Application of Machine Learning to Sporadic Experimental Data for Understanding Epitaxial Strain Relaxation

TL;DR

This paper uses machine learning to analyze sporadic experimental data, revealing key physical features influencing strain relaxation in epitaxial thin films, and validates the approach with experimental data.

Contribution

It introduces a data analytics framework combining correlation analysis and machine learning to understand epitaxial strain relaxation from limited experimental data.

Findings

Poisson's ratio and lattice mismatch are key features.

ML models achieve decent accuracy in predicting critical thickness.

Framework validated with experimental data.

Abstract

Understanding epitaxial strain relaxation is one of the key challenges in functional thin films with strong structure-property relation. Herein, we employ an emerging data analytics approach to quantitatively evaluate the underlying relationships between critical thickness (hc) of strain relaxation and various physical and chemical features, despite the sporadic experimental data points available. First, we have collected and refined reported hc of perovskite oxide thin film/substrate system to construct a consistent sub-dataset which captures a common trend among the varying experimental details. Then, we employ correlation analyses and feature engineering to find the most relevant feature set which include Poisson's ratio and lattice mismatch. With the insight offered by correlation analyses and feature engineering, machine learning (ML) models have been trained to deduce a decent accuracy, which has been further validated experimentally. the demonstrated framework is expected to be efficiently extended to the other classes of thin films in understanding hc.