos, M. R. 1990. Sorption kinetics of sulfur dioxide on yellow dent corn.

Eckoff, S. R., and Tso, C. C. 1991a. Wet milling of corn using gaseous sulfur dioxide addition

Eckoff, S. R., and Tso, C. C. 1991b. Starch recovery from steeped corn grits as affected by

(J. Preiss, ed.), Vol. 3, pp. 473-500. Academic Press, New York.

40-41, 82.

AIChE J . 21,292-297.

Trans. ASAE 33,855-861.

before steeping and the addition of lactic acid on steeping. Cereal Chem. 68, 248-251.

drying temperature and added commercial protease. Cereal Chem. 68,319-320.


Eckoff, S. R., Rausch, K. D., Fox, S. R., Tso, C. C., Wu, X., Pan, Z., and Buriak, P. 1993. A laboratory wet milling procedure to increase reproductivity and accuracy of product yields. Cereal Chem. 70,723-727.

Fan, L. T., Chen, H. C., Shellenberger, J. A., and Chung, D. S . 1965. Comparision of the rates of absorption of water by corn kernels with and without dissolved sulfur dioxide. Cereal Chem. 42,385-396.

Fox, S. R., Johnson, L. A., Hurburgh, C. R., Jr., Dorsey-Redding, C., and Bailey, T. B. 1992. Relations of grain proximate composition and physical properties to wet milling characteristics of maize. Cereal Chem. 69, 191-197.

Franzke, C. I., and Wahl, G. 1970. Biochemisch-technologische studien tiber die Nassverarbei- tung von mais. 2. Mitt. Die verlnderung der quellwasserzusammensetzung wahrend des quellprozesses. StarcWStaerke 22,64-67.

Gehman, H., and Osman, E. M. 1954. The chemistry of the sugar sulfite reaction and its relationship to food problems. Adv. Food Res. 5,53-96.

Goossens, J., Derez, F., and Bahr, K. H. 1988. Presence of starch/protein aggregates in the endosperm of corn. StarcWStaerke 40,327-333.

Hassanean, A., and Abdel-Wahed, A. 1986. A new method to short the steeping period of corn grains. StarcWStaerke 38,417-419.

Hebeda, R. E. 1987. Corn sweeteners. In ‘‘Corn: Chemistry and Technology” (S. A. Watson and P. E. Ramstad, eds.), pp. 377-397. Am. Assoc. Cereal Chem., St. Paul, MN.

Hoseney, R. C. 1986. “Principles of Cereal Science and Technology,” pp. 2-19. Am. Assoc. Cereal Chem., St. Paul, MN.

Jacks, T. J., Yatsu, L. Y., and Altschul, A. M. 1967. Isolation and characteriztion of peanut spherosomes. Plant Physiol. 42,585-597.

Jackson, D. S., Waniska, R. D., and Rooney, L. W. 1992. Molecular weight patterns of Nageli amylodextrins. StarcWStaerke 44(2), 59-61.

King, A. D., Ponting, J. D., Sanshuck, D. W., Jackson, R., and Mehara, K. 1981. Factors affecting death of yeast by sulfur dioxide. J. Food Prot. 44, 92-97.

Krochta, J. M., Look, K. T., and Wong, L. G. 1981. Modification of corn wet milling steeping conditions to reduce energy consumption. J. Food Process. Preserv. 539-46.

Liggett, R. W., and Koffler, H. 1948. Corn steep liquor in microbiology. Bacteriol. Rev.

Ling, D., and Jackson, D. S. 1991. Corn wet milling with a commercial enzyme preparation. Cereal Chem. 68,205-206.

MacMasters, M. M., Finkner, M. D., Holzapfel, M. M., Ramster, J. H., and Dungan, G. H. 1959. A study of the effect of drying conditions on the suitability for starch production of corn artificially dried after shelling. Cereal Chem. 36,247-260.

May, J. B. 1987. Wet milling: Process and products. In “Corn: Chemistry and Technology” (S. A. Watson and P. E. Ramstad, eds.), pp. 377-397. Am. Assoc. Cereal Chem., St. Paul, MN.

