The Power Series as Water Sorption Isotherm Models

In this study, a power series which can generate different families of new water sorption isotherm models were presented. Experimentally obtained values for equilibrium moisture content of pear for three different temperatures, 15, 30 and 45C, and water activities, from 0.112 to 0.920, as well as literature experimental values for equilibrium moisture content of potato for three different temperatures, 30, 45 and 60C, and water activities, from 0.112 to 0.900, were fitted with the newly generated sorption isotherm models plus the referent Anderson model known in the literature as Guggenheim-Anderson-de Boer (GAB) model. In order to find which model gives the best results for approximation of experimental sorption data, several statistical criteria proposed in scientific literature were used. For each model and experimental data set, the average performance index was calculated and models were ranked afterwards. After that, some statistical rejection criteria were checked (D'Agostino-Pearson test of normality, single-sample run test, confidence intervals of estimated parameters, significance and precision of the model parameters). The performed statistical analysis shows that the two newly generated three-parameter models, M32 and M34, give the better fit to the sorption data of pear than the referent three-parameter Anderson model, while M32 gives the best fit to the sorption data of potato. Practical ApplicationsThe sorption isotherms of food materials are of great importance in the food industry and technology, especially for the design and optimization of the drying equipment and in the approach to the prediction of shelf life stability of material. The new generated three-parameter models, M32 and M34, can be successfully used in practical calculations of the equilibrium moisture content, which is an important parameter in storage conditions of dry food materials. With the incorporation of M32 or M34 in the drying model, more accurate values of transient moisture and temperature profiles on the dried materials will be obtained.