A Universal Machine Learning Model for Elemental Grain Boundary Energies

TL;DR

This paper presents a machine learning model that accurately predicts grain boundary energies across multiple metals using only four geometric features, demonstrating universality and extrapolation capabilities.

Contribution

The study introduces a universal ML model for GB energies based on geometric features, capable of extrapolating to high $\Sigma$ boundaries with high accuracy.

Findings

Predicts GB energies with 0.13 J/m² MAE

Model generalizes to high $\Sigma$ GBs

Highlights importance of physics-informed features

Abstract

The grain boundary (GB) energy has a profound influence on the grain growth and properties of polycrystalline metals. Here, we show that the energy of a GB, normalized by the bulk cohesive energy, can be described purely by four geometric features. By machine learning on a large computed database of 361 small $\Sigma$ ($\Sigma < 10$) GBs of more than 50 metals, we develop a model that can predict the grain boundary energies to within a mean absolute error of 0.13 J m$^{-2}$. More importantly, this universal GB energy model can be extrapolated to the energies of high $\Sigma$ GBs without loss in accuracy. These results highlight the importance of capturing fundamental scaling physics and domain knowledge in the design of interpretable, extrapolatable machine learning models for materials science.