PC-SAFT predictive modeling of eutectic solvent and water mixtures: diol-based systems

A predictive framework based on the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) was developed to model the thermophysical behavior of eutectic solvents (ES) and their aqueous mixtures, providing a computationally efficient tool for solvent screening and design. Experimental densities and excess molar volumes ( V E ) were measured for binary aqueous mixtures of urea, L-proline, and D-fructose with 1,2-propanediol, 1,3-propanediol and 1,2-butanediol to validate the framework. Fully predictive PC-SAFT calculations revealed that the model alone could not accurately reproduce experimental density and V E . The inclusion of a single binary interaction parameter, fitted exclusively to neat ES density data, substantially improved density predictions while maintaining the predictive capability for aqueous systems, although V E remained less accurately described. Remarkably, introducing a single transferable interaction parameter enabled accurate prediction of both densities and V E across different hydrogen bond acceptors (HBA) and donors (HBD), including external literature systems. This transferable and minimal-parameter approach demonstrates that reliable predictions can be achieved without requiring experimental mixture data, offering a generically applicable, physically grounded framework that can accelerate rational solvent selection and process optimization in chemical engineering applications.