×

Indexed by

Efficient Training of ANN Potentials by Including Atomic Forces via Taylor Expansion and Application to Water and a Transition-Metal Oxide

April Cooper1, Johannes Kästner1, Alexander Urban2, Nongnuch Artrith2*

1 Institute for Theoretical Chemistry, University of Stuttgart, 70569 Stuttgart, Germany

2 Department of Chemical Engineering, Columbia University, 500 West 120th Street, New York, NY 10027, USA

* Corresponding authors emails: nartrith@atomistic.net
DOI10.24435/materialscloud:2020.0037/v1 [version v1]

Publication date: Apr 14, 2020

How to cite this record

April Cooper, Johannes Kästner, Alexander Urban, Nongnuch Artrith, Efficient Training of ANN Potentials by Including Atomic Forces via Taylor Expansion and Application to Water and a Transition-Metal Oxide, Materials Cloud Archive 2020.0037/v1 (2020), doi: 10.24435/materialscloud:2020.0037/v1.

Description

This data set contains atomic structures of water clusters, bulk water and rock-salt Li8Mo2Ni7Ti7O32 in the XCrySDen [1] structure format (XSF), and total energies are included as additional meta information. The extended XSF format is compatible with the atomic energy network (aenet) package [2,3] for artificial neural network potential construction and application. The structures were generated using ab initio molecular dynamics (AIMD) simulations performed with the Vienna Ab Initio Simulation Package (VASP) [4,5] and projector-augmented wave (PAW) [6] pseudopontentials. For the bulk water system the revised Perdew-Burke-Ernzerhof density functional [7] with the Grimme D3 van-der-Waals correction [8] (revPBE+D3) was used. The AIMD simulations of the Li-Mo-Ni-Ti-O system employed the strongly constrained and appropriately normed (SCAN) semilocal density functional [9]. For both periodic systems, the plane-wave cutoff was 400 eV, and Gamma-point only k-point meshes were employed. A time step of 1 fs was used for the integration of the equation of motion, and a Nosé-Hoover thermostat [10,11] was used to maintain the temperature at 400 K. The energies and interatomic forces of the water cluster structures were calculated using the BLYP density functional [12,13] with additional Grimme D3 correction as implemented in the Turbomole software [14]. Further details can be found in the associated research article. [1] A. Kokalj, J. Mol. Graphics Modell. 17, 176–179 (1999). [2] N. Artrith, A. Urban, Comput. Mater. Sci. 114, 135–150 (2016). [3] N. Artrith, A. Urban, G. Ceder, Phys. Rev. B 96, 014112 (2017). [4] G. Kresse, J. Furthmüller, Phys. Rev. B 54, 11169–11186 (1996). [5] Kresse, J. Furthmüller, Comput. Mater. Sci. 6, 15–50 (1996). [6] P. E. Blöchl, Phys. Rev. B 50, 17953–17979 (1994). [7] Y. Zhang, W. Yang, Phys. Rev. Lett. 80, 890–890 (1998). [8] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 132, 154104 (2010). [9] J. Sun, A. Ruzsinszky, J. Perdew, Phys. Rev. Lett. 115, 036402 (2015). [10] S. Nosé, J. Chem. Phys. 81, 511–519 (1984). [11] W. G. Hoover, Phys. Rev. A 31, 1695–1697 (1985). [12] A. D. Becke, Phys. Rev. A 38, 3098–3100 (1988). [13] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 37, 785–789 (1988). [14] F. Furche, R. Ahlrichs, C. Hättig, W. Klopper, M. Sierka, F. Weigend, WIREs Comput Mol Sci 4, 91–100 (2014).

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive record.

Files

File name Size Description
liquid-64water-AIMD-RPBE-D3-validation-data.tar.bz2
MD5md5:dc02af62b70d980cfcfba8146788fb47
10.8 MiB Independent validation set with additional structures of bulk liquid water.
LMNTO-SCAN-validation-data.tar.bz2
MD5md5:50b330b981999815b2e0cc7af2a44292
2.6 MiB Independent validation set with additional LMNTO structures.
water-clusters-BLYP-D3.tar.bz2
MD5md5:faecb1334d286e988df0bc6c475982d3
2.3 MiB Structures of water clusters.
README.txt
MD5md5:ced800fe8a59976c484fbcb6950b1e6a
453.2 KiB README.txt
liquid-64water-AIMD-RPBE-D3-train-test-data.tar.bz2
MD5md5:7faa5686b5263fa8745c2bef814220e4
3.6 MiB Structures of bulk liquid water used for training and testing ANN potentials.
LMNTO-SCAN-train-data.tar.bz2
MD5md5:2261f39e33d7844175e26d34f2b1ec86
1.2 MiB LMNTO structures used for training ANN potentials.

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

Journal reference (Peer-reviewed version of the article that makes use of this data set.)
Preprint (Preprint of the article that makes use of this data set.)

Keywords

VASP lithium transition metal oxide aenet water AIMD

Version history:

2020.0037/v1 (version v1) [This version] Apr 14, 2020 DOI10.24435/materialscloud:2020.0037/v1