Molecular Dynamics Simulations of Molten Magnesium Chloride Using Machine‐Learning‐Based Deep Potential. Issue 12 (2nd November 2020)

• Record Type: Journal Article
• Authors: Liang, Wenshuo; Lu, Guimin; Yu, Jianguo
• Is Part Of: Advanced theory and simulations. Volume 3:Issue 12(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-11-02
• Subjects: deep potentials -- machine learning -- magnesium chloride -- molecular dynamics simulations; Science -- Simulation methods -- Periodicals; Science -- Methodology -- Periodicals; Engineering -- Simulation methods -- Periodicals; Engineering -- Methodology -- Periodicals; 507.21
• DOI: 10.1002/adts.202000180
• Languages: English
• ISSNs: 2513-0390

Abstract

Abstract: In previous work, molten magnesium chloride has been investigated using first‐principles molecular dynamics (FPMD) simulations based on density functional theory (DFT). However, such simulations are computationally intensive and therefore are restricted in terms of simulated size and time. In this work, a machine learning‐based deep potential (DP) is trained to accelerate the molecular dynamics simulation of molten magnesium chloride. The trained DP can accurately describe the energies and forces with the prediction errors in energy and force being 1.76 × 10 −3 eV/atom and 4.76 × 10 −2 eV Å −1, respectively. Applying the deep potential molecular dynamics (DPMD) approach, simulations can be performed with more than 1000 atoms, which is infeasible for FPMD simulations. Additionally, the partial radial distribution functions, angle distribution functions, densities, and self‐diffusion coefficients predicted by DPMD simulations are also in reasonable agreement with FPMD or experimental results. This work shows that the DP enables higher efficiency and similar accuracy relative to DFT, exhibiting a bright application prospect in modeling molten salt systems. Abstract : Machine‐learning‐based deep potential (DP) can provide results with accuracy that is comparable to density functional theory (DFT) and efficiency that is similar to empirical potentials. DP can be used to study a number of physical properties that can only be calculated accurately in a larger system or on a longer timescale, which is outside the capability of DFT.