A STUDY OF CERTAIN ABSORPTION .* ; PROPERTIES RELATING TO THE LOSS 0F DIACE‘WL FROM SELECTED - - FOOD SYSTEMS ‘ Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY PAUL KELLY WHEELER. 1972 16 I T1301 |\\\\IIIIIII\IIIIIIIIIIIIIIIIIIIIII\IIIIIIIIII\\\\| * M ., .. I: .. I 3 1293 10524 7823 M"? 1"“ .' I. flnwcrsuy {A I IJII} I‘ll-I'll", i _ _ ..___._ ABSTRACT A STUDY OF CERTAIN ADSORPTION PROPERTIES RELATING TO THE LOSS OF DIACETYL FROM SELECTED FOOD SYSTEMS By Paul Kelly Wheeler This research study was pr0posed to investigate the effect of exposure in selected relative humidities upon the retention or loss of diacetyl from a number of food systems. A major segment of this study involved the analysis of moisture adsorptive characteristics of the food systems. Diacetyl was added to NFDM, casein, sucrose and cream prior to freeze-drying. Samples of the freeze-dried systems were subsequently exposed to various constant relative humidities for a period of time, or until the samples ceased to change in weight. The dehydrated food samples were then removed from the relative humidity environments and were analyzed for diacetyl to ascertain what losses, if any, occurred during moisture equilibration. Studies were also made to evaluate the capacities of the four products to adsorb water under these conditions. Data acquired from this study allow- the four food systems to be compared on the bases of: (a) rates of water adsorption at various relative humidities; (b) characteristics of their reSpective moisture adsorption ACKNOWLEDGEMENTS The author wishes to extend sincere gratitude to his major professor, Dr. C. M. Stine, for his friendship and guidance throughout the course of this graduate program. Professors R. C. Nicholas, A. L. Rippen and John W. Allen also receive thanks for serving on the guidance committee and for their review of this manuscript. Thanks are extended to Mr. Allen Kirleis for assistance in the preparation of samples. Acknowledgement is also given to the Department of Food Science and Human Nutrition for providing support through a graduate assistantship. The author's parents, Mr. and Mrs. Paul E. Wheeler receive a most deserved acknowledgement for their encouragement to pursue a career in Food Science, and for their many years of sacrifice. And lastly, to my wife, Susan, for her support and unselfish hours of assistance in the preparation of this manuscript, this thesis is respectfully dedicated. ii Paul Kelly Wheeler isotherms and conventional BET analyses; and (c) their respective diacetyl desorption curves. It was found that both casein and NFDM adsorbed a significant amount of moisture over the relative humidity range from 11% to 90%. Sucrose and freeze-dried cream, however, adsorbed substantially smaller amounts of moisture. All products exhibited characteristic sigmoid moisture adsorption isotherms, with monolayer moisture values corresponding to approximately 12% relative humidity and from 2% to 5% moisture (dry basis). A BET analysis of sucrose and cream exhibited values for l/YmC approaching zero and prohibited any reliable conclusions. Both NFDM and casein gave positive values for the monolayer moisture and the constant C. When the monolayer moisture values for NFDM and casein were compared on an equal surface area basis they were found to be equal, indicating that the protein fraction of NFDM plays a major role in determining the sorptive properties and stability of that product. All of the food systems tended to desorb the added diacetyl with increasing relative humidity. No positive relation between the adsorption of moisture and the desorption of diacetyl was indicated for any of the four food systems. The data would indicate that for 75% retention of added diacetyl in products such as NFDM, casein, or sucrose, storage conditions should be no greater than 10% relative humidity. However, in high fat products such as freeze-dried cream, a comparable retention of diacetyl may be obtained at much higher relative humidities, perhaps as high as h0%. These higher humidities would seldom be desired, however, as microbic growth and various degradative reactions would then be encouraged. A STUDY OF CERTAIN ABSORPTION PROPERTIES RELATING TO THE LOSS OF DIACETYL FROM SELECTED FOOD SYSTEMS By Paul Kelly Wheeler A 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 1972 TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . . . . . REVIEW OF LITERATURE . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . Sorption Theories and Models . . . . . . . . . The Kinetic Concept . . . . . . . . . . . The Potential Model . . . . . . . The Capillary Condensation Model . . . . . Secondary Sorption Theories and Models . . . . The Smith Equation . . . . . . . . . . . . The Henderson Equilibrium Equation . . . . The Young—Nelson Equation . . . . . . . . The Chung-Pfost Equation . . . . . . . . . Moisture Sorption in Foods . . . . . . . . . . Sorption Isotherm Techniques . . . . . . . Weighing bottle and humidity chamber Direct measurement . . . . . . . . . High vacuum measurement . . . . . . . Moisture Sorption and Stability . . . . . Sorptive Properties of Food Systems . . . Quantitative Determination of Diacetyl . . . . METHODOLOGY . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . Preparation of Samples . . . . . . . . . . . . Exposure to Selected Relative Humidities . . . Analytical Procedures . . . . . . . . . . . . . Moisture . . . . . . . . . . . . . . . . . Fat . . . . . . . . . . . . . . . . . . . Diacetyl . . . . . . . . . . . . . . . . . iii Page vii 10 11 12 13 13 15 16 17 19 19 20 21 21 23 25 28 28 28 29 32 32 32 32 RESULTS AND DISCUSSION . Examination of Recovery and Analysis of Diacetyl Preliminary Analyses Primary Data Adsorption of Water by Selected Food Systems Rates of Water Adsorption of Selected SUMMARY AND CONCLUSIONS Water Adsorption Isotherms of Selected BET Analysis of Water Adsorption Data Desorption of Diacetyl from Selected Food Systems Effect of Prolonged Exposure in Various Relative Humidities Food Systems . Food Systems . LITERATURE CITED . APPENDIX 0 iv Page 3h 3h 38 1T1 ll? 56 63 71 7h 76 78 86 Table 9. 10. ll. l2. 13. LIST OF TABLES Saturated salt solutions used to produce selected relative humidities . . . . . . . . . . . . . . . . . . . Recovery of diacetyl using the method of Westerfeld . . . The ability of the method of Westerfeld to detect small variances in the amount of diacetyl present in the sample (Diacetyl concentration in the samples of NFDM equals 7.7ug/g NFDM). . . . . . . . . . . . . . . . . . . Loss of diacetyl from selected food systems during freeZE-drying (FD) o g o a o o c Q o o o o o o o o o o 0 Moisture content (wb) and fat content (cream.only) of selected food systems after freeze-drying . . . . . . . . Dry basis moisture values for NFDM equilibrated in controlled relative humidity environments (25C) . . . . . Dry basis moisture values for casein equilibrated in controlled relative humidity environments (25C) . . . . . Dry basis moisture values for sucrose equilibrated in controlled relative humidity environments (250) . . . . . Dry basis moisture values for FD cream equilibrated in controlled relative humidity environments (25c) . . . . BET constants and derived values for NFDM and casein . . Loss of diacetyl from NFDM during prolonged exposure in selected relative humidities (25C) . . . . . . . . . . Moisture content (db) of NFDM during exposure in .selected relative humidities (250). . . . . . . . . . . . Moisture content (db) of casein during exposure in selected relative humidities (250) . . . . . . . . . . . Page 30 36 37 39 AO A2 A3 Ah AS 70 75 86 87 Table Page 1h. Moisture content (db) of cream during exposure in selected relative humidities (25C) . . . . . . . . . . . . 88 15. Moisture content (db) of sucrose during exposure in selected relative humidities (250) . . . . . . . . . . . . 89 16. Determination of BET plot for cream.. . . . . . . . . . . . 90 17. Determination of BET plot for sucrose . . . . . . . . . . . 91 18. Determination of BET plot for NFDM . . . . . . . . . . . . 92 19. Determination of BET plot for casein . . . . . . . . . . . 93 20. Desorption of diacetyl from NFDM (25c) . . . . . . . . . . 9h 21. Desorption of diacetyl from casein (25C) . . . . . . . . . 95 22. Desorption of diacetyl from sucrose (25C) . . . . . . . . . 96 23. Desorption of diacetyl from cream (250) . . . . . . . . . . 97 vi LIST OF FIGURES Figure 1. General moisture adsorption isotherm . . . . . . . . . . 2. Standard curve for diacetyl . . . . . . . . . . . . . . 3. Water adsorption by NFDM at various relative humidities(2SC).................... h. Water adsorption by casein at various relative hmidities (25C) 0 o o o o 0 Q o o o o o o o o o o o o o 5. Water sorption by sucrose at various relative hmdities (25C) 0 o o o c o o o o o o o o o o o o o o o 6. Water adsorption by freeze-dried cream at various relative humidities (250) . . . . . . . . . . . . . . . 7. Water adsorption by NFDM at various relative humidities (250) plotted with semilogarithmic coordinates . . . . . 8. Water adsorption by casein at various relative humidities (25C) plotted with semilogarithmic coordinates . . . . . 