Deep neural networks for accurate predictions of crystal stability.
basic_science · Level V
Where this comes from
- Record sourced from PubMed, PMID 30228262.
- Also identified by DOI 10.1038/s41467-018-06322-x and PMC identifier 6143552.
- Licence recorded as CC BY.
- The licence permits redistribution, so the abstract is shown in full and the full text is available from the publisher.
Abstract
Predicting the stability of crystals is one of the central problems in materials science. Today, density functional theory (DFT) calculations remain comparatively expensive and scale poorly with system size. Here we show that deep neural networks utilizing just two descriptors-the Pauling electronegativity and ionic radii-can predict the DFT formation energies of C<sub>3</sub>A<sub>2</sub>D<sub>3</sub>O<sub>12</sub> garnets and ABO<sub>3</sub> perovskites with low mean absolute errors (MAEs) of 7-10 meV atom<sup>-1</sup> and 20-34 meV atom<sup>-1</sup>, respectively, well within the limits of DFT accuracy. Further extension to mixed garnets and perovskites with little loss in accuracy can be achieved using a binary encoding scheme, addressing a critical gap in the extension of machine-learning models from fixed stoichiometry crystals to infinite universe of mixed-species crystals. Finally, we demonstrate the potential of these models to rapidly transverse vast chemical spaces to accurately identify stable compositions, accelerating the discovery of novel materials with potentially superior properties.