Jonathan Schmidt^1*, Love Pettersson^2*, Claudio Verdozzi^2*, Silvana Botti^3*, Miguel Marques^1*

1 Institut für Physik, Martin-Luther-Universität Halle-Wittenberg, Halle 06120, Germany

2 Department of Physics, Lund University, Box 118, 221 00 Lund, Sweden

3 Institut für Festkörpertheorie und -optik and European Theoretical Spectroscopy Facility Friedrich-Schiller-Universität Jena, D-07743 Jena, Germany

* Corresponding authors emails: jonathan.schmidt@student.uni-halle.de, lo4642pe-s@student.lu.se, claudio.verdozzi@teorfys.lu.se, silvana.botti@uni-jena.de, miguel.marques@physik.uni-halle.de

DOI10.24435/materialscloud:j9-bf [version v2]

Publication date: Dec 16, 2021

How to cite this record

Jonathan Schmidt, Love Pettersson, Claudio Verdozzi, Silvana Botti, Miguel Marques, Crystal-graph attention networks for the prediction of stable materials, Materials Cloud Archive 2021.222 (2021), doi: 10.24435/materialscloud:j9-bf.

Description

Graph neural networks have enjoyed great success in the prediction of material properties for both molecules and crystals. These networks typically use the atomic positions (usually expanded in a Gaussian basis) and the atomic species as input. Unfortunately, this information is in general not available when predicting new materials, for which the precise geometrical information is unknown. In this work, we circumvent this problem by predicting the thermodynamic stability of crystal structures without using the knowledge of the precise bond distances. We replace this information with embeddings of graph distances, allowing our networks to be used directly in high-throughput studies based on both composition and crystal structure prototype. Using these embeddings, we combine the newest developments in graph neural networks and apply them to the prediction of the distances to the convex hull. To train these networks, we curate a dataset of over 2 million density-functional calculations of crystals with consistent calculation parameters from various sources. The new dataset allows for the creation of a high quality convex hull and a large scale transfer learning approach. We apply the resulting model to the high-throughput search of 15 million tetragonal perovskites of composition ABCD2. As a result, we identify several thousand potentially stable compounds and demonstrate that transfer learning from the newly curated dataset reduces the required training data by 50%.

Materials Cloud sections using this data

Files

File name	Size	Description
aflow_ids.tar.xz MD5md5:6e48440ed4f8b17d3a7f5efd7e634870	18.8 MiB	Compressed json file containing the IDs of all materials from aflow used in the work as well as space group, distance to the convex hull and formation energies.
matproj_ids.tar.xz MD5md5:13e2b3859a5d4cda7179f6524d10cfb6	1.8 MiB	Compressed json file containing the IDs of all materials from the materials project used in the work as well as space group, distance to the convex hull and formation energies.
computed_entries.tar.gz MD5md5:1adc932b253db4a9045262d109d5a77b	267.3 MiB	.tar.gz archive of computed_structure_entries.json containing the information to build computed structure entries (pymatgen object containing knowledge of the structure, energy and some calculation parameters) for all calculations from our group used in this work
predicted_mixed_perovskites_computed_entries.tar.gz MD5md5:e473354906bf6e77d478d650d50b9ab3	10.2 MiB	.tar.gz archive of computed_structure_entries.json containing the information to build computed structure entries (pymatgen object containing knowledge of the structure, energy and some calculation parameters) for the predicted mixed perovskites in the paper
README.txt MD5md5:93bd3c3bf87e0f9cafe9e0be01bc8c1e	2.0 KiB	README

License

Files and data are licensed under the terms of the following license: Creative Commons Attribution 4.0 International.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

External references

Keywords

density-functional theory high-throughput PBE