v—W— _.'-——, __-_____._ __ ' .‘ , . THE EFFECT OF HIGH MOLECULAR WEIGHT CARBOHYDRATES 0N ASCORBIC ACID STABILITY IN A MODEL FOOD SYSTEM Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY STEPHEN CHARLES SECREST 1977 ABSTRACT THE EFFECT OF HIGH MOLECULAR HEIGHT CARBOHYDRATES 0N ASCORBIC ACID STABILITY IN A MODEL FOOD SYSTEM Product composition can affect the water activity, reactant mobility, and moisture content of a product, thus affecting ascorbic acid degradation. This study was designed to determine the effect. of fiber on ascorbic acid stability as a function of its water binding capacity. Six model systems were deweloPed, each containing a different commercial fiber. The commercial fibers included microcrystalline cellulose (MEG), cellulose floc with 2% carboxymethyl cellulose, cellulose floc, NCO/pectin mixture (70/30), NCO/pectin mixture (85/15) and cake flour. The six model food systems were equilibrated to 0.10, 0.40, and 0.65 av at 20°, 30., and 37°C, giving a total of 54 experimental conditions. The rate of ascorbic acid degradation followed first-order kinetics under all storage conditions. Vitamin C half-life values were extended in the high water binding capacity food systems. water activity, temperature, fiber and all interactions between the single sources of variation had statistically significant effects on ascorbic acid stability. THE EFFECT or HIGH MOLECULAR HEIGHT CARBOHYDRATES 0N ASCORBIC ACID STABILITY IN A MODEL FOOD SYSTEM By Stephen Charles Secrest THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1977 To J ena ii ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to Dr. James Kirk, his major professor, for continued guidance and encouragement throughout this study. Gratitude is also extended to Dr. V. Chenoweth, Dr. C. M. Stine and Dr. L. L. Bieber for reviewing this manuscript. A special word of thanks is extended to Mr. Dan Dennison, whom was a continual source of advise throughout the project. The author expresses his most sincere appreciation to his parents and wife, Jena, for their continual support and encouragement. iii TABLE OF CONTENTS Page Acmommms ....... .0.........OOOOOOOOOOOOOOOOO. m LIST OF TABLES 9.000.000.0.0.0000000000000000000000. Vii LIST OF FIGURES 0.0.0....0......OOOOOOOOOOOOOOOOOOOOO Vii INTRODUCTION .00....0.0.0.0........OOOOOOOOOO....... 1 REVIEW OF LITERATURE .......................... ..... 2 Nutritional Fortification of Cereals and Flour . 2 State of Vater as it Affects Microorganisms ... 4 State of Water as it Affects Enzymatic Reactions O0.0.0.000...OOOOOOOOOOOOOOOOOOOOOO 5 State of Water as it Affects The Chemical Stability of Food ..................... ...... 7 Lipid Deterioration Reactions as Affected Byw‘tor O......OOOOOO...00.000.000.000...... 8 Nonenzymatic Browning as Related to water Activity .....OOOOOOOOOOOO......OOOOOOOOOOOOO 9 State of Water as it Affects Vitamin Stability . 1O Ascorbic Acid Stability as Affected by Water Activity md water content ....OOOOOOOOOOOOOO 12 The Effect of Product Composition on Water Binding .0.........OOOOOOOOOOO00.0.0.0....... 15 Factors Affecting Water Activity of Foods Ha'ing Capillaryv“ter OOOIOOOOOOOOOOOOOOOOOO 19 EXPERIMENTAL PROCEDURES ............................ 22 Model System Constituents ..................... 22 Model System Processing Procedure ............. 23 Temperature - Relative Humidity Equilibration . 24 iv TABLE OF CONTENTS (cont.) Page Experimental Variables ........................ 26 Packaging of Model Systems .................... 26 Measurement of Ascorbic Acid Loss ............. 27 Moisture Content Determination ................ 28 Determination of Ascorbic Acid Half Life ...... 29 RESULTS 0............OOOOOOOOOOOOOOOOO...0.0.0.0.... 31 Ascorbic Acid Stability as a Function of Regulated Variables ................. ........ 31 Ascorbic Acid Stability as Affected by Interactions Between Single Sources of variation ......OOOOOOO.......OOOOOOOOOOOOOOO. 43 Fiber-Temperature Interaction ............. 44 Fiber-Water Activity Interaction .......... 45 Temperature-Water Activity Interaction .... 49 Affect of Heisture Content on Ascorbic Acid St‘bility ......OOOOOOOOO......OOOOOOOOOOIOOO. 50 DISCUSSION ......................................... 53 CONCLUSION ......................................... 61 RESEARCH SUGGESTIONS .................... ...... ..,.. 62‘ REFERENCES ......................................... 63 APPENDIXA .....OOOOOOOOOOOOOOOOO......OOOOOO....... 65 Total and Mean Ascorbic Acid Half Life Values ‘saFunction Of: .....OIOOOOOOOOOOOOOOOOOOO. 65 Fiber Group 0.0.0.000........OOOOOOOOOOOOO. 65 Temperature Group ......................... 65 V TABLE OF CONTENTS (cont.) Page Vater Activity Group ..... ....... .... ..... .. 66 Fiber-Temperature Interaction .... ....... ... 66 Fiber-Water Activity Interaction ........... 67 Temperature4Vater Activity Interaction ..... 68 APPENDIXB 0.00.0.0...OOOOOOOOOOOOOOOOOOI.0.00.0.0... 68 Analysis of Variance ...... ...... .... ..... ...... 68 Energy of Activation Values .. ...... . ...... ..... 69 APPENDIXC eeeeeeeeeeee e eeeeeeeeee eeeeee eeeeeeee eeeee 70 Fiber-Temperature Interaction .... ............ .. 70 Fiber-Water'Activity Interaction ............... 74 Temperature-water Activity Interaction ......... 78 LIST OF TABLES Table Page 1. Composition of Model Systems ................. 22 2. Commercial Fibers Utilized in Ascorbic tn Acid St‘bility Study ......OOOOOOOOOOOOOO... 23 3. Degradation Rates For Model Systems .......... 32 4. Ascorbic Acid Half Lives in Model Systems .... 35 5. Analysis of Variance For Ascorbic Acid Half- .3 Life Mean Values For All Model Food Systems As A Function 0f Single And MUltiple Experimental Variables ..................... 39 6. Moisture Contents 0f the Model Food Systems .. 50 LIST OF FIGURES Figure Page 1. Stability Map of Foods as a Function of water ActiVity ............OOOOOOOOOOIOOOOO. 6 2. Type II Isotherm Showing Sorption Hysteresis . 13 3. Variation of Spin-Lattice Relaxation Time (T2 ) Vith Moisture Content For Four mcr°m°1°cules ..........COOOCOOCCCOOOCOOC.0 17 4. Variation of Spin-LattiCe Relaxation Time (T2 ) with Water Activity For Four mcr°m°19cu1°8 ......O.......0.0.........CCO 17 5. Variation of Spin-Spin Relaxation Time Vith Water Activi y For Four Macromolecules ..... 18 6. First Order Kinetec Plotting Procedure ....... 29 vii List 10. 11. 12. 13. 14. of Figures (Cont.) Mean Ascorbic Acid Stability as a Function of Temperature 0.00..........OOOOOOOOCOOOOOO.... Mean Ascorbic Acid Stability as a Function of Water Activity .............................. Effect of Fiber on the Fiber-Temperature Interaction at 20 C ......................... Mean Ascorbic Acid Half Lives at Constant Fiber and Plotted as a Function of Temperature ......OOOOOOOOOOOOCOO0.00.0000... Significant Differences in Fiber-Water Activity Interactions as a Function of Fiber at a Constant av of 0.10 ......................... Effect of the FiberéVater Activity Interaction on Ascorbic Acid Stability ................... Significant differences in Fiber-Water Activity Interactions as a Function of Fiber at a constmtaw010e40e ......OOOQIOOOOOOIOOOOOO Effect of Temperature-Vater Activity Interaction on Ascorbic Acid Stability ...... viii Page 37 38 44 46 47 48 47 51 ‘EZZYT INTRODUCTION Presently, information concerning the stability of ascorbic acid as affected by water activity, moisture content, temperature and product composition is at best of limited value to the food processor. This is attri- butable in many instances to the use of artifically composed model systems which in no way resemble a food product. Trends, of course, can be noted when utilizing these simplistic systems, but do not always apply to an actual food product. This study was designed to look at the above stated variables and determine what affect they would have on ascorbic acid stability in a product composed of dehydrated food components. The temperature and water activity variables were standardized to represent various storage conditions which may be found through out the United States. Six model systems, each containing a different commercial fiber, were analyzed to measure what affect the physiochemical properties of each fiber had on the ascorbic acid degradation rate. All variables and interactions between the experimental variables were statistically analyzed to indicate which variables had the greatest and least influence on ascorbic acid stability. REVIEW OF LITERATURE NUTRITIONAL FORTIFICATION 0F CEREALS AND FLOUR Wheat flour, cereal based products, milk and salt *1 have historically served as carriers for vitamins and minerals. The contributions of the vitamin enriched ready- to eat breakfast cereals and fortified flours to the dramatic decrease of beri beri, pellagra, and ariboflavinosis serve HI as testimony to the importance of these cereal foods as nutrient carriers. The addition of vitamins and minerals to food products has long been viewed as a public health issue. The U. 8. Food and Drug Administration, the Council on Food and Nutrition of the American Medical Association, and the Food and Nutrition Board of the National Academy of Sciences- National Research Council have therefore attempted to establish a standard nutrient fortification policy. To date, fortification policies have not been standardized because of conflicting philosophies concerning nutrient fortification. The conflicting philosOphies are whether fortification should center on health needs with no concern to the composition of the original product or restoration of foods to their original nutrient