Quantifying Disorder One Atom at a Time Using an Interpretable Graph Neural Network Paradigm

TL;DR

This paper introduces an interpretable graph neural network approach to quantify atomic disorder in materials, enabling detailed analysis of local structural environments and their impact on material properties.

Contribution

The authors develop a novel, physically interpretable metric for local atomic disorder using graph neural networks, applicable to various material interfaces and microstructures.

Findings

Successfully applied to aluminum to track interface evolution

Extracted physics-preserved gradients for predicting material behavior

Quantified disorder spectrum between solid and liquid phases

Abstract

Quantifying the level of atomic disorder within materials is critical to understanding how evolving local structural environments dictate performance and durability. Here, we leverage graph neural networks to define a physically interpretable metric for local disorder. This metric encodes the diversity of the local atomic configurations as a continuous spectrum between the solid and liquid phases, quantified against a distribution of thermal perturbations. We apply this novel methodology to three prototypical examples with varying levels of disorder: (1) solid-liquid interfaces, (2) polycrystalline microstructures, and (3) grain boundaries. Using elemental aluminum as a case study, we show how our paradigm can track the spatio-temporal evolution of interfaces, incorporating a mathematically defined description of the spatial boundary between order and disorder. We further show how to extract physics-preserved gradients from our continuous disorder fields, which may be used to understand and predict materials performance and failure. Overall, our framework provides an intuitive and generalizable pathway to quantify the relationship between complex local atomic structure and coarse-grained materials phenomena.