Meuser, F., German, H., and Huster, H. 1985. The use of high pressure disintegration technique for the extraction of starch from corn. I n “New Approaches to Research on Cereal Carbohydrates” (R. D. Hill and L. Munck, eds.), pp. 161-180. Elsevier.

Mistry, A. H., and Eckoff, S . R. 1992. Characteristics of alkali extracted starch obtained from corn flour. Cereal Chem. 69,296-303.

Muthukumarappan, K., and Gunasdekaran, S. 1990. Vapor diffusivity and hygroscopic expansion of corn kernels during adsorption. Trans. ASAE 33,1637-1641.

Neryng, A., and Reilly, P. J. 1984. Laboratory wet milling of ensiled corn kernels. Cereal Chem. 61,8-14.

Nierle, W. 1972. Schwefeldioxidaufnahme bei der maisquellung. StarcWStaerke 24,345-347.

12,297-311.


Oomiya, M., and Imazato, S. 1982. Corn seed hemicelluloses. In “Maize: Recent Progress in Chemistry and Technology” (G. E. Inglett, ed.), pp. 19-32. Academic Press, New York.

Paiva, E., Kriz, A. L., Peixoto, M. J., Wallace, J. C., and Larkins, B. A. 1991. Quantitation and distribution of y-zein in the endosperm of maize kernels. Cereal Chem. 68,276-279.

Paulsen, M. R., Hill, L. D., White, D. G., and Spraugue, G. F. 1983. Breakage susceptibility of maize belt genotypes. Trans. ASAE 26,1830-1836.

Pointing, J. D., and Johnson, G. 1945. Determination of sulfur dioxide in fruits. Znd. Eng. Chem., Anal. Ed. 17, 682-686.

Roushdi, M., Ghali, Y., and Hassanean, A. 1979. Factors improving the steeping process of corn grains. StarcWStaerke 31,7-9.

Sandstead, H., Munzo, J., Jacob, R., Klevey, L., Reck, S., Logan, G., Dintzis, F., Inglett, G., and Shuey,W. 1978. Influence of dietary fiber on trace element balance. Am. J. Clin. Nutr. 31, S180.

Seckinger, H. L., Wolf, M. J., and MacMasters, M. M. 1960. Hemicellulases of the cementing layer and of some cell walls of the corn kernel. Cereal Chem. 37,121-128.

Shandera, D. L., Parkhurst, A. M. and Jackson, D. S. 1995. Interactions of sulfur dioxide, lactic acid, and temperature during simulated corn wet milling. Cereal Chem. 72,371-378.

Song, H. P., Litchfield, J. B., and Morris, H. D. 1992. Three dimensional microscopic MRI of maize kernels during drying. J. Agric. Eng. Res. 53, 51-69.

Steffe, J. F., and Singh, R. P. 1980. Diffusivity of starchy endosperm and bran of fresh and rewetted rice. J. Food Sci. 45, 356-361.

Steinke, J., and Johnson, L. 1991. Steeping maize in the presence of multiple enzymes. Cereal Chem. 68,7-17.

Swan, J. N. 1957. Thiols, disulfides, and thiosulphates: Some new reactions and possibilities in peptide and protein chemistry. Nature (London) 180,643-645.

Tan, S . L., and Morrison, W. R. 1979. Lipids in the germ, endosperm, and pericarp of the developing maize kernel. J. Am. Oil Chem. SOC. 56,759-764.

Tolaba, M. P., Suarez, C., and Viollaz, P. E. 1990. The use of a diffusional model in determining the permeability of corn pericarp. J. Food Eng. 12,53-66

Wagoner, J. A. 1948. Microscopic observations and procedures used in a study of industrial corn steeping. Cereal Chem. 25, 354-361.

Wahl, G. 1970. Biochemisch-technologische studien iiber die Nassverarbeitung von mais. 3. Mitteilung. Die milieubedingungen im maiskorn wlhrend des quellprozesses. Starch/ Staerke 22,77-84.

