The structure and catalytic behaviour of supported rhodium and rhodium/palladium nanoparticles

Energy Dispersive Extended X-ray Absorption Fine Structure (EDE), Transmission Electron Microscopy (TEM) and X-ray Photoelectron Spectroscopy (XPS) have been employed as principle techniques to investigate the local structure and catalytic properties of Rh and Rh/Pd nanoparticles supported on Y-AI2O3. The Rh catalyst systems were prepared with a metal loading & om 2.5wt% to 10wt%, and included systems prepared with chlorinated and nitrated precursors. The Rh/Pd systems were prepared with a combination of the metal components that did not exceed 5wt%. All of the Rh only systems exhibited a rapid room temperature oxidation in their '6esh' states, which was seen to be a recurring trend with the chlorinated systems, irrespective of particle size depicted by TEM. The oxidation, also delineated by XPS, under these conditions was seen on the highly dispersed systems, which would not be predicted by the low dispersed, bulk case. The nitrated Rh systems exhibited a reduced propensity for the oxidation, and EXAFS and TEM data reported that particle size effects were not the reason for the observed structural differences. Re-oxidation of the systems after exposure to H2 did appear to vary with metal loading, pointing to a structural factor. However, the dynamic equilibrium between the oxidised and reduced Rh phases was still present. These differences were not propagated under the catalytic conditions employed, which was the reduction of NO by H2. All of the systems exhibited virtually the same activity and selectivity, and the consequent production of N2O could therefore be assigned to be intrinsic to the bi-stable Rh nanoparticles. After 'light off and the collapse of the oxidised Rh phase, particulate Rh was seen to be very selective for the reduction of NO to N2. The introduction of an adjunct metal, palladium, to these systems to potentially remove the bi-stable Rh phase and unwanted production of N2O resulted in alloying between the components and a segregation of Pd to the surface of the particles, shown by XPS. Both the EDE and XPS results showed varying the loading of Pd could subsequently change the phase of the Rh, and therefore the Rh could be insulated against the aforementioned oxidation. However, although the phase was eliminated, the production of N2O was found to be an intrinsic catalytic property of Pd, and no 'intermediate' phase was found to exist. The study of a physically mixed sample found the system, somewhat surprisingly, to be the 'desired' catalyst in the scope of this study. The Pd component was found to 'communicate' with the Rh via what was most likely H2 spillover. This meant that although the oxidation of the discrete Rh component still transpired, the promotion of this phase to the highly selective metallic phase occurred at much lower light off temperatures compared to the alloyed systems. The absence of 'Pd-like' behaviour, when compared to the 'Pd-only' system, showed a physical result of the proposed Hz spillover. However, for the elimination of consequent N2O production and an understanding of the complex processes occurring over all of the systems studied, further work and concepts need to be evoked. In summary, the m time resolved, double edge EDE/MS experiments has allowed the observation of a range of dynamic processes occurring over the catalyst systems during the reduction of NO by H2. The complimentaiy probes utilised showed the physical behaviour of the systems in unprecedented detail, both statically and in time- resolved studies.