Rational Solvent Selection for Pharmaceutical Impurity Purge

This study focuses on the development of computational impurity purge factor to enable rational solvent selection for purification of pharmaceutical compounds via recrystallization. Three purge factors have been proposed under SLE thermodynamic consideration based on two properties: impurity partition coefficient between solvent and crystalline pharmaceutical, and thermodynamic solubility of the crystalline pharmaceutical. A limited testing of the purge factors against experimental observations was performed, using COSMO-RS approach as implemented in the COSMOtherm software. Good and superior performance of the purge factor proportional to a difference between the impurity partition coefficient and the crystalline pharmaceutical solubility was demonstrated. As a result, a virtual solvent screening based on that factor is proposed to be used to guide selection of preferred solvents for impurity purge via recrystallization.