A unified machine-learning framework for ab initio multiscale modeling of liquids

Abstract

Understanding and predicting the behavior of liquid matter across length scales, using only the microscopic interactions encoded in the Schrödinger equation, remains a central challenge in the physical sciences. Achieving this goal requires not only an accurate and efficient description of intermolecular forces but also a consistent framework that bridges the micro-, meso-, and macroscales. Here, by combining machine-learned interatomic potentials (MLIPs) with neural classical density functional theory (neural cDFT), we present such a framework. The underlying idea is simple: MLIPs trained on quantum-mechanical energies and forces are used to generate inhomogeneous microscopic density profiles, which in turn serve as the training data for neural cDFT. The resulting ab initio neural cDFT is not only significantly more computationally efficient than molecular simulations, but also provides a conceptually transparent route to the thermodynamics of both homogeneous and inhomogeneous systems. We demonstrate the approach for both water and carbon dioxide using several exchange-correlation functionals. Beyond accurately reproducing bulk equations of state and liquid-vapor phase diagrams, ab initio neural cDFT predicts, from first principles, how confinement modifies liquid-vapor coexistence in water. It also captures complex behavior in supercritical carbon dioxide such as the Fisher-Widom and Widom lines. Ab initio neural cDFT establishes a general first-principles route to multiscale modeling of fluids within a single unified conceptual framework.

Main: 15 pages, 4 figures. SI: 8 pages, 10 figures

References in corpus (47)

• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
• Anomalously low dielectric constant of confined water
• Relation Between the Widom line and the Strong-Fragile Dynamic Crossover in Systems with a Liquid-Liquid Phase Transition
• Square ice in graphene nanocapillaries
• Molecular transport through capillaries made with atomic-scale precision
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Ab initio thermodynamics of liquid and solid water
• First-principles kinetic Monte Carlo simulations for heterogeneous catalysis, applied to the CO oxidation at RuO2(110)
• Capillary condensation under atomic-scale confinement
• The first-principles phase diagram of monolayer nanoconfined water
• A deep potential model with long-range electrostatic interactions
• 2D ice from first principles: structures and phase transitions
• Self-Consistent Determination of Long-Range Electrostatics in Neural Network Potentials
• On the physisorption of water on graphene: Sub-chemical accuracy from many-body electronic structure methods
• Liquid-like behavior of supercritical fluids
• Machine learning and density functional theory
• Dissolving salt is not equivalent to applying a pressure on water
• Molecular Density Functional Theory of Water
• Transition in the supercritical state of matter: experimental evidence
• Interplay of local hydrogen-bonding and long-ranged dipolar forces in simulations of confined water
• The fate of carbon dioxide in water-rich fluids at extreme conditions
• Quantum-mechanical exploration of the phase diagram of water
• The standard mean-field treatment of inter-particle attraction in classical DFT is better than one might expect
• In-plane dielectric constant and conductivity of confined water
• Local molecular field theory for the treatment of electrostatics
• Classical quantum friction at water-carbon interfaces
• Neural functional theory for inhomogeneous fluids: Fundamentals and applications
• Ab-initio Structure and Thermodynamics of the RPBE-D3 Water/Vapor Interface by Neural-Network Molecular Dynamics
• Dielectric response with short-ranged electrostatics
• A Deep Potential model for liquid-vapor equilibrium and cavitation rates of water
• Disjoining Pressure of Water in Nanochannels
• Thermodynamic crossovers in supercritical fluids
• Understanding high pressure hydrogen with a hierarchical machine-learned potential
• Learning classical density functionals for ionic fluids
• Fast and flexible long-range models for atomistic machine learning
• Hyper-density functional theory of soft matter
• Machine-Learning Interatomic Potentials for Long-Range Systems
• Learning Neural Free-Energy Functionals with Pair-Correlation Matching
• Neural force functional for non-equilibrium many-body colloidal systems
• Revisiting the Green-Kubo relation for friction in nanofluidics
• Learning the bulk and interfacial physics of liquid-liquid phase separation with neural density functionals
• Neural Density Functional Theory in Higher Dimensions with Convolutional Layers
• A first principles approach to electromechanics in liquids
• Dielectrocapillarity for exquisite control of fluids
• Determining the chemical potential via universal density functional learning
• The roles of bulk and surface thermodynamics in the selective adsorption of a confined azeotropic mixture