Wall, J. S., and Paulis, J. M. 1978. Corn and sorghum grain proteins. Adv. Cereal Sci. Technol.

Wallace, J. C., Lopes, M. A,, Paiva, E., and Larkins, B. A. 1990. New methods for extraction and quantitation of zeins reveal a high content of y-zein in modified opaque-2 maize. Plant Physiol. 92, 191-196.

Walton, L. R. , White, G. M., and Ross, I. J. 1988. A cellular diffusion-based drying model for corn. Trans. ASAE 31,279-283.

Watson, S . A. 1967. Manufacture of corn and milo starches. In “Starch: Chemistry and Technology” (R. L. Whistler and E. F. Paschall, eds.), Vol. 2, pp. 1-51. Academic Press, New York.

Watson, S. A. 1984. Corn and sorghum starches: Production. In “Starch: Chemistry and Technology” (R. L. Whistler, J. M. BeMiller, and E. F. Paschall, eds.), pp. 417-467. Academic Press, Orlando, FL.

Watson, S. A. 1987. Structure and composition. In “Corn: Chemistry and Technology” (S. A. Watson andP. E. Ramstad,eds.), pp. 53-82. Am. Assoc. CerealChem.,St. Paul, MN.

2, 135-219.


Watson, S. A., and Sanders, E. H. 1961. Steeping studies with corn endosperm sections. Cereal Chem. 38,22-33.

Watson, S. A., Williams, C. B., and Wakely, R. D. 1951. Laboratory steeping procedures used in a wet milling research program. Cereal Chem. 28,105-118.

Watson, S . A., Hirata, Y., and Williams, C. B. 1955. A study of the lactic acid fermentation in commercial corn steeping. Cereal Chem. 32,382-394.

Wehling, R. L., Jackson, D. S., Hooper, D. G., and Ghaedian, A. R. 1993. Prediction of wet- milling starch yield from corn by near-infrared spectroscopy. Cereal Chem. 70,720-723.

Weller, C. L., Paulsen, M. R., and Steinberg, M. P. 1988. Correlation of starch recovery with assorted quality factors of four corn hybrids. Cereal Chem. 65,392-397.

Wolf, M. J., Buzan, C. L., MacMasters, M. M., and Rist, C. E. 1952. Structure of the mature corn kernel Cereal Chem. 29,321-382.

Zhang, W., and Jackson, D. S. 1992. Retrogradation behavior of wheat starch gels with differing molecular profiles. J . Food Sci. 57(6), 1428-1432.

Zipf, R. L. 1951. Wet milling of cereal grains. In “Crops in Peace and War: The Yearbook of Agriculture 1950-1951” U.S. Dept. Agric., Washington, DC.

INDEX

A

Acceptor molecules, effect on

Activation transgalactosylase activity, 76-79

energy, 204-205 lactase, 41 transgalactosylase activity of lactases,

82-83 Agglomeration, 209-210 Aglycones, specificity of transgalactosylase

Algin, immobilization of lactases, 43 Allolactose

reaction, 81

production K . lactis P-galactosidase velocity, 65, 70 potassium and magnesium effects, 82

in yogurts, 84 Antiplasticization, by sugars, 156-159 Arrhenius behavior, compared to WLF

behavior, 188-200 Arrhenius equation, explicit reference

temperature, 207 Arrhenius kinetics, contrast with WLF

kinetics, 204-212 Arrhenius plots, 205-206 Arrhenius relaxation behavior, 193 Aspergillus fonsecaeus, lactase, 59 Aspergillus niger

lactase, potential application in HTST

product inhibition, 9-10 saccharide production and temperature,

transgalactosylase products, 79-80

immobilized enzyme system, column

pasteurization of milk, 89

70-71

Aspergillus oryzae

reactor using, 47, 49

lactase, temperature effects, 34-35 saccharide production and temperature,

transgalactosylase products, 79-80 70-71

Assay methods, transgalactosylic activity,

Axial-annular flow reactor, 45-46 64-69

B

Bacillus circulans, P-galactosidase from,

Bacillus sterarothermophilus, lactase, 58 Bifidobacterium, growth promoters, 85

86, 89

C Caking, 209-210 Calcium, intake, milk avoidance and, 4 Carbohydrates, low-molecular-weight