levels. Presently, nutrient fortification is carried out in a large segment of our processed foods. For most foods, fortification above the label claim is required because of 2 imperfect distribution of nutrients in the product, analytical error, and nutrient degradation during processing and storage. The fortification level varies with each 'nutrient, product, and method of application. By regulating factors such as processing temperature, method of applica- tion, product packaging material and storage temperature, 5 maximum stability of the added nutrients can be promoted. ‘ The food technologist is often criticized for not directing more attention to the nutitional quality of food, ;J as affected by processing. However, the economic ease of nutrient supplementation versus changing a processing method to preserve the natural nutrients, monetarily benefits the consumer. Presently, flour, bread, corn grits and corn meal are enriched with vitamins B1, 82 and niacin. Foods without standards of identity, such as ready-to-eat breakfast cereals, have been fortified with a more diverse array of micro- nutrients. Breakfast cereals have been fortified with as many as ten vitamins. The Food and Nutrition Board of the National Academy of Sciences - National Research Council has proposed that cereal grains ( wheat, corn and rice) be fortified with vitamin A, thiamin, riboflavin, niacin, vitamin B6’ folic acid, iron, calcium, magnesium and zinc. Extensive information concerning vitamin stability has been accumulated for selected foods. However, 13.“:f1c- ient detail as to the effect of pH, moisture content, water activity and other product and package variables make extrapolation to other food products difficult. Further work in this area is required. STATE OF WATER AS IT AFFECTS MICROORGANISMS Scott (1957) studied the effect of water on microbial growth and determined that ‘w' rather than moisture content, determines the lower limit of availability of water for microbial growth. His studies indicate that most bacteria do not grow below a “w of 0.91, most molds cease to grow below a a" of 0.80 and xerOphilic fungi are inhibited at aw's less than 0.65 (Figure 1). Scott (1957) also concluded that environmental factors affect the specific aw required for microbial growth. Generally, the less favorable the other environmental factors, such as nutritional adequacy, pH, oxygen pressure and temperature, the higher the minimum water activity for growth of microorganisms. Labuza (1972) reported that, in addition to the above factors, microorganisms are affected in the following ways by the physiochemical state of the water. 1) A' modifies the sensitivity of microorganisms to heat, light, and chemicals. In general, micro- organisms are most sensitive at high aw's (dilute solutions). Minimum sensitivity occurs in the av range of intermediate moisture foods (av 0.50 - 0.65; water content 20%-40$). 2) Minimum water activities for production of toxins are often higher than for microbial growth. 3) Microbial growth is dependent upon the method of establishing equilibrum relative humidity of a food. Meat foods show a moisture sorption hysteresis ~ phenomenon, that is their moisture content at a given aw FT is lower on the adsorption loop of their sorption isotherm than on the desorption loop. Thus, the av limiting microbial growth may be higher for foods reaching their equilibrium moisture content by adsorption rather than by desorption of moisture. This supports the theory that microbial growth is controlled not by av alone but also by total moisture content. Some consideration has been given to the theory that differences in water content or microbial water binding may account for difference in heat resistance observed between closely related microorganisms. Tjoa (1973) noted that certain strains of Clostriduim botulinum bind more water than do other strains. The thermal resistance observed for the Clogtgiduig botulinum is greater in the strains that bind larger amounts of water at a given aw. STATE OF WATER AS IT AFFECTS ENZYMATIC REACTIONS Acker (1962) reported that a minimum amount of water is necessary for enzymatic activity. Acker (ibid) noted that it is not the absolute water content that is the 33:04 .033 0.0 0.0 .00 0.0 0.0 v.0 0.0 «.0 ..0 q N .2 . . . \\.WHIIIIHH¥ . \t\.\ \\\MO¢ N. \t s A? h0- s\\ \\\.0¢m¢9 \ \\\. at! Ill \ 9 \I \m?_ \\€%$.\\\ //z s I \aw. \ae ? / .haubuvee seems no momvomsu s as mecca no men homuwmsamli.— emsmwm BIVH BAIiV'IBM 7 controlling factor but the manner in which the water is bound. Thus aw is more suitable than moisture content in analyzing the physio-chemical state of water and its affect on enzymatic activity. Acker (ibid) also reported that enzymatic reactions occur above the monomolecular moisture level, where water is available for reactant mobility to the active site of the enzyme. Below the monomolecular moisture level, enzymatic reactions do not normally occur because of a lack of a transport medium (Figure 1). Other factors which affect enzymatic activity as a function of aw are: 1) The molecular weight and mobility of the substrate. 2) The heat stability of the enzyme, which is inversely related with moisture content and aw. STATE OF WATER AS IT AFFECTS THE CHEMICAL STABILITY OF FOOD It is an accepted fact that the correlation between chemical reactivity of food constituents and water is best represented by the av of the food rather than total moisture content. An increase in availability of water usually results in an increase in chemical reaction rates, although there are some exceptions to this rule. In chemical reactions, water can act in one or more of the following roles, as reported by Karel (1973): 1) Solvent for reactants, catalysts and inhibitors 2) Reactant (eg. hydrolysis reaction) 3) Product of reactions (eg. nonenzymatic browning condensation reactions) 4) Modifier of the catalytic or inhibitory substances (eg. inactivation of metal catlayst) PI... LIPID DETERIORATION REACTIONS AS AFFECTED BY WATER ! Lipid oxidation, unlike most chemical and biological : reactions is most rapid at low aw's. The accelerated rate ij of oxidation at low aw's is attributed to the absence of water's protective affect. Hydration of metal catalysts and hydrogen bonding to peroxides slows the oxidation reaction. The rate of oxidation decreases as the equilibrium moisture content is increased to the range of intermediate moisture foods (IMF). Salwin (1959) reported that many dehydrated products are less sensitive to lipid oxidation at av's above the monomolecular moisture level. The reaction rate slows with an increase in aw (s 0.50 aw) where upon it will accelerate upon further addition of water. Labuza (1970) has stated that the protective effects of water on lipid oxidation are due to: 1) Hydrogen bonding of water to hydrOperoxides which are produced during the free radical reaction. This inhibits the hydrOperoxides from decomposing, thus slows the rate of initiation. 2) Water lowers the catalytic activity of certain metals. 9 3) Vater can react with trace metals to produce insoluble metal hydroxides, which removes them from the reaction phase. At aw's above the IMF range the rate of oxidation increases (Chou and Labuza, 1973). The acceleration of the reaction rate at high water activities (over 0.50 av) is ' ”5‘ ___—. due to an increase in the soluble solids content (reactants), viscosity reduction, and swelling of the polymeric matrix .,‘ (Labuza, 1971). The swelling of the polymeric matrix LJ exposes new catalyst sites thus increasing the reaction rate. At water activities approaching 0.90 the dilution of reactants and catalysts will eventually reduce the reaction rate. The method by which foods reach a given aw also affects the rate of lipid oxidation. Oxidation rates have been reported to be 1.5 times greater for foods equilibrated on the desorption leap of the sorption isotherm versus the adsorption loop. This difference in the rate of oxidation, when comparing the adsorption and desorption equilibrated systems, is dependent on the degree of hysteresis the food system exhibits (Chou and Labuza, 1973). - NONENZYMATIC BROWNING AS RELATED TO WATER ACTIVITY Unlike lipid oxidation, nonenzymatic browning, which involves the reaction between the carbonyl and amino groups, increases as humidity increases to a maximum in the 10 av range of IMF and then decreases upon further dilution (Labuza, 1970). Water is both a solvent and a product of the non- enzymatic browning reaction. At low aw's the limiting factor is inadequate reactant mobility, however, even at aw's corresponding to the monomolecular moisture level browning has been shown to occur (Karel and Nickerson, 1964). At very high aw's (0.80 - 1.0) the dilution of reactants and the inhibition of browning by water condensation predom- inates, thus reducing the reaction rate (Labuza and Tannenbaum, 1970). Water not only accelerates the rate of browning, but also shortens the induction period of this complex reaction (Karel and Nickerson, 1964). The effect on induction time may indicate that formation of melanoidans proceeds by a different pathway at the low aw's or that sufficient intermediates must be formed before pigments are produced (Labuza, 1970). STATE OF WATER AS IT AFFECTS VITAMIN STABILITY Except for a few individual cases, extensive information relating the physiochemical state of water to vitamin stability is unavailable. However, correlations can be made between vitamin stability and the degradation of other components within the food system, which are affected by