9. Water adsorption isotherms for casein, NFDM, cream and sucrose (25C) . . . . . . . . . . . . . . . . . . . 10. Water adsorption isotherm for NFDM (250) . . . . . . 11. Water adsorption isotherm. for casein (250) . . . . . 12. Water adsorption isotherm; for FD cream (250) . . . . . 13. Water adsorption isotherm; for sucrose (250) . . . . . . 1h. BET plot for sucrose (250) . . . . . . . . . . . . . . . 15. BET plot for FD cream (250) . . . . . . . . . . . . . . vii Page 18 35 AB A9 50 51 52 5h 57 58 59 60 61 6h 65 Figure Page 16. 17. 18. BET plot for NFDM (25C) 0 o o o o I o o o o o o o o o o o o 66 BET plot for casein (250) . . . . . . . . . . . . . . . . . 67 Loss of diacetyl from NFDM, casein, cream, and sucrose during equilibration at various RH . . . . . . . . . . . . 72 viii INTRODUCTION This research study was proposed to investigate the effect of exposure in selected relative humidities upon the retention or loss of diacetyl from a number of food systems. A.major segment of this study involved the analysis of the moisture adsorptive characteristics of the food systems. This relation between a food product and its moisture content is of great value in predicting the behavior of the food in storage. Moisture contents determine the minimum levels at which microorganisms may grow and the relative rates of such reactions as Maillard browning and oxidation. Recent research has further suggested that the properties of various hygrosc0pic materials in equilibrium with the atmosphere influence the retention or loss of volatile materials. An attempt was therefore made in this study to relate the water adsorptive capacities of the selected food systems to the loss, if any, of diacetyl. Diacetyl was chosen as it is found in numerous food products, including cheese, sour cream, butter, coffee, cocoa, beer, and it is often included in many artifical flavor compounds. REVIEW OF LITERATURE Introduction Recent research in food science and technology has included an increasing amount of attention to the sorptive properties of foods. Investigation in the area of sorption phenomena has been especially concerned with examining the storage stability of dehydrated foods, retention and release of volatile food flavor compounds, and deter- mination of various parameters related to the diffusion of water vapor in food systems. This study was undertaken to investigate the retention or loss of diacetyl in a number of food systems, as influenced by the adsorption of various amounts of moisture. Flavor is recognized as one of the main acceptability attributes of a food. Consequently an understanding of the mechanism of formation of food flavors from their precursors is important. However, another equally important factor is the subsequent release or retention of flavor compounds in foods. For example, in many food processing operations a high degree of flavor retention is essential for the product to be acceptable. Several studies have measured the loss of food volatiles during the concentration of some food products. Saravacos and Moyer (1968) measured the loss of aroma compounds during the vacuum drying of aqueous solutions of pectin and glucose. The retention of volatile compounds during spray drying (Menting and Hoogstad, 1967) and centrifugal film evaporation (Malkki and Veldstra, 1967) have also been investigated. Gray and Roberts (1970) investigated factors controlling the adsorption and retention of volatile food flavor compounds on selected substrates. Parameters such as the activation energy of desorption and heats of preferential sorption were determined. The relation between a food product and its moisture content is of great value in predicting the behavior of the food in storage. Labuza (1968) mentioned the importance of an adequate knowledge of sorptive prOperties for prediction of the stability of a dehydrated product. Similarly, numerous authors have studied the relation of sorption phe- nomena in foods to lipid oxidation, browning reactions, growth of micro- organisms, enzyme activities, and various physiochemical reactions in- fluencing the stability of food products. Several models which describe the adsorption-desorption phenomenon in foods are now presented. Those models most applicable to foods have been stressed and the relationship of water activity to the stability of foods is pointed out. Sorption Theories and Models In view of the extensiveness of sorption phenomena, an exhaustive coverage of the subject is not intended. Rather, the principal models or theories which have been advanced to account for the sorptive behavior of food materials, or from which such theories are directly or indirectly derivable, will be presented. Labuza (1968) in a review of sorption phenomena in foods, points out that the theoretical treatment of sorption can be classed into three modelistic frameworks, namely: (a) The Kinetic Concept of Langmuir, (b) Polanyi's Adsorption Potential Theory, (c) Zsigmondy's Capillary Condensation Theory. The Kinetic Concept Langmuir (1918) proposed the classical kinetic concept based on his belief that adsorption was induced by unbalanced chemical forces on the surface of crystals leading to a unimolecular layer of the adsorbate. Assuming that: (a) adsorbed molecules are localized, (b) colliding adsorbate molecules are reflected elastically, and (c) the heat of ad- sorption for every adsorbate molecule striking the bare surface of a solid adsorbent is the same and equal to the heat of vaporization, Langmuir equated the rate of evaporation to the rate of condensation at the surface under conditions of equilibrium. The resulting isotherm equation is usually expressed in the form: P/Vm = l/bV + P/Ym, where V is the volume adsorbed in cc/gram solid or gram/gram solid at vapor pressure P, Vm is the volume adsorbed at the monolayer point, and b is a constant dependent on both the heat of adsorption and the isotherm temperature. Labuza (1968) has pointed out that for most food materials, the Langmuir model does not hold for the following reasons: (a) heat of adsorption is, in fact, not a constant, but dependent on specific reaction sites, (b) adsorbed molecules undoubtedly interact, whereas Langmuir predicts no such interaction, (c) maximum adsorption is much larger than the theorized unimolecular layer. Freundlich (1926) in attempting to improve upon the Langmuir model, postulated that adsorption involves a series of monolayers on a surface composed of heterogeneous sites. Young and Crowell (1962) concluded that the models of Langmuir and Freundlich are best suited for predicting the adsorption of gases on metals, glass an other such solids, and are limited to exPlaining only the adsorption of a single layer of molecular moisture in food systems. In Spite of its inherent limitations, the Langmuir isotherm equation is perhaps the most important single equation in the field of adsorption, serving in many cases as the starting point in the deri- vation of other equations. Perhaps its greatest merit lies in the fact that it forms the basis of the more universally accepted BET equation of multimolecular adsorption proposed by Brunauer, Emmett, and Teller in 1938. Stitt (1958) has summarized the assumptions of the BET model as follows: (a) interactions of the adsorbate on the solid surface of an adsorbent are products of Van der Waal's forces, (b) more than one layer of sorbate molecules may be present on the surface, (c) the heat of sorption for all molecules in the first layer is constant and equal to the total heat of vaporization, (c) the energy of sorption for molecules in all other layers is equal to the heat of condensation for the pure sorbate. The most familiar form of the BET equation is derived from a thermodynamic basis (Adamson, 1963): V/Ym = [Ca/(l—a)] ' [l + (C—l)a], where V and Vm are respectively the volume adsorbed and the monolayer value, a is the water activity, and where C = k exP (QB/RT), QS being the heat of sorption. The thermodynamic basis has been extensively applied to the studies of sorption by foodstuffs and provides a flexible analytical tool for a better understanding of the mechanism of the sorptive process. Young and Crowell (1962) and Brunauer (19h5) have explained that adsorption of gaseous molecules serves to partially restore the unbalanced forces which normally exist at the surface plane. Adsorption decreases the free energy of a system, which is a measure of the work done to accomplish the adsorption process. The heat of sorption indicates the binding energy between adsorbed molecules, such as water, and the surface of the adsorbent (Chung and Pfost, 1967a). Free energy and heat of sorption of water vapor by proteins and high polymers have been evaluated by Dole and McLaren (19h6). Bull (l9hh) studied the sorption isotherms of a number of proteins and nylon, and calculated the free energy. The heat and free energy of adsorption of ‘water vapor by cellulose have been calculated by Babbitt (l9h2). The heats of sorption of water vapor on wheat flour, starch, and gluten were determined from experimental data by Bushuk and Winkler (1957). Becker and Sallans (1956) obtained similar values on wheat.. Heats of hydration have been