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

Abstract

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.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 68

1. 1.

   McCammon JA, Gelin BR, Karplus M. Dynamics of folded proteins. Nature. 1977;267:585-590.

   Not in our library10.1038/267585a0

2. 2.

   Jorgensen WL, Ravimohan C. Monte Carlo simulation of differences in free energies of hydration. J. Chem. Phys. 1985;83:3050-3054.

   Not in our library10.1063/1.449208

3. 3.

   Behler J. Perspective: machine learning potentials for atomistic simulations. J. Chem. Phys. 2016;145:170901.

   Not in our library10.1063/1.4966192

4. 4.

   Botu V, Batra R, Chapman J, Ramprasad R. Machine learning force fields: construction, validation, and outlook. J. Phys. Chem. C. 2017;121:511-522.

   Not in our library10.1021/acs.jpcc.6b10908

5. 5.

   Deringer VL, Caro MA, Csányi G. Machine learning interatomic potentials as emerging tools for materials science. Adv. Mater. 2019;31:1902765.

   Not in our library10.1002/adma.201902765

6. 6.
   Bypassing the Kohn-Sham equations with machine learning · 2017

   No notice
7. 7.

   Noé F, Tkatchenko A, Müller K-R, Clementi C. Machine learning for molecular simulation. Annu. Rev. Phys. Chem. 2020;71:361-390.

   Not in our library10.1146/annurev-physchem-042018-052331

8. 8.

   Blank TB, Brown SD, Calhoun AW, Doren DJ. Neural network models of potential energy surfaces. J. Chem. Phys. 1995;103:4129-4137.

   Not in our library10.1063/1.469597

9. 9.

   Behler J, Parrinello M. Generalized neural-network representation of high-dimensional potential-energy surfaces. Phys. Rev. Lett. 2007;98:146401.

   Not in our library10.1103/physrevlett.98.146401

10. 10.

    Schütt KT, Sauceda HE, Kindermans P-J, Tkatchenko A, Müller K-R. SchNet-A deep learning architecture for molecules and materials. J. Chem. Phys. 2018;148:241722.

    Not in our library10.1063/1.5019779

11. 11.

    Unke OT, Meuwly M. PhysNet: a neural network for predicting energies, forces, dipole moments, and partial charges. J. Chem. Theory Comput. 2019;15:3678-3693.

    Not in our library10.1021/acs.jctc.9b00181

12. 12.

    Smith JS, Isayev O, Roitberg AE. ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost. Chem. Sci. 2017;8:3192-3203.

    Not in our library10.1039/c6sc05720a

13. 13.

    Bartók AP, Payne MC, Kondor R, Csányi G. Gaussian approximation potentials: the accuracy of quantum mechanics, without the electrons. Phys. Rev. Lett. 2010;104:136403.

    Not in our library10.1103/physrevlett.104.136403

14. 14.

    Shapeev AV. Moment tensor potentials: a class of systematically improvable interatomic potentials. Multiscale Model. Simul. 2016;14:1153-1173.

    Not in our library10.1137/15m1054183

15. 15.

    Thompson AP, Swiler LP, Trott CR, Foiles SM, Tucker GJ. Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials. J. Chem. Phys. 2015;285:316-330.

    Not in our library

16. 16.

    Drautz R. Atomic cluster expansion for accurate and transferable interatomic potentials. Phys. Rev. B. 2019;99:014104.

    Not in our library10.1103/physrevb.99.014104

17. 17.

    Balabin RM, Lomakina EI. Support vector machine regression (LS-SVM)-an alternative to artificial neural networks (ANNs) for the analysis of quantum chemistry data? Phys. Chem. Chem. Phys. 2011;13:11710.

    Not in our library10.1039/c1cp00051a

18. 18.

    Behler J. Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations. Phys. Chem. Chem. Phys. 2011;13:17930-17955.

    Not in our library10.1039/c1cp21668f

19. 19.

    Handley CM, Popelier PL. Potential energy surfaces fitted by artificial neural networks. J. Phys. Chem. A. 2010;114:3371-3383.

    Not in our library10.1021/jp9105585

20. 20.

    Prodan E, Kohn W. Nearsightedness of electronic matter. Proc. Natl. Acad. Sci. 2005;102:11635-11638.

    Not in our library10.1073/pnas.0505436102

21. 21.

    Deng Z, Chen C, Li X-G, Ong SP. An electrostatic spectral neighbor analysis potential for lithium nitride. NPJ Comput. Mater. 2019;5:75.

    Not in our library10.1038/s41524-019-0212-1

22. 22.

    Artrith N, Morawietz T, Behler J. High-dimensional neural-network potentials for multicomponent systems: applications to zinc oxide. Phys. Rev. B. 2011;83:153101.

    Not in our library10.1103/physrevb.83.153101

23. 23.