av. For example, the fat soluble vitamins (A,E) degrade as a function of aw, in a manner similar to unsaturated T741. LI 11 lipids (Figure 1). In general, the accelerating and inhibitory effects of water on lipid oxidation reactions apply to the oxidation of fat soluble vitamins. This relationship may be an over simplification of the effects of aw and moisture content on the lipid soluble vitamins, however, kinetic data describing the loss of fat soluble vitamins has not been completed and further conclusions can not be made at this time. Thiamine (Mulley and Stumbo, 1975) and ascorbic acid (Lee and Labuza, 1975; Kirk and Dennison, 1977) have been studied extensively in regard to the effect of water on their rate of degradation. Their overall instability within food products can be a limiting factor in the nutritional shelf-life of foods. The degradation rates of thiamin and ascorbic acid generally follow first-order kinetics and increase with increasing aw's. Thiamin is generally considered to be stable in most foods. Riboflavin is stable in food under ordinary conditions, however, when eXposed to light this vitamin readily degrades to lumiflavin which is a strong oxidizing agent and can catalyze destruction of a number of other vitamins, particularly ascorbic acid (Tannenbaum, 1976). Ascorbic acid stability will be discussed in the next section. Not enough is known about the other water soluble vitamins to comment on what one might expect in terms of degradation kinetics, although they are presently assumed to be more stable than ascorbic acid. 12 ASCORBIC ACID STABILIT;_AS AFFECTED BY WATER ACTIVITY AND WATER CONTENT Vater activity (av) is a measurement of the partial water vapor pressure of a substance divided by the vapor pressure of pure water at a constant temperature. The water absorbed by a hydrOphillic food substance exists in three different states (Kupranoff, 1958). 1) The monomolecular moisture level, referring to the fraction of water which is strongly bound to individual polar groups of the food product (carboxyl or amine groups). 2) The multilayer, referring to additional layers of water molecules which are hydrogen bonded to the monomolecular layer, aldehyde and hydroxyl groups. 3) Capillary region, referring to water condensed in the interstitial pores of the food product. The lowering of water activity in this region is affected by the concentration of soluble solids (Raoults Law) and capillary action restriction. The manner in which water is found in a dehydrated food product corresponds to three regions of the sorption isotherm of the product (Figure 2). Region A (Figure 2) corresponds to the strongly bound monomolecular moisture layer, B corresponds to the hydrogen bonded region and region C relates to the capillary condensed water. Thus a sorption isotherm would differ for each individual product 13 I- Z l.l.l l- O u o l m I 0: I D A . I B ‘_____. B—u———.i to 5 I 2 I - ----—- L L 20 4O 60 80 I00 % RELATIVE HUMIDITY Figure 2.-—Type II isotherm showing sorption hysteresis. 14 due to the fluctuation in water binding properties (Labuza, 1968). Because these regions overlap, we obtain the characteristic curve of the sorption isotherm. The destruction of ascorbic acid is most rapid in the range of av's which include region C of the sorption isotherm (Figure 2). However, all water present, including that adsorbed in the monomolecular moisture layer appears to be available for ascorbic acid destruction (Karel and Nickerson, 1964). At a given moisture content, the degree of water binding affects the reactant and catalyst mobility. A product having a higher preportion of ionic groups (at a given moisture content) would supposedly exhibit a lower ‘w’ resulting in a lower rate of ascorbic acid degradation. Likewise, at a constant aw, the food system with a greater proportion of ionic groups would have the higher moisture content. Reactant and catalyst mobility are a function of the av rather than moisture content, thus at a constant aw there is a greater loss of ascorbic acid in foods equilibrated on the desorption loop of the sorption isotherm versus the adsorption loop. This phenomena is reportedly due to the higher moisture content on the desorption loop (Lee and Labuza, 1975). As noted earlier, an increase in aw results in an increased degradation rate of ascorbic acid. Lee and Labuza (1975) have interpreted this increase in ascorbic acid destruction to be the result of dilution in the aqueous phase, resulting in a decreased viscosity and increased 15 mobility of reactants. It has also been reported that an increase in aw or moisture content can result in swelling of the polymeric matrix, thus exposing new catalytic sites and ascorbic acid (Labuza, 1970). The decreased viscosity resulting from increased aw may also affect the level of dissolved oxygen in the food system. Kirk and Dennison (1977) have reported that as the av of a food system increases, the molar ratio of dissolved oxygen to ascorbic acid increases. This may be the primary factor responsible for the increase in the rate of ascorbic acid degradation as a function of aw. THE EFFECT OF PRODUCT COMPOSITION ON HATER BINDING As stated earlier, the sorption isotherm of a food is affected by the degree and type of water binding as influenced by the food constituents. Theoretically, different food constituents could reduce or accelerate a chemical, enzymatic, or microbial reaction by altering the availability of water. The complexity of a food product results in a mixture of water binding properties, giving a continuous curve for the sorption isotherm rather than clear divisions at specific water activities. Rockland (1969) pointed out the complexity of water activity in food systems as affected by carbohydrates, proteins, lipids and minor constituents of foods. Recent work by Leung (1976) using NMR technique, has shown that at both constant aw's and water contents 16 there are dramatic differences in the tenacity with which the water is bound by different food products (Figures 3 and 4). As shown by the data in Figures 3 and 4, pectin has a relatively strong water binding effect, whereas, water bound by cellulose was so loosely associated that it was not plotted. Leung (1976) has indicated that as a"r increases water mobility increases (Figure 5). However, at a given aw, there are differences in the tenacity with which the water is bound. This differs from the former theory that aw is the ultimate measurement of water mobility. The following list points out the degree of water binding exhibited by selected types of food components. 1) Proteins and starches adsorb more water at low aw's than fatty materials or crystalline substances (Labuza, 1968). 2) Carboxymethyl cellulose is a better water binder than casein (Karmas and Chen, 1975). 3) Cellulose exhibits the poorest water binding capacity other than crystalline sugars (Leung, 1976). 4) Pectin binds less water than starch but holds on to the water more tenaciously (Leung, 1976). As noted earlier the amount of water bound by a substance does not always correlate with the strength of binding. Thus product composition not only affects the amount of water bound by the system but also affects the no 00 3. TI (meec) 20 17 Corn Starch ............ Sodium Alginate -——.... Casein -—- ..... Pectin .... 0.2 0.4 0.6 0.8 I.0 Water Activity Figure 4.-Variation of spin-lattice relaxation time (T2) with water activity for four macromolecules. IOO T'ImeecI Corn Starch ............ Sodium Alginate —-— /" Casein ....... "’,”:/ Pectin ,’ // l 0.2 0.4 0.6 0.8 1.0 .9 water per'g "DM Figure 3.-Variation of spin-lattice relaxation time (T2) with moisture content for four macromolecules. 18 Corn Starch -------- - Sodium Alginate —— _ .... C‘s.1n-_- -—-- ”a / 2.5 - Pectin / , 2.0 - / ," LS * {X /’ [.0 - ’ / 0.5 - -.';'/’ T2 (mse c) l 0.2 0.4 0.6 0.8 l.0 Water Activity Figure 5.--Variation of spin-spin relaxation time with water activity for five macromolecules. 19 water mobility at both constant aw's and moisture contents. The effect of processing on the aw of foods is outside the realm of this literature review, however, it should be noted that all food constituents are not affected in the same manner or degree by a single treatment. Heat, for example, has little or no effect on protein hydration, but has a dramatic effect on starch (Labuza, 1968). FACTORS AFFECTING WATER ACTIVITY OF FOODS HAVING CAPILLARY WATER The many differences in the chemical and physical composition of food can affect the water activity, thus influencing the degradation rate of nutrients within a food system. The capillary condensed water is most avail- able for reactant and catalyst mobility. Therefore, factors which affect the av of the capillary water will have a dramatic effect on the degradation rate of ascorbic acid. The depression of av in the capillary condensed region is due mainly to dissolved solutes, which ideally follows Raoult’s Law. RAOULT'S LAV aw = NV Nw + Ns Where aV = water activity w = number of moles of water number of moles of solute 20 Deviations from Raoult's Law occur for one or more of the following reasons: 1) Not all of the water in food is capable of acting as a solvent for the solutes (eg., bound to ionic groups). 2) Not all of the solute is in solution (eg., bound to insoluble food components). 