Prediction of Small-molecule Pharmaceuticals Solubility Parameters Using a Thermodynamic SAFT-based Equation of State.

Abstract

Accurate prediction of drug solubility parameters plays a crucial role in optimizing pharmaceutical formulations. In this study, the solubility parameters of pharmaceutical compounds are estimated using the Perturbed Chain Statistical Associating Fluid Theory (PC-SAFT) equation of state (EoS). It should be noted that experimental values of drug solubility parameters are scarce, and group contribution (GC) methods have several significant limitations. Solubility is influenced by factors such as steric hindrance and intramolecular hydrogen bonding, which are not captured by GC approaches. As well, most GC tables are based on common organic functional groups, whereas many drugs contain rare or novel groups, making GC-based estimates either unavailable or unreliable. It is the first attempt to predict the solubility parameter of pharmaceuticals using a SAFT-based EoS. The PC-SAFT EoS parameters were calculated from binary experimental solubility data and subsequently applied to predict drug solubility parameters. To enhance model performance, the association interactions between drug-drug and drug-solvent molecules were explicitly considered. The effect of each contribution (hard-chain, dispersion, and hydrogen bonding) on solubility parameter prediction has been investigated. The results demonstrate that hydrogen-bonding interaction plays a critical role in accurately predicting solubility parameters. In addition to using the PC-SAFT EoS, an unconstrained regression approach (URA) was employed as a complementary method. In the URA, experimental solubility data were incorporated to establish correlations for the Hansen solubility terms. The predictions obtained with PC-SAFT EoS were compared to those from regression model and GC approaches, showing that the PC-SAFT approach provides satisfactory accuracy for drug solubility parameter estimation. This approach provides a tool for pre-designing new drug candidates by optimizing solvent selection in chemical processes.