The flexible surface or the rigid surface?

Abstract

In this paper we discuss aspects of the concept described by Somorjai as the “flexible surface”, and whether some surfaces can be considered to be inflexible, or rigid. We present STM results which appear to manifest both types of behaviour for surfaces, depending on their oxidation state. Copper metal surfaces can be classed as flexible, showing facile reconstruction in the presence of oxygen, whereas an oxidised Pd(110) surface shows no apparent diffusivity, even at 500 K. We go on to show data for a bulk oxide which indicates that sub-stoichiometry in the sample induces an element of flexibility in the surface, especially during reaction with oxygen. Finally, this is related to the direct observation of spillover on model catalysts of Pd nanoparticles supported on TiO₂. It must be recognised that flexibility relates to surface diffusivity and hence length- and time-scales. Surfaces which appear inflexible at short times may be flexible at long times. In relation to catalysis then, surface flexibility depends on the relationship between the time-scale of diffusive events on the surface and the catalytic turnover number.