3) Interactions between solute molecules cause deviations from ideal relations. Another factor affecting the vapor pressure at high aw foods is depression by capillary forces. The extent to which these forces are Operative at very low aw's is not fully understood (Karel, 1973). However, one would expect definite capillary effects above 0.90 aw if the capillaries are present in extensive numbers (Labuza and Rutman, 1968). As earlier indicated, reactant mobility is at its maximum at the capillary condensed region. Therefore, reactant mobility could dramatically affect the stability of ascorbic acid in food systems which do not swell with water adsorption. Although reactant mobility is a rate determining factor for ascorbic acid degradation in a swelling food system, the swelling system will expose more catalytic sites upon addition of water therefore conceiv- ably making mobility of reactants a secondary consideration (Karel, 1973). 21 Since mobility of reactants is affected by water binding capacity (HBO), theoretically an increase in WBC should result in a decrese in aw. Such is not always the case. For example, in aqueous sodium caseinate mixtures there is no significant correlation between water binding and water activity (Karmas and Chen, 1975). As the concen- tration of sodium caseinate increases, water binding correspondingly increases, however, av was depressed very little. Karel (1973) concluded that this was a result of two physiochemical effects: 1) Collodial proteins do not have as much interaction with water as a completely water-soluble solute and 2) the molecular weight of proteins are too high to change the molar fraction of water and produce a significant effect on aw (Raoult's Law). EXPERIMENTAL PROCEDURES MODEL SYSTEM CONSTITUENTS Six dehydrated food model systems, differing only by the variability of commercial fiber added, were prepared according to the composition shown. Table 1. Composition of Model Systems Constituents Per Cent Flour a 30 Corn 011 b 10 NaCl 1 Sucrose 29 Commercial Fiber 30 The commercial fibers which distinguish one model system from another are listed in Table 2. a (60% extraction) Minnill Mill, Fostoria Ohio b Mazola Corn Oil 22 Table 2. 23 Commercial Fibers Utilized In Ascorbic Acid Stability Study Food Model System Number and Description #1 #2 #3 #4 #5 #6 MicrocrystallineCellulose (MCC) - FMC Corporation Low water Retention Cellulose Floc plus 2% Carboxy Methyl Cellulose Large partical size - FMC Corporation - Prototype # 174-1 High water Retention Cellulose Floc Large particle size - FMC Corporation - Prototype # 174-2 Low Water Retention Microcrystalline Cellulose/Pectin (70/30 ratio) FMC Corporation High Hater Retention Microcrystalline Cellulose/Pectin (85/15 ratio) FMC Corporation High Vater Retention Vheat Flour (60% extraction) Minnill Mills Fostoria, Ohio High Water Retention MODEL SYSTEM PROCESSING PROCEDURE All food model systems were processed under identical condition in order to eliminate the processing procedure as an experimental variable. The model system components (excluding fiber) were slurried to a concentration of 35% total solids in deionized water, at 60°C. The warmed slurry was homogenized at 2,500 psig on a two stage Manton-Gaulin homogenizer (2000 psig 1st stage; 500 psig 2nd stage). After 24 the slurry had cooled to room temperature, reduced L-ascorbic acid was added to the system to ensure homogenous distri- bution of the water soluble vitamin. Ascorbic acid was added at a level of 25% the NAS RDA/ 100 grams of model system (dry weight basis) or 11.25 mg/100 grams. The slurry was then poured into freeze-drying trays (approx. 1.3 cm layer), frozen to a temperature of -50°C and freeze-dried to an absolute pressure of 5 microns at a shelf temperature of 38°C. After drying, the system was crushed and blended with the various commercial fibers using a Hobart mixer. Homogeneity of the model systems was deter- mined by monitoring ascorbic acid concentration in various aliquots of the model system. Following mixing, the six model systems were stored in air tight containers at 4°C until samples were equilibrated to various water activities. TEMPERATURE-RELATIVE HUMIDITY EQUILIBRATION An Aminco-Aire unit was used to equilibrate the model systems to the desired water activities (0.10, 0.40, and 0.65). The model systems were equilibrated to their corresponding equilibrium relative humidities by placing thin slabs of the freeze-dried model system in the Aminco— Aire equilibration chamber and forcing conditioned air over the product as described by Kirk and Dennison (1977). Since all samples absorbed water during equilibration, the water activities of the sample were on the adsorption loop 26 of the sorption isotherm. For samples equilibrated to water activities-of 0.10, a dehumidifying system in.conjunction with the Aminco-Aire equilibration chamber was required to achieve the desired moisture content in the atmosphere. A Hygrodynamics Hygrometer was used to sense the relative humidity of the air being forced through the equili- bration chamber. EXPERIMENTAL VARIABLES Three temperatures, three water activities, and six fiber systems were used as the experimental variables in this study. For each of the six model systems there were nine storage conditions. 1) 20°c/o.1o aw 4) 30°c/o.1o av 7) 37°c/o.1o av 2) 20°c/o.4o aw 5) 30°C/0.40 aw 8) 37°C/o.4o aw 3) 20°C/o.65 aw 6) 30°C/0.65 aw 9) 37°C/0.65 aw PACKAGING OF MODEL SYSTEMS Approximately 20 grams of equilibrated model system were placed in a 303 X 406 enameled metal can (303 can). Ample headspace was left in each container to ensure that oxygen was not a rate limiting factor in the degradation of ascorbic acid. Package variability, such as permeability to light, moisture vapor and gas transmission, was eliminated by using the metal containers. After canning, the equili- brated samples were placed in storage cubicals which 26 of the sorption isotherm. For samples equilibrated to water activities of 0.10, a dehumidifying system in.conjunction with the Aminco-Aire equilibration chamber was required to achieve the desired moisture content in the atmosphere. A Hygrodynamics Hygrometer was used to sense the relative humidity of the air being forced through the equili- bration chamber. EXPERIMENTAL VARIABLES Three temperatures, three water activities, and six fiber systems were used as the experimental variables in this study. For each of the six model systems there were nine storage conditions. 1) 20°C/o.1o aw 4) 30°c/o.1o aw 7) 37°c/o.1o aw 2) 20°C/0.40 av 5) 30°C/0.4O aw 8) 37°C/0.4O aw 3) 20°C/0.65 av 6) 30°C/0.65 aw 9) 37°C/O.65 aw PACKAGING 0F MODEL SYSTEMS Approximately 20 grams of equilibrated model system were placed in a 303 X 406 enameled metal can (303 can). Ample headspace was left in each container to ensure that oxygen was not a rate limiting factor in the degradation of ascorbic acid. Package variability, such as permeability to light, moisture vapor and gas transmission, was eliminated by using the metal containers. After canning, the equili- brated samples were placed in storage cubicals which 27 corresponded to the temperature at which the model system was equilibrated. Ten samples were canned for each experi- mental condition in order to obtain an accurate measurement of the rate of ascorbic acid destruction as a function of time. MEASUREMENT OF ASCORBIC ACID LOSS The model systems were removed from the storage cubicals and analyzed for ascorbic acid concentration at various time intervals depending on the rate of ascorbic acid degradation. Sample preparation included the following steps: 1) Five grams of sample were placed in a Waring blender jar (250 ml capacity) to which 100 mls of a meta- phosphoric-acetic acid solution was added. 2) Samples were blended for approximately 30 seconds. 3) Blended solutions were filtered through Whatman #42 filter paper. Model food systems containing fibers #4 and #5 required filtration of the ascorbic acid extracts as outlined below: Food system #4 - Filter the blended solution through Vhatman #1 filter paper followed by Millipore filtration (HA 5u membrane). Food system #5 — Filter extract through Vhatman #1 filter paper followed by Vhatman #42 filter paper. 28 Total, reduced, and dehydroascorbic acid were measured by the continuous flow o-phenylenediamine with dehydroascorbic acid resulting in the formation of a blue fluor0phor, which has an absorption maximum at 350 nm and an emission maximum at 430 nm. Blank values were determined by complexing boric acid to the dehydroascorbic acid to prevent the condensation of o-phenylenediamine with dehydro- ascorbic acid. Total ascorbic acid was measured by oxidizing the reduced ascorbic acid to dehydroascorbic acid with 2,6 dichloroindophenol and following the same procedure described for dehydroascorbic acid. The first ascorbic acid analysis of the model system after canning was designated as zero time when determining the degradation rate. MOISTURE CONTENT DETERMINATION The moisture contents of the model systems were determined by placing approximately two grams of sample into a tared moisture dish and drying in vacuum at 100°C for 24 hours- A cold trap (-80°C) and the addition of dry air (bubbled through concentrated sulfuric acid) at a rate of approximately 15-20 mls/minute were used to aid in drying. Percent water was determined from the difference in weight before and after drying. 29 DETERMINATION OF ASCORBIC ACID HALF LIFE The model systems were analyzed for ascorbic acid concentration as a function of time and the rate of destruction was calculated using the first order kinetics equation. The kinetic determinations were supported by data extending over at least two half lives of ascorbic acid. The logarithim of the ratio of ascorbic acid remaining versus the initial ascorbic acid concentration was plotted against time to determine the degradation rate (k). Figure 6.