A Universal Graph Deep Learning Interatomic Potential for the Periodic Table

Abstract

Interatomic potentials (IAPs), which describe the potential energy surface of atoms, are a fundamental input for atomistic simulations. However, existing IAPs are either fitted to narrow chemistries or too inaccurate for general applications. Here, we report a universal IAP for materials based on graph neural networks with three-body interactions (M3GNet). The M3GNet IAP was trained on the massive database of structural relaxations performed by the Materials Project over the past 10 years and has broad applications in structural relaxation, dynamic simulations and property prediction of materials across diverse chemical spaces. About 1.8 million materials were identified from a screening of 31 million hypothetical crystal structures to be potentially stable against existing Materials Project crystals based on M3GNet energies. Of the top 2000 materials with the lowest energies above hull, 1578 were verified to be stable using DFT calculations. These results demonstrate a machine learning-accelerated pathway to the discovery of synthesizable materials with exceptional properties.

References in corpus (20)

• Restoring the density-gradient expansion for exchange in solids and surfaces
• Generalized gradient approximation for solids and their surfaces
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• CALYPSO: a method for crystal structure prediction
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Machine Learning of Accurate Energy-Conserving Molecular Force Fields
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Atomistic Line Graph Neural Network for Improved Materials Property Predictions
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• AiiDA: Automated Interactive Infrastructure and Database for Computational Science
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• Embedded Atom Neural Network Potentials: Efficient and Accurate Machine Learning with a Physically Inspired Representation
• Learning Properties of Ordered and Disordered Materials from Multi-fidelity Data
• A Novel Approach to Describe Chemical Environments in High Dimensional Neural Network Potentials
• Bridging the Gap Between Simulated and Experimental Ionic Conductivities in Lithium Superionic Conductors

Cited by in corpus (132)