reported.by Winkler and Geddes (1931) and by Schrenk gt_§l, (l9h9). Heats of vaporization have been determined for shelled corn by Rodriguez-Arias gt_al. (1963). A detailed study relating sorptive processes by cereal grains to free energy changes have been reported by Chung and Pfost (1967a). Sophisticated techniques for determining isosteric heats of adsorption employing frontal analysis chromatography were developed by Beebe e§_§g, (1966). Gray and Roberts (1970) employed a Stanton thermal balance and a flow microcalorimeter to measure activation energies and heats of preferential sorption in a pectinzethylamine model system. The BET equation is commonly rearranged to the following linear form: a/(l-a)V = 1/vmc + [a(C—1)/VmC], (Labuza, 1968; Chung and Pfost, 1967b; Karel and Nickerson, 196A; and Salwin, 1959). A plot of a/(l—a)V vs. a, gives a straight line. From the intercept and slope of this line, the monolayer value and the factor C can be calculated. Caution, however, must be observed in the BET plotting of sorption data. Many authors have exclaimed the BET equation is "unrealistic" (Pickett,.l9h5). Hill (1960) pointed out that BET's assumption of localized multilayers leads to erroneous entrOpy values. Cassel (19hh) has written that the BET model falsely predicts infinite adsorption at saturation. These two main errors have led to the general belief that the BET equation is applicable only to relative humidities of 50% or less (Becker and Sallans, 1956; Bushuk and Winkler, 1957; Dole and Foller, 1950; Hall and Rodriguez-Arias, 1970; Mellon.gt;§l:, 19h7; Pauling, l9h5; and.Smith, 19h7). However, data acquired from the 50% or less relative humidity range aresufficient to calculate a monolayer coverage of water. Further, the water surface area can be measured assuming the area of a'water molecule to be 10.6 32. The surface area So’ in (meter)2/gram.of solid, is determined by the following equation: 30 = vm ° 1/Mw - NO ~ Aw, where 7m is the monolayer value, Mw is the molecular weight of water, No is Avagadro's number, and Aw is the area of a water molecule (Labuza, 1968). Surface area values determined by water adsorption analyses usually bear little relation to those determined by nitrogen BET methods. Berlin §£_al, (1966), using low temperature (~19SC) adsorption data reported surface areas for vegetable, fruit, meat, and fish products less than one square meter per gram, Conventional all-glass volumetric adsorption apparatus was used to measure the nitrogen adsorption (Barr and Anhorn, 1959; Orr and Dalla Valle, 1959). Fox SE.§£: (1963) showed similar results for milk powders. In contrast, water surface areas are in the realm of 100-300 M2/g, several orders of magnitude higher. Stitt, in 1958, explained this difference, pointing out the unique ability of a water molecule to structurally alter long chain polymers, exposing interior sites for adsorption. Further, the water molecule is smaller and is able to enter smaller pores than the nitrogen molecule. Various modifications of the BET model have been developed to account for its obvious shortcomings. Hfittig's (19h8) modification is expressed in the following equation: v/V;m = (c ' aI/[l - C - a)(1 + a)]. A plot of this equation produces a straight line beyond 50% relative humidity. However, in view of the fact that the previously mentioned criticisms apply to Huttig's equation and indeed to the various modifications of the BET equation, the Opinion of Gorter and Frederikse in 19h9 that the kinetical BET theory gives a simple and valuable first picture of the phenomenon of adsorption, but it seems difficult to correct its obvious shortcomings without destroying the simplicity which perhaps constitutes its chief attraction, seems very appropriate. Halsey (1950) concludes that the BET isotherm should be regarded merely as convenient algebraic tools for locating "point B," which he feels marks the monolayer stage. Labuza (1968) has called the BET equation the most useful in predicting the monolayer value and the heat of adsorption, which are of concern to processing and storage. 10 The Potential Model At approximately the same time that Langmuir developed his monomolecular theory, Polanyi (191A, 1916, 1920) formulated an entirely different concept known as the potential theory. This concept recognizes the existence of multilayer adsorption, and considers that there is a potential field at the surface of a solid into which adsorbate molecules fall. The adsorbed layer thus resembles the atmosphere of a planet, being "compressed" most at the surface, and decreasing in density out— ward. Polanyi defined the adsorption potential at a point on the solid, as the work done in bringing an adsorbate molecule from the vapor phase to that point. It is fundamental to Polanyi's theory that the adsorption potential at any given point is characteristic of the adsorbent alone and is temperature independent, being unaffected by the presence of foreign molecules. Therefore, since the adsorbent—adsorbate complex remains a puzzle, functions derived under this model can only be considered approximate. The major disadvantage is that it cannot be used to predict the monolayer value which is of prime importance to the food field. Harkins and Jura in l9hh attempted to modify the Polanyi model by considering the distribution of surface forces to cause the adsorbed film to behave as a two-dimensional liquid. While this method has the advantage of not having to assume the area of a water molecule (as does BET), the basic equation does not hold for relative humidities above h0%. Considering this limitation, the original formalism of Polanyi and its subsequent modifications including those by Frenkel (19h6), ll Halsey (l9h8), Macmillon and Teller (1951), and Harkins and Jura (l9hh), must be regarded as somewhat primitive in specific applications. The Capillarngondensation Model It has long been rec0gnized that the vapor pressure over the meniscus of a liquid contained in a narrow capillary is lower than the vapor pressure of the free liquid at the same temperature. In other words, a vapor is liable to condense in a capillamy at a lower pressure than it would on a plane surface (McLaren and Bowen, 1951). This phenomenon is called capillary condensation. The quantitative relationship between the vapor pressure, P, over a liquid confined in a capillary, and the corresponding saturation vapor inva free space at the same temperature was given by Lord Kelvin (1871) in the form: P = .— a P/ 0 exp [( 20cos€VO)/ngT], where o is the surface tension of the liquid, V0 is the molecular volwme of the adsorbate, r is the radius of the capillary, RE is the universal gas constant, T is the temperature in degrees Kelvin, and 6 is.the contact angle. Zsigmondy (1911) and later Foster (1932) applied the capillary condensation theory to relationships between adsorption and prestructure in porous adsorbents. They argued that in porous structures, the same relation between vapor pressure and meniscus radius exists as in the case of ordinary capillaries. As the equilibrium relative humidity is increased in an adsorption experiment, condensation occurs in l2 successively larger pores. This rationalization leads to the calculation of pore size from adsorption data. Examples of these equations have been described by Adamson (1963), Labuza and Rutman (1967), and Roberts. (1967). Labuza (1968) has mentioned that knowledge of pore distribution has limited usefulness in the food field. Problems encountered include: (a) The Kelvin equation is not applicable when the pores are of molecular dimension. (b) There is difficulty in accurately determining 6, the contact angle. (c) The surface tension in the pores of foods is variable and not the same as that of pure water. Kuhn (196A) attempted to correct for the weak features of the Kelvin equation by combining the capillary theory with portions of other theories, establishing an empirical isotherm equation. To this date no practical application using his theory has been reported in the literature. Secondary_Sorption Theories and Models The models to be taken up in this section of the review are designated secondary in so far as they are subordinate to one or more of the primary models previously discussed. Although as early as 1882, Muller had prOposed an equation to predict the adsorption of water by textile fibers. His equation turned out to be quite valueless (Swan and Urquhart, 1927) because his arbit- rary assumptions-~(a) a linear relation between the amount adsorbed and the relative vapor pressure, and (b) zero adsorption at the boiling point of water-~were unsound. 