Accelerated materials property predictions and design using motif-based fingerprints

TL;DR

This paper introduces hierarchical motif-based topological fingerprints for materials, enabling rapid property prediction and inverse design, thus accelerating materials discovery processes.

Contribution

It proposes a novel motif-based fingerprinting method for materials and demonstrates its effectiveness in property prediction and inverse design tasks.

Findings

Fingerprints effectively predict material properties.

Learning model can be inverted for targeted material design.

Method applies to molecules and crystals.

Abstract

Data-driven approaches are particularly useful for computational materials discovery and design as they can be used for rapidly screening over a very large number of materials, thus suggesting lead candidates for further in-depth investigations. A central challenge of such approaches is to develop a numerical representation, often referred to as a fingerprint, of the materials. Inspired by recent developments in chem-informatics, we propose a class of hierarchical motif-based topological fingerprints for materials composed of elements such as C, O, H, N, F, etc., whose coordination preferences are well understood. We show that these fingerprints, when representing either molecules or crystals, may be effectively mapped onto a variety of properties using a similarity-based learning model and hence can be used to predict relevant properties of a material, given that its fingerprint can be defined. Two simple procedures are introduced to demonstrate that the learning model can be inverted to identify the desired fingerprints and then, to reconstruct molecules which possess a set of targeted properties.