-First order kinetic plotting procedure ascorbic acid concentration at various times 3» II A = initial ascorbic acid concentration at O 0 time _1 k = rate constant, days 30 A rate constant (k) was determined for each of the 54 variable conditions. All 54 conditions followed first- order rate kinetics. The relationship between the half life of ascorbic acid (ti) and the rate constant is as follows: The half life values, which are a function of the degradation rates, were used for the statistical analysis of the experimental data. RESULTS ASCORBIC ACID STABILITY AS A FUNCTION OF REGULATED VARIABLES The variable conditions within the model system included six fibers, three storage temperatures, and three water activities, giving a total of 54 different conditions. As previously stated, the distinguishing factor between the six model food systems were the commercial fibers incorp- orated into the products (Table 2). The sources of variation in the study included fiber, temperature, aw, fiber-temper- ature interaction, fiber-aw interaction and temperature-av interaction. The rate of total ascorbic acid degradation for each of the 54 conditions followed first order kinetics. The rate constants, standard deviations and correlation coeffecients are listed in Table 3. The mean correlation coefficient for the 54 rate constants was .9792. Half-life values calculated as a function of the degradation rate constants are listed in Table 4. By studying the ascorbic acid half-lives (Table 4), one can point out specific trends due to the sources of variation. Data in Tables 3 and 4 show specific effects of each experimental parameter on the stability of total ascorbic acid in the model food systems. 31 32 voowouuueoo :ofiasuoanoo c or N nowadm>oc cuscasam n N N mam. can. Nae.» ohm. one. mvp.h whm. 5mm. b»¢.m —hm. owe. $00.0 mam. On—. —oo.~ mom. phw. omn.n was. soo. Nao.n Adan mac. Poa.~ omm. EN". nov.m mam. ow—. eno.m mmm. N~—. mo—.p mam. va. can.— pam. who. Nun.p Nam. moo. mpm.p mam. hvo. vmp.p mam. ”no. FNN.F «as. one. own. .wom. ooo. saw. on no em on no as mfi So¢ahm pfi Seesaw mzmsmwm ammo: mom mmaqm onadfidmuQn OF N ..vfldfinflcc 0.5.6." 6 as.\ooam oo.\ooan o_.\ooan mo.\ooon oo.\oo0m o..\ooon mo.\ooo~ . oo.\ooo~ o..\ooo~ mowaudmoo .n ounce 33 acououuueoo soaeeaoaaoo o o. a ooaoo.>oo oooooooo a or K amsenooo cash a N m and. Ame. 4o..o “so. New. ohm.» new. no“. ms..n awo. .on. .mn.» mam. sou. sn~.~ and. cam. nap." and. ~o.. mm..~ Add. aao. oao.~ .sm. m4.. ooo.. one. amm. .aa.~ and. an.. .4... sea. a... .oo.. and. .m.. mnn.. osm. moo. a.o.. odd. «no. woo. woo. .ao. Nan.. ass. ooo. own. moo. moo. coo. ok no em oa be ex wk Bovahm 600% n% aeaehm woom armamwm ammo: mom mfladm on9 .efiao owe o. .. nmo.n mon.o wm mmfixmz season a mmqm Q492H=Hmmmxm HamHsADz 324 Havana mo onHOZDm < m4 mzmamwm Room ammo: mom mmDA<> 24H! flmHAIMA ho mHmHA no ooaoom Add .n ounce 4O variation have been shown to have a significant effect on the ascorbic acid stability. As indicated by the magnitude of the f-ratios, the Temperaturea'ater Activity Interaction has a greater influence on ascorbic acid stability than do the Temperature-Fiber or Fiber-Water Activity Interactions. As previously stated, the different fibers in the model systems appear to influence the ascorbic acid degrad- ation rate (Table 4). Analysis of Variance confirmed that these differences are statistically significant, however, it does not indicate which fibers or why these fibers extend the nutrient half life. The Fiber Group Totals (Appendix A) indicate that the fiber in food system #4 has the greatest stabilizing effect on total ascorbic acid, followed by #5, #2, #6, #1 and #3. Food systems #2, #4, #5 and #6 all contain high water binding capacity fibers which appear to induce a more favorab e e vironment for ascorbic acid than do the low water binding capacity fiber systems. It is of interest to note the difference between the ascorbic acid half-life values for food systems #2 and #3 (Fiber Group Totals, Appendix A). This difference could only be attributable to the 0.6% carboxymethyl cellulose which was present in model food system #2. In order to determine why some fibers stabilized ascorbic acid to a greater extent than others, a Bonferroni t test (Miller, 1966) was utilized. 2) 3) 4) 41 The questions that arise include the following: Is there a statistically significant difference (total ascorbic acid half-life values) among the high VBC fibers (#2, #4, #5 and #6) and the low WBC fibers (#1 and #3)? Is there a significant difference (total ascorbic acid half-life values) in the low pH model food systems (#4 and #5) versus the high pH model food systems (#1, #2, #3, and #6)? Is there a statistically significant difference in total ascorbic acid half-life values between the two low water binding food systems? Is there a significant difference in total ascorbic acid half-life values between the two low pH food systems? Bonferroni t test (Miller 1966) Question #1 tb=2(y1+y3) - (2+y44-y5-t- ) \Izcz (m3) / r C 2 questions r a repetition MSE a means square error I; = mean half life of ascorbic acid in corresponding fiber system (nine conditions) 42 tb = 2(28.1 + 36.2) - (42.7 + 47.2 + 43.0 + 41.9) J‘ 12 12.4) 9 vs. 4 tb 0.025 tb = -13.97 2.74 (corresponds to 95% confidence level) Because 13.97 is greater than 2.74, the high water binding food constituents significantly enhance the enhance the environment for ascorbic acid stability. Question #2 Low versus High pH Food Systems vs. a tb 0.025 tb = 11.68 2.74 Because 11.68 is greater than 2.74, polygalacturonic acid did significantly stabilize the ascorbic acid in the model food system. Question #3 Food System #1 versus #3 vs. 4 t 0.025 t = 1.17 2.74 b b In this case, 1.17 is less than 2.74, indicating no significant differences between the two low water binding food systems. Question #4 Food System #4 versus #5 vs. a t 0.025 t = 2.35 2.74 b b No significant difference could be determined between the two pectin based food systems. 43 These data confirm that by increasing the water binding capacity (UBO) of a product at a given water activity, the half life of ascorbic acid can be increased. Likewise, the addition ofaa food component which lowers the pH of the system may result in an extended half life for ascorbic acid. Using these two parameters it is pessible [Vi to speculate that the pectin based model systems, which have "‘5—1 . lower pH's and higher VBC's, would have a greater stabilizing effect than the other fiber model systems. Experimental [:‘Ild ...“.Vd data confirms that the pectin based fiber systems stabilize the ascorbic acid more effectively than the remaining fiber systems. In contrast, fiber systems which had a high pH and low VBC preperties were shown to reduce ascorbic acid stability (Fiber Group Totals, Appendix A). ASCORBIC ACID STABILITY AS AFFECTED BY INTERACTIONS BETWEEN SINGLE SOURCES OF VARIATION Analysis of Variance data in Table 5 confirms that the fiber-temperature interaction has a significant effect on ascorbic acid stability. However, which of these para- meters, temperature or fiber, has the greater influence on this interaction? The Scheffe test (Miller, 1966) was used to determine whether the fiber has a significant affect at a constant temperature and whether temperature has a significant effect at a constant fiber (Fiber-Temperature Interactions, Appendix A). 44 Fiber-Temperature Interaction Congtant Temperature 20°C As can be seen by the data (Appendix C), the effect of fiber on the fiber-temperature interaction at 20°C indicates that temperature fluctuations had a greater effect on the ascorbic acid degradation rate than did fiber vari- ability. Data in Figure 9 indicate the food systems which are significantly different from one another (fiber-temp- erature interaction) as a result of fiber variability. Figure 9.-Effectoof fiber on the fiber—temperature interaction at 20 C. -1' ““1' F_—-7' A- #4 #5 2 6 1 . L [#3 # 4 in 'T'I: "T1— ’ '7 Most Stable Least Stable Fiber-Temperature Interaction As Affected by Fiber At 20 C Fiber-Temperature Interactions Coggtant Tgmpggature 30°C and 37°C The Scheffe test (Miller, 1966) indicates that none of the rate constants at 30°C or 37°C were significantly different as a function of fiber variability (Appendix C). 45 Fiber-Temperature Interaction Constant Fibgr With fiber as a constant and temperature as the dependent variable, it is possible to determine the degree of influence exerted by temperature on the fiber-temperature interaction. The Scheffe procedure (Miller, 1966) confirms that all of the fiber-temperature interactions are signifi- cantly different from one another as a function of temper- ature (Appendix C). Therefore, the effect of a fiber- temperature interaction on ascorbic acid is due more to temperature variation than fiber variability. The effect of fiber on ascorbic acid stability was shown to be more pronounced at the lower temperatures (Figure 10). Fiber-Hater Activity Interaction Constagt Hater Activity 0.10 a" The fiber-av interaction is also shown to have a significant affect on ascorbic acid stability. Utilizing the Scheffe procedure (Miller, 1966), with water activity as a constant, the effect of fiber variability on the fiber-av interaction was determined (Appendix 0). Food systems which are significantly different at their fiber- water activity interaction point, as a function of fiber variability, are indicated by attached lines. Mean Half Live 8 (days) (H OH 00 Cd 46 Figure 10.