A novel method to predict physicochemical properties of mixed solvents for chemical engineering application: Effective fragment potential version 2 – molecular dynamics (EFP2-MD) simulation

Solute-solvent interactions are the governing parameters of many physicochemical properties such as viscosity and solute solubility. Previous studies have already revealed that the intermolecular interactions vary non-linearly with molar fraction, which implies that mixed solvents could be designed to be chemically functional solvents, for example, the novel electrolytic solutions and reaction fields. However, predicting the controllable non-linear intermolecular interactions in mixed solvents becomes essentially impractical without experimental verification. Therefore, there is need for a physicochemical theory that can simulate mixed solvents’ properties without depending directly on experimentally derived parameters. In line with this, the ab initio effective fragment potential technique has been largely investigated. Unfortunately, most of the published reports on this topic only focus on micro solvation.

To this note, Professor Hirotoshi Mori and his PhD student Nahoko Kuroki at Ochanomizu University in Japan investigated the applicability of effective fragment potential version 2 – molecular dynamics (EFP2-MD) simulations in predicting thermochemical excess properties of unknown solvent mixtures. Explicitly, they purposed to explore the systematic prediction of excess volume for water-methanol binary mixture and its viability. Their work is currently published in the research journal, Chemical Physics Letters.

The research technique employed commenced with the optimization of the molecular structures of water and methanol. Next, at the optimum structures for each basis set, ab initio polarizable EFP2 force fields were determined. The researchers then went ahead and performed the EFP2-MD simulations for water-methanol binary liquid mixtures with 11 different methanol molar fractions at specified intervals. Lastly, a set of 150 picoseconds equilibration and 100 picoseconds production runs were performed to evaluate the volume and the density of each water-methanol binary mixture.

The authors observed that reproducing the experimental excess molar volume (a mixed solvent property) was very difficult by classical molecular dynamics simulations that employ pre-optimized force field parameters. Additionally, a comparison of the accuracy of the EFP2-MD simulations with effective fragment potentials expanded by four different basis sets was seen to reveal the significance of obtaining superior accuracy for describing intermolecular interactions in liquid mixtures.

The authors’ study successfully reported on the applicability of EFP2-MD simulations to mixed solvent systems with a set of numerical examples for various water-methanol binary liquid mixtures. Generally, it was observed that contrary to classical molecular dynamics simulations, the ab initio EFP2-MD simulations were promising from both an accuracy and computational point of view. The authors have also published an article representing the applicability of EFP2-molecular dynamics simulations to ionic liquids, which is promising as CO₂ absorbate and non-volatile electrolyte. Altogether, this work has shown that it is possible to predict mixed solvents’ properties by EFP2-MD simulations. In the field of chemical engineering, it is very important to design chemical function of mixed liquids. The EFP2-MD method will open novel insights for chemical engineering in the future.