The influence of a chlorinated precursor on the redox properties of ceria was studied by comparing two Pt/CeO2-Al2O3 catalysts prepared from hexachloroplatinic acid and platinum acetylacetonate. In absence of chlorine, a dispersion value higher than 100% has been measured by irreversible hydrogen chemisorption. To interpret this result, the interaction of H-2 with the support and the two catalysts were studied by magnetic measurements. The variations of the Ce3+ content have given evidence that different redox processes may occur. The hydrogen spillover is clearly observed in presence of platinum with the phenomenon being more pronounced for the ex-acetylacetonate catalyst. The limited amount of hydrogen irreversibly adsorbed at 294 K on the support for this free chlorine reduced catalyst is sufficient to explain the excessive value obtained for the platinum dispersion. The results show a fixation of chlorine on the supported ceria phase which inhibits the mobility of hydroxyl species responsible for the hydrogen spillover. Moreover, independently to the presence of chlorine, a reduction at 573 K seems to stabilise a significant fraction of Ce3+ ions against reoxidation under air at room temperature. This was not the case for bulk ceria and probably originates from ceria/alumina interactions occurring after reduction. (C) 1999 Elsevier Science B.V. All rights reserved.

The accessible metallic area measured by hydrogen chemisorption on model ceria–alumina supported platinum or(and) rhodium catalysts has been compared to the results obtained in cyclohexane aromatization activity tests. The benzene formation was studied at 573 K after reduction by hydrogen at the same temperature. The turnover frequency, calculated on the basis of hydrogen chemisorption data, was found to depend on the type of support, ceria or alumina, in interaction with the precious metal particles. However, for the catalysts supported on ceria–alumina, the activity per metallic platinum or rhodium site was almost the same for both fresh catalysts or after hydrothermal aging at 1273 K. The possible use of the cyclohexane aromatization activity measurement for the determination of the metallic area of three-way catalysts is examined. Compared to hydrogen chemisorption, this method presents the advantage of a better sensitivity for measuring very low dispersions of aged catalysts, typically dispersions inferior to 3%. In the discussion, the possible influence of hydrogen spill-over effects and of the connected role of chlorine are considered. Although not directly studied, the possible effect of sulphur present in commercial catalysts aged in vehicles is also discussed.

The ceria surface area of a commercial Pt-Rh three-way catalyst was determined after laboratory hydrothermal aging at 1173-1373 EC and after 200 h on engine bench. It was measured by X-ray diffraction (XRD) line broadening analysis and by a method based on the exploitation of the hydrogen temperature programmed reduction (TPR) profiles. In this case, the hydrogen uptakes below about 900 K include the ceria surface reduction and that of the oxidized noble metals. They are analyzed and discussed, assuming two possiblities for the metals oxidation state.Compared to the fresh catalyst, the TPR profiles are deeply modified by the aging treatments. The ceria seems to sinter more than alumina, particularly between 1173 K and 1273 K. After aging at 1273-1373 K, the calculated ceria surface area is only 15-10 m(2)g(-1) washcoat, which represents 20% of the BET area, instead of 40% initially. A stabilization treatment at 823 K under reactants leads to an additional ceria sintering, even for the more aged system. Finally, the measurements on the engine bench aged catalysts seem to indicate a better resistance of ceria to sintering in working conditions. The presence of a pollutant layer, containing phosphorus, zinc and calcium, did not modify the accessible ceria surface area measured by TPR.

The surface area of cerium oxide was tentatively determined in model three-way catalysts using (i) a methodology based on the exploitation of the hydrogen temperature-programmed reduction (TPR) profiles and (ii) the adsorption of CO2on the hydroxyl groups of alumina followed by FTIR spectroscopy. These two methods were performed on a ceria–alumina support after impregnation with RhCl3and H2PtCl6and calcination under nitrogen at 773 K. Whereas the two methods gave agreeing results on the initial support, the results were not so straightforward for catalysts. In this case, the TPR profiles are deeply modified with respect to the support. Due to the presence of platinum and rhodium, the reduction peak corresponding to the ceria surface reduction is shifted toward lower temperatures. It also includes the quantity of hydrogen necessary for the reduction of the precious metal oxide. After examining the metal mean oxidation degree, it has been possible to calculate the ceria surface area in the catalyst. A 20–30% decrease was found compared to the initial support. On the contrary, with the CO2-FTIR method, the calculated ceria surface area increases after impregnation. It can be attributed to the fact that the density of OH species responsible for CO2adsorption is modified by the presence of chlorine introduced during the preparation. These results are discussed and compared to those for the support alone. A reexamination of the TPR curves has permitted the extent of ceria surface not in contact with precious metal to be taken into account. It appears that 10 to 30% of the initial ceria surface would remain uncovered by the precious metals after impregnation. Finally, the TPR method can be considered as the most reliable and potentially rich source of informations on the catalyst surface state.

