Interactions of metal particles with oxide supports can radically enhance the performance of supported catalysts. At the microscopic level, the details of such metal-oxide interactions usually remain obscure. This study identifies two types of oxidative metal-oxide interaction on well-defined models of technologically important Pt-ceria catalysts: (1) electron transfer from the Pt nanoparticle to the support, and (2) oxygen transfer from ceria to Pt. The electron transfer is favourable on ceria supports, irrespective of their morphology. Remarkably, the oxygen transfer is shown to require the presence of nanostructured ceria in close contact with Pt and, thus, is inherently a nanoscale effect. Our findings enable us to detail the formation mechanism of the catalytically indispensable Pt-O species on ceria and to elucidate the extraordinary structure-activity dependence of ceria-based catalysts in general.

Electronic interactions between metal nanoparticles and oxide supports control the functionality of nanomaterials, for example, the stability, the activity and the selectivity of catalysts. Such interactions involve electron transfer across the metal/support interface. In this work we quantify this charge transfer on a well-defined platinum/ceria catalyst at particle sizes relevant for heterogeneous catalysis. Combining synchrotron-radiation photoelectron spectroscopy, scanning tunnelling microscopy and density functional calculations we show that the charge transfer per Pt atom is largest for Pt particles of around 50 atoms. Here, approximately one electron is transferred per ten Pt atoms from the nanoparticle to the support. For larger particles, the charge transfer reaches its intrinsic limit set by the support. For smaller particles, charge transfer is partially suppressed by nucleation at defects. These mechanistic and quantitative insights into charge transfer will help to make better use of particle size effects and electronic metal-support interactions in metal/oxide nanomaterials.

Platinum is the most versatile element in catalysis, but it is rare and its high price limits large-scale applications, for example in fuel-cell technology. Still, conventional catalysts use only a small fraction of the Pt content, that is, those atoms located at the catalyst's surface. To maximize the noble-metal efficiency, the precious metal should be atomically dispersed and exclusively located within the outermost surface layer of the material. Such atomically dispersed Pt surface species can indeed be prepared with exceptionally high stability. Using DFT calculations we identify a specific structural element, a ceria "nanopocket", which binds Pt(2+) so strongly that it withstands sintering and bulk diffusion. On model catalysts we experimentally confirm the theoretically predicted stability, and on real Pt-CeO2 nanocomposites showing high Pt efficiency in fuel-cell catalysis we also identify these anchoring sites.