13 The Smith Equation It was not until l9h7 that a partially successful treatment of water sorption by biological materials was formulated by Smith who recognized the existence of two principal classes of sorbed moisture: (a) bound moisture, Mb’ and (b) normally condensed moisture, MC. He assumed that the relation between Mb and a can be approximated by the Langmuir equation. Accepting the basic concepts of the BET theory for the framework of Mc’ he derived the following expression for MC: Mc = —V'ln(l—a), and summed Mb and Me to obtain: M = Mb + Mc = Mb = V'ln(l—a), where V' is the absorbed mass in a monolayer of condensed moisture, and M is the total adsorbed mass. While the Smith equation has been used to fit experimental isotherm data (Becker and Sallans, 1956), it has been shown by other authors (Strohman and Yoerger, 1967; Chung and Pfost, 1967b) to be basically empirical. The Henderson Equilibrium Equation Perhaps the best known and most widely used equation for predicting the equilibrium moisture content of food materials is the semiempirical equation of Henderson (1952). Henderson derived an 1h equation which can account for the temperature dependence of the experimental curve. His equation is of the form: (l-a) = exp (-KTMn), where T is absolute temperature in degrees Rankine, K and n are empirical constants, and M is the per cent equilibrium.moisture content, dry basis. Henderson evaluated the two constants, n and K, on the basis of two.experimental points arbitrarily chosen at about 20% and 75% relative humidity. He applied his equation to a series of 18 miscellaneous hygroscopic products. For most of these, satisfactory agreement was obtained between the derived theoretical curves and experimental observations. When appropriate values of the parameters K and n are available, the Henderson equation or its modifications by Day and Nelson (1965) and Thompson §t_al, (1967) have been found to fit isotherm data for cereal grains fairly well (Rodriguequrias, 1956). However, at extreme relative humidities, Henderson's experimental data deviated significantly from calculated theoretical curves for corn, sorghum and other products. Karel and Proctor (1955) applied Henderson's equation to the calculation of theoretiCal curves for shredded coconut, starch, rice, gelatin dessert, and flaked oats. These authors also observed significant deviations between the derived experimental curves and observed experimental curves, and suggested that more than two curves be employed for the evaluation of the constants. 15 Other authors (see for example, Pickler, 1956; Bakker-Arkema, 1961; and Day and Nelson, 1965) also found the Henderson equation inadequate for certain biological products. Rockland, in 1957, recognized that Henderson's equation could be converted to the more useful linear form: log log (l—a) = nlog M + K. The constants n and K could be more easily and precisely evaluated by graphical or least squares analysis. However, a straight line was seldom obtained; rather, the experimental points generally described three straight lines. This led Rockland to formulate his "local isotherm" concept which suggests that moisture sorption.may not necessarily be a uniform, continuous process, but a series of two or more independent processes. Rockland (1969) concluded that Henderson's initial success in fitting Observed data to theoretical isotherms was due to "fortuitous choices of hygroscopic systems and the unavailability of reliable data..."* The Younngelson Equation An effort to construct a theory of adsorption for biological materials to reflect their basic cellular nature was made by Young and Nelson (1967). These investigators considered the cell as the ultimate basis for adsorption and recognized the existence of three modes of adsorbed moisture: (a) untmolecularebound moisture of Langmuir, *The main deficiency of the Henderson equation is that it is totally based on thermodynamics, it is therefore not founded on a model and gives no information about the nature of the adsorbent or its surface. 16 (b). normally condensed or multi-layer moisture of Brunauer, Emmett,. and Teller, and (c) adsorbed moisture which results from a diffusion of moisture into the inner cell and which, due to the irreversibility of the diffusion process, is responsible for the occurrence of hysteresis. Although these authors developed a comparatively straightforward representation of hysteresis, the simplifications and reasoning leading . to their explicit eXpression for the adsorbed moisture effectively destroyed their model. If the ultimate cell of a biological product is taken as the basis of sorption, it appears logical that moisture transfer across the semipermeable cell wall can take place only as a result of osmosis. Ngoddy (1969) presented proof that moisture transfer across the membrane cannot be justified prior to saturation, due to the osmotic pressure which could not be exceeded or even balanced at lesser relative humidities. The Chung—Pfost Equation The general framework of the potential theory was utilized by Chung and Pfost (1967b) to develop an isotherm equation for cereal grains and their derivatives. The equation is of the form: 1n (a) = -A/RgR exp (-BM), where M is the specific adsorbed.mass, and the parameters A and B are respectively product and temperature dependent constants. Bradley (1936) earlier developed a related equation, also based on potential theory: ln (8.) = KQKl". 17 where K2 and K1 are constants, and a is the amount adsorbed at a specific pressure. Hoover and Mellon (1950) found Bradley's equation fitted experimental data for relatiVe humidities from 6% to 90% for a variety of high polymers. Moisture Sorptign_ig_Foods Since derived moisture sorption isotherms serve a number of useful functions, moisture sorption data have been obtained for a wide variety of food materials. They may be employed to: define limits for the dehydration of various foods (Makower and Dehority, 19h3); estimate moisture content changes under any given condition of temperature or humidity (Makower and Myers, l9h3); evaluate processing variables and to distinguish differences between grades of varieties of agricultural commodities (Houston and Kester, 195A; Karon and Adams, 19h9); aid in the selection of packaging materials (Houston and Kester, 195h; Makower and Dehority, 19h3); and, define moisture or humidity conditions under which product deterioration (Brockington p_I_:_ 22-.- , 19h9; Cuendet gt_ 31; , 195h; Henderson, 1952; Salwin, 1963) and microbial growth can be inhibited (Breese, 1955; Mossel, 1955; and Mosel and van Kuijk, 1955). A plot of the amount of water adsorbed as a function of the relative humidity of the vapor space around the material, describes a sorption isotherm (Labuza, 1968). Most foods exhibit a sigmoid isotherm described as Type II Isotherm according to the classification of Brunauer (l9h5). (See Figure 1.) Figure 1 shows a general moisture isotherm. The isotherm can be divided into several regions: Area I in Figure 1 represents the 18 .Eumruomw COquHOmUm ousumwos Hmuocoo .H muswwm as Cain: making 8. oo oo ow 1 1 J O N ---------..d wumfiou msoHum> um znmz ho soHDQMOmom nouns .m ouswwm is. as: o. m m e N mm s __ W J . . . \ 1: sh» J .J J rm $50 9 J - .. I: .82. .7 n J rm $3 J J J . :m Sow v .r J . 2 O N .LNBLNOO BHOLSIOW ('Q'P “Id 0 t0 A9 .Aommv mowuHoHEss o>wumfiop m30wum> om :Hommo so coHDQMOmom noun: .263 as: o. m o .q ouswwm xxx... xx «rm» 1m $hn rages Ix $nm rm $Om U BBDLSIOW 1N31N00 ("I 'P '96) 0 l0 50 .AUva mowufipwEDL o>wumaou mooHpm> um omouosm so ooHDQMOm poumz 2:2: as: O. m m w N d d 1 1 d is: J, a J 1m *2. J 1mgnm 0 o J It .800 ('Q'P‘%) lNBiNOO BHHLSIOW 51 .Aova mowuwowesn o>wumfiou msoHum> um Emouo oowuonouooum xo soHDQMOmom nouns .o madman 363 oz: O. o 0 1v N 11* _. )1 0 JJ 0 7 is...» J. m )\.\\ 1‘850 0 J :m—«vnb 0 JJ - :msnm . . 1¢$om J« J .LNBLNOO BHOLSIOW ('Q'P ' %) 52 83% RH 33% RH 90% RH J J L l O I 2 3 4 TIME (days) Figure 7. Water adsorption by NFDM at various relative humidities (25C) plotted with semilogarithmic coordinates. 53 moisture content (db) before exposure. The plots for NFDM in the relative humidities from 11% to 83% are.superimposed upon one another, ‘with the plot of the NFDM samples exposed in the 90% chamber being slightly removed from the other lines. From these graphs one could conclude that water adsorption by NFDM is independent of time and is solely a function of the product. As can'be seen from the graphs in Figure A, the rates of water adsorption by casein are similar to those of NFDM. Most of the gained moisture is adsorbed by the fourth day at all relative humidities studied. The samples of casein at 11%, 33% and 57% RH adsorbed very close to maximum moisture content in two days, with little additional moisture being adsorbed from the second through the tenth days. The samples at 75%, 83% and 90% RH gained the majority of the total adsorbed moisture by the second day; however, a significant amount of moisture was adsorbed, at a decreased rate, from the second to the fifth day of exposure. The constant weight pafise was attained by the samples of casein by the fifth day. A 10g M22228: , versus time plot for casein (Figure 8) shows no relation in the rates of water adsorption at various relative humidities. This would perhaps indicate an altering or rear- rangement of the casein structure upon adsorption of large amounts of water. The fact that the plots of the casein exposed to the 11%, 33% and 57% RH are very similar would suggest an alteration of the adsorption capacity.of casein during exposure at the higher relative humidities. As mentioned previously, sucrose tended to lose moisture during exposure in the relative humidities, from 11% to 75%. At higher humidities above 83% rapid adsorption of moisture was observed, 54 .5 +- 90%RH 83%RH - 75% RH .4 - 2 2° 3 - I I O C 2 2 CD 2 .2 - I J- 0 l 2 3 4 TIME (days) Figure 8. Water adsorption by casein at various relative humidities (25C) plotted with semilogarithmic coordinates. 