Conductivity and Diffusivity of Ions in Aqueous MgCl₂ from Equilibrium and Nonequilibrium Simulations

Advanced electrochemical energy storage technologies require new electrolytes to be considered, so efficient computational characterization of ionic conductivity in a range of systems is of importance. In this manuscript we compare different equilibrium (EMD) and nonequilibrium molecular dynamics (NEMD) simulation algorithms to determine ionic conductivities. Aqueous magnesium ion batteries utilizing magnesium chloride as the electrolyte are a promising alternative to conventional lithium-ion batteries, so we focus on magnesium chloride electrolytes to demonstrate our results. We show the importance of accounting for ionic correlations and find that NEMD algorithms can provide more efficient calculations of ionic conductivity than EMD algorithms when ionic correlations need to be accounted for. In contrast, diffusivities and Nernst-Einstein conductivities can be determined more efficiently with EMD algorithms in the highly conductive systems considered here. We also demonstrate that the alignment of the water molecules due to the applied field in NEMD simulations is small at typical field strengths and has no impact on the calculated conductivities. Comparison of the results for the conductivity of different ions and their coupling provides insight into how force fields might be improved.