Published November 4, 2020 | Version v1
Dataset Open

A fourth-generation high-dimensional neural network potential with accurate electrostatics including non-local charge transfer

  • 1. Institut für Physikalische Chemie, Theoretische Chemie, Universität Göttingen, Tammannstraße 6, 37077, Göttingen, Germany
  • 2. Department of Physics, Universität Basel, Klingelbergstrasse 82, 4056 Basel, Switzerland

* Contact person

Description

Machine learning potentials have become an important tool for atomistic simulations in many fields, from chemistry via molecular biology to materials science. Most of the established methods, however, rely on local properties and are thus unable to take global changes in the electronic structure into account, which result from long-range charge transfer or different charge states. In this work we overcome this limitation by introducing a fourth-generation high-dimensional neural network potential that combines a charge equilibration scheme employing environment-dependent atomic electronegativities with accurate atomic energies. The method, which is able to correctly describe global charge distributions in arbitrary systems, yields much improved energies and substantially extends the applicability of modern machine learning potentials. This is demonstrated for a series of systems representing typical scenarios in chemistry and materials science that are incorrectly described by current methods, while the fourth-generation neural network potential is in excellent agreement with electronic structure calculations.

Files

File preview

files_description.md

All files

Files (37.9 MiB)

Name Size
md5:832437d3cd7102d0a601fba567d7c9b8
233 Bytes Preview Download
md5:0aa11ea305b3af6a47bb3c577c2729bc
37.9 MiB Download
md5:8a7eb0061688d4014848bc37009210f1
2.5 KiB Preview Download

References

Preprint (Paper in which the data is used)
T. W. Ko, J. A. Finkler, S. Goedecker, J. Behler, arXiv:2009.06484 (2020).