55 leading to the formation of a saturated sucrose solution (Figure 5). Losses of moisture from the samples of sucrose occurred.at relatively the same rate.and to the same degree in.ll%, 33%. 57% and 75% relative humidities. Maximum moisture loss was achieved in two to three days in those humidities. A Slight gain in moisture by the sucrose samples was experienced in the 83% relative humidity.chamber. In the extreme relative humidity of 90%, a rapid linear adsorption of moisture was observed, with the smaller crystals of the sucrose samples going into solution after only a few days of exposure. As shown by the curves in Figure 6, the samples of freeze—dried cream.sdsorbed very little moisture during exposure at relative humidities less than 57%. In the atmosphere with 11% relative humidity the cream.samples lost a small amount (approximately 0.1% (db)) of moisture. During exposure in atmospheres of 75% RH, and greater, the cream samples rapidly gained moisture for two days. After two days of exposure in.75+% RH, however, the samples of cream lost about 6.5% of the adsorbed moisture. The loss was rapid, occurring in a twenty—four hour period. In all samples of cream exposed to the various relative humidities, constant weight was not achieved until after about the eighth day of exposure. From these data it could be concluded that freeze—dried cream.has a limited capacity to adsorb moisture from its environment. Further, at moderate to low relative humidities, freeze- dried cream adsorbs a small amount of moisture, only about 1% (db). The decrease in adsorbed moisture from samples held at the higher RH remains unexplained at this time. 56 water Adsorption Isotherms of Selected Food Systems ' Water adsorption isotherms for NFDM, casein, sucrose, and freeze—dried cream were obtained.by plotting per cent moisture (db), calculated after the samples reached constant weight, versus the per cent RH. TheSe isotherms are shown in Figures 10 through 13, and reapective tabular data are presented in the Appendix. As displayed by the isotherms in Figure 9, the four systems all exhibit characteristic sigmoid isotherms (Labuza, 1968). The isotherms for casein and NFDM are very similar, indicating that the casein fraction of NFDM plays a major role in determining the adsorptive . prOperties of that product. The isotherms for sucrose and cream indicate that these products tend to adsorb relatively little moisture over the range of relative humidities examined. Definite and extended plateaus are.seen in the isotherms for cream.snd sucrose. Had the sucrose samples been initially exposed in the relative humidities as anhydrous materials, it is believed that the isotherm would have commenced with a positive slope. It is the Opinion of many investigators (Salwin, 1959; Rockland, 1957; Stitt, 1958) that Area I (See Figure 1) or the first inflection point of the sigmoid isotherm represents the moisture content at which a product has greatest stability. This moisture content corresponds to the adsorption of a single monolayer of'moisture. The monolayer of moisture has been theorized to protect the product from harmful degradative reactions. At moisture levels above this value, destructive chemical, enzymatic and microbial activity reactions are accelerated. 57 00. .Aommv omouoom pow Emouo .zomz .coomso pom menonuOmw COHDaMOmom nous: 3S >._._Q_SDI m>_._.<._mm ow ON a 1J4 - omomosm . ’ ‘ 2t922... m>_._.<._mm 00 ON .c_ enemas d I o. O '0 BUOLSION iNBlNOO ('9 'P ‘96) 59 co. .Aummv awommo now EnoLDOmH co«uau0mom nous: 3£ >._.5.23I m>.._.<._wm 00 ON d 1 .ss magmas Q BHRLSION .LNBLNOO C) n ('9? ‘70) 60 00. .Aova Emouo on now Eponuomw coHuauomom soumz 3.1 . 5.922.. 92.53% om om .Ns spawns d 1! N. .LNBLNOO BERLSIOW N'P "/o) 61 .Aommv omouosm pom EuoLDOmH coHuaoOmom noum3 .ma osswwa GS >._.E_2:I m>.._.<4mm 00 00 ow 0N 1N BLNOO BERLSIOW ('9'? ”96) 62 At moisture levels below the monolayer moisture value, certain reactions such as oxidation are accelerated (Acker, 1969; Block, 1953; Cuendet g£_§l,, 195A) It would seem logical then that storage specifications for a particular product would tentatively be the same as those relative humidities corresponding to the monolayer moisture value. The monolayer moisture content would also be a good first target in dehydration when specific stability data are lacking (Salwin, 1963). Both NFDM and casein (Figures 10 and 11) have monolayer moisture values corresponding to about 12% relative humidity and 5% moisture (db). Therefore, these products should be dried to at least 6% moisture (db) and exposed to a maximum relative humidity of 12% during storage for maximum stability. Also of interest is Area 11 (See Figure 1) of the isotherm, or the portion between the first and second inflection points of the curve. In this area, hysteresis is most pronounced and the adsorbed water is highly mobile (Salwin, 1959). Area II of the NFDM and casein isotherms corresponds to those relative humidities between 12% and 57%. Freeze-dried cream (Figure 12) also exhibits a monolayer moisture value corresponding to about 12%. However, the reSpective moisture content is much lower than either NFDM or casein, being around A% (db). Very little water is adsorbed in Area 1183 indicated by the isotherm of freeze-dried cream suggesting that other factors, more important than the presence of moisture may influence the stability of the product. From the isotherm of sucrose (Figure 13) a monolayer moisture content of approximately 2% db can be observed. This correSponds to a 63 relative humidity of 10%. There was virtually no difference in the final moisture content for sucrose exposed in relative humidities from 11% to 75%. This would indicate.that at relative humidities less than 75%, the monolayer moisture value is approximated by sucrose and significant stability due to its limited adsorptive capacity is attained. At relative humidities greater than 75%, sucrose tends to rapidly adsorb moisture, approaching a saturated solution. BET Analysis of Water Adsorption Data The BET theory is most often applied when dealing with foodstuffs because it enables prediction of the monolayer moisture value and the determination of important thermodynamic parameters. From a plot of a/(l-a)m versus a, where a is the relative humidity and m is the moisture fraction, values for the following can be determined: (a) the monolayer moisture content, (b) the net binding energy of the monolayer, (c) the total binding energy, and (d) the surface area per gram of solid material. (Macmillon and Teller, 1951 and Labuza, 1968) The BET plots for NFDM, casein, cream, and sucrose are presented in Figures 1A through 17, with respective tabular data included in the Appendix. For relative humidities below 50%, the graph of a/(l-a)m versus a, usually produces a straight line. From the y-intercept and lepe of this line the monolayer coverage value can be calculatedeTmlthC 64 IO- ( a solid: I 9 water) _9_ (I-o)m A l 0 .20 .60 RELATIVE HUMIDITY Figure 14. BET plot for sucrose (25C). —-°-—— (g cancel a wow) (l-o)m 64 I0- A ’ l 0 .20 .60 RELATIVE HUMIDITY Figure 14. BET plot for sucrose (250). __q_ (9 solIds / 9 water) (I-o)m 40 DD 0 Figure 15. 65 I L .20 .60 RELATIVE HUMIDITY BET plot for FD cream (250). (g solids/q wafer) (I-a)m I5 Figure J6. 66 1 l .20 .40 _go RELATIVE HUMIDITY BET plot for NFDM (25C). (g solids lg water) a (I-a)m O) J} 67 ‘ l 1 J .20 .40 .60 RELATIVE HUMIDITY Figure 17. BET plot for casein (25C). 68 following equation: a/(l—a)m = l/VmC + [a(C-l)]/7mC, where Mn is the monolayer moisture value and C is a thermodynamic constant (Chung and Pfost,.1967b; Rockland, 1969). The constant C is equal to. C = k exp (QS/RT), where Q8 is the net binding energy of the monolayer moisture. (Labuza, 1968) The heat of adsorption, Q5, can then be broken down into its component parts thusly, Q3 = El - Hv (250). where E1 is the total binding energy and Hv is the heat of vaporization of water (Dole and McLaren, 19A6). In addition to the thermodynamic data, the surface area, So’ in (meter)2/gram of solid can be found by the following equation: . 