Studies of metal–support interactions with “real” and “inverted” model systems: reactions of CO and small hydrocarbons with hydrogen on noble metals in contact with oxides

Two types of model catalysts are compared: thin film catalysts consisting of polyhedral noble metal nanocrystals (Rh and Pt) supported by reducible and non‐reducible oxides, and their inverted pendants, submonolayers of titania and vanadia deposited under UHV conditions on the respective metal surfaces (Pd and Rh(111) and Rh (polycrystalline)). The structure and composition of the inverse catalysts were examined in situ by LEED and AES and the nanoparticles were characterized by HRTEM. The activity of thin film and inverse catalysts was studied in a series of reactions, such as the ring opening of methylcyclopentane and methylcyclobutane, the dissociation of CO and the CO methanation. Reaction conditions comprise atmospheric pressure but also molecular beam experiments. The reaction rates are related to the oxidation state of the supporting oxide, to the free metal surface area and to the number of sites at the interface between metal and support. This revised version was published online in June 2006 with corrections to the Cover Date.     

Steam reforming of n-butane on Pd/ceria

We have examined the steam reforming of n-butane on ceria, 1 wt% Pd/ceria, 1 wt% Pd/alumina, and 15 wt% Ni/silica between 573 and 873 K, with H₂O:C ratios between 1.0 and 2.0. No stable rates could be observed on Ni/silica due to rapid coking under these conditions. While rates were stable on the other catalysts, Pd/ceria showed a much higher activity than either Pd/alumina or ceria individually. Of additional interest, CO₂:CO ratios were much higher on Pd/ceria and approached equilibrium. The reaction order for n-butane on Pd/ceria was 0.15. For H₂O, reaction order changed from 0.6 to zero at the stoichiometric, n-butane:H₂O ratio. It is suggested that the high activity of Pd/ceria for this reaction is due to a dual-function mechanism, in which ceria can be oxidized by H₂O and then supply oxygen to the Pd.     

Investigation of CO oxidation and NO reduction on three-way monolith catalysts with transient response techniques

The kinetics of CO oxidation and NO reduction reactions over alumina and alumina-ceria supported Pt, Rh and bimetallic Pt/Rh catalysts coated on metallic monoliths were investigated using the step response technique at atmospheric pressure and at temperatures 30–350°C. The feed step change experiments from an inert flow to a flow of a reagent (O₂, CO, NO and H₂) showed that the ceria promoted catalysts had higher adsorption capacities, higher reaction rates and promoting effects by preventing the inhibitory effects of reactants, than the alumina supported noble metal catalysts. The effect of ceria was explained with adsorbate spillover from the noble metal sites to ceria. The step change experiments CO/O₂ and O₂/CO also revealed the enhancing effect of ceria. The step change experiments NO/H₂ and H₂/NO gave nitrogen as a main reduction product and N₂O as a by-product. Preadsorption of NO on the catalyst surface decreased the catalyst activity in the reduction of NO with H₂. The CO oxidation transients were modeled with a mechanism which consistent of CO and O₂ adsorption and a surface reaction step. The NO reduction experiments with H₂ revealed the role of N₂O as a surface intermediate in the formation of N₂. The formation of NN bonding was assumed to take place prior to, partly prior to or totally following to the NO bond breakage. High NO coverage favors N₂O formation. Pt was shown to be more efficient than Rh for NO reduction by H₂.     

Support effects in the dissociation of CO on Rh/CeO2(111)

The adsorption and reaction of CO on Rh particles supported on stoichiometric and partially reduced CeO₂(111) surfaces was studied using a combination of HREELS and TPD. A fraction of the CO adsorbed on the supported Rh particles was found to undergo dissociation to produce adsorbed C and O atoms. TPD results for isotopically labeled CO demonstrated that O atoms produced by CO dissociation rapidly exchange with the oxygen in the ceria lattice. The fraction of adsorbed CO which dissociated was found to increase significantly with the extent of reduction of the CeO₂(111) surface, suggesting that oxygen vacancies on the surface of the support play a direct role in the CO dissociation reaction. This revised version was published online in July 2006 with corrections to the Cover Date.     

