Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt
Abstract
Abstract Atomistic modeling of chemically reactive systems has so far relied on either expensive ab initio methods or bond-order force fields requiring arduous parametrization. Here, we describe a Bayesian active learning framework for autonomous “on-the-fly” training of fast and accurate reactive many-body force fields during molecular dynamics simulations. At each time-step, predictive uncertainties of a sparse Gaussian process are evaluated to automatically determine whether additional ab initio training data are needed. We introduce a general method for mapping trained kernel models onto equivalent polynomial models whose prediction cost is much lower and independent of the training set size. As a demonstration, we perform direct two-phase simulations of heterogeneous H 2 turnover on the Pt(111) catalyst surface at chemical accuracy. The model trains itself in three days and performs at twice the speed of a ReaxFF model, while maintaining much higher fidelity to DFT and excellent agreement with experiment.
- Authors:
- Publication Date:
- Research Org.:
- Harvard Univ., Cambridge, MA (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Basic Energy Sciences (BES); National Science Foundation (NSF)
- OSTI Identifier:
- 1885166
- Alternate Identifier(s):
- OSTI ID: 1904468
- Grant/Contract Number:
- SC0020128; SC0012573; AC02-05CH11231; 1808162; 2003725; DGE1745303
- Resource Type:
- Published Article
- Journal Name:
- Nature Communications
- Additional Journal Information:
- Journal Name: Nature Communications Journal Volume: 13 Journal Issue: 1; Journal ID: ISSN 2041-1723
- Publisher:
- Nature Publishing Group
- Country of Publication:
- United Kingdom
- Language:
- English
- Subject:
- 42 ENGINEERING; 71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS; computational methods; molecular dynamics
Citation Formats
Vandermause, Jonathan, Xie, Yu, Lim, Jin Soo, Owen, Cameron J., and Kozinsky, Boris. Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt. United Kingdom: N. p., 2022.
Web. doi:10.1038/s41467-022-32294-0.
Vandermause, Jonathan, Xie, Yu, Lim, Jin Soo, Owen, Cameron J., & Kozinsky, Boris. Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt. United Kingdom. https://doi.org/10.1038/s41467-022-32294-0
Vandermause, Jonathan, Xie, Yu, Lim, Jin Soo, Owen, Cameron J., and Kozinsky, Boris. Fri .
"Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt". United Kingdom. https://doi.org/10.1038/s41467-022-32294-0.
@article{osti_1885166,
title = {Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt},
author = {Vandermause, Jonathan and Xie, Yu and Lim, Jin Soo and Owen, Cameron J. and Kozinsky, Boris},
abstractNote = {Abstract Atomistic modeling of chemically reactive systems has so far relied on either expensive ab initio methods or bond-order force fields requiring arduous parametrization. Here, we describe a Bayesian active learning framework for autonomous “on-the-fly” training of fast and accurate reactive many-body force fields during molecular dynamics simulations. At each time-step, predictive uncertainties of a sparse Gaussian process are evaluated to automatically determine whether additional ab initio training data are needed. We introduce a general method for mapping trained kernel models onto equivalent polynomial models whose prediction cost is much lower and independent of the training set size. As a demonstration, we perform direct two-phase simulations of heterogeneous H 2 turnover on the Pt(111) catalyst surface at chemical accuracy. The model trains itself in three days and performs at twice the speed of a ReaxFF model, while maintaining much higher fidelity to DFT and excellent agreement with experiment.},
doi = {10.1038/s41467-022-32294-0},
journal = {Nature Communications},
number = 1,
volume = 13,
place = {United Kingdom},
year = {Fri Sep 02 00:00:00 EDT 2022},
month = {Fri Sep 02 00:00:00 EDT 2022}
}
https://doi.org/10.1038/s41467-022-32294-0
Works referenced in this record:
U1 snRNP regulates cancer cell migration and invasion in vitro
journal, January 2020
- Oh, Jung-Min; Venters, Christopher C.; Di, Chao
- Nature Communications, Vol. 11, Issue 1
Generalized Gradient Approximation Made Simple
journal, October 1996
- Perdew, John P.; Burke, Kieron; Ernzerhof, Matthias
- Physical Review Letters, Vol. 77, Issue 18, p. 3865-3868
Using metadynamics to build neural network potentials for reactive events: the case of urea decomposition in water
journal, March 2021
- Yang, Manyi; Bonati, Luigi; Polino, Daniela
- Catalysis Today
Heterogeneous Catalysis: A Central Science for a Sustainable Future
journal, March 2017
- Friend, Cynthia M.; Xu, Bingjun
- Accounts of Chemical Research, Vol. 50, Issue 3
Origins of structural and electronic transitions in disordered silicon
journal, January 2021
- Deringer, Volker L.; Bernstein, Noam; Csányi, Gábor
- Nature, Vol. 589, Issue 7840
Atomic cluster expansion for accurate and transferable interatomic potentials
journal, January 2019
- Drautz, Ralf
- Physical Review B, Vol. 99, Issue 1
The atomic simulation environment—a Python library for working with atoms
journal, June 2017
- Hjorth Larsen, Ask; Jørgen Mortensen, Jens; Blomqvist, Jakob
- Journal of Physics: Condensed Matter, Vol. 29, Issue 27
Adsorption of hydrogen on a Pt(111) surface
journal, February 1976
- Christmann, K.; Ertl, G.; Pignet, T.
