Universal fragment descriptors for predicting properties of inorganic crystals

TLDR

Materials discovery has shifted from laborious trial‑and‑error to knowledge‑driven design enabled by machine learning and materials databases, with universal applicability attributed to Property‑Labelled Materials Fragments. The authors use AFLOW ab initio data with Quantitative Materials Structure‑Property Relationship models, employing Property‑Labelled Materials Fragments that need only minimal structural input for simple heuristic design rules. The prediction accuracy matches the quality of the training data and aligns with experimental thermomechanical data for virtually any stoichiometric inorganic crystal.

Abstract

Abstract Although historically materials discovery has been driven by a laborious trial-and-error process, knowledge-driven materials design can now be enabled by the rational combination of Machine Learning methods and materials databases. Here, data from the AFLOW repository for ab initio calculations is combined with Quantitative Materials Structure-Property Relationship models to predict important properties: metal/insulator classification, band gap energy, bulk/shear moduli, Debye temperature and heat capacities. The prediction’s accuracy compares well with the quality of the training data for virtually any stoichiometric inorganic crystalline material, reciprocating the available thermomechanical experimental data. The universality of the approach is attributed to the construction of the descriptors: Property-Labelled Materials Fragments. The representations require only minimal structural input allowing straightforward implementations of simple heuristic design rules.