-Mean ascorbic acid half lives at constant fiber and plotted as a function of temperature. ~ A 20 Food Food Food Food Food Food l 30 0 Temperature C System #1 System #2 System #3 System #4 System #5 System #6 l 37 II 0) D' 5 (3 C) 47 Figure 11.—-Significant differences in fiber-water activity interactions as a function of fiber at a constant aw of 0.10. Most Stable Least Stable Fiber-Water Activity Interaction (0.10 aw) The high water binding capacity food systems (#2, #4, #5 and #6) are significantly more protective toward ascorbic acid than the low water binding capacity food systems. Fiber-Water Activity Interaction Consta t V ter Activit 0.40 a" Figure 13.-Significant differences in fiber-water activity interactions as a function of fiber at a constant a" of 0.40. In; L_#6 #3 Most Stable Least Stable FiberaVater Activity Interaction (0.40 aw) Mean Half Lives (days) 48 74 . Food System #1 0 Food System #2 D I Food System #3 ‘ 6 4 Q Food System #4 A Food System #5 e Food System #6 I C 54 ’ 4 4 . 34 " 24 J: L701 Ufa oss- Water Activity Figure 12.-Effect of the Fiber-Water Activity Interaction on ascorbic acid stability. 49 As noted from the data in Figures 11 and 13, there are fewer food systems at 0.40 aw which exhibit significantly different rate constants for total ascorbic acid destruction, than at 0.10 aw. At 0.65 aw the stability of ascorbic acid was not significantly different in the model food systems as a function of fiber. The high VBC fiber systems did, however, correspond to longer half-life values than did the low YBC fiber systems. PIBERAWATER ACTIVITY INTERACTIONS AS A FUNCTION OF WATER ACTIVITY Vith water activity as the experimental variable, statistical analysis using the Scheffe procedure (Miller, 1966) indicates the effect of water activity on the fiber- water activity interaction (Appendix A). As shown by these data (Appendix C), all values are greater than the minimum significant difference and therefore, all points (at a constant fiber) are significantly different from one another as a function of aw. TEMPERATUREAVATER ACTIVITY INTERACTION The temperature-water activity interaction did exhibit a trend in stabilizing ascorbic acid in the model food systems (Figure 11: Appendix A). Analysis of Variance of the rate constants for ascorbic acid degradation showed statis~. tically significant interactions between temperature and 50 av, although not for all points. Using the Scheffe procedure (Miller, 1966) the effect of av and temperature were separately analyzed to determine their effect on the temperature-aw interaction (Appendix C). As can be seen from the data in Appendix C, all points at a given temp- erature were significantly different with the exception of 0.40 and 0.65 av at 37°C. AFFECT 0F MOISTURE CONTENT ON ASCORBIC ACID STABILITY Moisture contents of the model system were measured at each of the nine water activity-temperature conditions. Table 6.-Moisture contents of the model food systems (fl) V Food System 20°c/.10 av 20°c/.40 aw 20°c/.65 aw 30°c/.10 a #1 2.36 3.54 5.46 2.56 #2 2.16 3.93 5.60 2.71 #3 2.30 3.74 5.55 2.98 #4 2.65 3.87 6.06 2.99 #5 2.32 3.74 5.77 2.99 #6 3.51 4.99 6.75 3.51 Mean Half Lives (days) 51 I20 ‘ l00 01 O 0 0 «lb 0 N O A 20 30 o 40 Temperature C Figure 14 .---Effect of temperature-water activity interaction on ascorbic acid stability. 52 Table 6 continued Food 0 0 0 0 System 30 c/.40 av 30 c/.65 av 37 c/.10 av 37 c/.40 aw #1 4.37 5.36 2.25 3.21 #2 4.79 5.40 2.14 3.15 #3 4.60 5.50 2.08 3.23 #4 5.17 5.98 2.40 3.68 #5 4.52 5.63 2.08 3.60 #6 5.87 7.19 2.97 5.61 Food 0 Mean of the System 37 C/.65 aw nine conditions #H 3.84 3.66 #2 3.82 3.75 #3 3.68 3.74 #4 4.43 4.14 #5 4.55 3.91 #6 5.46 5.10 Moisture contents of the model systems varied with the type of fiber incorporated. The four highest moisture contents correspond to the four high water binding capacity fibers. As stated earlier, the high VBC fibers extended the half life of ascorbic acid in the dehydrated model food systems. The dramatic difference in water content might be expected to effect the ascorbic acid in the dehydrated model food systems. However, in food system #6 the higher moisture content did not adversely affect the stability of ascorbic acid. DISCUSSION The rate of degradation for total ascorbic acid, followed first-order kinetics in all 54 experimental conditions. An Analysis of Variance indicated that all single sources of variation (temperature, fiber and av) significantly influenced the ascorbic acid degradation rate. The degree of influence each parameter induced is a function of the range of that particular parameter. For example, if av had ranged from 0.00 to 1.00, it rather than temperature could have exhibited the greater influence on ascorbic acid degradation. Of all single sources of variation, fiber variability was shown to have the least effect on ascorbic acid stability. However, as is noted by the Analysis of Variance, fiber variability does have a statistically significant effect on the rate of ascorbic acid degradation, at a constant a" and storage temperature (Table 4). The effect of fiber on the stability of ascorbic acid appears to be related to the water binding capacity (VBC) exhibited by each fiber system. The high VBCinber systems contained higher moisture contents than did the low VBC systems (Table 6). Not only did the higher NBC systems bind more water at a given aw, but also significantly extended the half-life values of ascorbic acid over the low VBC food systems (Fiber Group Total, Appendix A; Bonferroni t test). This phenomenon may be due to a decreased mobility of 53 54 catalysts and ascorbic acid. This theory is supported by Leung's work (1976) which indicates that at a constant av, water mobility varies as a function of the food components. His work indicated that pectin drastically reduced water mobility as compared to other components, even at higher moisture contents. Thus, the same effect could apply to the pectin containing food systems in this study. Even at higher moisture contents, the pectin based food systems were most effective in extending the ascorbic acid half-lives (Tables 4 and 6). A significant difference in the rate of ascorbic acid destruction was found to exist between the low pH model food systems (#4 and #5) and the high pH model food systems (#1, #2, #3 and #6). Vhether this difference was due to the lowering of pH by the polygalacuronic acid units, an increase in VBC or metal chelation by the organic acid can not be determined from this study. No statistically significant differences in ascorbic acid degradation rates were found to exist between the two low pH food systems, nor between the two low VBC food systems (Bonferroni t Test). The effect of fiber variability on the rate of ascorbic acid destruction was much more pronounced at low av's than at high aw's. Likewise, the effect of fiber on ascorbic acid stability was more profound at low temperatures than at high temperatures. The method by which fiber affected the ascorbic acid half life values may not have changed as a function of variable temperature and aw, but 55 became more noted as the influence of the dominating variables decreased. This theory is supported by the fact that at high aw's, differences in mobility as affected by fiber vari- ability would be less of a differentiating factor, due to the presence of capillary water. When capillary condensed water exist, differences in water mobility as affected by product composition would be diminished. At these higher aw's, all catalysts and reactants would be mobile, thus equilizing the reaction rates. The ability of water at low av's to exert vapor pressure but inability to act as a solvent for reactants and catalysts, may explain the larger fluctuations in half life values at the low aw, as a function of YBC. It is of interest to note that fiber variability at 0.40 aw, is more influencial in determining the ascorbic acid degradation rate, than at 0.10 or 0.65 aw, as indicated by the percent variation in ascorbic acid half- life values. However, at 0.10 the half-life variation in total days was greater. At 0.40 aw all ascorbic acid and reaction catalysts may be concentrated in solution, which could influence chelation of metal catalysts by organic acids (pectin fiber). Chelation of metal ions would promote ascorbic acid stability, whereas solubilixa- tion in the concentrated form would increase the degrada- tion of ascorbic acid. These opposing effects could be the cause for the more diverse ascorbic acid half-life values 56 at 0.40 av as a function of fiber. At 0.65 aw, it can be postulated that only a small amount of additional catalyst and/or reactant molecules are being drawn into solution and a dilution of reactants would occur. Thus a leveling off of the degradation rate would be anticapated. The fact that ascorbic acid degradation was less in the high HBO fiber systems (higher moisture content at a given ‘w)’ may further indicate that the dilution of ascorbic acid and reactants as a function of fiber variability is important in the determination of the ascorbic acid degradation rate. From these experimental data, it could be inferred that the high VBC fiber systems may produce a much more stable environment for ascorbic acid at a constant water content than a low VBC fiber system. Differences in ascorbic acid half life values, as affected by fiber variability, ranged from 27 days at favorable storage conditions (low temperature and low aw) to only 2 days at the unfavorable storage condition (high temperature and high av). whether these differences in ascorbic acid stability, as influenced by fiber variability, are sufficient for industrial application is questionable. VBC of a product, as it effects moisture content and nutrient stability, should not be underscored as a tool for the food industry, because significant economic gains can be made by selling a product with a higher moisture content which meets av specification. 