    Morawietz T, Sharma V, Behler J. A neural network potential-energy surface for the water dimer based on environment-dependent atomic energies and charges. J. Chem. Phys. 2012;136:064103.

    Not in our library10.1063/1.3682557

24. 24.

    Yao K, Herr JE, Toth DW, Mckintyre R, Parkhill J. The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics. Chem. Sci. 2018;9:2261-2269.

    Not in our library10.1039/c7sc04934j

25. 25.

    Bereau T, Andrienko D, Von Lilienfeld OA. Transferable atomic multipole machine learning models for small organic molecules. J. Chem. Theory Comput. 2015;11:3225-3233.

    Not in our library10.1021/acs.jctc.5b00301

26. 26.

    Hoshino T, et al. First-principles calculations for vacancy formation energies in Cu and Al; non-local effect beyond the LSDA and lattice distortion. Comp. Mat. Sci. 1999;14:56.

    Not in our library10.1016/s0927-0256(98)00072-x

27. 27.

    Parsaeifard, B., Finkler, J. A. & Goedecker, S. Detecting non-local effects in the electronic structure of a simple covalent system with machine learning methods, arXiv:2008.11277 (2020).

    Not in our library

28. 28.

    Rappe AK, Goddard WA. Charge equilibration for molecular dynamics simulations. J. Phys. Chem. 1991;95:3358.

    Not in our library10.1021/j100161a070

29. 29.

    van Duin ACT, Dasgupta S, Lorant F, Goddard WA. ReaxFF: a reactive force field for hydrocarbons. J. Phys. Chem. A. 2001;105:9396-9409.

    Not in our library10.1021/jp004368u

30. 30.

    Zhou XW, Wadley HNG. A charge transfer ionic–embedded atom method potential for the O–Al–Ni–Co–Fe system. J. Phys.: Condens. Matter. 2005;17:3619.

    Not in our library

31. 31.

    Gasteiger J, Marsili M. Iterative partial equalization of orbital electronegativity–a rapid access to atomic charges. Tetrahedron. 1980;36:3219-3228.

    Not in our library10.1016/0040-4020(80)80168-2

32. 32.

    Ghasemi SA, Hofstetter A, Saha S, Goedecker S. Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network. Phys. Rev. B. 2015;92:045131.

    Not in our library10.1103/physrevb.92.045131

33. 33.

    Faraji S, et al. High accuracy and transferability of a neural network potential through charge equilibration for calcium fluoride. Phys. Rev. B. 2017;95:104105.

    Not in our library10.1103/physrevb.95.104105

34. 34.

    Amsler, M. et al. FLAME: a library of atomistic modeling environments. Comput. Phys. Commun.256, 107415 (2020)

    Not in our library

35. 35.

    Hafizi R, Ghasemi SA, Hashemifar SJ, Akbarzadeh H. A neural-network potential through charge equilibration for WS2: From clusters to sheets. J. Chem. Phys. 2017;147:234306.

    Not in our library10.1063/1.5003904

36. 36.

    Faraji S, Ghasemi SA, Parsaeifard B, Goedecker S. Surface reconstructions and premelting of the (100) CaF2 surface. Phys. Chem. Chem. Phys. 2019;21:16270-16281.

    Not in our library10.1039/c9cp02213a

37. 37.

    Rasoulkhani R, et al. Energy landscape of ZnO clusters and low-density polymorphs. Phys. Rev. B. 2017;96:064108.

    Not in our library10.1103/physrevb.96.064108

38. 38.

    Xie X, Persson KA, Small DW. Incorporating electronic information into machine learning potential energy surfaces via approaching the ground-state electronic energy as a function of atom-based electronic populations. J. Chem. Theory Comput. 2020;16:4256-4270.

    Not in our library10.1021/acs.jctc.0c00217

39. 39.

    Zubatyuk, R., Smith, J., Nebgen, B.T., Tretiak, S. & Isayev, O. Teaching a neural network to attach and detach electrons from molecules, ChemRxiv 12725276.v1 (2020).

    Not in our library10.1038/s41467-021-24904-0

40. 40.

    Hirshfeld FL. Bonded-atom fragments for describing molecular charge densities. Theor. Chim. Acta. 1977;44:129-138.

    Not in our library10.1007/bf00549096

Cited by papers we hold 6

• AIMNet2: a neural network potential to meet your neutral, charged, organic, and elemental-organic needs · 2025
  No notice
• SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentials · 2023
  No notice
• Machine Learning Interatomic Potentials and Long-Range Physics · 2023
  No notice
• Gaussian Process Regression for Materials and Molecules · 2021
  No notice
• Implicit Solvation Methods for Catalysis at Electrified Interfaces · 2021
  No notice
• SpookyNet: Learning force fields with electronic degrees of freedom and nonlocal effects · 2021
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.