No abstract

A reformulation of the classical Brunauer, Emmett and Teller multimolecular adsorption equation resulted in a unimolecular equation in which a third parameter was introduced into general statements. The third parameter represented adsorption centres on the unimolecular surface. This parameter modified the constant C in the BET equation to C 1/n which, on analysis, was found to measure the density of the most strongly adsorbed molecules. The new equation, which was shown to apply to all isotherm shapes, delineated limits between which three states of unimolecularly adsorbed molecules could be identified along the isotherm. It also provided a new method for calculating the net heat of adsorption, which is identical with the Clausius-Clapeyron equation at low values of Aw. The equation also explained the phenomenon of adsorption compression in sorption studies and this was because of the increased density of primary adsorbate molecules at low values of Aw.

It has been explained that hysteresis is a sorption phenomenon, which rests on temperature, Aw, moisture content and on surface energy Q. As temperature and Aw as environmental variables have opposite effects on moisture content the isobar and isotherm methods are bound to have opposite effects on the hysteresis loop. Using Caurie's equation it has been shown that the desorption isotherm of the hysteresis loop is always at a higher energy level than the adsorption isotherm. This energy difference is stated to arise from physical changes in the adsorbent matrix which expose new energetic sites which adsorb moisture on return to lower Aws rather than desorb moisture and this is indicated to be responsible for the hysteresis phenomenon. It has also been argued that hysteresis may be used as an index of food quality.

A simple method for integrating and extending the binary form of the Gibbs-Duhem equation for routine application in food technology is given. The integrated equation was used to show that considerable solute-solute interaction occurs in mixed solutions which, when corrected for in the Ross equation, results in a revised equation whose predictions of the a, of mixed solutions more closely agree with measured values. The revised equation indicates that one can obtain accurate estimates of the a,,, of a mixed solution by combining estimated water activities with estimated interactions based on the number and concentrations of the solution components. INTRODUCTIONPRACTICAL SOLUTIONS as they exist in foods are far from ideal due to solute interactions in solution. A useful method for predicting both the a, of simple and complex solutions is the Gibbs-Duhem equation. Working with this equation Ross (1975) made the simplifying assumption that the interaction occurring between solute components in a mixed solution do not change the activity coefficients of the solutes in the mixture from their values in binary solutions. With this assumption he derived a simple a, product method for predicting the a, of mixed solutions. However, when applied to practical solutions, the Ross equation is found to overestimate observed values (Bone et al., 1975;Chuang and Toledo, 1976). A simple method for integrating the binary form of the Gibbs-Duhem equation is presently not available and therefore Ross (1975) in his derivation used the symbolic form of the equation.We propose in this study to suggest a simple method to fully integrate the binary form of the Gibbs-Duhem equation and to use it to test the Ross (1975) equation to identify and eliminate any factors which reduce its prediction value.

Parameters of Caurie's [International Journal of Food Science and Technology 40 (2005) 283] unimolecular adsorption equation have been used to calculate total bound water to equal the square of the primary water capacity or m 0 2 grams. Current freezing methods predict bound water up to nm 0 grams which leaves a fraction of the total bound water with limited freezing properties unaccounted for. From these studies three types of bound water have been identified at room temperature along a decreasing energy gradient. It has been shown that the stability of processed and blended foods will improve with formula modifications consistent with expansion of type II bound water molecules and processed foods will be more stable the smaller the fractional ratio of type III to type II bound water molecules.