57 Lee and Labuza (1975) proposed an increased mobility of reactants and catalysts was the major reason for increased ascorbic acid degradation, as a function of aw. Using NMR, it was shown that an increase in water content resulted in decreased viscosity, which increased the mobility of water in the system. Data in Tables 4 and 6 supports the theory that viscosity plays a major role in regulating the ascorbic acid degradation rate, since even at higher moisture contents, the high VBC fibers extended the ascorbic acid half life. As stated in the literature review, lipid oxidation is more prevalent at low av’s than at the higher av's. The free radical formation which occurs during lipid oxida- tion, could possibly have accelerated the ascorbic acid degradation rate at the lower aw's. However, as is evident from the data in Table 4, if lipid oxidation affected the rate of ascorbic acid destruction, it did not appear to be the predominate factor. As expected, an increase in a' within a single food system held at a constant temperature resulted in an increased moisture content (Table 6). Likewise, at a constant aw and fiber, an increase in temperature resulted in a decreased moisture content (Table 6). The major differences in total moisture content at a given a" did not noticeably effect the ascorbic acid degradation rate (Fiber Group Totals, Appendix A), indicating that a is a more V accurate method of measuring the effect of water on the 58 vitamin, than is total moisture content. However, as noted by the deviations of ascorbic acid half life values, aw is only one parameter in determining the nutrient half life in a product. Water binding capacity of the product is also of importance in predicting the ascorbic acid half life value. Av in the range measured (0.10-0.65 av) had a greater influence on the degradation rate of ascorbic acid than did fiber variability (Analysis of Variance), but less of an effect than temperature. An increase in av at a constant temperature and fiber resulted in a decreased ascorbic acid half life. This coincides with studies by Vojnovich (1970), Karel and Nickerson (1964), Jensen (1967) and Kirk (1977) which indicated that an increase in moisture or aw resulted in an increased ascorbic acid degradation rate. Lee and Labuza (1975) have noted that the mechanism involved in ascorbic acid degradation, did not change in the ascorbic acid degradation rate. Lee and Labuza's data are supported by this study which found no significant differ- ences in the E‘ for ascorbic acid destruction in the 0.10 - 0.65 aw range (Appendix B). Lee and Labuza (1975) speculated that with an increase in a' above a certain level, a dilution of the reactants may be occurring. This was suspected since above a critical moisture level, the rate increase was very small and almost constant. These findings are supported by data from this study (Table 4), which indicates that above 0.40 aw only a 59 slight increase in degradation rates were observed. An increase in storage temperature at a constant aw and fiber source was shown to have the greatest effect on the stability of ascorbic acid in the model food systems. Characteristic of most biological systems, a 10°C increase in storage temperature represented an approximate doubling of the rate of ascorbic acid destruction regardless of the fiber added to the model system (Table 4). As noted by the constant Ea’ (Appendix B) the mechanism by which temperature accelerated the degradation rate did not appear to change by the increase on storage temperature. INTERACTIONS BETWEEN SINGLE SOURCES OF VARIATION The Analysis of Variance also points out that all interactions between the single sources of variation had a significant effect on the ascorbic acid degradation rate. The temperature-aw interaction had a greater influence on ascorbic acid degradation than did the other interactions. The fiber-temperature and fiber-aw interactions, influenced the degradation rate to a lesser degree due to the minimal effect fiber variability had on ascorbic acid stability. All temperature-aw interactions were significantly different from one another except at 0.40 and 0.65 aw at 37°C. This may be due to the lowered oxygen level dissolved in the moisture of the product at these higher temperatures. The influence of the fiber-temperature interaction on the rate of ascorbic acid degradation was affected more 60 by temperature variability than by fiber variability (Appendix 0; Figure 9). Likewise, the effect of the fiber-aw interactions on ascorbic acid degradation was influenced more by the av than fiber variability (Appendix C; Figure 10, 11, 12). Although fiber had an effect on the degradation of the vitamin, as shown by the inter- actions of single variables, it was negligible when compared to effect of temperature and av. CONCLUSION Ascorbic acid stability was shown to fluctuate as a function of the induced variable conditions. Temperature and water activity variability, as affecting ascorbic acid stability has been well documented in previous studies (Lee and Labuza, 1975; Kirk et. al. (1977). This study has shown that water activity is not a perfect tool in analyzing ascorbic acid stability in a food product. Even at a given water activity, which is presently used as a measure of water availability and reactant mobility, differences in the half life values exist due to the presence of variable high molecular weight carbohydrates in the model systems. These differences in half life values would have been much more dramatic at a constant moisture content. Due to the significant differences between the fiber systems at a given water activity, one can justifiably state that water activity should not be the sole criteria upon which ascorbic acid stability is evaluated. Utilizing various sources of fiber to increase the shelf-life of ascorbic acid might be justified at low aw- low temperature storage conditions. However, this practice would be impractical at high aw's stored at temperatures greater than 20°C. A more important aspect of this research might be the stabilization of water activity by fiber addition in order to increase moisture content and profit margin. 61 1) 2) 3) 4) RESEARCH SUGGESTIONS NMR studies could be utilized to determine whether the high water binding capacity fiber systems actually did reduce water mobility. NMR could be used to adjust food systems to a given water mobility factor. The rate of ascorbic acid destruction should be measured as a function of water mobility within the various systems. A correlation between water binding capacity (NMR), moisture content, and aw, and their effect on the ascorbic acid degradation rate to provide a method with which to accurately predict ascorbic acid stability in a food product. Use of highly unsaturated fat in a model system, with and without antioxidants, to demonstrate the possible effects of lipid oxidation on ascorbic acid stability and the relative effectiveness of various antioxidants. 62 REFERENCES Acker, L. V. 1962. Enzymic reactions in foods of low moisture content. Adv. Food Res. 11: 263. Acker, L. V. 1969. Water activity and enzyme activity. Food Tech. 32: 1257. Chou, H. E. and Labuza, T. P. 1973. Sorption hysteresis and chemical reactivity: Lipid oxidation. J. Food Sci. 28: 316. Jensen, A. 1967. Tocopherol content of seaweed and seameal. 3: Influence of processing and storage on the content of tocopherol, carotenoids and ascorbic acid in seaweed meal. J. Sci. Food Agric. 39: 622. Karel, H. and Nickerson, J. T. 1964. Effect of relative humidity, air and vacuum on browning of dehydrated orange juice. Food Tech. 18: 1214. Karel, M. 1973. Recent research and development in the field of lowbmoisture and intermediate-moisture foods. CBC Critical Reviews in Food Technology February: 329. Karmas, E. and Chen, C. C. 1975. Relationship between water activity and water binding in high and intermediate moisture foods. J. Food Sci. 19: 806. Kirk, J. R. and Ting, N. 1975. Fluorometric assay for total vitamin C using continuous flow analysis. J. Food Sci. 19: 463. Kirk, J. 3., Dennison, D., Kokoczka, P., Heldman, D. and Singh, 3. 1977. Degradation of ascorbic acid in a dehydrated food system. Accepted for publication. J. Food Sci. 1977. Kuprianoff, J. 1958. Bound water foods. Fundamental As ects of the Deh dration of Food Stuffs. ociety of Chemical Industry, The Macmillian Co., New Iork. Labuza, T. P. 1968. Sorption phenomenon in foods. Food Tech. 23: 263. Labuza, T. P. and Rutman, M. 1968. The effect of surface active agents on sorption isotherms of a model food system. Can. J. Chem. Eng. 46: 364. 63 64 Labuza, T. P., Tannenbaum, S. R. and Karel, M. 1970. Water content and stability of low-moisture and intermediate-moisture foods. Food Tech. 24: 543. Labuza, T. P. 1971. Kinetics of lipid oxidation in foods. Critical Rev. Food Tech. 2: 355. Labuza, T. P., Cassil, S., and Sinskey, A. J. 1972. Stability of intermediate moisture foods. 