Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Benchmarking machine learning interatomic potentials via phonon anharmonicity

Journal Article · · Machine Learning: Science and Technology
 [1];  [2];  [1]
  1. Columbia Univ., New York, NY (United States)
  2. Idaho National Laboratory (INL), Idaho Falls, ID (United States)
+ Show Author Affiliations
Machine learning approaches have recently emerged as powerful tools to probe structure-property relationships in crystals and molecules. Specifically, machine learning interatomic potentials (MLIPs) can accurately reproduce first-principles data at a cost similar to that of conventional interatomic potential approaches. While MLIPs have been extensively tested across various classes of materials and molecules, a clear characterization of the anharmonic terms encoded in the MLIPs is lacking. Here, we benchmark popular MLIPs using the anharmonic vibrational Hamiltonian of ThO2 in the fluorite crystal structure, which was constructed from density functional theory (DFT) using our highly accurate and efficient irreducible derivative methods. The anharmonic Hamiltonian was used to generate molecular dynamics (MD) trajectories, which were used to train three classes of MLIPs: Gaussian approximation potentials, artificial neural networks (ANN), and graph neural networks (GNN). The results were assessed by directly comparing phonons and their interactions, as well as phonon linewidths, phonon lineshifts, and thermal conductivity. The models were also trained on a DFT MD dataset, demonstrating good agreement up to fifth-order for the ANN and GNN. Our analysis demonstrates that MLIPs have great potential for accurately characterizing anharmonicity in materials systems at a fraction of the cost of conventional first principles-based approaches.
Research Organization:
Idaho National Laboratory (INL), Idaho Falls, ID (United States)
Sponsoring Organization:
USDOE; USDOE Laboratory Directed Research and Development (LDRD) Program; USDOE Office of Nuclear Energy (NE); USDOE Office of Science (SC), Basic Energy Sciences (BES). Scientific User Facilities (SUF)
Grant/Contract Number:
AC02-05CH11231; AC07-05ID14517; SC0016507
OSTI ID:
2429470
Alternate ID(s):
OSTI ID: 2406510
OSTI ID: 2587261
Report Number(s):
INL/JOU--25-86312-Rev000
Journal Information:
Machine Learning: Science and Technology, Journal Name: Machine Learning: Science and Technology Journal Issue: 3 Vol. 5; ISSN 2632-2153
Publisher:
IOP PublishingCopyright Statement
Country of Publication:
United States
Language:
English

References (43)

Temperature‐Dependent Anharmonic Phonons in Quantum Paraelectric KTaO3 by First Principles and Machine‐Learned Force Fields journal February 2023
Novel machine learning framework for thermal conductivity prediction by crystal graph convolution embedded ensemble journal October 2021
A Deep Neural Network Potential to Study the Thermal Conductivity of MnBi2Te4 and Bi2Te3/MnBi2Te4 Superlattice journal April 2023
Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set journal July 1996
Developing machine learning potential for classical molecular dynamics simulation with superior phonon properties journal February 2022
TeaNet: Universal neural network interatomic potential inspired by iterative electronic relaxations journal May 2022
Accelerated computation of lattice thermal conductivity using neural network interatomic potentials journal August 2022
Screening outstanding mechanical properties and low lattice thermal conductivity using global attention graph neural network journal October 2023
Thermal conductivity modeling using machine learning potentials: application to crystalline and amorphous silicon journal August 2019
A unified deep neural network potential capable of predicting thermal conductivity of silicon in different phases journal March 2020
Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error journal October 2017
Library-Based LAMMPS Implementation of High-Dimensional Neural Network Potentials journal January 2019
E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials journal May 2022
Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning journal March 2022
Machine learning potential assisted exploration of complex defect potential energy surfaces journal January 2024
CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling journal September 2023
A universal graph deep learning interatomic potential for the periodic table journal November 2022
Gaussian approximation potentials for accurate thermal properties of two-dimensional materials journal January 2023
Atom-centered symmetry functions for constructing high-dimensional neural network potentials journal February 2011
Commentary: The Materials Project: A materials genome approach to accelerating materials innovation journal July 2013
Transferability of neural network potentials for varying stoichiometry: Phonons and thermal conductivity of Mn x Ge y compounds journal June 2020
A deep neural network interatomic potential for studying thermal conductivity of β -Ga 2 O 3 journal October 2020
Recent advances in lattice thermal conductivity calculation using machine-learning interatomic potentials journal December 2021
Accuracy of Machine Learning Potential for Predictions of Multiple-Target Physical Properties* journal December 2020
f90wrap: an automated tool for constructing deep Python interfaces to modern Fortran codes journal May 2020
Thermal conductivity of h-BN monolayers using machine learning interatomic potential journal December 2020
The stochastic self-consistent harmonic approximation: calculating vibrational properties of materials with full quantum and anharmonic effects journal July 2021
Group theoretical approach to computing phonons and their interactions journal July 2019
Compressive sensing lattice dynamics. I. General formalism journal November 2019
Generalized quasiharmonic approximation via space group irreducible derivatives journal July 2022
Anharmonic phonon behavior via irreducible derivatives: Self-consistent perturbation theory and molecular dynamics journal March 2023
Precisely computing phonons via irreducible derivatives journal May 2023
Ab initiomolecular dynamics for liquid metals journal January 1993
On representing chemical environments journal May 2013
Self-consistent phonon calculations of lattice dynamical properties in cubic SrTiO 3 with first-principles anharmonic force constants journal August 2015
Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the Electrons journal April 2010
Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning journal January 2012
Lattice Anharmonicity and Thermal Conductivity from Compressive Sensing of First-Principles Calculations journal October 2014
Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces journal April 2007
Machine-learning-based interatomic potential for phonon transport in perfect crystalline Si and crystalline Si with vacancies journal July 2019
Quantum paraelectricity and structural phase transitions in strontium titanate beyond density functional theory journal March 2023
Regression shrinkage and selection via the lasso: a retrospective: Regression Shrinkage and Selection via the Lasso journal April 2011
Simulating lattice thermal conductivity in semiconducting materials using high-dimensional neural network potential journal August 2019

Similar Records

Benchmarking structural evolution methods for training of machine learned interatomic potentials
Journal Article · Thu Jul 21 20:00:00 EDT 2022 · Journal of Physics. Condensed Matter · OSTI ID:1982442

Related Subjects

interatomic potentials
machine learning
phonon