(189ar) Comparison of Interatomic Potentials for Interfacial Studies of Ionic Liquid Systems

Authors: 
Tiet, F., Vanderbilt University
Thompson, M., Vanderbilt University
Cummings, P. T., Vanderbilt University
Molecular simulation is now a well-established too for understanding the structure, dynamics, and thermodynamics of ionic liquids at interfaces and in confinement. Some theoretical issues, however, are pervasive in the literature, largely owing to the selection and application of force fields. The most popular classical force field, the so-called CL&P force field of Canongia Lopes and Padua, has well-documented issues including slow transport properties and underprediction of hydrogen bonding behavior. Various corrections to the original force field have been explored and other similar force fields have been developed. The improvements are largely evaluated on the accuracy of their fundamental physiochemical properties, i.e. density, molecular diffusivity, enthalpy of vaporization, etc. Unclear, however, are the effects on high-level properties that drive the performance of ionic liquid applications. Additionally, the lack of a general protocol for handling fluid-solid interactions, controlling for any errors in the fluid-fluid interactions, provides further of uncertainty in the validity of these molecular models. To this end, we consider two systems that attempt to bridge low- and high-level properties: nanodroplets of ionic liquids on a graphene surface and a channel of ionic liquids interacting with charged graphene surfaces. In droplet simulations, we observe remarkably different wetting behavior across the different force fields and fluid-solid mixing rules. In channel simulations, we observe largely similar differential capacitances, despite notable differences in the structure of the ions at the interface and dynamics everywhere. These results imply that large inaccuracies in fundamental molecular properties may still be able to reasonably predict broader physical properties at larger scales.