More than twenty years ago it was shown that gold, despite its limited adsorption properties, catalyzes reactions such as CO oxidation, [1] alkene hydrogenation, [2] and ethylene hydrochlorination.[3] However, strong interest in gold for catalysis emerged only recently owing to the seminal work of Haruta et al. [4] These authors showed that when gold is deposited on a metal-oxide support by a coprecipitation method, the metal exhibits high catalytic activity for CO oxidation at low temperatures. Moreover, they found that the distinctive behavior of the gold catalysts was due to the presence of nanosized Au particles.In a recent review, Bond and Thompson [5] showed that, in the case of Au-supported catalysts, factors such as Au particle size, preparation method, pretreatment conditions, and choice of the support play an important role in the CO oxidation activity of the final catalyst. Indeed, when nanoparticles of Au are deposited on TiO 2 , [4b, 6a] Co 3 O 4 , [4b, 6b] ZnO, [7] Fe 2 O 3 , [4b, 6a] or MgO, [8] the catalysts are particularly active for the above-mentioned reaction. In contrast, nanoparticles of Au were much less active when other supports were used. [5,9] These results suggest that a synergetic effect between the metal oxide support and Au may exist at the interface such that the metal oxide does not simply act as an inert carrier, but intervenes in the catalytic process.We thought that if gold nanoparticles on the order of 3 to 5 nm were essential for the catalysis (owing to the presence of a large number of surface gold atoms with properties different from those of bulk gold) and if a synergetic mechanism occurs at the gold and metal oxide interface, the characteristics of the metal oxide surface should be of paramount importance. In other words, if the support could be prepared in the form of discrete, well-defined nanoparticles, this should have an influence on the surface electronic properties and consequently on the gold-support interaction.
