Formula graph self-attention network for representation-domain independent materials discovery

TL;DR

This paper introduces a novel formula graph self-attention network that unifies different material representations, enabling transferable embeddings and improved property prediction, including dielectric functions and epsilon-near-zero materials.

Contribution

It proposes a new formula graph concept and a self-attention GNN architecture that bridges structure-based and stoichiometry-only material descriptors.

Findings

Outperforms previous structure-agnostic models.

Achieves better sample efficiency and faster convergence.

Successfully predicts dielectric properties and identifies epsilon-near-zero materials.

Abstract

The success of machine learning (ML) in materials property prediction depends heavily on how the materials are represented for learning. Two dominant families of material descriptors exist, one that encodes crystal structure in the representation and the other that only uses stoichiometric information with the hope of discovering new materials. Graph neural networks (GNNs) in particular have excelled in predicting material properties within chemical accuracy. However, current GNNs are limited to only one of the above two avenues owing to the little overlap between respective material representations. Here, we introduce a new concept of formula graph which unifies stoichiometry-only and structure-based material descriptors. We further develop a self-attention integrated GNN that assimilates a formula graph and show that the proposed architecture produces material embeddings transferable between the two domains. Our model can outperform some previously proposed structure-agnostic models and their structure-based counterparts while exhibiting better sample efficiency and faster convergence. Finally, the model is applied in a challenging exemplar to predict the complex dielectric function of materials and nominate new substances that potentially exhibit epsilon-near-zero phenomena.