(366y) Understanding the Role of Potential and Ion Concentration on the Structure of the Electric Double Layer Using Machine-Learning Interatomic Potential Simulations

Research Interests: Machine Learning for Property Prediction, Application of Machine Learning Interatomic Potentials to Materials Discovery, Application of Machine Learning Interatomic Potentials to Heterogenous Catalysis, Homogeneous Catalysis

Electrode-electrolyte interactions such as the formation of an electrical double layer (EDL) play a crucial role in the performance of electrocatalytic systems. Depending on the specific application, the EDL can either promote or hinder electrochemical reactions at the interface. For example, the EDL can facilitate charge transfer between the electrode and electrolyte in fuel cells, but can also influence kinetics and selectivity in electrochemical CO₂ reduction by affecting the adsorption and activation of solvated CO₂. Despite the importance of the EDL, its molecular structure under different potential and ion concentration conditions are not well-understood in general. In this work, we have developed a general machine-learning interatomic potential framework which enables the simulation of equilibrated EDLs under different potential conditions while maintaining DFT-level accuracy. As a case study, we model a Na₂SO₄(aq) electrolyte solution in contact with suspended graphene electrodes and study how varying ion concentrations and potential affect the structure of the EDL. To validate the computational results, we calculate the sum frequency vibrational spectra of the EDL using our computational framework and compared it with benchmark measurements. We show that MLIPs using an equivariant architecture can reproduce experimental sum frequency vibrational spectra and MLIP-driven molecular dynamics simulations give insights into the segregation of ions that lead to the formation of the EDL under different ion concentration and potential conditions.