Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture

Park, Cheol Woo; Kornbluth, Mordechai; Vandermause, Jonathan; Wolverton, Chris; Kozinsky, Boris; Mailoa, Jonathan P.

doi:10.1038/s41524-021-00543-3

Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture

Journal Article · 21 May 2021 · npj Computational Materials

DOI:https://doi.org/10.1038/s41524-021-00543-3· OSTI ID:1848250

Park, Cheol Woo ^[1]; ; ; ; ;

OSTI

Abstract

Recently, machine learning (ML) has been used to address the computational cost that has been limiting ab initio molecular dynamics (AIMD). Here, we present GNNFF, a graph neural network framework to directly predict atomic forces from automatically extracted features of the local atomic environment that are translationally-invariant, but rotationally-covariant to the coordinate of the atoms. We demonstrate that GNNFF not only achieves high performance in terms of force prediction accuracy and computational speed on various materials systems, but also accurately predicts the forces of a large MD system after being trained on forces obtained from a smaller system. Finally, we use our framework to perform an MD simulation of Li₇P₃S₁₁, a superionic conductor, and show that resulting Li diffusion coefficient is within 14% of that obtained directly from AIMD. The high performance exhibited by GNNFF can be easily generalized to study atomistic level dynamics of other material systems.

Research Organization:: Robert Bosch LLC, Farmington Hills, MI (United States); UT-Battelle LLC/ORNL, Oak Ridge, TN (United States)

Sponsoring Organization:: USDOE Advanced Research Projects Agency - Energy (ARPA-E)

DOE Contract Number:: AR0000775; AC05-00OR22725;

OSTI ID:: 1848250

Journal Information:: npj Computational Materials, Journal Name: npj Computational Materials Journal Issue: 1 Vol. 7; ISSN 2057-3960

Publisher:: Nature Publishing Group

Country of Publication:: United States

Language:: English

References (47)

