(123d) Molecular Simulation of Aqueous and Non-Aqueous Electrolyte Solutions

Electrolyte solutions play an important role in many industrial processes. In the present work, we apply molecular modeling and simulation on the basis of classical force fields with explicit solvent models to determine the thermodynamic properties of aqueous and non-aqueous electrolyte solutions. The focus of this work is on alkali cations and anions from the halide group. The ions are modeled as one Lennard-Jones (LJ) site with a point charge of ± 1e in its center of mass. For aqueous solutions, SPC/E water is used which is of the LJ + partial charge type. The force fields of the non-aqueous solvents, namely methanol and ethanol, are taken from previous work of our group. These models consist of two and three LJ sites, respectively, and three partial charges to model both polarity and hydrogen bonding. The simulations were performed with the molecular simulation program ms2. The long range interactions of the charges were considered by Ewald summation.

A comprehensive set of new force fields for the ions of the groups specified above is developed. They are optimized for describing basic thermodynamic data of aqueous solutions. The size parameter is adjusted to the reduced liquid solution density at 293.15 K, whereas the energy parameter (which does not significantly influence the density) is fitted to self-diffusion coefficient data at 298.15 K. A global fitting approach is used. The new models perform well over a wide range of ionic strength for all combinations of cations with anions. They predict well structural properties like the hydration number and transport properties like the electric conductivity in aqueous electrolyte solutions. The models are also studied with respect to the temperature dependence of the reduced density of the electrolyte solutions up to 333.15 K. The predictions from molecular simulation are in good agreement with experimental data, part of which is measured in the present work.

Basic thermodynamic properties of non-aqueous electrolyte solutions with the solvents methanol and ethanol are predicted by molecular simulation for all cation/anion combinations from the list above. The Lorentz-Berthelot combining rule is used for modeling the unlike LJ interactions between the ions and the solvent molecules. No adjustments to data for electrolytes in non-aqueous solvents are made. The predictions of the reduced liquid solution density from molecular simulation are in good agreement with experimental data, part of which is measured in the present work. Furthermore, structural properties like the radial distribution function of the solvent molecules around the ions and transport properties like the self-diffusion coefficient and the electric conductivity in non-aqueous electrolyte solutions are investigated.