Abstract:Understanding ionic behaviour under external electric fields is crucial to develop electronic and energy-related devices using ion transport. In this study, we propose a neural network (NN) model to predict the Born effective charges of ions along an axis parallel to an applied electric field from atomic structures. The proposed NN model is applied to Li$_3$PO$_4$ as a prototype. The prediction error of the constructed NN model is 0.0376 $e$/atom. In combination with an NN interatomic potential, molecular dynamics (MD) simulations are performed under a uniform electric field of 0.1 V/angstrom, whereby an enhanced mean square displacement of Li along the electric field is obtained, which seems physically reasonable. In addition, the external forces along the direction perpendicular to the electric field, originating from the off-diagonal terms of the Born effective charges, are found to have a nonnegligible effect on Li migration. Finally, additional MD simulations are performed to examine the Li motion in an amorphous structure. The results reveal that Li migration occurs in various areas despite the absence of explicitly introduced defects, which may be attributed to the susceptibility of the Li ions in the local minima to the electric field. We expect that the proposed NN method can be applied to any ionic material, thereby leading to atomic-scale elucidation of ion behaviour under electric fields.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to predict the migration behavior of ions in materials under the action of an external electric field. Specifically, the author proposes a method based on neural network (Neural Network, NN) to predict the Born effective charge of ions, and then studies the migration path and kinetic behavior of ions under an external electric field. Taking Li₃PO₄ as the prototype material, the effectiveness of the proposed neural network model is verified by molecular dynamics (Molecular Dynamics, MD) simulations. ### Main research problems 1. **Predicting Born effective charge**: Develop a neural network model that can predict the Born effective charge of ions according to the atomic structure. These charges are crucial for understanding the behavior of ions under an external electric field. 2. **Ion migration path and kinetics**: Use the predicted Born effective charge, combined with the traditional neural network potential energy model, to conduct molecular dynamics simulations to study the migration path and kinetic behavior of Li⁺ ions under an external electric field. 3. **Influence of off - diagonal terms**: Explore the influence of the off - diagonal terms of the Born effective charge on the ion migration under an external electric field, especially in crystalline and amorphous Li₃PO₄. ### Research methods - **Neural network model**: A neural network model is designed. The input is the characteristics of the atomic structure, and the output is the Born effective charge. The model uses vector atomic fingerprint (Vector Atomic Fingerprint, VAF) as input features to retain directional information. - **Molecular dynamics simulation**: Combined with the traditional neural network potential energy model, molecular dynamics simulations are carried out. The simulation conditions include different temperatures and external electric field intensities. - **NEB calculation**: The nudged elastic band (NEB) method is used to calculate the potential energy path of Li⁺ ions under an external electric field, verifying the influence of the electric field on the ion migration energy barrier under an external electric field. ### Main results - **Model performance**: The proposed neural network model performs excellently in predicting the Born effective charge, with a prediction error of 0.0376 e/atom. - **Enhanced ion migration**: Under the action of an external electric field, the mean square displacement (Mean Square Displacement, MSD) of Li⁺ ions increases significantly, indicating that the electric field promotes the migration of Li⁺ ions. - **Influence of off - diagonal terms**: The off - diagonal terms of the Born effective charge have a slight but effective influence on the ion migration path under an external electric field, further enhancing the migration of Li⁺ ions. - **Amorphous state research**: In amorphous Li₃PO₄, the migration of Li⁺ ions under an external electric field is more extensive and not limited by specific defects, indicating that Li⁺ ions in the amorphous structure are more sensitive to the external electric field. ### Conclusion The paper proposes a neural network - based model, which successfully predicts the Born effective charge under an external electric field in Li₃PO₄, and verifies the effectiveness of the model through molecular dynamics simulations. The research results show that the Born effective charge and its off - diagonal terms have an important influence on the ion migration behavior under an external electric field, especially in amorphous materials, the migration behavior of Li⁺ ions is very sensitive to the external electric field. These findings provide new insights for understanding the behavior of ions under an external electric field and may be applied to the research of other ion - conductive materials.

Prediction of Born effective charges using neural network to study ion migration under electric fields: applications to crystalline and amorphous Li$_3$PO$_4$

Investigating Ionic Diffusivity in Amorphous Solid Electrolytes using Machine Learned Interatomic Potentials

An Electrostatic Spectral Neighbor Analysis Potential (eSNAP) for Lithium Nitride

Electronic charge density as a fast approach for predicting Li-ion migration pathways in superionic conductors with first-principles level precision

High-throughput investigation of stability and Li diffusion of doped solid electrolytes via neural network potential without configurational knowledge

Simulating Diffusion Properties of Solid‐State Electrolytes via a Neural Network Potential: Performance and Training Scheme

Predicting oxidation and spin states by high-dimensional neural networks: Applications to lithium manganese oxide spinels

Exploring Li-ion Transport Properties of Li$_3$TiCl$_6$: A Machine Learning Molecular Dynamics Study

Size Dependent Lithium-Ion Conductivity of Solid Electrolytes in Machine Learning Molcular Dynamics Simulations

Universal Machine Learning Interatomic Potentials: Surveying Solid Electrolytes

High-Dimensional Neural Network Potential for Liquid Electrolyte Simulations.

Prediction of NMR, X-ray and Mössbauer experimental results for amorphous Li-Si alloys using a novel DFTB model

Computational study of Na diffusion and conduction in P2- and O3-Na2x[NixTi1-x]O2 materials with machine-learning interatomic potentials

On the role of long-range electrostatics in machine-learned interatomic potentials for complex battery materials

Molecular dynamics simulation with finite electric fields using Perturbed Neural Network Potentials

Lithium Localization by Anions in Argyrodite Solid Electrolytes from Machine‐Learning‐based Simulations

Accurate description of ion migration in solid-state ion conductors from machine-learning molecular dynamics

Machine Learning Enhanced Electrochemical Simulations for Dendrites Nucleation in Li Metal Battery

Nature of the Amorphous–Amorphous Interfaces in Solid-State Batteries Revealed Using Machine-Learned Interatomic Potentials

Machine learning 3D-resolved prediction of electrolyte infiltration in battery porous electrodes

CHGNet: Pretrained universal neural network potential for charge-informed atomistic modeling