2: Microbiology. J. Food Sci. 21: 160. E! e: Labuza, T. P. 1976. In Hater Relations of Foods. Duckworth, R. 8., ed, p. 155. Academic Press, London. Lee, S. H. and Labuza, T. P. 1975. Destruction of ascorbic acid as a function of water activity. J. Food Sci. 49: 370. W Leung, H. Ki, Steinberg, M. P., Wei, L. S. and Nelson, A. I. H 1976. Water binding of macromolecules determined by pulsed NMR. J. Food Sci. 41: 297. Miller, B. G. 1966. Simultaneous Statistical Inference. McGraw Hill, New York. Mulley, E. A., Stumbo, C. R. and Hunting, V. M. 1975. Kinetics of thiamine degradation by heat. J. Food Sci. 39: 985. National Academy of Sciences, 1974. Pr0posed fortification policy for cereal-grain products. Printing and Publishing Office, National Academy of Sciences, 2101 Constitution Avenue, N. W., Washington, D. C. Rockland, L. B. 1969. Water activity and storage stability. Food Tech. 2;: 1241. Salwin, H. 1959. Defining minimum moisture contents for dehydrated foods. Food Tech. 12: 549. Scott, Y. J. 1957. Water relations of food spoilage organisms. Adv. Food Res. 1: 83. Tannenbaum, S. R. 1976. Principles of Food Science. Part 1 Food Chemistry, p. 347. Marcel Dekker, Inc. New Iork. Tjoa, V. L., Hoffman, C. and Grecz, N. 1973. Abstr. Ann. Meet. Am. Soc. Microbiol. G 215, p. 62. Vojnovich, C. and Pfeifer, V. F. 1970. Stability of ascorbic acid in blends with wheat flour, CSM and infant cereals. Cereal Sci. Today 12: 317. APPENDIX A The Analysis of Variance was arrived by the following method. All half life values are derived from Table 4. Sources of Variation: Ax = fiber Bx = temperature Cx = water activity x indicates individual condition Fiber Group Half Life Totals and Half Lifg Means (days) Totals Mean (r=9) pH. Food System #1 (A1) 349 39 5.14 Food System #2 (A2) . 384 43 ' 4.87 Food System #3 (A3) 326 36 5.10 Food System #4 (A4) 425 47 3.75 Food System #5 (A5) 387 43 4.31 Food System #6 (A6) 377 41 4.90 Temperature Group Half Life Totals and Half Life Means Totals Mean (r218) Temperature B1 (20°C) 1314 73 Temperature B2 (30°C) 654 36 Temperature B3 (37°C) 280 14 65 66 Hater Activity Group Half Life Totals and Half Life Means Totals Means (r=18) water Activity 01 (.10 av) 1165 65 Water Activity C2 (.40 av) 605 34 Water Activity 03 (.65 av) 478 27 Fibep-Temperaturg Interaction Half Life Totals an Half Life M ans da s Fiber-Temperature Totals Means (r=3) (AB)1,1 205 68 (213),,2 102 34 (413),,3 42 14 (413),,1 224 75 (48),,2 112 37 (413),,3 48 16 (413)3’1 186 62 (43),,2 99 33 (AB)3,3 41 14 (1111),“1 249 83 “13),,2 121 40 (43),,3 55 18 (413).},1 229 76 (AB)5,2 109 36 1.3),,, 49 16 (313%,1 221 74 (1113),“2 111 37 (413)6’3 45 15 67 Fiberéwater Activity Interaction Half Life Totals and Half Life Means da s Fiber-Water Activity Totals Mean (r=3) (110)“1 174 58 (110),,2 94 31 (110)“3 81 27 (AC)2’1 205 68 (AC)2,2 103 34 (AC)2,3 76 25 (40),,1 173 58 (110),,2 80 27 (AC)3’3 73 24 (.10)“,1 219 73 (AC)4,2 122 41 (110),”3 84 28 (110)5’1 198 66 (AC)5’2 107 36 (sc)5’3 82 27 (AC)6,1 196 65 (1M6,2 99 33 (AC)6,3 82 27 68 Temperature-Vater Activity Interaction Group Totals and Means da s TemperatureAWater Activity Totals Mean (r=3) (13(3),,1 670 112 (80),,2 365 61 (130),,3 279 47 (13c)2’1 338 56 (130),,2 173 29 (BC)2'3 143 24 (BC)3,1 159 27 (3°)3,2 67 11 (130)3’3 56 9 Appendix B, ANALYSIS OF VARIANCE 54 54 2 2 ssyzjffl yi " E (23,1) /54] SSy = 144,126 - 93,583 = 50,543 54 C. F. (correction factor) = (E y1)2/54 = 93,583 5 2 88A = (fi1Ai) /9 - C.F. = 84890 6 - 93,583 = 645 S SS SS SS SS SS S AC BC 69 3 2 ()2 Bi) /18 - C.F. i=1 (1314)2 + (65412 + (280)2 - 93,583 30,457 3 2 i=1 (1165)2 +3605)2 + (478)2 — 93,583 1 _.__ _ _-__ _—__l.._l 14,847 A B 18 2 (z (AB)i /3) - 0.17. — SS - SS 181 II: J 'l’)_-L(1III_"1 2113§12,- 93,583 - 645 - 30,456 273 18 2 (z (Ac)1 /3) - 0.3. — SS 88 i=1 A ‘ c 2 6o - 93,583 — 645 - 14,847 3 245 9 2 (§;1 (Bc)i /6) - 0.9. - SSB - SSC 8213424 - 93,583 - 30,456 - 14,847 4,029 SSI - (SSA + SSB + SSC + SSAB + SSAC + SSBC) 50,543 - (645 + 30,456 + 14,847 + 273 + 245 + 4029) 48 2 Error SS 70 The sum of squares was used in the Analysis of Variance. The half life means were used to draw con- clusions using graphs and different statistical tests (Scheffe and Bonferroni t). Energy of Activation as Calculated by Arrhenius Equation Arrhenius Equation 1n k = In A — Ea if “w Fiber 1 Fiber 2 Fiber 3 Fiber 4 Fiber 5 Fiber 6 0.10 18.36 15.31 14.13 14.36 19.31 21.52 0.40 17.60 17.29 15.84 17.12 19.13 17.61 0.65 15.97 16.19 ' 15.25 15.82 18.49 17.39 Ea values in K cals/mole. Appendix C Fiber-Temperature Interaction Scheffe Test 0 Constant Tepperpture 20 C min. Sign. Diff. (yAB - 7A3 ) = 68.3 - 74.7 = -6.4 7.09 1,1 2,1 (yAB - yAB ) = 68.3 - 62.0 = 6.3 7.09 (yAB - ypn ) = 68.3 - 83.0 -14.7 7.09 71 Continued (;9B1,1- §335,1) = 68.3 - (5,31’1- 3,36’1) = 68.3 - (;932,1- 5933,1) = 74.7 - (3932,1- §;B4'1) - 74.7 - (;932,1- 3has,” a 74.7 - (Fng’1- 5136’1) = 74.7 - (§;B3,1- 5,34’1) = 62.0 — (5933,1- ;935,1) g 62.0 - (5933.1- §936,1) = 62.0 - (5:34’1- 5335’1) = 83.0 - (5,34,1- 3;36,1) = 83.0 - (5,35’1- 5,36’1) = 76.3 - 76.3 73.7 62.0 83.0 76.3 73.7 83.0 76.3 73.7 76.3 73.7 73.7 min. sign. diff. -8.0 1.0 -21.0 -14.3 -11e7 6.7 9.3 3.6 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 Fiber-Temperature Interaction Scheffe Test Constant Tepperature (5? -3" ) = AB1,2 AB2,2 5‘ ll £51 5'1. £5] 51. I U I V 1| £51 51. a?) 5'11 651 é' £1 .5 30°C 34.0 34.0 34.0 34.0 34.0 37.3 37.3 37.3 37.3 34.0 34.0 34.0 40.3 40.3 36.3 37.3 33.0 40.3 36.3 37.0 33.0 40.3 36.3 37.0 40.3 36.3 37.0 36.3 37.0 37.0 min. sign. diff. -3e3 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 73 Fiber-Temperature Interaction Scheffe Test Constant Fiber Food System #1 ('y' -3; ) = 68.3 - “31,1 AB1,2 G -'§ ) = 68.3 - “31,1 AB1,3 (gm "' _.AB ) = 34.0 '- 1,2 1,3 Food System #2 AB2,1 AB2,2 AB2,1 A132,3 AB2,2 “32,3 Food System #3 (5’ - §’ ) = 62.0 - AB3,1 AB3,2 A33,1 “33,3 ('3: - 37' ) = 33.0 - ”3,2 AB3,3 Food System #4 A34,1 M34,2 . 6 -§ ) = mm- “4,1 AB4,3 ('3' -§' ) = 40.3 - ”4,2 AB4,3 34.0 14.0 14.0 37.3 16.0 16.0 33.0 13.7 13.7 40.3 18.3 18.3 min. sign. diff. 34.3 54.3 20.0 37.4 58.7 21.3 29.0 48.3 19.3 42.7 64.7 22.0 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 Food System #5 Food System #6 (FAB -yAB (ya " yAB 15“,, - 57' ) “36,3 Fiber-Water Activity Scheffe Test 76.3 76.3 36.3 73.7 73.7 37.0 74 - 36.3 ‘16e3 -16e3 - 37.0 - 15.0 - 15.0 Interaction Constant Water Activity 13",, (5,, (3",, (5",, (3",, (5110 ('5 A02 58.0 58.0 58.0 58.0 58.0 68.3 68.3 0.10 68.3 57.7 73.0 - 66.0 "' 6503 - 57e7 73.0 40.0 60.0 20.0 36.7 58.7 22.0 -10.3 0.3 -15e0 -800 10.6 .407 3.34 3.34 3.34 3.34 3.34 3.34 7.09 7.09 7.09 7.09 7.09 7.09 7.09 .P‘r'“. “-.V lie: on I _ Continued (37 A02,1 (3 .402,1 (5 A03,1 yAc yAc yAc (7A0 ' yAc 3,1 (yAc ' yA0 3,1 (yAC ' yAc 4,1 (7A0 ' yAc (3 AC5’1 yAc 5,1 6,1 4,1 5,1 6,1 6,1 6,1 ) ) ) ) ) ) ) ) 68.3 68.3 57.7 57.7 57.7 73.0 73.0 66.0 75 73.0 65.3 73.0 66.0 65.3 66.0 65.3 65.3 -4.7 3.0 —15e3 -8.3 7.0 8.3 0.7 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 Fig. VI. 76 Fiber-Water Activity Interaction Scheffe Test Constant Water Agtivity (5,31 ’ 2- 6101 ,2- ($01 ’ 2- 6,01 ’ 2- 6‘01 , 2- (73102 , 2- (3,102 ’ 2- 6‘02 , 2- 6‘92 ’ 2- 6",“;3 ' 2- 6,63 ’ 2- 6103 ’2- (31%,; (5,04 , 2- ($105 ’ 2- §AC §AC "fie 3‘s: 5".c 5m: 5.1: yAC 5' AC 37.1: 5m: 37m: 5m: 3"Ac 3"110 2,2 3,2 5,2 6,2 3,2 4,2 5,2 6,2 4,2 5,2 6,2 5,2 6,2 ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) 31.3 31.3 31.3 31.3 31.3 34.3 34.3 34.3 34.3 26.7 26.7 26.7 40.7 40.7 35.7 0.40 34.3 26.7 40.7 35.7 33.0 26.7 40.7 35.7 33.0 40.7 35.7 33.0 35.7 33.0 33.0 min. sign. diff. 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 7.09 L) 77 Fiber-Water Activity Interactions Scheffe Test Constant Fiber Food System #1 min. sign. diff. (yAc - yAC ) = 58.0 - 31.3 = 26.7 3.34 1,1 1,2 (TLC - FAG ) = 58.0 - 27.0 s 31.0 3.34 (3' - 5' ) = 31.3 - 27.0 = 4.3 3.34 Food System #2 <§,C2’1- §4°2,2) = 68.3 - 34.3 = 34.0 3.34 (5,02,1- 5302.3) = 68.3 - 25.3 = 43.0 3.34 (3,02’2- 5,02’3) = 34.3 - 25.3 = 9.0 3.34 Food System #3 (§,c3,1- 7203’2) = 57.7 - 27.7 = 30.0 3.34 (§,03,1- $403,3) = 57.7 - 24.3 a 33.4 3.34 (5,03,2- 5403,3) = 27.7 - 24.3 = 3.4 3.34 Food System #4 (5,04’1- 5,04,2) = 73.0 - 40.7 = 32.3 3.34 (§,C4,1- 5,64,3) = 73.0 - 28.0 = 45.0 3.34 (7A0 - 5A0 ) = 40.7 - 28.0 = 12.7 3.34 4,2 4,3 78 Food System #5 (5' - 5' ) = 66.0 - 35.7 = 30.3 3.34 A05,1 A05,2 (5 -5 ) = 66.0-27.3 = 38.7 3.34 A05,1 AC5,3 (5' - 5' ) = 35.7 - 27.3 = 8.4 3.34 A05,2 A05,3 Food System #6 (5 -5 ) = 65.3-33.0 = 32.3 3.34 AC6,1 ‘°6,2 (5' - 5' ) = 65.3 - 27.3 s 36.0 3.34 AC6'1 AC6,3 — 33.0 - 27.3 = 5.7 3.34 (5' - 5' ) — Temperatureawater Activity Interaction Scheffe Test Constant Tempgrature 20°C min. sign. diff. (ync - ’80 ) = 111.7 - 60.8 = 46.0 2.363 1,1 1,2 (720 - 5,0 ) = 111.7 - 46.5 = 60.3 2.363 1,1 1,3 (5' - 5' ) = 60.8 - 46.5 = 14.3 2.363 BC1,2 BC1,3 Constant Temperaturg 30°C (5' - 5' ) = 56.3 - 28.8 a 27.5 2.363 zacz,1 302,2 (5' - 5' ) = 56.3 - 23.8 = 32.5 2.363 1302,1 302,3 (5'3c - 580 ) = 28.8 - 23.8 = 5.0 2.363 79 Constant Temperature 37°C min. sign. diff. (5' - 5' ) = 26.5 - 11.2 = 15.3 2.363 1303,1 BC3’2 (5' - 5' ) = 26.5 - 9.3 = 17.2 2.363 1303'1 BC3’3 (5 -5 ) = 11.2- 9.3 = 1.9 2.363 BC3,2 BC3,3 Temperature—Hater Activity Interaction Scheffe Test Constant Water Agtivity 0.10 (530 - Vac ) = 111.7 - 56.3 = 55.4 2.363 (5'3c - 550 ) = 111.7 - 26.5 = 85.2 2.363 1,1 3,1 (5' - 5' ) = 56.3 - 26.5 =‘ 30.3 2.363 BC2,1 BC3,1 Constant Water Agtivity 0.40 (5' - 5' ) = 60.8 - 28.8 = 32.0 2.363 BC1,2 BC2,2 . (5' - §’ ) = 60.8 - 11.2 = 49.6 2.363 BC1,2 BC3,2 (5' - 5' ) = 28.8 - 11.2 = 17.6 2.363 BC2,2 BC3,2 Constant Water Activity 0.65 (5' - 5' ) = 46.5 - 23.8 = 22.7 2.363 3C1,3 802,3 (5 -5 ) = 46.5- 9.3 = 35.2 2.363 BC1,3 BC3,3 (§BC " 5.30 ) = 23.8 - 9.3 '—" 14.5 2.363 2,3 3,3 MICHIGAN STQTE UNIV LIBRRRIES 293010888075