Learning intermolecular forces at liquid-vapor interfaces
arXiv:2107.06208 · doi:10.1063/5.0067565
Abstract
By adopting a perspective informed by contemporary liquid state theory, we consider how to train an artificial neural network potential to describe inhomogeneous, disordered systems. We find that neural network potentials based on local representations of atomic environments are capable of describing some properties of liquid-vapor interfaces, but typically fail for properties that depend on unbalanced long-ranged interactions which build up in the presence of broken translation symmetry. These same interactions cancel in the translationally invariant bulk, allowing local neural network potentials to describe bulk properties correctly. By incorporating explicit models of the slowly-varying long-ranged interactions and training neural networks only on the short ranged components, we can arrive at potentials that robustly recover interfacial properties. We find that local neural network models can sometimes approximate a local molecular field potential to correct for the truncated interactions, but this behavior is variable and hard to learn. Generally, we find that models with explicit electrostatics are easier to train and have higher accuracy. We demonstrate this perspective in a simple model of an asymmetric dipolar fluid where the exact long-ranged interaction is known, and in an ab initio water model where it is approximated.
arXiv commentSmall edits for clarity; added appendices
References in corpus (28)
- Adam: A Method for Stochastic Optimization
- Separable Dual Space Gaussian Pseudo-potentials
- Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
- 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
- PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
- Machine learning for molecular simulation
- Machine Learning Unifies the Modelling of Materials and Molecules
- How van der Waals interactions determine the unique properties of water
- Ab initio thermodynamics of liquid and solid water
- Symmetry-Adapted Machine-Learning for Tensorial Properties of Atomistic Systems
- Accurate Interatomic Force Fields via Machine Learning with Covariant Kernels
- Incorporating long-range physics in atomic-scale machine learning
- Silicon liquid structure and crystal nucleation from ab-initio deep Metadynamics
- End-to-end Symmetry Preserving Inter-atomic Potential Energy Model for Finite and Extended Systems
- Connecting local structure to interface formation: a molecular scale van der Waals theory of nonuniform liquids
- Interplay of local hydrogen-bonding and long-ranged dipolar forces in simulations of confined water
- Mass Density Fluctuations in Quantum and Classical descriptions of Liquid Water
- Local molecular field theory for the treatment of electrostatics
- Liquid-Like Interfaces Mediate Structural Phase Transitions in Lead Halide Perovskites
- Dissecting Hydrophobic Hydration and Association
- Ab-initio Structure and Thermodynamics of the RPBE-D3 Water/Vapor Interface by Neural-Network Molecular Dynamics
- Attraction Between Like-Charged Walls: Short-Ranged Simulations Using Local Molecular Field Theory
- Deconstructing classical water models at interfaces and in bulk
- Microscopic dynamics of charge separation at the aqueous electrochemical interface
- Dielectric response with short-ranged electrostatics
- Isomerization and Decomposition Reactions of Acetaldehyde Relevant to Atmospheric Processes from Dynamics Simulations on Neural Network-Based Potential Energy Surfaces
- Rate constants in spatially inhomogeneous systems