Cg’ values, 167-168

glass transition temperature, 148-149 solute-specific subzero glass transition

Carbohydrate-water systems, kinetically metastable, WLF behavior, 189

Caseinate, 37 Continuous stirred tank reactors, lactase

Cookies

disparity, 229

temperature, 139

immobilization, 18-19

absence of polymerization and glutenins

glass transition, 185

chemical components, 271-272 quality protein maize, 274 structure, 273-278

and gliadins, 224

Corn

301

302 INDEX

caryopsis, 274 endosperm, 275-277 enzymatic degradation, 291-292 germ, 277-278 protein bodies, 275, 277 seed coat, 274-275 steeping agent effects, 283-286

Corning process, hydrolysis of lactose,

Corn sweeteners, 294-295 Corn wet milling, 271-297

biochemical effects, 287 cleaning, 278 industry scale, 273 kernel degradation, 283-286 laboratory versus commercial milling,

lactic acid role, 281-282 milling and final processing, 287-288 nonstarch by-products, 296 process flow, 272 products, derived from starch, 292-295

types used, 273

47,51

288-290

modified starches, 293-294 native starches, 292-293 using fermentation or enzymatic

technologies, 294-295 research to improve, 290-292 steeping, 278-280 steepwater absorption, 280-281 sulfur dioxide role, 282-283, 285-286 temperature role, 286-287

Crackers, glass transition, 185 Cryostabilization, 162-163, 165-166 Crystallization

kinetics, 210-212 time, WLF- versus Arrhenius-type

temperature dependence, 181-182 Curing, material states, 214

D

Dairy industry, waste utilization, 20-22 Degerminating mill, 287-288 Disaccharides, formed by K. la& lactase,

Dynamics map, 129, 134, 159-177 77,79

Cg’ values, 166-169 solute-specific subzero glass transition

temperature, 161, 163-166

water and glass dynamics, 160 WLF equation, 160, 177-188

E Energy, activation, 204-205

F

Food amorphous powder, state diagram,

temperature versus water content, 209

aqueous glasses, diffusion of water and solute, 169-177

flour-based products, viscoelasticity,

glass transition temperature measurement, 114-118

labile components, encapsulation, 171 nonequilibrium nature of products and

processes, references, 115-117 patents using /3-galactosidase activity

for galactooligosaccharide production, 87

184-185

transgalactosylase activity, 84-95 Food molecule-water systems,

characterization, 120, 124-125 Food polymer science, 106-143; see also

Dynamics map; Glass transition temperature

evolution of approach, 109-114 glasses and glass transitions, 107-109

ice-melting onset transition, 164-165 plasticization

temperature measurement, 114-118

effect on glass transition temperature,

by water, 150-159 state diagrams, 135-143 structure-property relationships,

references, 107 thermosetting, amorphous polymers,

water and glass dynamics, 127, 129-135

118, 120-125

212-226

Food-polymer systems, aqueous,

Fragility plot, 193 Freezing point, depression, oligosaccharide

viscoelastic behavior, 183-188

formation and, 30

INDEX 303

Fringed micelle model, crystalline-amorphous structure of partially crystalline polymers, 129, 133

Fuel alcohol, from corn wet milling, 295

G

Galactobiose, in yogurts, 84 Galactose

concentration changes and allolactose

effect on lactase activity during milk

enzymatic modification, 10

from B. bifdium, 85 catalysis of lactose hydrolysis, 24-25 hydrolysis, 23-24 patents using, for production of

galactooligosaccharides, 87 purified, 23 specificity for lactose portions, 63

production, 83

processing, 12

P-Galactosidase, see also Lactase

Galactosyl, enzymatic transfer reactions, 60 Gelatinization, sugar role, 158 Gelation, curing reaction and, 214 Gel point, 214 Gidley endotherm, 172 Glass dynamics, 133-135, 160 Glass-forming liquids

behavior classification, 192-195 super-structuring effects, 197-198

definition, 108 ice-melting onset transition, 164-165 recent publication, 111-113 research needs, 226-233 studies using methods other than DSC or

texture of cereal and, 185 Glass transition temperature

diluent-monomer blend, 175-176 effects of water as plasticizer, 151-159 as function of

