Atomistic modeling of the mechanical properties: the rise of machine learning interatomic potentials

Abstract

This minireview highlights the superiority of machine learning interatomic potentials over the conventional empirical interatomic potentials and density functional theory calculations for the analysis of mechanical and failure responses.

Promising novel thermoelectric materials: two-dimensional penta-like PtPX (X = S, Se, Te) nanosheets

“…58 It has been reported that MLIP also plays a crucial role in examining the mechanical properties of materials by measuring the mechanical response, piezoelectricity, and photocatalysis of two-dimensional materials. 59–61 These findings demonstrate that MLIP offers a special opportunity for examining structural features and merits further development.…”

The thermoelastic properties of monolayer covalent organic frameworks studied by machine-learning molecular dynamics

“…Recently, machine-learned potential (MLP) has been shown to be a promising on-demand approach for investigating the mechanical properties of 2D materials. 29 For example, machine-learning interatomic potentials (MLIPs) were developed by Mortazavi et al for studying the mechanical behaviors and properties of various 2D materials. 30,31 It was shown that MLIPs could enable the efficient use of classical MD simulations to evaluate the mechanical properties of relatively large 2D material systems with the DFT level of accuracy.…”

Molecular dynamics simulations of liquid gallium alloy Ga–X (X = Pt, Pd, Rh) via machine learning potentials

“…Machine learning potentials (MLPs) have recently been proposed to address the limitations of both AIMD and CMD simulations. [15][16][17][18][19][20][21][22] Typically, MLPs can be obtained by training the datasets from AIMD simulations. Thus, MD simulations with MLPs can not only maintain the accuracy of AIMD but also achieve the time-scale of CMD.…”