Neural network and ReaxFF comparison for Au properties: Comparison of ReaxFF and BPNN Potentials Boes, Jacob R.; Groenenboom, Mitchell C.; Keith, John A. International Journal of Quantum Chemistry, Vol. 116, Issue 13 https://doi.org/10.1002/qua.25115	journal	March 2016
Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set Kresse, G.; Furthmüller, J. Computational Materials Science, Vol. 6, Issue 1, p. 15-50 https://doi.org/10.1016/0927-0256(96)00008-0	journal	July 1996
Super-Ionic Conduction in Solid-State Li ₇ P ₃ S ₁₁ -Type Sulfide Electrolytes Chang, Donghee; Oh, Kyungbae; Kim, Sung Joo Chemistry of Materials, Vol. 30, Issue 24 https://doi.org/10.1021/acs.chemmater.8b03000	journal	November 2018
Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals Chen, Chi; Ye, Weike; Zuo, Yunxing Chemistry of Materials, Vol. 31, Issue 9 https://doi.org/10.1021/acs.chemmater.9b01294	journal	April 2019
High-Dimensional Neural Network Potentials for Organic Reactions and an Improved Training Algorithm Gastegger, Michael; Marquetand, Philipp Journal of Chemical Theory and Computation, Vol. 11, Issue 5 https://doi.org/10.1021/acs.jctc.5b00211	journal	April 2015
Automated Discovery and Refinement of Reactive Molecular Dynamics Pathways Wang, Lee-Ping; McGibbon, Robert T.; Pande, Vijay S. Journal of Chemical Theory and Computation, Vol. 12, Issue 2 https://doi.org/10.1021/acs.jctc.5b00830	journal	January 2016
Insights into the Performance Limits of the Li ₇ P ₃ S ₁₁ Superionic Conductor: A Combined First-Principles and Experimental Study Chu, Iek-Heng; Nguyen, Han; Hy, Sunny ACS Applied Materials & Interfaces, Vol. 8, Issue 12 https://doi.org/10.1021/acsami.6b00833	journal	March 2016
ReaxFF: A Reactive Force Field for Hydrocarbons van Duin, Adri C. T.; Dasgupta, Siddharth; Lorant, Francois The Journal of Physical Chemistry A, Vol. 105, Issue 41 https://doi.org/10.1021/jp004368u	journal	October 2001
Potential Energy Surfaces Fitted by Artificial Neural Networks Handley, Chris M.; Popelier, Paul L. A. The Journal of Physical Chemistry A, Vol. 114, Issue 10 https://doi.org/10.1021/jp9105585	journal	March 2010
Discovering chemistry with an ab initio nanoreactor Wang, Lee-Ping; Titov, Alexey; McGibbon, Robert Nature Chemistry, Vol. 6, Issue 12 https://doi.org/10.1038/nchem.2099	journal	November 2014
Quantum-chemical insights from deep tensor neural networks Schütt, Kristof T.; Arbabzadah, Farhad; Chmiela, Stefan Nature Communications, Vol. 8, Issue 1 https://doi.org/10.1038/ncomms13890	journal	January 2017
A general-purpose machine learning framework for predicting properties of inorganic materials Ward, Logan; Agrawal, Ankit; Choudhary, Alok npj Computational Materials, Vol. 2, Issue 1 https://doi.org/10.1038/npjcompumats.2016.28	journal	August 2016
Graph dynamical networks for unsupervised learning of atomic scale dynamics in materials Xie, Tian; France-Lanord, Arthur; Wang, Yanming Nature Communications, Vol. 10, Issue 1 https://doi.org/10.1038/s41467-019-10663-6	journal	June 2019
A universal strategy for the creation of machine learning-based atomistic force fields Huan, Tran Doan; Batra, Rohit; Chapman, James npj Computational Materials, Vol. 3, Issue 1 https://doi.org/10.1038/s41524-017-0042-y	journal	September 2017
On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events Vandermause, Jonathan; Torrisi, Steven B.; Batzner, Simon npj Computational Materials, Vol. 6, Issue 1 https://doi.org/10.1038/s41524-020-0283-z	journal	March 2020
A fast neural network approach for direct covariant forces prediction in complex multi-element extended systems Mailoa, Jonathan P.; Kornbluth, Mordechai; Batzner, Simon Nature Machine Intelligence, Vol. 1, Issue 10 https://doi.org/10.1038/s42256-019-0098-0	journal	September 2019
Quantum chemistry structures and properties of 134 kilo molecules Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias Scientific Data, Vol. 1, Issue 1 https://doi.org/10.1038/sdata.2014.22	journal	August 2014
ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost Smith, J. S.; Isayev, O.; Roitberg, A. E. Chemical Science, Vol. 8, Issue 4 https://doi.org/10.1039/C6SC05720A	journal	January 2017
Molecular dynamics simulation of O2 sticking on Pt(111) using the ab initio based ReaxFF reactive force field Valentini, Paolo; Schwartzentruber, Thomas E.; Cozmuta, Ioana The Journal of Chemical Physics, Vol. 133, Issue 8 https://doi.org/10.1063/1.3469810	journal	August 2010
Atom-centered symmetry functions for constructing high-dimensional neural network potentials Behler, Jörg The Journal of Chemical Physics, Vol. 134, Issue 7 https://doi.org/10.1063/1.3553717	journal	February 2011
Ab initio molecular dynamics simulation of the solvation and transport of hydronium and hydroxyl ions in water Tuckerman, M.; Laasonen, K.; Sprik, M. The Journal of Chemical Physics, Vol. 103, Issue 1 https://doi.org/10.1063/1.469654	journal	July 1995
Study of Li atom diffusion in amorphous Li3PO4 with neural network potential Li, Wenwen; Ando, Yasunobu; Minamitani, Emi The Journal of Chemical Physics, Vol. 147, Issue 21 https://doi.org/10.1063/1.4997242	journal	December 2017
SchNet – A deep learning architecture for molecules and materials Schütt, K. T.; Sauceda, H. E.; Kindermans, P. -J. The Journal of Chemical Physics, Vol. 148, Issue 24 https://doi.org/10.1063/1.5019779	journal	June 2018
