Heinz J. Robota

Physicochemical properties of Ce-containing three-way catalysts and the effect of Ce on catalyst activity

The activity of Pt,Rh,Ce/γ-Al2O3 and Pt,Rh/CeO2 catalysts has been studied in a full complement synthetic exhaust gas mixture consisting of H2, CO, C3H6, C3H8, NO, O2, N2, CO2, H2O, and SO2. Direct interaction between Pt and CeO2 was shown to lead to large improvements in catalyst performance after activation of the catalyst in the synthetic exhaust gas. Catalyst activation was shown to be due to reduction of the noble metals and surface Ce and this state of the catalyst was found to be particularly effective for CO oxidation. The key component responsible for activation was shown to be H2; even at the low H2 level present in the exhaust gas (≤.26 vol%), reduction is shown to be very easy. The degree of Pt/Ce interaction, and thus activity, after catalyst activation could be controlled by varying the CeO2 crystallite size. Decreasing the CeO2 crystallite size led to greater Pt/Ce interaction as shown by TPR and STEM analysis and resulted in greater activity for both fresh and laboratory aged catalysts. Direct Pt/Ce interaction also led to a synergistic reduction of Pt and surface Ce which we found to qualitatively correlate with catalyst performance after activation.

Deactivation in Cu-ZSM-5 lean-burn catalysts

Abstract Cu-ZSM-5 catalysts that exhibit fair fresh performance for nitric oxide reduction by hydrocarbons under oxidizing conditions also exhibit facile deactivation. Brief periods of catalytic use at temperatures of 600–800°C result in substantial deactivation. The higher the temperature, the more severe the deactivation. Besides causing performance deactivation, these catalytic treatments result in substantial losses of micropore volume. Under our conditions, these pore volume losses appear not to be caused either by carbon (which could accumulate as “coke”, conceivably) or by dealumination. Surface decomposition of the zeolite could occur but the data suggest this should be slow (if it occurs at all) with respect to deactivation. Substantial sintering of copper species to CuO, and perhaps also to Cu2O, is observed. Appearance of these phases is accompanied by losses in zeolite crystallinity as probed by X-ray diffraction. We believe CuO crystallites grow primarily in interior surfaces of the zeolite. As these crystallites exceed critical sizes imposed by zeolite structural constraint, copper sintering requires local destruction of the zeolite. This destruction may be the main cause of the loss of micropore volume. Sintering of active copper into inactive phases, such as CuO, together with resultant disruption of zeolitic crystallinity and porosity, appears to be the primary cause of catalyst deactivation under our conditions. Thus, the catalytically active component of ZSM-5, not the supporting zeolitic framework, ages and deactivates.

In situ XANES characterization of the Cu oxidation state in Cu-ZSM-5 during NO decomposition catalysis

The oxidation state of Cu in Cu-ZSM-5 has been investigated by the X-ray absorption near-edge structure (XANES) spectroscopic method during NO decomposition catalysis. We designed an in situ reactor system with which we can measure the relative NO decomposition rate while taking XANES spectra. We observed that the 1s→4p electronic transition of Cu(I) in Cu-ZSM-5 appears as a narrow, intense peak which is an effective measure of changes in the population of copper oxidation states. This transition is quite intense after Cu-ZSM-5 is activated in inert gas flow. However, its intensity decreases but by no means disappears after the admission of a NO/N2 gas mixture. We conducted the reaction in a temperature cycle around the optimum conversion temperature of 773 K and recorded the XANES at each temperature. We observed that the integrated intensity of the Cu(I) 1s→4p transition, which is proportional to the cuprous ion concentration in Cu-ZSM-5, was well correlated with the NO decomposition rate. This finding supports the conjecture that Cu(I) participates in a redox mechanism during catalyzed NO decomposition in Cu-ZSM-5 at elevated temperature.

