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.

arXiv comment13 pages, 11 figures

References in corpus (16)

• An overview of gradient descent optimization algorithms
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• SchNet - a deep learning architecture for molecules and materials
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Machine learning for molecular simulation
• By-passing the Kohn-Sham equations with machine learning
• Nearsightedness of Electronic Matter
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network
• Transferable atomic multipole machine learning models for small organic molecules
• The effect of ionization on the global minima of small and medium sized silicon and magnesium clusters
• FLAME: a library of atomistic modeling environments
• Detecting non-local effects in the electronic structure of a simple covalent system with machine learning methods

Cited by in corpus (86)

• Machine Learning Force Fields
• Recent Advances and Applications of Deep Learning Methods in Materials Science
• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• Towards self-driving laboratories: The central role of density functional theory in the AI age
• A deep potential model with long-range electrostatic interactions
• Deep Potentials for Materials Science
• Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning
• Self-Consistent Determination of Long-Range Electrostatics in Neural Network Potentials
• Neural Network Potentials for Chemistry: Concepts, Applications and Prospects
• Learning neural network potentials from experimental data via Differentiable Trajectory Reweighting
• End-to-End Differentiable Molecular Mechanics Force Field Construction
• Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations
• High-Dimensional Neural Network Potentials for Magnetic Systems Using Spin-Dependent Atom-Centered Symmetry Functions
• Tutorial: How to Train a Neural Network Potential
• Thermodynamics and dielectric response of ${\text{BaTiO}}_3$ by data-driven modeling
• Modelling atomic and nanoscale structure in the silicon-oxygen system through active machine learning
• BIGDML: Towards Exact Machine Learning Force Fields for Materials
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulations
• Insights into lithium manganese oxide-water interfaces using machine learning potentials
• Spin-Dependent Graph Neural Network Potential for Magnetic Materials
• High-accuracy thermodynamic properties to the melting point from ab initio calculations aided by machine-learning potentials
• Predicting tensorial molecular properties with equivariant machine-learning models
• Deep Coarse-grained Potentials via Relative Entropy Minimization
• Scalable Hybrid Deep Neural Networks/Polarizable Potentials Biomolecular Simulations including long-range effects
• Force-Field-Enhanced Neural Network Interactions: from Local Equivariant Embedding to Atom-in-Molecule properties and long-range effects
• Lifelong Machine Learning Potentials
• Molecular dynamics simulation with finite electric fields using Perturbed Neural Network Potentials
• Towards Linearly Scaling and Chemically Accurate Global Machine Learning Force Fields for Large Molecules
• Accelerated identification of equilibrium structures of multicomponent inorganic crystals using machine learning potentials
• ArcaNN: automated enhanced sampling generation of training sets for chemically reactive machine learning interatomic potentials
• Robustness of Local Predictions in Atomistic Machine Learning Models
• Advances in modeling complex materials: The rise of neuroevolution potentials
• On the design space between molecular mechanics and machine learning force fields
• Predicting the Charge Density Response in Metal Electrodes
• Multilayer atomic cluster expansion for semi-local interactions
• Machine Learning and Data-Driven Methods in Computational Surface and Interface Science
• Lightweight and Effective Tensor Sensitivity for Atomistic Neural Networks
• Modeling electronic response properties with an explicit-electron machine learning potential
• EspalomaCharge: Machine learning-enabled ultra-fast partial charge assignment
• Impact of the characteristics of quantum chemical databases on machine learning predictions of tautomerization energies
• High-dimensional neural network potentials for accurate vibrational frequencies: The formic acid dimer benchmark
• Fast and flexible long-range models for atomistic machine learning
• Machine-Learning Interatomic Potentials for Long-Range Systems
• Accelerating fourth-generation machine learning potentials by quasi-linear scaling particle mesh charge equilibration
• Machine Learning for Exploring Small Polaron Configurational Space
• Accuracy of Charge Densities in Electronic Structure Calculations
• Accurate quantum Monte Carlo forces for machine-learned force fields: Ethanol as a benchmark
• Accelerating QM/MM simulations of electrochemical interfaces through machine learning of electronic charge densities
• FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentials
• Efficient Composite Infrared Spectroscopy: Combining the Doubly-Harmonic Approximation with Machine Learning Potentials
• Mean-Field Density Matrix Decompositions
• Understanding High-Temperature Chemical Reactions on Metal Surfaces
• How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm
• Machine Learning Models Capture Plasmon Dynamics in Ag Nanoparticles
• A Hessian-Based Assessment of Atomic Forces for Training Machine Learning Interatomic Potentials
• Prediction rigidities for data-driven chemistry
• Shadow molecular dynamics and atomic cluster expansions for flexible charge models
• Machine learning potentials for complex aqueous systems made simple
• Metatensor and metatomic: foundational libraries for interoperable atomistic machine learning
• Gaussian Process Regression Adaptive Density-Guided Approach: Towards Calculations of Potential Energy Surfaces for Larger Molecules
• Iterative charge equilibration for fourth-generation high-dimensional neural network potentials
• Neural Network Approach for Predicting Infrared Spectra from 3D Molecular Structure
• Efficient moment tensor machine-learning interatomic potential for accurate description of defects in Ni-Al Alloys
• Fortnet, a software package for training Behler-Parrinello neural networks
• A Fully Quantum-Mechanical Treatment for Kaolinite
• Equivariant Machine Learning Interatomic Potentials with Global Charge Redistribution
• A neural network potential with self-trained atomic fingerprints: a test with the mW water potential
• Reconstructions and Dynamics of $\beta$-Lithium Thiophosphate Surfaces
• Modeling Chemical Exfoliation of Non-van der Waals Chromium Sulfides by Machine Learning Interatomic Potentials and Monte Carlo Simulations
• Atomistic Simulations of Oxide-Water Interfaces using Machine Learning Potentials
• A neural-network-backed effective harmonic potential study of the ambient pressure phases of hafnia
• Lifelong Machine Learning Potentials for Chemical Reaction Network Explorations
• Water Isotope Separation using Deep Learning and a Catalytically Active Ultrathin Membrane
• Transferability of the chemical bond-based machine learning model for dipole moment: the GHz to THz dielectric properties of liquid propylene glycol and polypropylene glycol
• Teaching oxidation states to neural networks
• Noise Tolerant Force Calculations in Density Functional Theory: A Surface Integral Approach for Wavelet-Based Methods
• Efficient Long-Range Convolutions for Point Clouds
• Machine-Learned Potentials for Solvation Modeling
• A unified machine-learning framework for ab initio multiscale modeling of liquids
• Graph Convolutional Neural Networks for (QM)ML/MM Molecular Dynamics Simulations
• Macro-Dipole-Constrainted Learning of Atomic Charges for Accurate Electrostatic Potentials at Electrochemical Interfaces
• Scalable Machine Learning Models for Predicting Quantum Transport in Disordered 2D Hexagonal Materials
• Learning Atomic Multipoles: Prediction of the Electrostatic Potential with Equivariant Graph Neural Networks