3 '- V . l . N = o 10 V , So m /(MH20) o AH20‘ 3 5 x m whereFm is the calculated monolayer value, MH 0 is the molecular weight 2 of water,.NO is Avagadro's number, and AH 0 is the area of a water ' 2 ~20m2 (Labuza, 1968). molecule = 10.6 x 10 From examination of the BET plots for sucrose and cream, one can see that the linear portion intercepts the y—axis at the origin giving a value for 1/VmC of zero. While the linear portion appears to 69 intercept at the origin, the lack of graphical precision can only allow the conclusion that the value for l/VmC is, indeed, very small and approaching zero. It would seem unlikely that either the monolayer moisture or the heat of binding for either cream or sucrose would equal zero. As demonstrated previously, the monolayer moistures (db) for cream and sucrose were quite small. It is therefore quite possible that a low monolayer moisture value, perhaps coupled with a low heat of adsorption (which is what the low adsorptive capacities of sucrose and cream might indicate) would yield a value for 1/VhC approaching zero. Considering this, it is very difficult to subject the data for cream or sucrose to a BET treatment with any hape of reliability. Therefore, only the data for NFDM and casein will receive further BET analysis. As can be seen from the graphs in Figures 16 and 17 the linear portions of the BET plots for NFDM and casein intersect the y-axis above the origin, giving positive values for 1/VmC. The values for the BET constants as well as derived values for QS and E1, and So for NFDM and casein are presented in Table 10. As can be seen from the data in Table 10, the values for the monolayer moistures of NFDM and casein are rather similar. When comparing these values, it must be remembered that Vm is expressed on a per gram of solids basis, which is related to the surface areas of the two products. From the examination of the 80 column of Table 10, it can be observed that the surface area of the casein is L2 times that of 70 Table 10. BET constants and derived values for NFDM and casein. Product Vm C Qs El 2 So (g/g solid) (1n QS/RT) (Cal/mol) (Cal/mol) (m /g) NFDM 0.060 18.35 1,700 12,200 210 Casein 0.0T2 11.22 1,400 12,000 252 71 NFDM. Therefore, if the monolayer moisture values for NFDM and casein were recalculated on an equal surface area basis, they would be equal. Without further data as to the influence of the various other components of NFDM on its adsorptive capacity and stability, definite conclusions are difficult. It appears, however, that the casein fraction of NFDM plays a very major role in determining the stability of that product during exposure in various relative humidities. The very similar values of QS and E1 for NFDM and casein further support the importance of the casein fraction. Within the accuracy of this experiment, the values for the total moisture binding energy of the two products can virtually be regarded as equal. The very low adsorptive capacity of lactose, the other major component of NFDM, would indicate that it has little influence on the stability of NFDM. Desorption of Diacetyl from Selected Food Systems The losses of diacetyl from NFDM, casein, sucrose and cream are diaplayed graphically in Figure 18. Reapective tabular data are presented in the Appendix. Samples of the freeze-dried food systems were exposed in various relative humidities until they attained constant weight. The samples were then removed from the humidity chambers and the loss, if any, of diacetyl from the product was determined. As can be seen from the graphs in Figure 18, a curvilinear relation is exhibited by NFDM, casein, and sucrose. The loss of diacetyl from cream was linear with increasing relative humidity. IOO q CI 50 DIACETYL REMAINING (96) “D (I 72 CREAM CASEIN SUCROSE ' NFDM Y 1 J j l .20 .40 .60 .80 RELATIVE HUMIDITY Figure 18. Loss of diacetyl from NFDM, casein, cream and sucrose during equilibration at various RH. 73 The three curvilinear graphs are composed of a linear slope, where the loss of diacetyl from NFDM, casein, and sucrose was quite rapid with an increase in relative humidity; and a plateau, where the loss of diacetyl was rather constant with increasing humidities. In NFDM, casein, and sucrose, the plateau phase or attainment of maximum diacetyl loss, was reached around 35% relative humidity. Casein continued to lose small amounts of diacetyl with increasing relative humidities, while sucrose lost virtually all of the added diacetyl in relative humidities greater than 80% due to the formation of a saturated solution. Maximum loss of diacetyl in 90% relative humidity varied from 70% by the samples of freeze-dried cream to virtually 100% in the sucrose samples. One may question whether the ability of a food product to adsorb moisture is related to the desorption of diacetyl from that product. This supposition is supported by the data for cream, NFDM and casein. The samples of cream adsorbed little moisture and lost relatively little diacetyl. NFDM and casein adsorbed a substantial amount of moisture and experienced a substantial loss of diacetyl. The data for sucrose, however, would tend to refute this supposition as that product adsorbed little moisture and lost a considerable amount of the added diacetyl. These data would indicate that for 75% retention of added diacetyl in products such as NFDM, casein, or sucrose, storage conditions should be no greater than 10% relative humidity. However, in high fat products such as the samples of freeze—dried cream, a comparable retention of diacetyl can be obtained at much higher relative humidities, 7h perhaps as high as h0%. These high hwmidities, however, would seldom be desired as they are much greater than those indicated from monolayer moisture values. Effect of Prolonged Exposure in Various Relative Humidities As a supplementary experiment, four accurately weighed samples of NFDM were exposed in each of three selected relative humidities for a prolonged period of time. The samples were allowed to reach moisture equilibrium for a period of seven days. One sample from each relative humidity environment was analyzed for diacetyl at the end of l, 2, 3, and h weeks, following moisture equilibrium. The results of this investigation are included in Table 11. As can be seen from the data, prolonged exposure did not cause further significant loss of diacetyl from the samples of NFDM. 75 Table 11. Loss of diacetyl from NFDM during prolonged exposure in selected relative humidities (250). Time After Amount of Diacetyl (ppm) Moisture Equilibration 11% RH 57% RH 90% RH (daYS) 0 5.70 0.85 0.70 7 5.70 0.80 0.70 1h 5.60 0.80 0.70 21 5.60 0.75 0.70 28 5.60 0.75 0-70 SUMMARY AND CONCLUSIONS From this study it has been found that the chemical nature of a food system greatly determines its particular water adsorptive capacities. The results would further indicate.that the protein fraction of a composite food system such as NFDM greatly influences the ability of NFDM to adsorb water while the influence of the carbohydrate or lipid portion is relatively slight. In products which were largely protein, carbohydrate, or fat in nature, as well as in a composite food, it was found that moisture equilibrium in any relative humidity was established within seven days. In all food systems examined, a monomolecular layer of moisture was adsorbed at approximately 12% equilibrium relative humidity. This moisture monolayer corresponded toia total product moisture content of ca. 6%. These data would then indicate the approximate storage relative humidity and product moisture content for maximum stability. In all food systems examined except sucrose, rates of water adsorption were generally very rapid in relative humidities less than 57%, with moisture equilibrium in the higher humidities taking several days longer to be established. Sucrose tended to lose moisture in humidities less than 83% and at higher humidities the crystals of sucrose tended to dissolve and eventually, to form a saturated solution. It was found that with NFDM the rates of water adsorption were a function 76 77 of a product and not related.to time. This was not the case, however, with the other food systems examined, as the rates of adsorptions generally differed within each relative humidity. From a BET analysis of the selected food systems, it appears that both freeze—dried cream and sucrose have very high heats of sorption and low values for the monolayer moisture content. The monolayer moisture values for casein and NFDM, calculated on an equal surface area baSis, were found to be equal, further indicating that casein largely determines the adsorptive capacity, and, perhaps, the storage stability of NFDM. No relation was found between the water adsorptive capacity of a food system and the ability of that product to bind the flavor compound diacetyl. The data would indicate that, in products high in protein or carbohydrates, the relative humidity during storage ought to be maintained at less than 10% RH for at least 75% retention of diacetyl. In products high in fat, however, the relative humidity may reach as high as hO% RH with only slight loss of diacetyl. LITERATURE CITED LITERATURE CITED LITERATURE CITED Acker, L. 1969. Water activity and enzyme activity. Food Technol. 23:1257. Adamson, A. W. 1963. Physical chemistry of surfaces. Interscience Publishing Company, New York. Babbitt, J. D. 19h2. On the adsorption of water vapor by cellulose. Canadian J. Res. 20(9):lh3-l72. Bakker—Arkema, F. W. 1961. Desorption isotherms for alfalfa of four growth stages in the temperature range of h0° to 120°F. Unpublished M.S. Thesis, Michigan State University, East Lansing, Mi. Barnicoat, C. R. 1935. The determination of diacetyl and acetyl methyl carbinol. Analyst. 60:653—662. Barr, W. E. and V. J. Anhorn. 1959. Scientific and industrial glass blowing and laboratory techniques. Instruments Publishing Company, Pittsburgh. Barritt, M. M. 1936. The intensification of the Voges-Proskauer reaction by the addition of a-naphthol. J. Path. Bact. h2:hh1-h53. Becker, H. A. and H. R. A. Sallans. 