Anomalous Diffusion of Metal Atoms on Oxide Surfaces: A Machine Learning Molecular Dynamics Study of Pt₁/TiO₂

We investigate the high-temperature evolution of the metal–support interface in atomically dispersed catalysts to probe the dynamical evolution of the metal atom. Since high computational costs limit the time scales of ab initio molecular dynamics (AIMD) trajectories to a few picoseconds, we deploy a machine learned interatomic potential (MLIP), trained on AIMD data, using the FLARE program. We generate multiple trajectories spanning at least 50 ps each for atomically dispersed Pt on rutile TiO2 (110), initiated at different positions on the stoichiometric surface. We find that the motion of Pt is subdiffusive (not Brownian) in most cases, quantified by the parameters constituting the fractional Fokker-Planck equation. The subdiffusive behavior originates in strong interactions of Pt with bridging oxygen atoms of the support. Several trajectories show that Pt mobility is quenched when it coordinates with two bridging oxygen atoms to form a near-linear O–Pt–O complex. Diffusion that resembles Brownian motion is observed only at the highest simulation temperature examined (1000 K) for sites at which Pt is far from a bridging oxygen. The study therefore shows that in the absence of surface defects or adsorbates, the thermal stability of the metal atom is determined by coordination with bridging oxygen atoms.