Structure-function properties in Cu-ZSM-5 no decomposition and NO SCR catalysts

Abstract Physicochemical characteristics of Cu-ZSM-5 relevant to its abilities to catalyze the decomposition of NO to the elements and to catalyze selective catalytic reduction (SCR) of NO by hydrocarbons are reviewed. An emphasis is placed on in situ characterization of the catalyst under bona-fide catalytic conditions by X-ray absorption methods. Copper chemistry has central importance. Critical review of literature results demonstrates that Bronsted acidity of the zeolite is not a necessary factor in NO SCR. A model rationalizing ‘excessive’ copper ion exchange levels suggests that the exchanging species are not Cu 2+ cations, as assumed formally, but rather [CuOH] + and its related oligomers. The working catalyst exhibits dynamic Cu(II) → Cu(I) conversions under catalytic conditions. Cu(I) can be the predominant oxidation state under SCR conditions and viable reaction mechanisms will have to take this into account. Reaction pathways for NO decomposition and NO SCR are similar but, in all likelihood, somewhat different. A brief resume is provided concerning adsorbate chemistry which may be important in the mechanisms of both the decomposition and the SCR reactions.

In situ characterization of Cu-ZSM-5 by X-ray absorption spectroscopy: XANES study of the copper oxidation state during selective catalytic reduction of nitric oxide by hydrocarbons

Abstract Reported here is our recent investigation of the mechanism of the selective nitric oxide reduction by hydrocarbons on Cu-ZSM-5 catalysts in an oxygen-rich gas mixture. We studied the copper oxidation state change during the catalytic reaction using the X-ray Absorption Near Edge Structure (XANES) method. We observe that even under strongly net oxidizing conditions, a significant fraction of the copper ions in ZSM-5 is reduced to Cu I at elevated temperature, when propene is present in the reactant stream. XANES spectra show that the Cu I 1s → 4p transition intensity, which is proportional to cuprous ion concentration, changes with the reaction temperture in a pattern similar to the NO conversion activity. For comparison purposes, we also studied the Cu I concentration change using a gas mixture in which propene was replaced by a stoichiometrically equivalent concentration of methane. Unlike propene, methane provides no NO selective reduction pathway over Cu-ZSM-5. No window of enhanced Cu I concentration was observed using methane as the reductant. Our study indicates that, even in a strongly oxidizing environment, cupric ion can be partially reduced by propene to form Cu I , possibly by way of allylic intermediate, which may be a crucial step for effective NO conversion through a redox mechanism.

Interaction of platinum and ceria probed by transmission electron microscopy and catalytic reactivity

Abstract The beneficial effects of ceria on automotive exhaust catalysis have been amply documented in the literature, but the mechanisms for ceria promotion are still being debated. Contact between the metal and ceria could lead either to electronic or geometric perturbations in the surface of the metal particles, or atom transfer (such as of oxygen) between ceria and the metal may be responsible for the enhanced oxidation activity. In this study, we have probed the nature of surface modifications in the metal by measuring catalyst activity in reactions such as the hydrogenolysis of n-butane and carbon monoxide hydrogenation. The activity in the probe reactions provides a measure of the interaction between the metal and ceria. Transmission electron microscopy (TEM) has been used to provide direct information on the nature of the metal-ceria contacts. We find that metal-ceria interaction leads to suppressed hydrogenolysis activity after high-temperature reduction (773 K), but the loss in activity can be reversed after oxidation at 773 K. The presence of ceria also leads to an enhancement in methanation activity which, unlike hydrogenolysis, is unaffected by the high-temperature pretreatments (oxidation or reduction). These results in conjunction with the TEM data have been used to propose a model for the beneficial effects of ceria in automotive catalysts.

Kinetics of ethylidyne formation on Pt(111) From time-dependent infrared spectroscopy

We have measured the rate of formation of ethylidyne on Pt(111) by monitoring the intensity increase of the symmetric CH3 bend at 1339 cm−1 as a function of time between 230 and 250 K. We obtain a first-order rate constant with an activation energy of 14±1 kcal/mol for a 0.3 L ethylene exposure at 100 K.

