Heterogeneous catalysts consisting of supported metallic nanoparticles typically derive exceptional catalytic activity from their large proportion of undercoordinated surface sites which promote adsorption of reactant molecules. Simultaneously, these high energy surface configurations are unstable, leading to nanoparticle growth or degradation and eventually a loss of catalytic activity. Surface morphology of catalytic nanoparticles is paramount to catalytic activity, selectivity, and degradation rates, however it is well-known that harsh reaction conditions can cause the surface structure to change. Still, limited research has focused on understanding the link between nanoparticle surface facets and degradation rates or mechanisms. Here, we study a model Au supported catalyst system over a range of temperatures using a combination of in situ transmission electron microscopy, kinetic Monte Carlo simulations, and density functional theory calculations to establish an atomistic picture of how variations in surface structures and atomic coordination environments lead to shifting evolution mechanisms as a function of temperature. By combining experimental results, which yield direct observation of dynamic shape changes and particle sublimation rates, with computational techniques, which enable understanding the fundamental thermodynamics and kinetics of nanoparticle evolution, we illustrate a two-step evolution mechanism in which mobile adatoms form through desorption from low-coordination facets and subsequently sublimate off the particle surface. By understanding the role of temperature in the competition between surface diffusion and sublimation, we are able to show how individual atomic movements lead to particle scale morphological changes and rationalize why sublimation rates vary between particles in a system of nearly identical nanoparticles.
Keywords: density functional theory; image segmentation; in situ transmission electron microscopy; kinetic Monte Carlo; particle morphology; supported catalyst.