Abstract:Unleashing the predictive power of molecular dynamics (MD), Neural Network Potentials (NNPs) trained on Density Functional Theory (DFT) calculations are revolutionizing our ability to simulate chemical systems with unprecedented accuracy and efficiency. Electrolyte solutions are a natural initial system to apply this tool to because they are critically important for a wide range of applications and their properties cannot currently be predicted. Unfortunately, however most DFT approximations are not sufficiently accurate to predict many practically relevant properties of electrolytes. Additionally, tracking the position of every atom in a system during molecular simulations is inherently limited, even with NNP-MD. Here, we use a state-of-the-art DFT approximation to demonstrate highly accurate all-atom NNPs with minimal training data. We demonstrate that NNPs can reliably be recursively trained on a subset of their own output to enable coarse-grained continuum solvent molecular simulations that can access much longer timescales. We apply our technique to simulate lithium chloride, potassium chloride, and lithium bromide in water. We reproduce key structural, thermodynamic, and kinetic properties of these solutions in agreement with experimental data. The formation of a previously unknown Li cation dimer is observed, along with identical anion-anion interactions of chloride and bromide. Finally, the coarse-grained model is capable of reproducing crystal phase behavior and infinite dilution pairing free energies despite being trained solely on moderate concentration solutions, disproving the notion that NNPs are only useful for interpolation. This approach should be scalable to determine the properties of electrolyte solutions over a much wider range of conditions and compositions than is possible experimentally.

First-principles molecular dynamics of liquid alkali metals based on the quantal hyper-netted chain theory

Probing ion-ion and electron-ion correlations in liquid metals within the quantum hypernetted chain approximation

Study of theoretical models for the liquid-vapor and metal-nonmetal transitions of alkali fluids

Capturing the Nuclear Quantum Effects in Molecular Dynamics for Lattice Thermal Conductivity Calculations: Using Ice As Example

Atomic Structure of Liquid Refractory Nb5Si3 Intermetallic Compound Alloy Based Upon Deep Neural Network Potential

Learning the Quantum Centroid Force Correction in Molecular Systems: A Localized Approach

Scalable molecular simulation of electrolyte solutions with quantum chemical accuracy

Towards first-principles understanding of the metal-insulator transition in fluid alkali metals

Long range interactions between like homonuclear alkali metal diatoms

Atomistic simulation of Mott transition in fluid metals: Combining molecular dynamics with dynamical mean-field theory

Nonperturbative ab initio approach for calculating the electrical conductivity of a liquid metal

Ground-state properties of metallic solids from ab initio coupled-cluster theory

A First-Principles Study On Quasi-1d Alkali Metal Chains Within Zeolite Channels

Gaussian-basis LDA and GGA Calculations for Alkali-Metal Equations of State

First-principles molten salt phase diagrams through thermodynamic integration

Density functional theory based molecular dynamics study of solution composition effects on the solvation shell of metal ions

Non-perturbative $ab$ $initio$ approach for calculating the electrical conductivity of a liquid metal

Nucleation of Supercooled Liquid Aluminum: an First-Principles Molecular Dynamics Study Based on Orbital-Free Density Functional Theory

Quasielastic neutron scattering measurements and ab initio MD-simulations on single ion motions in molten NaF

Molecular Dynamics Simulations of the Local Structures and Transport Coefficients of Molten Alkali Chlorides

Phase Behavior and Atomic Dynamics in Rb x Na 1- x : Insights from Machine Learning Interatomic Potentials based on Ab Initio Molecular Dynamics