Structure sensitive reactions over supported ruthenium catalysts during Fischer-Tropsch synthesis

Highly dispersed ruthenium catalysts can be prepared on alumina by aqueous impregnation of ruthenium. EXAFS at the K-edge showed that this type of catalyst, after calcination and reduction, consisted of ruthenium particles, which were about 0.8 nm in size. When highly dispersed on alumina, ruthenium appears to catalyze the water-gas shift reaction, which occurs subsequent to Fischer-Tropsch synthesis. The hydrocarbons produced had low olefinicity, possibly because ofin situ production of hydrogen via the water-gas shift reaction. Highly dispersed ruthenium was not stable on alumina during Fischer-Tropsch synthesis. The ruthenium agglomeration on the alumina surface, as well as overall ruthenium loss from the catalyst, was attributed to the formation of a volatile ruthenium carbonyl species.Catalysts with about 85% of the ruthenium in the form of 3–7 nm particles were prepared on alumina by reverse micelle impregnation of ruthenium. These larger particles were stable against ruthenium carbonyl formation and, therefore, did not exhibit ruthenium agglomeration or loss of ruthenium. Catalysts with 3–7 nm ruthenium particles displayed a higher turnover number for hydrocarbon synthesis, higher olefinicity, and chain-growth probability and did not exhibit water-gas shift activity in contrast to ruthenium particles which were about 0.8 nm in size.The CO disproportionation measurements showed much less CO dissociation over highly dispersed ruthenium relative to 3–7 nm ruthenium particles. This phenomenon is consistent with the low activity, the low chain-growth probability and may also relate to the tendency to form ruthenium carbonyl that is observed with small ruthenium particles. The apparent water-gas shift activity of highly dispersed ruthenium can be explained by the low CO dissociation efficiency as well as by the proposed ability to dissociate the water molecule.

Physico-Chemical Properties of Ce-Containing Three-Way-Catalysts and the Effect of Ce on Catalyst Activity

The activity of Pt,Rh,Ce/γ-Al2O3 and Pt,Rh/CeO2 catalysts has been studied in a full complement synthetic exhaust gas mixture consisting of H2, CO, C3H6, C3H8, NO, O2, N2, CO2, H2O and SO2. Direct interaction between Pt and CeO2 was shown to lead to large improvements in catalyst performance after activation of the catalyst in the synthetic exhaust gas. Catalyst activation was shown to be due to reduction of the noble metals and surface Ce and this state of the catalyst was found to be particularly effective for CO oxidation. The key component responsible for activation was shown to be H2, even at the low level present in the exhaust gas (0.26 vol.%), reduction is shown to be very facile. The degree of Pt/Ce interaction and thus activity after catalyst activation could be controlled by controlling the CeO2 crystallite size. Decreasing the CeO2 crystallite size led to greater Pt/Ce interaction as shown by TPR and STEM analysis and resulted in greater activity for both fresh and laboratory aged catalysts. Direct Pt/Ce interaction was also shown to lead to a synergistic reduction of Pt and surface Ce andthis feature of the catalyst was shown to qualitatively correlates with catalyst performance after activation.

Metal Particle Size Effects in Fischer-Tropsch Synthesis with Supported Ruthenium Catalysts

Abstract The effect of metal particle sizes on Fischer-Tropsch synthesis was studied over supported ruthenium catalysts with well defined metal particle sizes. Reverse micelle solutions of ruthenium were used to prepare catalysts with different size ruthenium particles at specified metal loading levels. Impregnation with conventional salt solutions and ion exchange techniques were also utilized. The catalysts were characterized by STEM, EXAFS, and gas adsorption techniques, before and after reaction. The smallest ruthenium metal particles agglomerated on alumina during the reaction. Particles larger than 4 nm were stable on alumina. The extent of ruthenium metal agglomeration also was affected by the support material. Larger ruthenium particles on alumina gave higher turnover frequencies for Fischer-Tropsch synthesis. Well dispersed alumina-supported ruthenium showed water gas shift activity while larger ruthenium particles did not. The ratio of olefinic to paraffinic products and the chain growth probability increased with an increase in metal particle size for alumina-supported ruthenium catalysts. The product distributions with, all catalysts tested obeyed the Anderson-Schulz-Flory distribution up to a carbon number of 150, irrespective of ruthenium metal particle size.