The practical application of single atom catalysts (SACs) is often limited by stability issues and the lack of understanding of how reaction conditions affect the mobility of the metal adatoms on the catalyst support. In this study, density functional theory (DFT) calculations are used to explore how the redox state of the catalyst surface, modulated by the gaseous atmosphere, impacts the stability of Ag and Cu adatoms on anatase TiO
2. Potential energy surfaces (PES) are constructed for single atom diffusion under various redox states of the catalyst. An inverse volcano effect is found between adatom diffusion barriers and surface redox state, with highly oxidizing or reducing conditions enhancing the single atom catalyst stability. Kinetic Monte Carlo (KMC) simulations translate the stochastic behavior of single atoms into observable macroscopic properties and are used to estimate the average diffusion coefficient of metal adatoms on the surface, highlighting the inhibitory effect of surface hydroxylation and oxygen vacancies on the mobility of supported SACs.
To further explore agglomeration dynamics, the nucleation and growth of metallic particles (Agn and Cun, with n up to 20 ) on titania are investigated using machine learning (ML) and DFT. ML is exclusively used to screen and pre-optimize plausible cluster geometries and transition states, and DFT is employed to fully relax the selected pre-optimized structures. The nucleation energies and energy barriers associated with cluster growth are input into the KMC model, enabling prediction of cluster size distributions as functions of temperature, metal loading, oxygen vacancy concentration, and surface hydroxylation degree. The final KMC model incorporating both diffusion and growth serves as a predictive tool for designing regeneration strategies and tailoring surface properties to promote long-term catalyst stability.