• DeePMD-kit v2: A software package for Deep Potential models
• Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics Simulations
• A practical guide to machine learning interatomic potentials -- Status and future
• General-purpose machine-learned potential for 16 elemental metals and their alloys
• Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science
• A critical examination of robustness and generalizability of machine learning prediction of materials properties
• DPA-2: a large atomic model as a multi-task learner
• Accelerating computational materials discovery with artificial intelligence and cloud high-performance computing: from large-scale screening to experimental validation
• On the redundancy in large material datasets: efficient and robust learning with less data
• Performance Assessment of Universal Machine Learning Interatomic Potentials: Challenges and Directions for Materials' Surfaces
• Deep learning generative model for crystal structure prediction
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Probing out-of-distribution generalization in machine learning for materials
• Accelerating CALPHAD-based Phase Diagram Predictions in Complex Alloys Using Universal Machine Learning Potentials: Opportunities and Challenges
• Machine learning accelerated discovery of corrosion-resistant high-entropy alloys
• JARVIS-Leaderboard: A Large Scale Benchmark of Materials Design Methods
• Spin-Dependent Graph Neural Network Potential for Magnetic Materials
• Developments and applications of the OPTIMADE API for materials discovery, design, and data exchange
• GPTFF: A high-accuracy out-of-the-box universal AI force field for arbitrary inorganic materials
• Lifelong Machine Learning Potentials
• Machine-learning structural reconstructions for accelerated point defect calculations
• Deep learning of experimental electrochemistry for battery cathodes across diverse compositions
• Developments and Further Applications of Ephemeral Data Derived Potentials
• Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models
• Universal Machine Learning Kohn-Sham Hamiltonian for Materials
• Recent progress in the JARVIS infrastructure for next-generation data-driven materials design
• DenseGNN: universal and scalable deeper graph neural networks for high-performance property prediction in crystals and molecules
• An equivariant graph neural network for the elasticity tensors of all seven crystal systems
• Con-CDVAE: A method for the conditional generation of crystal structures
• Deciphering Chemical Ordering in High Entropy Materials: A Machine Learning-Accelerated High-throughput Cluster Expansion Approach
• Advances in modeling complex materials: The rise of neuroevolution potentials
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Navigating the Evolution of Two-dimensional Carbon Nitride Research: Integrating Machine Learning into Conventional Approaches
• Model-free quantification of completeness, uncertainties, and outliers in atomistic machine learning using information theory
• Heat flux for semi-local machine-learning potentials
• Navigating chemical reaction space with a steering wheel
• Accelerating the prediction of inorganic surfaces with machine learning interatomic potentials
• Data efficiency and extrapolation trends in neural network interatomic potentials
• MolSets: Molecular Graph Deep Sets Learning for Mixture Property Modeling
• SCINE -- Software for Chemical Interaction Networks
• Creation of crystal structure reproducing X-ray diffraction pattern without using database
• MGNN: Moment Graph Neural Network for Universal Molecular Potentials
• Data-driven Design of High Pressure Hydride Superconductors using DFT and Deep Learning
• Modeling Extensive Defects in Metals through Classical Potential-Guided Sampling and Automated Configuration Reconstruction
• Recent Advances in Unconventional Ferroelectrics and Multiferroics
• Generator of Neural Network Potential for Molecular Dynamics: Constructing Robust and Accurate Potentials with Active Learning for Nanosecond-scale Simulations
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• Predicting ionic conductivity in solids from the machine-learned potential energy landscape
• Universal machine learning interatomic potentials poised to supplant DFT in modeling general defects in metals and random alloys
• Augmentation of Universal Potentials for Broad Applications
• Accelerating superconductor discovery through tempered deep learning of the electron-phonon spectral function
• How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm
• General time-reversal equivariant neural network potential for magnetic materials
• Modeling intercalation chemistry with multi-redox reactions by sparse lattice models in disordered rocksalt cathodes
• Accurate and efficient machine learning interatomic potentials for finite temperature modeling of molecular crystals
• Global optimization in the discrete and variable-dimension conformational space: The case of crystal with the strongest atomic cohesion
• DPmoire: A tool for constructing accurate machine learning force fields in moiré systems
• Exploration of amorphous V${}_2$O${}_5$ as cathode for magnesium batteries
• Prediction rigidities for data-driven chemistry
• Automated Microsolvation for Minimum Energy Path Construction in Solution
• Crystalline Material Discovery in the Era of Artificial Intelligence
• Pre-training, fine-tuning, and distillation (PFD): Automatically generating machine learning force fields from universal models
• Optical materials discovery and design with federated databases and machine learning
• Accelerated Discovery of Crystalline Materials with Record Ultralow Lattice Thermal Conductivity via a Universal Descriptor
• A literature-derived dataset of migration barriers for quantifying ionic transport in battery materials
• Machine Learning Potentials for Alloys: A Detailed Workflow to Predict Phase Diagrams and Benchmark Accuracy
• Identifying Crystal Structures Beyond Known Prototypes from X-ray Powder Diffraction Spectra
• A Comprehensive Assessment and Benchmark Study of Large Atomistic Foundation Models for Phonons
• Space Group Informed Transformer for Crystalline Materials Generation
