Are particulate noble-metal catalysts metals, metal oxides, or something in-between?

A good model? Noble-metal particulate catalysts often require small amounts of oxygen to obtain optimal activity. However, the structure and stoichiometry of the oxidized metal clusters involved remains obscure, even almost two hundred years after their discovery. A heteropolypalladate salt (see picture; Pd yellow, O red) now offers a view of how oxygen might be incorporated into small noble-metal clusters.