Surface Caracterisation of  Car Exhaust Catalysts

	Overview   All modern car exhausts are fitted with an exhaust catalyst to reduce harmful emissions of hydrocarbons, carbon monoxide and nitrogen oxides into the atmosphere. The role of these catalysts is to oxidise hydrocarbons and CO to CO2 while reducing NOx to N2. Ceria is widely used as a promoter for alumina-noble metal exhaust catalyst and more recently zirconium has been introduced into the ceria matrix to enhance the thermal stability of the catalyst. X-ray photoelectron spectroscopy is ideally suited to the surface characterisation of such materials, where the surface of the material will determine the catalytic activity.
Figure 1: X-Ray photoelectron survey scan from the "as introduced" catalyst surface.	The Catalyst Samples   The Ce-Zr mixed metal oxide catalysts were prepared by co-precipitation of the metal salts. After washing and drying, the resulting finely divided powders were pressed into 10mm diameter discs and annealed at 900°C to form robust pellets. These pellets were introduced into vacuum and analysed by XPS without further surface preparation. Ce1-x Zrx O2 samples are known to be insulators and therefore subject to charging during surface analysis. The configuration of the AMICUS instrument means that sample charging due to the photoelectron emission process is minimised. Where sample charging was observed, peaks have been shifted so as to align adventitious hydrocarbon C 1s photoemission to 285eV binding energy.
Element As Introduced (atomic %) After Sputter (atomic %) cerium 10 18 zirconium 10 11 oxygen 49 53 carbon 31 18   Table 1: Quantification results of Ce1-xZrxO2 surface as introduced and after 450 seconds Ar+ sputtering.	Determination of Surface Composition   Determination of the surface composition from the survey scan shown in Figure 1 and presented in Table 1, shows that the 1:1 Ce:Zr stoichiometry in the bulk is maintained at the surface. Carbon is also observed at the surface and is thought to be due to hydrocarbon contamination before introduction of the sample into vacuum.
Figure 2: High resolution scan over the Ce 3d region       (a) sample as introduced, and       (b) after a total of 450s Ar+ sputtering. Figure 3: Narrow scans over the O 1s region       (a) sample as introduced, and       (b) after a total of 450s Ar+ sputtering.	Chemical State Analysis of Cerium   There has been considerable debate in the literature on the assignment of the CeO2 3d photoelectron peaks due to the complexity of the spectra. The peaks are spin-orbit split into a doublet, with each doublet showing further structure due to final state effects. Using the notation of Burroughs et al.1 the peaks are assigned for Ce(IV)O2 as v, v'' and v''' for Ce 3d5/2, with the corresponding Ce 3d3/2 peaks labelled u, u'' and u'''. An additional doublet is also observed due to the presence of Ce2O3 (Ce in oxidation state (III)) and is assigned v' and u'. The "as introduced" sample shows structure in the Ce 3d photoemission spectrum due to both CeO2 and Ce2O3, implying that cerium is present at the surface in both Ce(IV) and Ce(III) oxidation states. The Ce 3d photoemission spectrum is shown in Figure 2(a) annotated to show peaks due to both CeO2 and Ce2O3 for the "as introduced" catalyst sample. Due to a rearrangement of bonding electrons, a progressive reduction of CeO2 to Ce2O3 will affect the Ce 3d spectrum causing a decrease in the u'', u''', v'', v''' peaks with an associated increase in v' and u'. This effect is observed as the sample is Ar+ ion sputtered due to the preferential removal of oxygen from the surface. The relative peak areas of the features assigned u' and v' are seen to increase in Figure 2 after Ar+ ion sputtering for 450 seconds, indicating an increase in the surface content of Ce2O3 The O 1s photoemission peak, shown in Figure 3, is composed of two peaks due to the nonequivalence of the surface oxygen chemical environments. Previously, the high binding energy feature has been assigned to extraneous surface materials remaining from sample preparation, such as a surface carbonate species. A further suggestion presented by Pfau et al.2 is the formation of Ce(III) related surface defects, where oxygen occupies additional lattice sites, resulting in an O 1s core level peak at 2.4eV higher binding energy. This suggestion is supported by this work as the proportion of the O 1s high binding energy feature increases relative to the primary O 1s peak as the Ce(III) surface content increases with increasing exposure of the surface to Ar+ ion sputtering. Further evidence that the O 1s high binding energy peak is not due to surface carbonate is demonstrated by considering the C 1s region. With each sputter cycle the C 1s signal is seen to decrease such that after 750 seconds sputtering the surface, the C 1s becomes negligible, whereas the high binding energy O 1s peak increases in intensity and is at a maximum after 750 seconds sputtering.
	Acknowledgements   Samples were provided by Prof. Wendy Flavell, UMIST.
	References   1. P. Burroughs, A. Hamnett, A.F. Orchard and G. Thornton, J. Chem. Soc. Dalton Trans., (1976) 1686. 2. A. Pfau and K.D. Schierbaum,Surf. Sci., 321 (1994) 71.
Summary   The surface characterisation the Ce1-xZrxO2 car exhaust catalysts has shown that the AMICUS instrument can easily be employed to provide quantitative information on sample composition. The features of AMICUS demonstrated in this application include: A high intensity X-ray source allowing good quality photoemission data to be collected within a few minutes Excellent energy resolution of the analyser allowing chemical state information to be determined Quantitative nature of X-ray photoelectron spectroscopy enabling surface composition to be deduced

previous Applications Index next