Accelerating and Enhancing Thermodynamic Simulations of Electrochemical Interfaces

Area of Science:

• Materials Science
• Electrochemistry
• Computational Chemistry

Background:

• Electrochemical interfaces are vital for catalysis, energy storage, and corrosion.
• Predicting stable surface structures is challenging due to complex interactions and limitations of current methods like Pourbaix diagrams and static ML potentials.
• Existing methods often neglect dynamic surface transformations and full thermodynamic equilibration with the environment.

Purpose of the Study:

• To extend the Virtual Surface Site Relaxation-Monte Carlo (VSSR-MC) method for autonomous sampling of surface reconstructions under aqueous electrochemical conditions.
• To accurately and efficiently predict surface energetics and electrochemical stability.
• To provide a scalable framework for understanding and designing materials for electrochemical applications.

Main Methods:

• Fine-tuning foundational Machine Learning (ML) force fields.
• Extending the VSSR-MC method to autonomously sample surface reconstructions.
• Modeling under aqueous electrochemical conditions, explicitly accounting for bulk-electrolyte equilibria.

Main Results:

• Accurate and efficient prediction of surface energetics.
• Successful recovery of known Pt(111) surface phases.
• Discovery of new LaMnO3(001) surface reconstructions.
• Enhanced predictions of electrochemical stability.

Conclusions:

• The VSSR-MC method, enhanced with fine-tuned ML force fields, provides a powerful tool for predicting electrochemical interface stability.
• The framework successfully models dynamic surface transformations and bulk-electrolyte equilibria.
• This approach offers a scalable and accurate method for designing advanced materials for electrochemical applications.