Molecular Dynamics Simulation for Electrochemical Systems - Challenges of Field Fluctuation

Explore molecular dynamics simulation techniques for electrochemical systems in this 45-minute conference talk that addresses the computational challenges of field fluctuation in ab initio molecular dynamics (AIMD) calculations. Learn about the complexities involved in applying macroscopic electric fields to study electrochemical processes at solid/liquid interfaces, including various methods for treating counter charges and boundary conditions that may not be available in standard DFT codes. Discover the implementation of electrified surface calculations under arbitrary fields using the VASP-python interface, which enables AIMD with applied electric fields through standard VASP code with enhanced control and flexibility. Examine recent developments in machine learning force fields (MLFFs) as cost-effective alternatives to expensive AIMD calculations, while understanding their limitations in accurately describing long-range electrostatic interactions. Investigate how traditional ML models trained on local charge distributions like Hirshfeld charge or Wannier centers fail to reproduce macroscopic electrostatics, and explore a proposed alternative model trained on macro dipole moments that can predict macroscopic electric fields for realistic electrochemical systems.