An Efficient Monte Carlo Algorithm for Determining the Minimum Energy Structures of Metallic Grain Boundaries

TL;DR

This paper introduces a fast Monte Carlo algorithm to efficiently explore and identify low-energy grain boundary structures in metallic systems across five-dimensional crystallographic space, aiding materials design.

Contribution

The paper presents a novel Monte Carlo method with atom insertion/removal moves for rapid convergence to minimum energy grain boundaries in metals.

Findings

Validated on over 1184 GBs in aluminum, nickel, and iron.

Achieved rapid convergence to low-energy structures.

Demonstrated robustness across different metallic systems.

Abstract

Sampling minimum energy grain boundary (GB) structures in the five-dimensional crystallographic phase space can provide much-needed insight into how GB crystallography affects various interfacial properties. However, the complexity and number of parameters involved often limits the extent of this exploration to a small set of interfaces. In this article, we present a fast Monte Carlo scheme for generating zero-Kelvin, low energy GB structures in the five-dimensional crystallographic phase space. The Monte Carlo trial moves include removal and insertion of atoms in the GB region, which create a diverse set of GB configurations and result in a rapid convergence to the low energy structure. We have validated the robustness of this approach by simulating over 1184 tilt, twist, and mixed character GBs in both fcc (Aluminum and Nickel) and bcc ($\alpha$-Iron) metallic systems.