Calculation of the melting point of alkali halides by means of computer simulations

TL;DR

This study uses computer simulations to calculate the melting points of alkali halides, comparing different models and methods, revealing that some models accurately predict NaCl's melting point while others overestimate, highlighting the need for improved force-fields.

Contribution

It introduces three independent computational routes to determine melting points and evaluates their consistency and accuracy across different alkali halide models.

Findings

The Tosi-Fumi model accurately predicts NaCl's melting point.

Smith-Dang and Joung-Cheatham models overestimate melting points.

Current force-fields need improvement for better property predictions.

Abstract

In this manuscript we study the liquid-solid coexistence of NaCl-type alkali halides, described by interaction potentials such as Tosi-Fumi (TF), Smith-Dang (SD) and Joung-Cheatham (JC), and compute their melting temperature (Tm) at 1 bar via three independent routes: 1) liquid/solid direct coexistence, 2) free-energy calculations and 3) Hamiltonian Gibbs-Duhem integration. The melting points obtained by the three routes are consistent with each other. The calculated Tm of the Tosi-Fumi model of NaCl is in good agreement with the experimental value as well as with other numerical calculations. However, the other two models considered for NaCl, SD and JC, overestimate the melting temperature of NaCl by more than 200 K. We have also computed the melting temperature of other alkali halides using the Tosi-Fumi interaction potential and observed that the predictions are not always as close to the experimental values as they are for NaCl. It seems that there is still room for improvement in the area of force-fields for alkaline halides, given that so far most models are still unable to describe a simple yet important property such as the melting point.