Crystal phase effect and site confinement effect of metals in heterogeneous catalysis

Abstract

The rapid development of nanotechnology facilitates the design and synthesis of unusual nanomaterials with well-controlled atomic structures, such as metal nanostructures with unconventional crystal phases and confined single metal atoms. An important application of these novel materials is heterogeneous catalysis. The aim of this thesis is to study the catalytic performance of them by using density functional calculations. Specifically, this thesis mainly focuses on oxygen activation and reduction, which is important in both thermal catalysis and electrocatalysis, such as selective oxidation reactions and oxygen reduction reactions (ORR). Gold nanostructures are well-known catalysts for selective oxidation reactions and the crystal phase effect is studied in this thesis. Atomically dispersed Me/N/C (Me = Fe, Co, etc.) are promising nonprecious catalysts for electrochemical ORR, and the reaction mechanism on these singe-atom sites is also studied here.

Crystal phase engineering is a promising strategy to tune the catalytic performance of metal nanomaterials. Generally, the crystal phase effect on catalysis is ascribed to distinct surface atomic arrangements of catalysts with different crystal phases. This thesis shows that even for similar surfaces, such as the close-packed surfaces, different crystal phases have considerably different surface reactivity due to their distinct intrinsic surface strains. From first-principles calculations, it is found that the close-packed surfaces of hexagonal close-packed (HCP) and double HCP (4H) gold have significantly smaller intrinsic strains (~1.3%) than that of face-centered cubic (FCC) gold (~2.3%). These distinct intrinsic surface strains result in various oxygen adsorption energies and O2 dissociation barriers on these close-packed gold surfaces, and the dissociation of O2 on different crystal phases and surfaces follows the Brønsted-Evans-Polanyi principle.

The electrochemical oxygen reduction reaction (ORR) mechanism was generally considered to be O2 → OOH* → O* → OH* → H2O (O* mechanism). This O* mechanism predicted reasonable ORR half-wave potential (E1/2) of Co/N/C but abnormally underestimated the one of Fe/N/C. This thesis highlights an unconventional 2OH* ORR mechanism (O2 → OOH* → 2OH* → OH* → H2O), which was often ignored because the free energies (ΔG) of 2OH* and O* are equal according to the famous scaling relation: 2ΔG(OH*) = ΔG(O*). This scaling relation is true for traditional catalysts with near-continuous active sites. This thesis shows a different scaling relation: ΔG(2OH*) = ΔG(O*) + 1.5 eV on single-atom catalysts (Me/N/C, Me = Fe, Co, etc.) because of the confinement of two OH* on the same active site, and suggests that the 2OH* mechanism should not be overlooked. In consideration of both O* and 2OH* mechanisms, the ORR E1/2 of Co/N/C and Fe/N/C are in good agreement with experimental results. This thesis reveals the site confinement effect on ORR reaction mechanisms and scaling relations in single-atom catalysis, and it is also heuristic for other reactions such as O2 evolution and N2 reduction on single-atom catalysts.