Speaker
Description
Plasmonic catalysts such as metal nanoparticles harness the energy transfer between light, electrons and phonons at interfaces to drive chemical reactivity at interfaces.
However, even on clean metal surfaces with a regular structure, it is unclear whether these dynamics are the result of mode-selective energy transfer or photothermal heating effects.
Molecular dynamics simulations with electronic friction (MDEF) offer a quantum-classical description of electron-phonon coupling and have previously been used to model ultrafast dynamics on metal surfaces.
Using machine learning surrogate models to accelerate MDEF simulations, we show that mode-selective energy transfer has a negligible influence on light-driven hydrogen evolution from copper surfaces at low coverages.
We expect mode-selective energy transfer to play a stronger role at higher coverage, and show preliminary results for surface coverage dependence in laser-driven desorption from ruthenium as a function of different electronic friction approximations.