(257ab) Employing Modeling Techniques to Predict the Solubility of Hesperetin in Binary Solvent Systems

Processes involving solubility, such as crystallization, are utilized throughout industry.  The efficiency of these processes significantly increases when optimal solute-solvent mixtures are used.  The composition of these mixtures can be determined both quickly and inexpensively with the use of predictive solubility models.  The Non-Random Two Liquid Segment Activity Coefficient (NRTL-SAC) thermodynamic solubility model, first proposed by Chen and Song, predicts solubility behavior more accurately than competing models, such as COSMO-SAC and UNIFAC.  Recently, however, Ferreira et al. (2013) suggested that NRTL-SAC is incapable of qualitatively predicting the solubility of hesperetin, a flavanone molecule, in some binary solvent systems.  In this work, we use NRTL-SAC, along with single-solvent solubility data, to accurately predict the solubility trends of hesperetin in the aforementioned binary solvent systems.  Additionally, we compare the predictions obtained by using COSMO-SAC with those obtained by using NRTL-SAC.  Finally, we demonstrate a representation of the system involving the use of COSMO-SAC in conjunction with NRTL-SAC