Glass transition, 107-109

DMA, 115, 119-120

log degree of polymerization, 138, 144 moisture, hand-washed and lyophilized

wheat gluten, 153-154 gluten, moisture effect, 219 low-molecular-weight carbohydrates,

148- 149

measurement in foods, 114-118 methods, 227-228

molecular weight effect, 138, 144-150 maltooligosaccharides, 144, 147

plasticization effect, 118, 120-125 polymer-polymer and

relaxation phenomena, 195-196 solute-specific subzero, 136-137, 139-143,

polymer-plasticizer blends, 145-150

161, 163-166 as function of dextrose equivalent,

as function of molecular weight,

low-molecular-weight carbohydrates,

temperature-composition domain above,

temperature location, determined by free

temperature-time conditions of

144-145

maltooligosaccharides, 144, 146

139-143

207-208

volume, 198

measurement, 227 Glassy state, recent publication, 111-113 Gliadins, 216

local intermolecular disulfide crosslinking, 217, 222

thermolabile disulfide bonds, 225 Glucose, concentration changes and

allolactose production, 83 Glucose oxidase, enzymatic acidification of

milk by, 16 Gluten

amorphous, thermosetting behavior,

aqueous systems, thermoplastic and

classification of proteins, 216 corn, 288,296 disulfide crosslinks, no change in

free-SH groups, 224 glassy state, 123, 126-127 hand-washed and lyophilized, glass

218-223

thermoset behavior, 223-226

transition temperature as function of moisture, 153-154

pathways leading to ultimate thermosets,

plasticization, bread-baking, 218-219 water-plasticized, thermosetting, 222-223

225-226

304 INDEX

wheat, as viscoelastic polymer system, 215-218

Glutenins, 216-217 long-range intermolecular disulfide

network formation, 217,222 thermolabile disulfide bonds, 225

Glycerol, solute-specific subzero glass transition temperature, 228-229

Glycoside hydrolases, 23-25 Glycosyl, enzymatic transfer reactions, 60 Gordon-Taylor equation, 150

H

Hevea rubber, local viscosity, 192 Hollow-fiber enzymatic reactor, 52-53

loaded on lumen side of membrane, 52-54

Hydrolysis lactose

chemical reactor, 45-46 in milk, 10-12

reverse, versus transgalactosylation,

whey, immobilized lactase reactors, 61-64

46-47

I

Ice cream, lactase in, 13-15 Industrial chemicals, from corn wet milling,

Inhibition 295

lactase, 42-43 transgalactosylase activity of lactases,

82-83

K Kluyveromyces fragilis, P-galactosidase, 25,

Kluyveromyces lactis 8-galactosidase

hydrolysis catalyzed by, 72-73 specificity, 81 transgalactosylase activity and

substrate components, 77-78 velocity for production of allolactose,

65, 70

78-79

lactase concentration effect on lactase-

catalyzed hydrolysis, 54-55 disaccharides and oligosaccharides

formed by, 77,79 temperature requirements, 34

enzyme concentration effect, 37-38 function of protease contamination, 40

lactase stability

product inhibition, 9-10

L

Lactase, see also P-Galactosidase

future potential, 56-59 properties, 30-39 purity, 39-41 sources, 5-6 technical data, 32 transgalactosylase assay conditions, 64,

commercial

66-68 deficiency, 3 effect on sensory properties of nonfat ice

enzyme modification, 57-58 in fermented milk, 14-17 fungus-derived, optimum temperatures,

hydrolase, assay methods and enzyme

milk, 14-15

33-34

activity chemically modified substrates, 27-29 colorimetric analysis, 26-27 hydrolysis rates, 30-31

