Best Practices for Fitting Machine Learning Interatomic Potentials for Molten Salts: A Case Study Using NaCl-MgCl2

TL;DR

This paper presents a methodology for developing transferable machine learning interatomic potentials for molten salts, demonstrating its effectiveness on NaCl-MgCl2 systems and comparing different DFT methods for accuracy.

Contribution

It introduces a robust, compositionally transferable ML potential for NaCl-MgCl2 molten salts using atomic cluster expansion and evaluates DFT methods for property prediction.

Findings

R2SCAN-D4 provides the most accurate thermophysical properties.

A potential trained on limited compositions is effective for the entire pseudo-binary system.

The developed potential enables better simulation of molten salt behavior.

Abstract

In this work, we developed a compositionally transferable machine learning interatomic potential using atomic cluster expansion potential and PBE-D3 method for (NaCl)1-x(MgCl2)x molten salt and we showed that it is possible to fit a robust potential for this pseudo-binary system by only including data from x={0, 1/3, 2/3, 1}. We also assessed the performance of several DFT methods including PBE-D3, PBE-D4, R2SCAN-D4, and R2SCAN-rVV10 on unary NaCl and MgCl2 salts. Our results show that the R2SCAN-D4 method calculates the thermophysical properties of NaCl and MgCl2 with an overall modestly better accuracy compared to the other three methods.