DOE OSTI2017
The surface atomic structure of metallic nanoparticles (NPs) plays a key role in shaping their physicochemical properties and response to external stimuli. Not surprisingly, current research increasingly focuses on exploiting its prime characteristics, including the amount, location, coordination, and electronic configuration of distinct surface atomic species, as tunable parameters for improving the functionality of metallic NPs in practical applications. The effort requires clear understanding of the extent to which changes in each of these characteristics would contribute to achieving the targeted functionality. This, in the first place, requires good knowledge of the actual surface of metallic NPs at atomic level. Through a case study on Au–Pd nanoalloy catalysts of industrial and environmental importance, we demonstrate that the surface atomic structure of metallic NPs can be determined in good detail by resonant high-energy X-ray diffraction (HE-XRD). Furthermore, using our experimental surface structure and CO oxidation activity data, we shed new light on the elusive origin of the remarkable catalytic synergy between surface Au and Pd atoms in the nanoalloys. In particular, we show that it arises from the formation of a specific “skin” on top of the nanoalloys that involves as many unlike, i.e., Au–Pd and Pd–Au, atomic pairs as possible given the overall chemical composition of the NPs. Moreover, unlike atoms from the “skin” interact strongly, including both changing their size and electronic structure in inverse proportions. That is, Au atoms shrink and acquire a partial positive charge of 5d-character whereas Pd atoms expand and become somewhat 4d-electron deficient. Accordingly, the reactivity of Au increases whereas Pd atoms become less reactive, as compared to atoms at the surface of pure Au and Pd NPs, respectively. Ultimately, this renders Au–Pd alloy NPs superb catalysts for CO oxidation reaction over a broad range of alloy compositions. Our findings are corroborated by DFT calculations based on a refined version of d-band center theory on the catalytic properties of late transition metals and alloys. Here, we discuss opportunities for improving the accuracy of current theory on surface-controlled properties of metallic NPs through augmenting the theory with surface structure data obtained by resonant XRD.