Accurate and efficient models to predict the phase-behavior of nonelectrolyte solids in a wide range of solvents are central to the design of novel separation processes. A promising design tool is the MOSCED limiting activity coefficient model which is parameterized for 133 solvents. However, before predictions may be made for a solute of interest, solute MOSCED parameters are required. For novel compounds of interest, the necessary MOSCED parameters are unavailable and sufficient reference data is likely unavailable to regress the necessary parameters.
In this study we explore the use of molecular simulation and electronic structure methods to generate the reference data necessary to obtain solute MOSCED parameters. Specifically, we use solvation free energy calculations to predict the limiting activity coefficient of the solute in a limited range of solvents. This reference data is then used to regress the necessary MOSCED parameters for the solute. The method is applied to a range of pharmaceutical compounds and environmental pollutants. The limiting activity coefficients computed using MOSCED may subsequently be used to parameterize Wilson’s equation and make equilibrium solubility predictions in pure and multicomponent solvents.
See more of this Group/Topical: Computational Molecular Science and Engineering Forum