Transferable Polarizable Model Development for Aqueous and Crystalline Alkali-Halide Salts

Electrolyte solutions are vital to the development of novel technologies such as new batteries and carbon sequestration methods. Accurate theoretical models that can predict the collective properties of electrolyte solutions (solubilities, activity coefficients, transport coefficients, etc.) are highly desirable as they can guide the rational design of new technologies that rely on these substances. However, theoretical modeling using molecular dynamics simulations is plagued by an efficiency vs. accuracy tradeoff in the description of the interatomic forces. In this work we investigate how the inclusion of atomic polarizability, by means of Drude oscillators, affects the ability of efficient classical force fields to model the temperature dependence of the solubility and activity coefficients of alkali-halide salts. To achieve this, we propose a new method to efficiently and accurately compute derivatives of the aqueous chemical potential with respect to force field parameters, enabling gradient-based fitting directly to chemical potential data. Using this method, we investigate the ability of polarizable models to predict pure solid – solution and hydrated solid – solution solubility limits as functions of temperature and show that predictions can be improved provided that empirical combining rules are relaxed. Overall, this work explores the limits of classical polarizable force fields in modeling electrolyte solutions across the solid-solution phase diagram.