1956. A study of desorption isotherms of wheat at 25°C and 50°C. Cereal Chem. 33:79. Beebe, R. A., P. L. Evans, T. C. W. Kleinsteuber, and L. W. Richards. 1966. Adsorption isotherms and heats of adsorption by frontal analysis chromatography. J. Phys. Chem. 70(h):1009. Berlin, E., P. G. Kliman and M. J. Pallansch. 1966. Surface areas and densities of freeze-dried foods. J. Agr. Food Chem. 1h(l):15-l7. Block, S. S. 1953. Humidity requirements for mold growth. Applied Microbiol. 1:287. Bradley, R. S. 1936. Polymolecular adsorbed films. I. The adsorption of argon on salt crystals at low temperatures and the determination of surface fields. J. Chem. Soc. lh67-1h7h. 78 79 Breese, M. H. 1955. Hysteresis in the hygroscopic equilibria of .rough rice at 25°C. Cereal Chem. 32zh81-h87. Brockington, S. F., H. C. Dorin and H. K. Howerton. 19h9. Hygroscopic equilibria of whole kernel corn. Cereal Chem. 26:166-173. Brunauer, Stephen. 19h5. Adsorption of gases and vapors. Princeton University Press, Princeton, N.J. Brunauer, S., P. H. Emmett and E. Teller. 1938. Adsorption of gases in multimolecular layers. J. Am. Chem. Soc. 60:309-319. Bull, H. B. l9hh. Adsorption of water vapor by proteins. J. Am. Chem. SOC. 66:1h99-1507. Bushuk, W. and C. A. Winkler. 1957. Sorption of water vapor on wheat flour, starch, and gluten. Cereal Chem. 3h(2):73—86. Cassell, H. M. l9hh. Cluster formation and phase transitions in the adsorbed state. J. Phys. Chem. h8zl95. Chung, D. S. and H. B. Pfost. 1967a. Adsorption and desorption of water vapor by cereal grains and their products. I. Heat and free energy changes of adsorption and desorption. Trans. Am. Soc. Agr. Engrs. 5h9—555. Chung, D. S. and H. B. Pfost. 1967b. Adsorption and desorption of water vapor by cereal grains and their products. II. Development of the general isotherm equation. Trans. Am. Soc. Agr. Engrs. 552-555. Chung, D. S. and H. B. Pfost. 1967c. Adsorption and desorption of water vapor by cereal grains and their products. III. ‘A hypothesis for explaining the hysteresis effect. Trans. Am. Soc. Agr. Engrs. 556—557. Cuendet, L. S., E. Larson, C. G. Norris and W. F. Geddes. l95h. The influence of moisture content and other factors on the stability of wheat flours at 37.8°C. Cereal Chem. 31:362—389. Day, D. L. and G. L. Nelson. 1965. Desorption isotherms for wheat. Trans. Am. Soc. Agr. Engrs. 82(2):2h3. Dole, M. and I. L. Foller. 1950. Water sorption by synthetic high polymers. J. Am. Chem. Soc. 72:h1h-h19. Dole, M. and A. D. McLaren. 19h7. The free energy, heat and entrophy of sorption of water vapor by proteins and high polymers. J. Am. Chem. Soc. 69:651-657. 80 Foster, A. G. 1932. The sorption of condensable vapors by porous solids. I. The applicability of the capillary theory. (Translated) Faraday Society 28:6h5. Fox, K. K., V. H. Holsiniger, M. K. Harper, N. Howard, L. S. Pryor and M. J. Pallansch. 1963. Measurement of the surface areas of milk powders by a permeability procedure. Food Technol. 17:127. Frenkel, Y. I. 19h6. Kinetic theory of liquids. Clarendon Press, Oxford. Freundlich, H. 1926. Colloid and capillary chemistry. Methuen and Cwmmw,LmflmL Gorter, C. J. and H. D. R. Frederikse. 19h9. A few remarks on physical adsorption. Physica. 15:891. Gray, J. I. and D. G. Roberts. 1970. Retention and release of volatile food flavor compounds. J. Food Technol. 5:231-239. Hall, C. W. and J. H. Rodriguez-Arias. 1970. Equilibrium moisture content of shelled corn. Agr. Eng. 36:h66-h70. Halsey, G. D. 19h8. Physical adsorption on non-uniform surfaces. J. Chem. Phys. 16:931. Halsey, G. D. 1950. The role of heterogeneity in adsorption and catalysis. Faraday Society, 8:5h. Harden, A. and D. J. Norris. 1911. J. Physiol. h2:332. Harkins, W. C. and G. Jura. l9hh. Surfaces of solids. XIII. A vapor adsorption model for the determination of area of solids without assumptions of a molecular area and areas occupied by nitrogen and other molecules on the surface of a solid. J. Am. Chem. Soc. 66:1366. Hart, J. R. 196A. Hysteresis effects in mixtures of wheats taken from the same sample but having different moisture contents. Cereal Chem. hl:3h0—350. Henderson, S. M. 1952. A basic concept of equilibrium moisture. Agr. Eng. 33:2h-29. Hill, T. C. 1960. Introduction to statistical thermodynamics. Addison-Wesley Publishing Company, London. Hoover, J. R. and E. F. Mellon. 1950. Application of polarization theory to sorption of water vapor by high polymers. J. Am. Chem. Soc. 72:195. 81 Houston, D. A. and E. B. Kester.- l95h. Hygroscopic equilibria of whole-grain edible forms of rice. Food Technol. 8:302—30h. Hubbard, J. E., F. R. Earle and E. R. Senti. 1957. MOisture relations in wheat and corn. Cereal Chem. 3hzh22-h33. Hfittig, G. F. 19h8. Eur austertung der adsorptions isotherm. (Translated) Montash Chemistry 78:177. Issenberg, P., G. Greenstein and M. Boskovfc. 1968. Adsorption of volatile organic compounds in dehydrated food systems. I. Measurement of sorption isotherms at low water activities. J. Food Sci. 33(6):621-625. Karel, M., Y. Aikawa and B. E. Proctor. 1955. New approach to humidity equilibria data. Modern Packag. 29(2):153-156. Karel, M. and J. T. R. Nickerson. 196k. Effects of relative humidity, air and vacuum on browning of dehydrated orange juice. Food Technol. 18(8):th—108. Karon, M. L. and M. E. Adams. 19h9. Hygroscopic equilibrium of rice and rice fractions. Cereal Chem. 26:1-12. Klotz, I. M. and R. E. Heiney. 1957. Changes in protein topography upon oxygenation. Proc. Nat. Acad. Sci. h3z717. Kuhn, Istvan. 196M. A new theoretical analysis of adsorption phenomena. Introductory part. The characteristic expression of the main regular types of adsorption isotherms by a single simple equation. J. Colloid Sci. 19:685-698. Labuza, T. P. 1968. Sorption phenomena in foods. Food Technol. 22(3):ls-2u. Labuza, T. P. and M. Rutman. 1967. The effect of surface active agents on sorption isotherms of model systems. 17th Ann. Canadian Chem. Eng. Conf. Oct. 18, 1967. Langmuir, Irving. 1918. The adsorption of gases on plane surfaces of glass, mica and platinum. J. Am. Chem. Soc. h0:1361—1373. Macmillon, W. G. and E. Teller, 1951. The assumptions of the BET theory. J. Phys. Chem. 55:17. Makower, B. and G. L. Dehority. 19h3. Equilibrium moisture content of dehydrated vegetables. Ind. Eng. Chem. 35:193-197. Makower, B. and W. B. Dye. 1956. Equilibrium moisture content and crystallization of amorphous sucrose and glucose. J. Agr. and Food Chem. h(l):72—77. 82 Makower, B. and S. Myers. l9h3. A new method for the determination of moisture in dehydrated vegetables. Proc. Insti. Food Technologists 156. Malamud, H.,R. Geisman and S. Lowell. 1967. Adsorption isotherms of xylene isomers on zinc oxide by gas chromatography. Analy. Chem. 39:1h68-1h70. Malkki, Y. and J. Veldstra. 1967. Flavor retention and heat transfer during concentration of liquids in a centrifugal film evaporation. Food Technol. 21:1179-118h. Maloney, J. F., T. P. Labuza, D. H. Wallace and M. Karel. 1966. Autoxidation of methyl linoleate in a freeze-dried model system. J. Food Sci. 31:878. Martinez, F. and T. P. Labuza. 1968. Rate of deterioration of freeze— dried salmon as a function of relative humidity. J. Food Sci. 33:2hl. McBain, J. W. and A. M. Bakr. 1926. Sorption of gases by solids. J. Am. Chem. Soc. h8:690. McLaren, A. D. and J. W. Rowen. 1951. Sorption of water vapor by proteins and polymers. J. Polymer Sci. 7:289-299. Mellon, E. F., A. H. Korn and S. R. Hoover. 19h6. Water adsorption of proteins. I. The effect of free amino groups in casein. J. Am. Chem. Soc. 69:827—831. Mellon, E. F., A. H. Korn and S. R. Hoover, 19h7 a. Water adsorption of proteins. II. Lack of dependence of hysteresis in casein on free amino groups. J. Am. Chem. Soc. 70:11hh-11h6. Mellon, E. F., A. H. Korn and s. R. Hoover, lQth. Water adsorption of proteins. III. Contribution of the peptide group. J. Am. Chem. Soc. 70:30h0-3ohh. Nellon, E. F. A. H. Korn and S. R. Hoover. 19h8. Water adsorption of proteins. IV. Effect of physical structure. J. Am. Chem. Soc. 71:2761-276h. Mellon, E. F., A. H. Korn, E. L. Kokes and S. R. Hoover. 1950. Water adsorption of proteins. VI. Effect of guanidino groups in casein. J. Am. Chem. Soc. 73:1870-1871. Menting, L. C. and B. Hoogstad. 1967. Volatiles retention during the drying of aqueous carbohydrate solutions. J. Food Sci. 32:87—90. 83 Mohr, W. and J. wellm. 1937. Kritsche Ubersicht uber die . Bestimmungsmethoden des Diacetyls als Aromatrager der Butter und Vorschlag zur Schaffung einer Einheitsmethode. Wissenschaftliche Berichte des XI Milchwirtschaftlichen Weltkongresses, Berlin. Published by Molkereizeitung, Hildesheim. 2:98-10h. Mossel, D. A. A. 1955. The physiology of the microbial spoilage of foods. J. Appl. Bacteriology 18:232-268. Mossel, D. A. A. and H. J. L. van KuiJk. 1955. A new and simple technique for the direct determination of the equilibrium relative humidity of foods. Food Research eozhls-h23. Muller, H. 1882. On the relation between moisture present in textile fabrics and that in the atmosphere. J. Soc. Chem. Industries 1:356. Ngoddy, P. 0. 