This study evaluates the effects of temperature (30–50 °C) and sucrose concentration (40–60%) on the osmotic dehydration (OD) of bayo beans (Phaseolus vulgaris), focusing on water loss ( \(\:WL\) ), solid gain ( \(\:SG\) ), and the estimation of diffusion coefficients (D). The Azuara and Page models were applied, and a new model was proposed. Phenomenological models with geometries of infinite slab, parallelepiped, finite cylinder, hemispherical, sphericity factor, and sphere were employed to analyze mass transfer, with the Bayesian Information Criterion ( \(\:BIC\) ) guiding model selection. The new empirical model demonstrated accuracy comparable to the Page model, with BIC values ranging from −204 to −126 and −207 to −128, respectively. D for \(\:WL\) , estimated using the hemispherical model, ranged from 0.709 × \(\:{10}^{-10}\) to 2.21 × \(\:{10}^{-10}\) m2/s, with this model providing the best fit. For \(\:SG\) , planar geometries such as the parallelepiped yielded better fits, with coefficients ranging from 0.181 × \(\:{10}^{-10}\) and 0.326 × \(\:{10}^{-10}\) m2/s, suggesting dominant surface interactions. Activation energy values indicated distinct mechanisms governing \(\:WL\) and \(\:SG\) , each influenced by temperature and sucrose levels. These findings underscore the importance of selecting suitable geometric models to describe mass transfer in OD and support the development of innovative bean-based products.