The accessible surface area of cerium oxide was measured on model ceria-alumina supported Pt and/or Rh catalysts submitted to an hydrothermal treatment at 1273 K (10% H2O in N-2). Two distinct methods, developed for fresh catalysts, were utilized. They are based respectively on the hydrogen temperature programmed reduction (TPR) profiles and the infrared spectra of the specific adsorption of CO2 on the hydroxyl groups of alumina. After aging, the TPR curves have become flattened and it is difficult to have a good reliability in the deconvolution of the peak corresponding to the ceria surface reduction. On the support alone, the results of both methods are in agreement and indicate a decrease of the ceria surface area by a factor close to 4, whereas the total surface area is only divided by 2. It appears that cerium oxide sinters more rapidly than alumina. For the metallic catalysts, the calculated ceria surface areas values depend on the hypothesis done on the oxidation state of the precious metals before the TPR. If reliable results can be obtained in the case of rhodium with O/Rh = 1.5, in presence of platinum, the mean oxidation state of platinum O/Pt = 1 results in underestimated values. The ceria surface areas deduced from the infrared results are higher than those deduced from TPR. Although the former are uncertain, they are in good agreement with a lower coverage of the alumina surface by sintered cerium oxide particles.

The influence of the reduction temperature on the accessibility of the metallic surface was studied on model ceria-alumina supported platinum or rhodium catalysts. For a 0.5% Pt-Ce/Al solid, the HM values, deduced from hydrogen irreversible chemisorption, decrease deeply with Tr, the reduction temperature, from 60% at Tr = 300°C to 19% at Tr = 500°C. This can be attributed to strong interactions between ceria and platinum, since, the initial H2 chemisorption could be restored after reoxidation. The presence of BaSO4 in the support accelerates the loss of metallic area, because of sulfur poisoning of the platinum surface. For Tr = 300°C, the dispersion values were in agreement with those deduced from FTIR spectroscopy of adsorbed CO. In the case of rhodium, a 37% H/M dispersion was obtained, which did not change when Tr was increased from 300 to 500°C. For two industrial Pt-Rh three-way catalysts, the behaviour was found similar to that of platinum, the amount of chemisorbed hydrogen decreasing for Tr > 350°C. Thus, in the three-way catalysts characterization, the maximum metal accessibility is obtained after a reduction at 300°C.

Gravimetric temperature programmed oxidation was used to study the combustion of a diesel soot mixed with copper catalysts supported on La2O3 or La2O2CO3. In a first step, different systems associating copper oxide with an other metal oxide were prepared and tested in presence of SO2. The association of copper and niobium was found the most active. The influence of alkali on the activity was also studied. It results that potassium is the most effective in lowering the combustion temperature domain in agreement with literature. Finally, CuNbK catalysts deposited on lanthanum oxide have an improved catalytic activity at low temperatures compared to CuVK or CuMoK/TiO2, reported in literature. For this catalyst, the maximum oxidation rate was observed at ca. 300°C with the combustion starting at about 250°C. A similar behaviour is obtained when replacing Nb by Ta or the support La2O3 by either La2O2CO3 or TiO2.

Cu/Al2O3 solids with various Cu loadings, between 0.3 and 6.4 wt %, are used for the reduction of NO by propane in the absence and in the presence of oxygen (up to 10 vol. %) in the 423 - 773 K temperature range. At a given temperature and for high Cu loadings, the introduction of oxygen induces a decrease in the activity in nitrogen formation. For low Cu loadings the activity increases with the oxygen content in the 1-2 vol. % range, then slightly decreases for higher oxygen amounts. The nature of the Cu species accessible to CO and NO is determined by infrared spectroscopy. High Cu loadings favor the formation of bulk oxides at the surface of the support whereas low Cu loadings favor the formation of isolated Cu species. The selective reduction of NO is thus related to the presence of these isolated copper species easily reduced and reoxidized.

In a general study of the role played by each component of three-way catalysts, a selective measurement of alumina and ceria surfaces area has been developed on model ceria-alumina supports. Three CeO2/Al2O3 solids (6.5, 13.4, 21 wt.% CeO2) were prepared by grafting Ce acetylacetonate on the surface of an alumina and calcination at 673 K. They were characterized by BET and X-ray diffraction measurements. The hydrogen temperature-programmed reduction (TPR) resulted in profiles similar to those of unsupported cerias, thus allowing an estimation of the equivalent surface area of ceria alone from the hydrogen uptake at low temperature. By adsorbing CO2 on the OH groups of the alumina surface, hydrogenocarbonates species were formed and the IR band at 1235 cm(-1) was selected to determine quantitatively the unpertubed alumina surface. In that case, by comparison with the BET surface area, it was possible to deduce an other value of the ceria surface area. The two sets of values obtained by these independent methods were found in good agreement, thus giving a good support for the reliability of each method. They show that the unpertubed alumina surface decreases when increasing the cerium content. Only 20% of the original alumina surface is preserved in the support containing 21 wt.% CeO2.

Copper ions were introduced in a Na-ZSM-5 zeolite by using exchange, impregnation and precipitation procedures. The catalytic activity of the three solids was measured in the reduction of NO by propane in the presence of oxygen (0 to 10 vol.-%) in the 423–773 K temperature range. At a given temperature the activity for nitrogen formation remarkably increased with the oxygen content in the 0.5–2 vol.-% range, then slightly decreased for higher amounts of oxygen. For an oxygen content exceeding 0.5 vol.-% and regardless of the reaction temperature, the Cu-ZSM-5 sample prepared by exchange was the most active catalyst, the catalytic activity being expressed per gram of introduced copper. The adsorption of carbon monoxide followed by FT-IR spectroscopy suggested that copper was present as isolated Cu+ ions in the three catalysts. The dispersion of cuprous ions was the highest for the sample prepared by the exchange process. The catalytic activity appeared related to a high dispersion of isolated Cu+ ions in the MFI framework.