Materials Graph Transformer predicts the outcomes of inorganic reactions with reliable uncertainties

TL;DR

This paper introduces a graph transformer model that predicts the main products of inorganic solid-state reactions, using learned reaction representations to improve accuracy and uncertainty estimation, aiding materials discovery.

Contribution

It presents a novel graph-based transformer approach with learned reaction representations, enhancing prediction accuracy and uncertainty quantification in inorganic reaction outcomes.

Findings

Outperforms baseline methods in predicting reaction products.

Provides more reliable uncertainty estimates.

Uses reaction graphs with message passing for representation learning.

Abstract

A common bottleneck for materials discovery is synthesis. While recent methodological advances have resulted in major improvements in the ability to predicatively design novel materials, researchers often still rely on trial-and-error approaches for determining synthesis procedures. In this work, we develop a model that predicts the major product of solid-state reactions. The cardinal feature of this approach is the construction of fixed-length, learned representations of reactions. Precursors are represented as nodes on a `reaction graph', and message-passing operations between nodes are used to embody the interactions between precursors in the reaction mixture. Through an ablation study, it is shown that this framework not only outperforms less physically-motivated baseline methods but also more reliably assesses the uncertainty in its predictions.