• Exploring the energy landscape of aluminas through machine learning interatomic potential
• Efficient Generation of Stable Linear Machine-Learning Force Fields with Uncertainty-Aware Active Learning
• Benchmarking CHGNet Universal Machine Learning Interatomic Potential Against DFT and EXAFS: Case of Layered WS2 and MoS2
• Materials discovery acceleration by using condition generative methodology
• EOSnet: Embedded Overlap Structures for Graph Neural Networks in Predicting Material Properties
• Machine-learning potentials for structurally and chemically complex MAB phases: strain hardening and ripplocation-mediated plasticity
• Crystal structure prediction with host-guided inpainting generation and foundation potentials
• Topological descriptors for the electron density of inorganic solids
• Multi-head committees enable direct uncertainty prediction for atomistic foundation models
• Predicting neutron experiments from first principles: A workflow powered by machine learning
• Multi-Objective Quality-Diversity for Crystal Structure Prediction
• Fine-Tuning Unifies Foundational Machine-learned Interatomic Potential Architectures at ab initio Accuracy
• Modeling Chemical Exfoliation of Non-van der Waals Chromium Sulfides by Machine Learning Interatomic Potentials and Monte Carlo Simulations
• High-throughput computational framework for lattice dynamics and thermal transport including high-order anharmonicity: an application to cubic and tetragonal inorganic compounds
• Dismai-Bench: Benchmarking and designing generative models using disordered materials and interfaces
• Lifelong Machine Learning Potentials for Chemical Reaction Network Explorations
• Data Fusion of Deep Learned Molecular Embeddings for Property Prediction
• Accurate Machine Learning Interatomic Potentials for Polyacene Molecular Crystals: Application to Single Molecule Host-Guest Systems
• Developing a Complete AI-Accelerated Workflow for Superconductor Discovery
• Identification of soft modes in amorphous Al${}_2$O${}_3$ via first-principles
• XtalOpt Version 14: Variable-Composition Crystal Structure Search for Functional Materials Through Pareto Optimization
• Reducing Self-Interaction Error in Transition-Metal Oxides with Different Exact-Exchange Fractions for Energy and Density
• Modeling phase transformations in Mn-rich disordered rocksalt cathodes with machine learning interatomic potentials
• Leveraging transfer learning for accurate estimation of ionic migration barriers in solids
• Machine Learning Phonon Spectra for Fast and Accurate Optical Lineshapes of Defects
• DeFecT-FF: a machine learning force field framework for high throughput defect modeling in CdTe-based solar cells
• Benchmarking foundation potentials against quantum chemistry methods for predicting molecular redox potentials
• Revealing interstitial energetics in Ti-23Nb-0.7Ta-2Zr gum metal base alloy via universal machine learning interatomic potentials
• Beyond Predicted ZT: Machine Learning Strategies for the Experimental Discovery of Thermoelectric Materials
• Machine learning metallic glass critical cooling rates through elemental and molecular simulation based featurization
• Search for thermodynamically stable ambient-pressure superconducting hydrides in GNoME database
• Accelerating point defect simulations using data-driven and machine learning approaches
• Radiation damage and phase stability of Al${}_x$CrCuFeNi${}_y$ alloys using a machine-learned interatomic potential
• FastCHGNet: Training one Universal Interatomic Potential to 1.5 Hours with 32 GPUs
• Combining graph deep learning and London dispersion interatomic potentials: A case study on pnictogen chalcohalides
• Reinforcement Fine-Tuning for Materials Design
• Do Graph Neural Networks Work for High Entropy Alloys?
• Bridging Text and Crystal Structures: Literature-driven Contrastive Learning for Materials Science
• A general formalism for machine-learning models based on multipolar-spherical harmonics
• Massive Discovery of Low-Dimensional Materials from Universal Computational Strategy
• SPEA -- an analytical thermodynamic model for defect phase diagram
• Accelerating Complex Materials Discovery with Universal Machine-Learning Potential-Driven Structure Prediction
• FastTrack: a fast method to evaluate mass transport in solid leveraging universal machine learning interatomic potential
• Scalable Training of Neural Network Potentials for Complex Interfaces Through Data Augmentation
• Phase transitions of correlated systems from graph neural networks with quantum embedding techniques
• Managing Expectations and Imbalanced Training Data in Reactive Force Field Development: an Application to Water Adsorption on Alumina
• Recent Breakthrough in AI-Driven Materials Science: Tech Giants Introduce Groundbreaking Models
• Interpretation of Crystal Energy Landscapes with Kolmogorov-Arnold Networks
• Active Δ-learning with universal potentials for global structure optimization
• Assessing zero-shot generalisation behaviour in graph-neural-network interatomic potentials
• KAN-Enhanced Contrastive Learning Accelerating Crystal Structure Identification from XRD Patterns
• Accelerated Machine Learning Force Field for Predicting Thermal Conductivity of Organic Liquids
• Exchange-Correlation Functionals in 2D Materials: Applications, Challenges, and Limitations
• Data-Driven Exploration and Insights into Temperature-Dependent Phonons in Inorganic Materials
• Atoms to Processes: The Role of Artificial Intelligence and Machine Learning in Chemical Engineering
• A charge-density machine-learning workflow for computing the infrared spectrum of molecules
• Electron-Informed Coarse-Graining Molecular Representation Learning for Real-World Molecular Physics
• GEWUM: General Exploration Workflow for the Utopia of Materials: A Unified Platform for Automated Structure Generation, Selection, and Validation
• High-throughput study of kagome compounds in the AV3Sb5 family
• A flexible kinetic Monte Carlo framework for GaN molecular beam epitaxy with adaptive on-the-fly barrier evaluation
• Inverse design of bespoke interatomic potentials via active learning by information-matching
• Lateral Graphene-Metallene Interfaces at the Nanoscale
• A generative machine learning model for designing metal hydrides applied to hydrogen storage