Extracting an Empirical Intermetallic Hydride Design Principle from Limited Data via Interpretable Machine Learning

Witman, Matthew; Ling, Sanliang; Grant, David M.; Walker, Gavin S.; Agarwal, Sapan; Stavila, Vitalie; Allendorf, Mark D.

Abstract

An open question in the metal hydride community is whether there are simple, physics-based design rules that dictate the thermodynamic properties of these materials across the variety of structures and chemistry they can exhibit. While black box machine learning-based algorithms can predict these properties with some success, they do not directly provide the basis on which these predictions are made, therefore complicating the a priori design of novel materials exhibiting a desired property value. In this work we demonstrate how feature importance, as identified by a gradient boosting tree regressor, uncovers the strong dependence of the metal hydride equilibrium H2 pressure on a volume-based descriptor that can be computed from just the elemental composition of the intermetallic alloy. Elucidation of this simple structure–property relationship is valid across a range of compositions, metal substitutions, and structural classes exhibited by intermetallic hydrides. This permits rational targeting of novel intermetallics for high-pressure hydrogen storage (low-stability hydrides) by their descriptor values, and we predict a known intermetallic to form a low-stability hydride (as confirmed by density functional theory calculations) that has not yet been experimentally investigated.

• Journal Article Type: Article
• Acceptance Date: Dec 7, 2019
• Online Publication Date: Dec 7, 2019
• Publication Date: Jan 2, 2020
• Deposit Date: Dec 21, 2019
• Publicly Available Date: Dec 8, 2020
• Journal: Journal of Physical Chemistry Letters
• Print ISSN: 1948-7185
• Electronic ISSN: 1948-7185
• Publisher: American Chemical Society
• Peer Reviewed: Peer Reviewed
• Volume: 11
• Issue: 1
• Pages: 40-47
• DOI: https://doi.org/10.1021/acs.jpclett.9b02971
• Public URL: https://nottingham-repository.worktribe.com/output/3612090
• Publisher URL: https://pubs.acs.org/doi/10.1021/acs.jpclett.9b02971
• Additional Information: This document is the Accepted Manuscript version of a Published Work that appeared in final form in Journal of Physical Chemistry Letters copyright © American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see https://pubs.acs.org/doi/10.1021/acs.jpclett.9b02971