Review of force fields and intermolecular potentials used in atomistic computational materials research Harrison, Judith A.; Schall, J. David; Maskey, Sabina Applied Physics Reviews, Vol. 5, Issue 3 https://doi.org/10.1063/1.5020808	journal	September 2018
Long-range Finnis-Sinclair potentials for f.c.c. metallic alloys Rafii-Tabar, H.; Sulton, A. P. Philosophical Magazine Letters, Vol. 63, Issue 4 https://doi.org/10.1080/09500839108205994	journal	April 1991
The classical equation of state of gaseous helium, neon and argon Buckingham, R. A. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, Vol. 168, Issue 933, p. 264-283 https://doi.org/10.1098/rspa.1938.0173	journal	October 1938
Embedded-atom method: Derivation and application to impurities, surfaces, and other defects in metals Daw, Murray S.; Baskes, M. I. Physical Review B, Vol. 29, Issue 12 https://doi.org/10.1103/PhysRevB.29.6443	journal	June 1984
Computer simulation of local order in condensed phases of silicon Stillinger, Frank H.; Weber, Thomas A. Physical Review B, Vol. 31, Issue 8 https://doi.org/10.1103/PhysRevB.31.5262	journal	April 1985
Empirical chemical pseudopotential theory of molecular and metallic bonding Abell, G. C. Physical Review B, Vol. 31, Issue 10 https://doi.org/10.1103/PhysRevB.31.6184	journal	May 1985
New empirical approach for the structure and energy of covalent systems Tersoff, J. Physical Review B, Vol. 37, Issue 12 https://doi.org/10.1103/PhysRevB.37.6991	journal	April 1988
Interaction potential for SiO 2 : A molecular-dynamics study of structural correlations Vashishta, P.; Kalia, Rajiv K.; Rino, José P. Physical Review B, Vol. 41, Issue 17 https://doi.org/10.1103/PhysRevB.41.12197	journal	June 1990
Empirical potential for hydrocarbons for use in simulating the chemical vapor deposition of diamond films Brenner, Donald W. Physical Review B, Vol. 42, Issue 15 https://doi.org/10.1103/PhysRevB.42.9458	journal	November 1990
Modified embedded-atom potentials for cubic materials and impurities Baskes, M. I. Physical Review B, Vol. 46, Issue 5 https://doi.org/10.1103/PhysRevB.46.2727	journal	August 1992
Ab initiomolecular dynamics for liquid metals Kresse, G.; Hafner, J. Physical Review B, Vol. 47, Issue 1, p. 558-561 https://doi.org/10.1103/PhysRevB.47.558	journal	January 1993
Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set Kresse, G.; Furthmüller, J. Physical Review B, Vol. 54, Issue 16, p. 11169-11186 https://doi.org/10.1103/PhysRevB.54.11169	journal	October 1996
From ultrasoft pseudopotentials to the projector augmented-wave method Kresse, G.; Joubert, D. Physical Review B, Vol. 59, Issue 3, p. 1758-1775 https://doi.org/10.1103/PhysRevB.59.1758	journal	January 1999
Modified embedded atom method potential for Al, Si, Mg, Cu, and Fe alloys Jelinek, B.; Groh, S.; Horstemeyer, M. F. Physical Review B, Vol. 85, Issue 24 https://doi.org/10.1103/PhysRevB.85.245102	journal	June 2012
Combinatorial screening for new materials in unconstrained composition space with machine learning Meredig, B.; Agrawal, A.; Kirklin, S. Physical Review B, Vol. 89, Issue 9 https://doi.org/10.1103/PhysRevB.89.094104	journal	March 2014
Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the Electrons Bartók, Albert P.; Payne, Mike C.; Kondor, Risi Physical Review Letters, Vol. 104, Issue 13 https://doi.org/10.1103/PhysRevLett.104.136403	journal	April 2010
Deep Potential Molecular Dynamics: A Scalable Model with the Accuracy of Quantum Mechanics Zhang, Linfeng; Han, Jiequn; Wang, Han Physical Review Letters, Vol. 120, Issue 14 https://doi.org/10.1103/PhysRevLett.120.143001	journal	April 2018
Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties Xie, Tian; Grossman, Jeffrey C. Physical Review Letters, Vol. 120, Issue 14 https://doi.org/10.1103/PhysRevLett.120.145301	journal	April 2018
Empirical Interatomic Potential for Carbon, with Applications to Amorphous Carbon Tersoff, J. Physical Review Letters, Vol. 61, Issue 25 https://doi.org/10.1103/PhysRevLett.61.2879	journal	December 1988
Generalized Gradient Approximation Made Simple Perdew, John P.; Burke, Kieron; Ernzerhof, Matthias Physical Review Letters, Vol. 77, Issue 18, p. 3865-3868 https://doi.org/10.1103/PhysRevLett.77.3865	journal	October 1996
Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces Behler, Jörg; Parrinello, Michele Physical Review Letters, Vol. 98, Issue 14 https://doi.org/10.1103/PhysRevLett.98.146401	journal	April 2007
Developing an improved crystal graph convolutional neural network framework for accelerated materials discovery Park, Cheol Woo; Wolverton, Chris Physical Review Materials, Vol. 4, Issue 6 https://doi.org/10.1103/PhysRevMaterials.4.063801	journal	June 2020
Machine learning of accurate energy-conserving molecular force fields Chmiela, Stefan; Tkatchenko, Alexandre; Sauceda, Huziel E. Science Advances, Vol. 3, Issue 5 https://doi.org/10.1126/sciadv.1603015	journal	May 2017
Long Short-Term Memory Hochreiter, Sepp; Schmidhuber, Jürgen Neural Computation, Vol. 9, Issue 8 https://doi.org/10.1162/neco.1997.9.8.1735	journal	November 1997

Similar Records

Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture

Journal Article · 2021 · npj Computational Materials · OSTI ID:1784359

Dynamic Monkey Bar Mechanism of Superionic Li‐ion Transport in LiTaCl ₆

Journal Article · 2023 · Angewandte Chemie · OSTI ID:2251523

Related Subjects

Chemistry
Materials Science

Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture

Citation Formats

References (47)

Similar Records

Related Subjects