An evaluation of in-silico methods for predicting solute partition in multiphase complex fluids – A case study of octanol/water partition coefficient

Abstract

Solute partition in multiphase fluids is an important thermodynamic phenomenon and performance attribute for a wide range of product formulations of foods, pharmaceuticals and cosmetics. Experimental evaluation of partition coefficients in complex product formulations is empirical, difficult and time consuming. In-silico methods such as fragment constant method and group contribution method require parameter fitting to the experimental data and are limited to relatively simple fluids. Recently, a method combining molecular dynamics (MD) and quantum chemical (QC) calculation of screening charge density function has been reported. The method does not only use fundamental properties of intermolecular force and charge density function, which does not require parameter fitting to the experimental data, but also applies to complex fluid structures such as micelles. In this work, the predictive accuracy of the combined method of MD and QC is evaluated. Using widely available octanol-water partition coefficients as a case study, the performance of the combined MD and COSMOmic for predicting octanol/water partition coefficients has been compared with those of the EPI Suite™ fragment constant method, UNIFAC group contribution method and COSMOtherm. The prediction of the combined MD/COSMOmic method is the closest to the best performing fragment constant method which was specifically designed for the octanol-water system. The combined MD/QC method proves to be the most promising and robust method applicable to a wide range of complex structures of multiphase fluid systems.