Universal Fragment Descriptors for Predicting Electronic Properties of Inorganic Crystals

TL;DR

This paper introduces universal property-labeled fragments (PLMF) as descriptors for inorganic crystals, enabling highly accurate predictions of multiple electronic and thermomechanical properties across diverse materials, thus accelerating materials discovery.

Contribution

The paper presents a new universal descriptor set (PLMF) that improves the prediction of various properties for inorganic crystals using minimal structural information.

Findings

QMSPR models achieve near-training accuracy for diverse properties

PLMF descriptors enable universal and interpretable predictions

Materials informatics can significantly speed up materials discovery

Abstract

Historically, materials discovery has been driven by a laborious trial-and-error process. The growth of materials databases and emerging informatics approaches finally offer the opportunity to transform this practice into data- and knowledge-driven rational design. By using data from the AFLOW repository for high-throughput ab-initio calculations, we have generated Quantitative Materials Structure-Property Relationship (QMSPR) models to predict eight critical electronic and thermomechanical materials properties, such as the metal/insulator classification, band gap energy, bulk and shear moduli, Debye temperature, and heat capacity. The prediction accuracy obtained with these QMSPR models approaches training data for virtually any stoichiometric inorganic crystalline material. The success and universality of these models is attributed to the construction of new materials descriptors---referred to as the universal Property-Labeled Materials Fragments (PLMF). The representation requires only minimal structural input and affords straightforward model interpretation in terms of simple heuristic design rules that guide rational materials design. This study demonstrates the power of materials informatics to dramatically accelerate the search for new materials.