Surrogate models for vibrational entropy based on a spatial decomposition

TL;DR

This paper introduces a machine learning surrogate model based on atomic site decomposition to efficiently predict vibrational entropy in crystalline structures, reducing computational costs and enabling better defect behavior analysis.

Contribution

It presents a novel approach combining atomic site entropy decomposition with machine learning to accurately predict vibrational entropy, improving computational efficiency.

Findings

Surrogate models accurately predict vibrational entropy for defects.

Atomic site decomposition effectively captures local entropy contributions.

Method reduces computational cost compared to traditional techniques.

Abstract

The temperature-dependent behavior of defect densities within a crystalline structure is intricately linked to the phenomenon of vibrational entropy. Traditional methods for evaluating vibrational entropy are computationally intensive, limiting their practical utility. We show that total entropy can be decomposed into atomic site contributions and rigorously estimate the locality of site entropy. This analysis suggests that vibrational entropy can be effectively predicted using a surrogate model for site entropy. We employ machine learning to develop such a surrogate models employing the Atomic Cluster Expansion model. We supplement our rigorous analysis with an empirical convergence study. In addition we demonstrate the performance of our method for predicting vibrational formation entropy and attempt frequency of the transition rates, on point defects such as vacancies and interstitials.