hydrolase activity, 22-59 activation and inhibition, 41-43 assay methods and, 25-31 enzyme mechanism, 23-26 future potential for commercial

sources, 56-59 immobilization mechanisms and

reactor systems, 43-51 ultrafiltration bioprocess reactors, 47,

50, 52-56 in ice cream, 13-15 immobilization, 18-20

commercially available technologies, 47-48

immobilized systems, 16, 18-20, 44 impact of microbial source, 80

INDEX 305

level decrease with age, 3-4 linked to modified corn grits, 45 low-lactose milk, 6-7

processing by consumers, 7-9 powders, lactase units, 28-29 pressure-induced immobilization, 19 product inhibition, 9-13 protease activity, 40-41 research needs, 89-90 selection

PH, 33 stability, 35-39 temperature, 33-35

pH change effects, 39 skim milk

stability

enzyme concentration effect, 37-38 media and temperature effect, 37

transgalactosylase activity, 59-89 activation and inhibition, 82-83 assay methods, 64-69 commercial source effect, 77-80 compounds formed, 86,88 donor plus acceptor molecules, 76-79 in food, 84-85 future sources, 86, 89 future utilization, 85-89 lactase concentration, 76

potential uses, 89-90 specificity, 80-82 substrate concentration, 76-70 temperature and reaction time, 70-76 transgalactosylation versus reverse

pH, 65,69-70

hydrolysis, 61-64 waste lactose utilization, 20-21 yeast

immobilization of whole cells, 45 stability, 35

Lactic acid in laboratory steeping, 289-290 role, in corn wet milling, 281-282

Lactobacillus bulgaricus, potential lactase

Lactose, 37 sources, 86

concentration effect on allolactose production, 65, 69-70 production of oligosaccharides by

transgalactosidation, 74-76 half-life in column reaction, 47, 50

hydrolysis catalyzed by P-galactosidase, 24-25 chemical reactor, 45-46 maximum product synthesis, 72

syrup production, 21-22 Lactose-reduced foods, 5 Lactulose, formation, 69

M Magnesium ions, effects on

P-galactosidase, 82 Maize, see Corn Maltodextrins, 295 Maltooligosaccharides

Cg' values, 168-169 glass transition temperature as function

Michaelis-Menten equation, analogy with

Microwave, reheating of baked products,

Milk

of molecular weight, 144, 146-147

WLF equation, 200-204

220-221

fermented, lactase in, 14-17 P-galactosidase stability in, 36 low-lactose, 6-7

skim processing by consumers, 7-9

half-life in column reaction, 47, 50 lactase stability

enzyme concentration effect, 37-38 media and temperature effect, 37

sugar, malabsorption, 2-4 MM equation, 202 Mobility

defined by WLF equation, 179 diffusional, above and below glass

water, in glassy materials, 169-170 transition temperature, 173-174

0 Oligosaccharides

formed by K. lactis, 77, 79 in low-lactose milk and whey products,

production by transgalactosidation, 84-85

lactose concentration. 74-76

P pH, lactase, 65.69-71

selection and, 33

306 INDEX

Plasticization effect on glass transition temperature,

by sugars, 156-159 by water, 150-159, 187-188

118, 120-125

breakfast cereals, crispness and, 185 distinguished from presence of water,

unfreezability, 170

vitrification, 214

156-157

Plasticizer, effect on curing and

Polyisobutylene, local viscosity, 192 Polymer-plasticizer blends, glass transition

temperature, 145-147, 149-150 Polymer-polymer blends

diffusional motion, 174-175 glass transition temperature, 145-147,

149-150 Po 1 y m e r s

amorphous, thermosetting, 212-226 curing, 213-214 wheat gluten, see Gluten

crystalline melting enthalpies, 190 glass-forming, T,,,/Tg ratio, 191-192 partially crystalline