- Surface Science, Vol. 54, Issue 2
On representing chemical environments
journal, May 2013
- Bartók, Albert P.; Kondor, Risi; Csányi, Gábor
- Physical Review B, Vol. 87, Issue 18
Fast Parallel Algorithms for Short-Range Molecular Dynamics
journal, March 1995
- Plimpton, Steve
- Journal of Computational Physics, Vol. 117, Issue 1
Less is more: Sampling chemical space with active learning
journal, June 2018
- Smith, Justin S.; Nebgen, Ben; Lubbers, Nicholas
- The Journal of Chemical Physics, Vol. 148, Issue 24
On-the-fly machine learning force field generation: Application to melting points
journal, July 2019
- Jinnouchi, Ryosuke; Karsai, Ferenc; Kresse, Georg
- Physical Review B, Vol. 100, Issue 1
Affinity and Selectivity Assessment of Covalent Inhibitors by Free Energy Calculations
journal, December 2020
- Mihalovits, Levente M.; Ferenczy, György G.; Keserű, György M.
- Journal of Chemical Information and Modeling, Vol. 60, Issue 12
A benchmark database for adsorption bond energies to transition metal surfaces and comparison to selected DFT functionals
journal, October 2015
- Wellendorff, Jess; Silbaugh, Trent L.; Garcia-Pintos, Delfina
- Surface Science, Vol. 640
Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials
journal, March 2015
- Thompson, A. P.; Swiler, L. P.; Trott, C. R.
- Journal of Computational Physics, Vol. 285
High-dimensional neural network potentials for metal surfaces: A prototype study for copper
journal, January 2012
- Artrith, Nongnuch; Behler, Jörg
- Physical Review B, Vol. 85, Issue 4
Quasielastic helium atom scattering measurements of microscopic diffusional dynamics of H and D on the Pt(111) surface
journal, July 1999
- Graham, Andrew P.; Menzel, Alexander; Toennies, J. Peter
- The Journal of Chemical Physics, Vol. 111, Issue 4
SchNet – A deep learning architecture for molecules and materials
journal, June 2018
- Schütt, K. T.; Sauceda, H. E.; Kindermans, P. -J.
- The Journal of Chemical Physics, Vol. 148, Issue 24
Atomistic Adsorption of Oxygen and Hydrogen on Platinum Catalysts by Hybrid Grand Canonical Monte Carlo/Reactive Molecular Dynamics
journal, April 2016
- Gai, Lili; Shin, Yun Kyung; Raju, Muralikrishna
- The Journal of Physical Chemistry C, Vol. 120, Issue 18
Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture
journal, May 2021
- Park, Cheol Woo; Kornbluth, Mordechai; Vandermause, Jonathan
- npj Computational Materials, Vol. 7, Issue 1
Moment Tensor Potentials: A Class of Systematically Improvable Interatomic Potentials
journal, January 2016
- Shapeev, Alexander V.
