Current Research

Machine-learning molecular dynamics applied to energy materials and structural ceramics — connecting atomistic simulations to electrochemical and mechanical observables.

• Pt corrosion in fuel cells. The full Pt degradation sequence in PEMFC cathodes, simulated with the PtOH GAP potential and TurboGAP molecular dynamics.
• Ta_1−xHf_xC ultra-high-temperature ceramics. Phase stability, mixing thermodynamics, and fracture of a solid-solution UHTC candidate for hypersonic leading edges and thruster nozzles.
• General-purpose CuAu alloy GAP. A machine-learning interatomic potential for the full Cu–Au binary system, with direct application to CO₂ reduction catalysis on nanoparticles.