Metal supported oxide nanostructures: model systems for advanced catalysis J. Schoiswohl a , M. Sock a , Q. Chen b , G. Thornton b , G. Kresse c , M. G. Ramsey a , S. Surnev a , and F. P. Netzer a, * a Institut fu ¨r Physik, Oberfla ¨chen- und Grenzfla ¨chenphysik, Karl-Franzens-Universita ¨t Graz, A-8010 Graz, Austria b London Centre for Nanotechnology and Chemistry Department, University College London, London, UK c Institut fu ¨r Materialphysik, Universita ¨t Wien, A-1090 Wien, Austria Metal supported oxide nanostructures are discussed within the framework of the ‘‘inverse model catalyst’’ concept. We show that oxide nanostructures on metal surfaces may be regarded as artificial oxide materials, which display novel properties as compared to bulk oxide compounds and are stabilised by interfacial interactions and two-dimensional confinement effects. This is illustrated for prototypical examples of vanadium oxide overlayers on Rh(111) and Pd(111) surfaces. Structure and morphological changes of the oxide phase on V-oxide/Rh and V-oxide/Pd inverse catalyst surfaces are discussed, and the mass transport problem in catalyst systems during oxidation-reduction cycles is addressed. We demonstrate that the diffusion of oxide cluster over the metal surface provides a effective means of mass transport. The role of metal-oxide interface in determining the oxide nanolayer structure on particular substrate surfaces is investigated, and interfacial chemistry and interfacial strain effects are identified as important parameters. KEY WORDS: vanadium oxide; Rh; Pd; inverse model catalyst; STM; XPS. 1. Introduction The role of the oxide phase of oxide-supported metal catalysts in the course of catalytic reactions has been associated with a number of different effects, amongst them are: providing a high-area support for the stabili- sation of the metal phase in the form of small particles; influencing the metal phase via electronic interactions; or providing special reaction sites at the metal-oxide phase boundary [1]. A particularly drastic interaction phenomenon between the oxide and the metal phase has been reported in the form of the so-called ‘‘strong metal support interaction’’ (SMSI) effect [2,3], in which cata- lysts containing transition metal oxide supports display radically altered chemisorption and reaction behaviour after high temperature reduction treatments. A key feature of SMSI is the reversibility of the effect—the catalyst in the SMSI state can be converted back to its normal state by an oxidising treatment. The origin of the SMSI effects has been the subject of much debate over a number of years [4], but consensus of opinion is now in favour of the so-called encapsulation model, where the metal particles are covered with a thin layer of oxide in a reduced oxidation state [5]. The encapsulation of metal particles by a migrating reduced oxide phase of the carrier is associated with an interesting materials trans- port problem across the catalyst surface, which has been addressed in a recent publication [6], where the wetting and spreading of a reduced oxide phase across a metal surface and the subsequent dewetting of the reoxidised phase has been observed at the atomic level. In all the cases mentioned above the metal-oxide interface plays a crucial role in the process, as it also does in the general reactivity behaviour of oxide-supported metal catalysts. An interesting approach in the study of the metal- oxide interface is the concept of the ‘‘inverse’’ or ‘‘inverted’’ catalyst. While real catalysts consist of metal particles supported by an oxide carrier phase, the ‘‘inverse catalyst’’ is a metal surface, preferably a single crystal surface, which is decorated by oxide nanostruc- tures. Of course, the ‘‘inverse catalyst’’ is a model system which does not occur in practical industrial catalysis, but it is well suited to address a number of interesting aspects of the metal-oxide interface at the atomic level. For example, the promoting character of transition metal oxide minority phases in the CO and CO 2 hydrogenation reaction catalysed by group VIII metals has been demonstrated on such oxide decorated ‘‘inverse catalyst’’ surfaces by the Somorjai group [7,8], and bonding sites at the oxide-metal interface related to the Lewis acidity character of the systems have been sug- gested as the promoting agents. More recently, Hayek et al. [9] have related the catalytic activity for the CO hydrogenation reaction with structural and composi- tional parameters of vanadium oxide and cerium oxi- de—Rh ‘‘inverse catalysts’’, and Sock et al. [10] have investigated the adsorption and oxidation of CO on vanadium oxide—Pd(111) ‘‘inverse catalysts’’ to probe the influence of the metal-oxide phase boundary on neighbouring metal adsorption sites. The ‘‘inverse cat- alyst’’ model system has a number of technical advan- tages as compared to the real catalyst systems, amongst them the ability to apply surface science techniques with atomic precision for their characterisation [10,11]. In this paper we will concentrate on the structural and electronic characterisation of metal supported oxide * To whom correspondence should be addressed. E-mail: falko.netzer@uni-graz.at Topics in Catalysis Vol. 46, Nos. 1–2, September 2007 (Ó 2007) 137 DOI: 10.1007/s11244-007-0324-6 1022-5528/07/0900-0137/0 Ó 2007 Springer Science+Business Media, LLC