- Multiscale Modeling & Simulation, Vol. 14, Issue 3
Guidelines to Achieving High Selectivity for the Hydrogenation of α,β-Unsaturated Aldehydes with Bimetallic and Dilute Alloy Catalysts: A Review
journal, October 2020
- Luneau, Mathilde; Lim, Jin Soo; Patel, Dipna A.
- Chemical Reviews, Vol. 120, Issue 23
On the dynamics of H 2 adsorption on the Pt(111) surface: KRAUS and FRANK
journal, May 2017
- Kraus, Peter; Frank, Irmgard
- International Journal of Quantum Chemistry, Vol. 117, Issue 17
Development and testing of a general amber force field
journal, January 2004
- Wang, Junmei; Wolf, Romain M.; Caldwell, James W.
- Journal of Computational Chemistry, Vol. 25, Issue 9
Atomic cluster expansion: Completeness, efficiency and stability
journal, April 2022
- Dusson, Geneviève; Bachmayr, Markus; Csányi, Gábor
- Journal of Computational Physics, Vol. 454
Assessing the performance of recent density functionals for bulk solids
journal, April 2009
- Csonka, Gábor I.; Perdew, John P.; Ruzsinszky, Adrienn
- Physical Review B, Vol. 79, Issue 15
The resurgence of covalent drugs
journal, April 2011
- Singh, Juswinder; Petter, Russell C.; Baillie, Thomas A.
- Nature Reviews Drug Discovery, Vol. 10, Issue 4
First Principles Neural Network Potentials for Reactive Simulations of Large Molecular and Condensed Systems
journal, August 2017
- Behler, Jörg
- Angewandte Chemie International Edition, Vol. 56, Issue 42
sGDML: Constructing accurate and data efficient molecular force fields using machine learning
journal, July 2019
- Chmiela, Stefan; Sauceda, Huziel E.; Poltavsky, Igor
- Computer Physics Communications, Vol. 240
De novo exploration and self-guided learning of potential-energy surfaces
journal, October 2019
- Bernstein, Noam; Csányi, Gábor; Deringer, Volker L.
- npj Computational Materials, Vol. 5, Issue 1
Extending the accuracy of the SNAP interatomic potential form
journal, June 2018
- Wood, Mitchell A.; Thompson, Aidan P.
- The Journal of Chemical Physics, Vol. 148, Issue 24
Phase Transitions of Hybrid Perovskites Simulated by Machine-Learning Force Fields Trained on the Fly with Bayesian Inference
journal, June 2019
- Jinnouchi, Ryosuke; Lahnsteiner, Jonathan; Karsai, Ferenc
- Physical Review Letters, Vol. 122, Issue 22
On-the-Fly Active Learning of Interatomic Potentials for Large-Scale Atomistic Simulations
journal, July 2020
- Jinnouchi, Ryosuke; Miwa, Kazutoshi; Karsai, Ferenc
- The Journal of Physical Chemistry Letters, Vol. 11, Issue 17
Machine Learning Estimates of Natural Product Conformational Energies
journal, January 2014
- Rupp, Matthias; Bauer, Matthias R.; Wilcken, Rainer
- PLoS Computational Biology, Vol. 10, Issue 1
Machine-learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide
journal, July 2020
- Sivaraman, Ganesh; Krishnamoorthy, Anand Narayanan; Baur, Matthias
- npj Computational Materials, Vol. 6, Issue 1
Evolution of Metastable Structures at Bimetallic Surfaces from Microscopy and Machine-Learning Molecular Dynamics
journal, August 2020
- Lim, Jin Soo; Vandermause, Jonathan; van Spronsen, Matthijs A.
- Journal of the American Chemical Society, Vol. 142, Issue 37
The ReaxFF reactive force-field: development, applications and future directions
journal, March 2016
- Senftle, Thomas P.; Hong, Sungwook; Islam, Md Mahbubul
- npj Computational Materials, Vol. 2, Issue 1
Quantum and classical dynamics of reactive scattering of H 2 from metal surfaces
journal, January 2016
- Kroes, Geert-Jan; Díaz, Cristina
- Chemical Society Reviews, Vol. 45, Issue 13
Hydrogen and Deuterium Exchange on Pt(111) and Its Poisoning by Carbon Monoxide Studied by Surface Sensitive High-Pressure Techniques
journal, October 2006
- Montano, Max; Bratlie, Kaitlin; Salmeron, Miquel
- Journal of the American Chemical Society, Vol. 128, Issue 40
A Molecular Dynamics Study for Dissociation of H2 Molecule on Pt(111) Surface
journal, January 2012
- Tokumasu, Takashi
- Advanced Materials Research, Vol. 452-453
A transferable active-learning strategy for reactive molecular force fields
journal, January 2021
- Young, Tom A.; Johnston-Wood, Tristan; Deringer, Volker L.