1969. A generalized theory of sorption phenomena in biological materials. Ph.D. Thesis, Michigan State University, East Lansing , Mi . Nickerson, T. A. and R. M. Dolby. 1971. Adsorption of diacetyl by lactose and other sugars. J. Dairy Sci. 5h(8):1212-121h. O'Meara, R. A. Q. 1931. J. Path. and Bact. 3h(h01). Orr, C. and J. M. Dalla Valle. 1959. Fine particle measurement. Macmillan, New York. Pauling, L. 19h5. The adsorption of water by proteins. J. Am. Chem. Soc. 67:555. Pickett, G. l9h5. Modification of the BET theory of multimolecular adsorption. J. Am. Chem. Soc. 67:1958. Pickler, H. J. 1956. Sorption isotherms of wheat and grape seed. Landtechnische Forschung 2:h7. Pien, J., J. Baisse and R. Martin. 1936. Contribution a 1'etude du diacetyle. Lait. 16:119-138; 2h3-255. Polanyi, M. 191A. Verhandl Deutsch. Physic. Ges. (Translated) 16:1012. Polanyi, M. 1916. Verhandl Deutsch. Physic. Ges. (Translated) 18:55. Polanyi, M. 1920. Verhandl Deutsch. z. Elektro (Translated) 26:370. 8h Prill, E. A. and B. W. Hammer.. 1938. A colorimetric method for the microdetermination of diacetyl. 385-395. Journal Paper J5h8 of the Iowa Agricultural Experiment Station, Ames, Iowa. Project No. 127. Rao, K. S. l9h1. Hysteresis in sorption. VI. Disappearance of the hysteresis loop. J. Phys. Chem. h5:53l-539. Roberts, B. F. 1967. Procedure for estimating pore volume and area distribution from sorption isotherms. J. Colloid and Interface Science 23:266. Rockland, L. B. 1957. A new treatment of hygroscopic equilibria: application to walnuts (Juglans regia) and other foods. Food Research 22:6oh-628. Rockland, L. B. 1960. Saturated salt solutions for static control of relative humidity between 5° and h0°C. Anal. Chem. 32(10):1375-1376. Rockland, L. B. 1969. Water activity and storage stability. Food Technol. 23(10):11—21. Rodriguez-Arias, J. H. 1956. Desorption isotherms and drying rates of shelled corn in the temperature range of hO-lh0°F. Ph.D. Thesis, Michigan State University, East Lansing, Mi. Rodriguez—Arias, J. H., C. W. Hall and F. W. Bakker—Arkema. 1963. Heat of vaporization for shelled corn. Cereal Chem. 38:676-683. Ruehe, H. A. and W. J. Corbett. 1937. Volumetric method for the determination of diacetyl. J. Dairy Sci. 20:h56-h57. Salwin, H. 1959. Defining minimum moisture contents for dehydrated foods. Food Technol. 13:59h-595. Salwin, H. 1963. Moisture levels required for stability in dehydrated foods. Food Technol. 17(9):3h-h1. Saravacos, G. D. and J. C. Moyer. 1968. Volatility of some flavor ‘compounds during freeze-drying of foods. Food Technol. 22:623-627. Schrenk, W. G., A. C. Andrews and H. H. King. 19h9. Heat of hydration of certain flours and gluten. Cereal Chem. 26:51—59. Scott, W. J. 1957. Water relations of food Spoilage microorganisms. Advan. Food Res. 7:83. Sharp, J. G. 1962. Non enzymatic browning deterioration in dehydrated meat., Recent advances in food science. Editors, Hawthorne and Leitch, Butterworth, London. 85 Smith, S. E. 19h7. The sorption of water vapor by high polymers. Stadtman, Earl R. 19h8. Non-enzymatic browning in fruit products. Advan. Food Res. 1:325. Stitt, F. 1958. Moisture equilibrium and the determination of water content of dehydrated foods. Society of Chemical Industry, London. Stokes, R. H. and R. A. Robinson. 19h9. Industrial Engineering chemistry. hlz20l3. Strohman, R. D. and R. R. Yoerger. 1967. A new equilibrium moisture content equation. Trans. Am. Soc. Agr. Eng. l9(5):675. Swan, E. and A. R. Urquhart, 1927. Adsorption equations-~a review of literature. J. Phys. Chem. 31:251. Taylor, A. A. 1961. Determination of moisture equilibrium in dehydrated foods. Food Technol. 15:536-5h0. Testoni, G. and W. Ciusa. 1931. Dosamento del diacetile nel burrl. Ann. Chim. applicata 21:1h7-150. Thompson, T. L., R. M. Peart and G. H. Foster. 1967. Mathematical simulation of corn drying—-a new model. Canadian Am. Soc. Agr. Eng. Conf., No. 67, 313. Tomson, (Lord) Kelvin W. 1871. On the equilibrium of vapor at a curved surface of a liquid. Phil. Maga. h2(282):hh8. Westerfeld, W. W. 19h5. A colorimetric determination of blood acetain. J. Biol. Chem. 16lzh95. Wink, W. A. 19h6. Equilibrium relative humidities above saturated salt solutions. Ind. Eng. Chem. 18:251. Wink, W. A. and G. R. Sears. 1950. Equilibrium relative humidities above saturated salt solutions. TAPPI (Tech. Assoc. Pulp and Paper Inst) 33(9):96a~99a. Winkler, C. A. and W. F. Geddes. 1931. The heat of hydration of flour and certain starches including wheat, rice and potato. Cereal Chem. 8:h55-h75. Young, D. M. and A. D. Crowell. 1962. Physical adsorption of gases Butterworth, Washington, D.C. Young, J. H. and G. L. Nelson. 1967. Research of hysteresis between sorption and desorption isotherms of wheat. Trans. Am. Soc. Agr. Eng. (10(2):260. Zsigmondy, R. 1911. Uber die Struktur des Gels der Kieselsaure. Z. Anorg. Chem. 71:356. APPENDIX 86 Table 12. Moisture content (db) of NFDM during exposure in selected relative humidities (250). Time Relative (days) Humidity 1 5 7 10 (z) 11 h.27 h.5h h.68 A.68 33 n.28 8.05 8.07 8.07 57 b.27 9.83 10.01 10.01 75 h.28 16.13 15.82 15.88 83 h.27 18.93 18.69 18.97 90 h.28 30.18 30.01 30.65 87 Moisture content (db) of casein during exposure in selected relative humidities (25C). Table 13. Time Relative (days) Humidity 1 2 6 7 8 (%) 11 3.8h A.68 A.68 A.68 A.68 4.68 33 3.8M 9.07 9.09 9.09 9.10 9.10 57 3.8h 11.83 11.90 11.90 11.90 11.98 75 3.8h 15.63 19.99 19.95 19.96 19.98 83 3.8h 21.65 22.72 23.07 23.21 23.21 90 3.8h 2h.19 28.95 29.83 29.56 29.56 88 Table 1h. Moisture content (db) of cream during exposure in selected relative humidities (25C). Time Relative (days) Humidity 1 2 3 8 10 (¢) 11 3.9h 3.81 3.82 3.86 3.87 3.87 33 3.95 h.22 h.30 h.29 h.33 h.33 57 3.95 h.1h h.1h h.h1 h.61 A.60 75 3.95 6.57 5.90 5.75 5.8)4 5.80 83 3.95 7.37 7.01 6.70 6.73 6.73 90 3.95 8.61 8.12 8.3]. 8.214 8.26 89 Table 15. Moisture content (db) of sucrose during exposure in selected relative humidities (25C). Time Relative (daYS) Humidity 1 2 3 5 7 (%) 11 3.50 2.07 2.07 2.07 2.08 33 3.h9 2.26 2.26 2.2h 2.2h 57 3.50 2.2h 2.26 2.26 2.26 75 3.50 2.27 2.30 2.30 2.30 83 3.50 3.91 3.92 3.92 3.92 90 3.50 6.25 7.29 9.86 ——— 90 Table 16. Determination of BET plot for cream. Relative Humidity Moisture Content .CEEERE (1) m(db) (g solids/g water) 11 0.236 0.52 33 0.255 1.93 57 0.271 b.89 75 0.3h6 8.67 83 0.396 12.33 90 0.h87 18.h8 91 Table 17. Determination of BET plot for sucrose. a Relative Humidity Moisture Content 71:275. (z) m(db) (g solid/g water) 11 2.08 5.911 33 2.214 22.00 57 2.26 58.65 75 2.30 130.h3 83 3.92 125.00 .92 Table 18. Determination of BET plot for NFDM. Relative Humidity Moisture Content TEES (%) m(db) (g solid/g water) 11 0.0117 2.61: 33 0.081 6.10 57 0.100 13.20 75 1.588 18.91: 83 1.897 25.77 90 3.065 29.36 93 Table 19. Determination of BET plot for casein. Relative Humidity Moisture Content 715:3E7 (%) m(db) (g solid/g water) 11 ' 0.047 2.6h 33 0.091 5.42 57 0.120 11.05 75 0.200 15.00 83 0.232 21.01 90 0.296 30.hl 94 Table 20. Desorption of diacetyl from NFDM (25C). Amount of Amount of Amount of Relative Diacetyl After Diacetyl Diacetyl Humidity Equilibration Desorbed Lost Mean (7.) (ppm) (ppm) (7.) ll 1) 5.70 2.00 26.0 2) 5.60 2.10 27.4 26 3) 5.70 2.00 26.0 4) 5.90 1.80 24.1 33 l) 0.85 6.85 89.0 2) 0.90 6.70 87.0 88 3) --- --- --- 4) 0.98 6.72 87.3 57 1) 0.85 6.85 89.0 89 75 l) 0.80 6.90 89.5 2) 0.75 6.95 90.3 90 3) 0.80 6.90 89.5 4) 0.70 7.00 91.0 83 l) 0.70 7.00 91.0 2) 0.75 6.95 90.3 90 3) 0.70 7.00 91.0 4) 0.80 6.90 89.5 90 l) 0.75 6.95 90.3 2) 0.70 7.00 91.0 90 3) 0.70 7.00 91.0 4) 0.75 6.95 90.3 95 Table 21. Desorption of diacetyl from casein (25C). Amount of Amount of Amount of Relative Diacetyl After Diacetyl Diacetyl Humidity Equilibration Desorbed Lost Mean (%) (ppm) (ppm) (%) ll 1) 6.21 5.09 h5.0 2) 5.85 5.h5 h8.2 h7 3) 5.94 5.36 47.4 33 1) 3.50 7.80 69.0 2) 3.65 7.65 67.7 69 3) 3.h8 7.82 99.2 57 1) 3.06 8.2h 72.9 73 75 1) 2.85 8.h5 7h.8 2) 2.79 8.51 75.3 75 3) 2.80 8.50 75.2 83 l) 2.65 8.65 76.5 2) 2.59 8.71 77.0 77 .3) 2.58 8.72 77.1 90 l) 2.60 8.70 76.9 2) 2.hh 8.86 78.h 77 3) 2.63 8.67 76.7 96 Table 22. Desorption of diacetyl from sucrose (25C). Amount of Amount of Amount of Relative Diacetyl After Diacetyl Diacetyl Humidity Equilibration Desorbed Lost Mean (%) (ppm) (ppm) (%) 11 1) 4.10 1.03 20.07 2) 3.85 1.28 24.95 23 3) 3.90 1.23 23.97 33 l) 1.05 4.08 79.53 2) 0.98 4.15 80.90 80 3) 0.99 4.14 80.70 57 l) 0.92 4.21 82.10 82 75 l) 0.87 4.26 83.04 2) 0.90 4.23 82.45 83 3) 0.86 4.27 83.23 83 1) 0.65 4.48 87.33 2) 0.67 4.46 86.94 87 3) 0.68 4.45 86.74 90 l) 0.32 4.81 93.76 2) 0.29 4.84 94.35 95 3) 0.20 4.93 96.10 97 Table 23. Desorption of diacetyl from cream.(25C). Amount of Amount of Amount of Relative Diac etyl After Di acetyl Di acetyl Humidity Equilibration Desorbed Lost Mean (%) (ppm) (ppm) (%) ll 1) 5.85 0.40 6.40 2) 5.92 0.33 5.28 6 3) 5.80 0.45 7.20 33 l) 4.78 1.47 23.52 2) 4.91 1.34 21.44 22 3) 4.85 1.40 22.40 57 1) 3.59 2.66 42.60 43 75 1) 2.69 3.56 56.96 2) 2.73 3.52 56.32 57 3) 2.70 3.55 56.80 83 1) 2.21 4.04 64.64 2) 2.19 4.06 64.96 65 3) 2.22 4.03 64.48 90 l) 2.07 4.18 66.88 2) 2.15 4.10 65.60 66 3) 2.06 4.19 67.04 HICHIGRN STRTE UNIV. LIBRQRIE llll3|IILHIlllzJIIELUIWILIUllllllllllllllHIIUIIMIIIIINHI 23 05247823