Abstract:

Small cluster and periodic surface models with low coverages of intermediates are frequently employed to investigate reaction mechanisms and identify active sites on nanoparticles (NPs) in density functional theory (DFT) studies. However, diverse active sites on NPs cannot be sufficiently represented by these simple models, hampering the in-depth insights into the catalytic behavior of NPs. This paper describes the crucial roles of both model and coverage effect on understanding the nature of active sites for CO₂ reduction over Au and Pd NPs using DFT calculations. Terrace sites exhibit higher selectivity for CO than edge sites on Au NPs, which is opposite to the results on Au periodic surfaces. This contradiction reveals the computational model effect on clarifying active site properties. For Pd catalysts, the coverage effect is more significant. On bare Pd NPs and periodic surfaces, the selectivity for CO at edge sites is nearly identical to that at terrace sites, whereas edge sites display higher selectivity for CO than terrace sites in the case of high CO coverages. Through considering the more realistic models and the coverage effect, we successfully describe the size effect of Au and Pd NPs on CO selectivity. More importantly, this work reminds us of the necessity of reasonable models in DFT calculations.

Key words: Density functional theory, CO₂ reduction, Coverage effects, Catalyst model, Nanoparticles