Take It With a Grain of Salt: Probing Effects of System Size, Cutoff Radius and Crystal Interface on NaCl Solubility Using Direct Coexistence Simulations

Abstract

Monte Carlo and molecular dynamics simulations were performed obtain the solubility of NaCl-H2O systems with respect to several key factors. The solubility of sodium chloride in water plays an essential role in various biological, geophysical, and environmental systems. Understanding the key processes that govern the equilibrium of ionic solutes in solid and solution phases is necessary to further develop technologies which exploit solubility such as carbon sequestration. While the solubility of aqueous NaCl systems has been well studied using Chemical Potential Methods (CPM), there is a need to better understand it from the lens of Direct Coexistence Methods (DCM). Both methods consistently predict differing solubilities for identical systems and understanding this discrepancy is an area of active research.

Limited CPM results were reproduced using isobaric-isothermal Monte Carlo simulations on NaCl crystals. In particular, the solid chemical potential was calculated using the SPC/E water model and force field parameters calculated by Joung and Cheatham. This model and the TIP4P-Ew model were used to perform extensive DCM studies using NPT molecular dynamics simulations. A system of a nanocrystal immersed in ionic solution was confirmed as the fastest to reach equilibrium and used to study the effects of system size, cutoff radius, and crystal interface on solubility. In addition, the buildup of crystal charge was observed and examined with respect to the independent variables and a system with initial charge bias was studied. As part of the goal to improve the understanding of DCM methods for studying phase equilibria, existing data on NaCl-H2O systems was verified and complemented. All reproduced DCM results concurred with reference data from past work within statistical error, thereby supporting the validity of this work's procedures. Strong evidence was found to suggest that solubility calculation is independent of system size and cutoff radius using 100 fcc crystals and no long range corrections. With respect to effect of crystal interface, no significant difference was observed between geometrically comparable fcc crystals with Miller indices of 100 and 110. Crystal charge buildup was observed in agreement with past computational and experimental work and shown to be strongly linearly related to system size. An initial charge bias on the crystal was shown to have no significant effect on the calculated solubility. The simulation results were largely consistent with observations from past work; DCM methods consistently predicted solubilities higher than CPM methods and uneven salting out of sodium ions over chloride ions was observed.