Macro-Dipole-Constrainted Learning of Atomic Charges for Accurate Electrostatic Potentials at Electrochemical Interfaces

2025/10/01 by Jing Yang, Yang, Jing, Bingxin Li +9
Chemistry · Engineering · Materials Science · #Chemical Physics (physics.chem-ph) #Electrochemical Analysis and Applications #Electron and X-Ray Spectroscopy Techniques #FOS: Physical sciences #Materials Science (cond-mat.mtrl-sci) #Non-Destructive Testing Techniques

paper · pdf · doi:10.48550/arxiv.2510.01042

openalex publication_date 2025/10/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01

Abstract

Large thermal fluctuations of the liquid phase obscure the weak macroscopic electric field that drives electrochemical reactions, rendering the extraction of reliable interfacial charge distributions from ab initio molecular dynamics extremely challenging. We introduce SMILE-CP (Scalar Macro-dipole Integrated LEarning - Charge Partitioning), a macro-dipole-constrained scheme that infers atomic charges using only the instantaneous atomic coordinates and the total dipole moment of the simulation cell - quantities routinely available from standard density-functional theory calculations. SMILE-CP preserves both the global electrostatic field and the local potential without invoking any explicit charge-partitioning scheme. Benchmarks on three representative electrochemical interfaces - nanoconfined water, Mg2+ dissolution in water, and a kinked Mg vicinal surface under anodic bias - show that SMILE-CP eliminates the qualitative errors observed for unconstrained charge decompositions. The method is computationally inexpensive and data-efficient, opening the door to charge-aware machine-learning potentials capable of bias-controlled, nanosecond-scale simulations of realistic electrochemical systems.

Citations

• Iterative charge equilibration for fourth-generation high-dimensional neural network potentials
• First principles approaches and concepts for electrochemical systems
• Charge-constrained Atomic Cluster Expansion
• Polarizable Water Model with Ab Initio Neural Network Dynamic Charges and Spontaneous Charge Transfer
• Machine Learning Potential for Electrochemical Interfaces with Hybrid Representation of Dielectric Response
• On the origin of univalent Mg⁺ ions in solution and their role in anomalous anodic hydrogen evolution
• Unified Differentiable Learning of Electric Response
• Active learning strategies for atomic cluster expansion models
• Implicit Solvation Methods for Catalysis at Electrified Interfaces
• Four Generations of High-Dimensional Neural Network Potentials
• Performant implementation of the atomic cluster expansion (PACE): Application to copper and silicon
• The MLIP package: Moment Tensor Potentials with MPI and Active Learning
• Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Maximally localized Wannier functions: Theory and applications
• Maximally-localized generalized Wannier functions for composite energy bands
• Efficient iterative schemes for<i>ab initio</i>total-energy calculations using a plane-wave basis set

Related

• Electrochemical Interfaces at Constant Potential: Data-Efficient Transfer Learning for Machine-Learning-Based Molecular Dynamics
• Learning charges and long-range interactions from energies and forces
• Machine learning accelerated finite-field simulations for electrochemical interfaces
• A constant potential reactor framework for electrochemical reaction simulations
• ELECTRODE: an electrochemistry package for atomistic simulations