- Chemical Science, Vol. 12, Issue 32
Computational aspects of many-body potentials
journal, May 2012
- Plimpton, Steven J.; Thompson, Aidan P.
- MRS Bulletin, Vol. 37, Issue 5
Ab initio molecular dynamics for liquid metals
journal, December 1995
- Kresse, G.
- Journal of Non-Crystalline Solids, Vol. 192-193
On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events
journal, March 2020
- Vandermause, Jonathan; Torrisi, Steven B.; Batzner, Simon
- npj Computational Materials, Vol. 6, Issue 1
A fast neural network approach for direct covariant forces prediction in complex multi-element extended systems
journal, September 2019
- Mailoa, Jonathan P.; Kornbluth, Mordechai; Batzner, Simon
- Nature Machine Intelligence, Vol. 1, Issue 10
ReaxFF: A Reactive Force Field for Hydrocarbons
journal, October 2001
- van Duin, Adri C. T.; Dasgupta, Siddharth; Lorant, Francois
- The Journal of Physical Chemistry A, Vol. 105, Issue 41
Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces
journal, April 2007
- Behler, Jörg; Parrinello, Michele
- Physical Review Letters, Vol. 98, Issue 14
Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the Electrons
journal, April 2010
- Bartók, Albert P.; Payne, Mike C.; Kondor, Risi
- Physical Review Letters, Vol. 104, Issue 13
Surface and Subsurface Hydrogen: Adsorption Properties on Transition Metals and Near-Surface Alloys
journal, March 2005
- Greeley, Jeff; Mavrikakis, Manos
- The Journal of Physical Chemistry B, Vol. 109, Issue 8
Parallel reactive molecular dynamics: Numerical methods and algorithmic techniques
journal, April 2012
- Aktulga, H. M.; Fogarty, J. C.; Pandit, S. A.
- Parallel Computing, Vol. 38, Issue 4-5
A systematic DFT study of hydrogen diffusion on transition metal surfaces
journal, September 2012
- Kristinsdóttir, Lilja; Skúlason, Egill
- Surface Science, Vol. 606, Issue 17-18
Optimal construction of a fast and accurate polarisable water potential based on multipole moments trained by machine learning
journal, January 2009
- Handley, Chris M.; Hawe, Glenn I.; Kell, Douglas B.
- Physical Chemistry Chemical Physics, Vol. 11, Issue 30
Hydrogen storage in carbon nanostructures via spillover
journal, November 2016
- Pyle, Darryl S.; Gray, E. MacA.; Webb, C. J.
- International Journal of Hydrogen Energy, Vol. 41, Issue 42
Molecular adsorption at Pt(111). How accurate are DFT functionals?
journal, January 2015
- Gautier, Sarah; Steinmann, Stephan N.; Michel, Carine
- Physical Chemistry Chemical Physics, Vol. 17, Issue 43
Machine learning based interatomic potential for amorphous carbon
journal, March 2017
- Deringer, Volker L.; Csányi, Gábor
- Physical Review B, Vol. 95, Issue 9
Active learning of uniformly accurate interatomic potentials for materials simulation
journal, February 2019
- Zhang, Linfeng; Lin, De-Ye; Wang, Han
- Physical Review Materials, Vol. 3, Issue 2
Active learning of linearly parametrized interatomic potentials
journal, December 2017
- Podryabinkin, Evgeny V.; Shapeev, Alexander V.
- Computational Materials Science, Vol. 140
Theoretical prediction of high melting temperature for a Mo–Ru–Ta–W HCP multiprincipal element alloy
journal, January 2021
- Hong, Qi-Jun; Schroers, Jan; Hofmann, Douglas
- npj Computational Materials, Vol. 7, Issue 1