fringed micelle model of crystalline-amorphous structure, 129, 133

viscoelastic behavior, 183-184 Potassium ions, effects on

P-galactosidase, 82 Product inhibition, lactase, 9-13 Prolamins

corn, 275,277 wheat. 216

R Recrystallization, 211-212 Relaxation

diffusion-limited, rate defined by WLF equation and Arrhenius equation, 205-206

enthalpic, 172-173 a Relaxation, 199 p Relaxation, 199-200 Rotational diffusion, 196-197

s Salt, effect on P-galactosidase stability, 36 Shortening, effect on four-water bread

doughs, 220

Solute, diffusion in aqueous food glasses,

Stability, P-galactosidase, 35-39 Starch

169-177

acid-modified, 294 crosslinked, 293 gelatinization and retrogradation studies,

glassy state, 123, 126-127 modified, products from, 293-294 native, products from, 292-293 stabilized, 293

123, 128-129

Starch-water system, gelatinized, glass transition temperature as function of mass fraction, 152-153

glass transition temperature State diagrams, 135-143

as function of mass fraction, 152-153 location, 199

references, 137-138 Steeping, corn, 278-280 Steepwater, absorption by corn, 280-281 Streptococcus thermophillus

lactase, 58 potential lactase sources, 86

Sucrose, synthesis, lactase usage, 22 Sugar

aqueous solutions, viscosity temperature

effect on P-galactosidase stability, 36 malabsorption, 2-4 plasticization and antiplasticization by,

dependence, 180-181

156-159 Sulfur dioxide

in laboratory steeping, 289 role in corn wet milling, 282-283,

285-286

T Temperature

lactase selection and, 33-35 reaction time and, lactase, 70-76 role in corn wet milling, 286-287

Thermosetting, amorphous polymers,

Time-temperature scaling parameter, WLF

Time-temperature transformation reaction

Transgalactosylase activity, see Lactase

212-226

plots, 189-190

diagram, 212-213

INDEX 307

Transgalactosylation, versus reverse hydrolysis, 61-64

Translational diffusion, 196-197 Trehalose, Cg’ value, 168

U Ultrafiltration

bioprocess reactors, lactose hydrolysis,

enzyme bioreactor, modified localized,

lactase immobilization, 18-20

47,50, 52-56

52,54

V Viscosity

as function of reduced temperature for glassy and partially crystalline polymers, 108-109

local, 198 temperature dependence, aqueous

Vogel-Tammann-Fulcher equation, 194 solutions, 180-181

W Water

as crystallizing plasticizer, 161 diffusion in aqueous food glasses,

mobility, in glassy materials, 169-170 as plasticizer, 150-159

169-177

breakfast cereals, crispness and, 185 distinguished from presence of water,

food and biological materials, 120-123 independent mechanical aspect,

mechanical relaxation kinetics, 186-187 unfreezability, 170

156-157

187-188

Water activity concept, 229-230 Water dynamics, 127, 129-133, 160

Water-food structure interactions, 103-233; see also Food polymer science; Glass transition temperature

studies, 104-105

/3-galactosidase stability in, 36 half-life in column reaction, 47, 50 hydrolysis, immobilized lactase reactors,

lactose syrup production, 21-22 waste utilization, 20-21

Williams-Landel-Ferry free-volume

WLF behavior

Whey

46-47

theory, 135, 177

compared to Arrhenius behavior,

effect on kinetics of diffusion-limited relaxation processes, 208-209

relaxation times, 189-191 WLF coefficients, intuitive implications, 200 WLF equation, 160, 177-188

188-200

analogy with Michaelis-Menten

application, 178, 230 aqueous food-polymer systems,

correct glass transition temperature

crystallization kinetics, 181-182, 210-212 explicit reference temperature, 207 mobility definition, 179

contrast with Arrhenius kinetics, 204-212

equation, 200-204

viscoelastic behavior, 183-188

reference state, 230-231

WLF kinetics, 135, 176

WLF shift factor, 189-190

Y Yogurt

allolactose and galactobiose in, 84 lactose content, 14-16

Z Zea mays, see Corn Zein, 275, 277

I S B N O-L2-0Lb438-8

advances in food and nutrition research volume 38

Documents