Rapid discovery of stable materials by coordinate-free coarse graining

Abstract

A fundamental challenge in materials science pertains to elucidating the relationship between stoichiometry, stability, structure, and property. Recent advances have shown that machine learning can be used to learn such relationships, allowing the stability and functional properties of materials to be accurately predicted. However, most of these approaches use atomic coordinates as input and are thus bottlenecked by crystal structure identification when investigating previously unidentified materials. Our approach solves this bottleneck by coarse-graining the infinite search space of atomic coordinates into a combinatorially enumerable search space. The key idea is to use Wyckoff representations, coordinate-free sets of symmetry-related positions in a crystal, as the input to a machine learning model. Our model demonstrates exceptionally high precision in finding unknown theoretically stable materials, identifying 1569 materials that lie below the known convex hull of previously calculated materials from just 5675 ab initio calculations. Our approach opens up fundamental advances in computational materials discovery.

Recent advances and applications of deep learning methods in materials science

“…Nevertheless, the model errors tend to be much higher than graph-based models. Similar coarse-grained representation using Wyckoff representation was also used by Goodall et al 128 . Alternatively, Zuo et al 129 started from the hypothetical structures without precise atom positions, and used a Bayesian optimization method coupled with a MEGNet energy model as an energy evaluator to perform direct structural relaxation.…”

Heterogeneous Catalysis in Grammar School

“…A powerful catalyst string representation must therefore be able to handle nonlocality and an enormous variety of elemental compositions. A recently proposed approach to overcome these challenges is to use a coordinate-free representation based on crystallographic Wyckoff positions . By avoiding the definition of bonds between atoms altogether, this is potentially a viable route toward powerful string representations of catalysts.…”

Data-augmentation for graph neural network learning of the relaxed energies of unrelaxed structures

“…The significant errors for unrelaxed structures are due to the limited sampling of the complex multi-dimensional configuration space of the potential energy surface (PES), with the relaxed structures only describing the minima of the surface. Since unrelaxed structures are not located at these minima, predicting a structure's formation energy at an unrelaxed configuration is a qualitatively different task 31 . To sample configurations near the minima of the PES, Smith et al 32 applied a data augmentation technique known as normal mode sampling to a large dataset of molecular structures, resulting in a high fidelity neural network potential.…”