Noble metal promoted Ni0.03Mg0.97O solid solution catalysts for the reforming of CH4 with CO2

Reforming of CH₄ with CO₂ to produce syngas was studied over Ni_0.03Mg_0.97O solid solution catalyst and its bimetallic derivative catalysts which contained small amounts of Pt, Pd or Rh (the atomic ratio M/(Ni + Mg) was about 2 × 10^–4, M = Pt, Pd or Rh). It was found that although the Ni_0.03Mg{0.97}O catalyst showed an excellent stability and activity at the reaction temperature of 1123 K, it lost its activity completely within 51 h when the reaction temperature was as low as 773 K. However, both the activity and the stability at 773 K were improved significantly by adding Rh, Pt, or Pd. This synergistic effect is rationally explained by the promoted reducibility of Ni. On all these catalysts, the amount of deposited carbon during the reaction was very low, suggesting that carbon deposition was not the main cause of the deactivation. Also, the catalytic activity of bimetallic catalysts increased gradually with the noble metal loading, but after passing through a maximum, it decreased with superfluous addition. The maximum was found to be located at around the atomic ratio of M/(Ni + Mg) 0.02% (M = Pt, Pd and Rh). This phenomenon could most probably be attributed to the different composition of Pt-Ni alloy particles formed after the reduction.     

Structure Sensitivity of CO Dissociation on Rh Surfaces

Using periodic self-consistent density functional calculations it is shown that the barrier for CO dissociation is 120 kJ/mol lower on the stepped Rh(211) surface than on the close-packed Rh(111) surface. The stepped surface binds molecular CO and the dissociation products more strongly than the flat surface, but the effect is considerably weaker than the effect of surface structure on the dissociation barrier. Our findings are compared with available experimental data, and the consequences for CO activation in methanation and Fischer–Tropsch reactions are discussed.     

The Mechanism of N2O Decomposition on Fe-ZSM-5: An Isotope Labeling Study

The role of Fe promoters has been investigated on Pd/ceria, Pt/ceria and Rh/ceria catalysts for the water–gas shift (WGS) reaction in 25 torr of CO and H₂O under differential reaction conditions. While no enhancement was observed with Pt and Rh, the activity of Pd/ceria increased by as much as an order of magnitude upon the addition of an optimal amount of Fe. Similarly, the addition of 1 wt% Pd to an Fe₂O₃ catalyst increased the WGS rate at 453 K by a factor of 10 over that measured on Fe₂O₃ alone, while the addition of Pt or Rh to Fe₂O₃ had no effect on rates. The amount of Fe that was necessary to optimize the rates increased with Pd loading but was independent of the order in which Fe and Pd were added to the ceria. Increased WGS activity was also observed upon the addition of Fe to Pd supported on Ce_0.5Zr_0.5O₂. XRD measurements, performed after running the catalyst under WGS conditions, show the formation of a Fe–Pd alloy, even though similar measurements on an Fe/ceria catalyst showed that Fe₃O₄ was the stable phase for Fe in the absence of Pd. Possible implications of these results on the development of new WGS catalysts are discussed.     

Effect of SO2 on the oxygen storage capacity of ceria-based catalysts

We have examined the effect of SO₂ poisoning on a series of catalysts having Pd supported on ceria, alumina, and ceria–zirconia. For pre-exposure of 20 ppm SO₂ at 673 K, we observed no changes in the light-off curves for CO oxidation on Pd/alumina. This pre-exposure of SO₂ to Pd/ceria resulted in a significant upward shift in the light-off curve, so that the poisoned Pd/ceria catalyst exhibited similar rates to that of Pd/alumina. Similar upward shifts were observed for the water–gas-shift reaction upon exposure of Pd/ceria or Pd/ceria–zirconia samples to SO₂. However, pulse-reactor data with alternating CO and O₂ pulses showed that SO₂ poisoning actually increased the amount of oxygen that could be transferred to and from the catalyst over the entire temperature range that was examined. The implication of these results for understanding the effect of SO₂ poisoning and the measurement of OSC are discussed.     

A comparison of CO oxidation on ceria-supported Pt,Pd, and Rh

Steady-state, CO-oxidation kinetics have been studied at differential conversions on model, ceria-supported, Pt, Pd, and Rh catalysts, from 467 to 573 K, and the results compared to the alumina-supported metals. On each of the ceria-supported metals, there is a second mechanism for CO oxidation under reducing conditions which involves oxygen from ceria reacting with CO on the metals. The rates of this second process are independent of which metal is used. The process has a significantly lower activation energy (14±1 kJ/mol compared to 26±2 kJ/mol on alumina-supported catalysts) and different reaction orders for both CO (zeroth-order compared to –1) and 02 (0.40 to 0.46 compared to first-order). This second process leads to significant rate enhancements over alumina-supported catalysts at low temperatures, especially for Pt. The implications of these results for automotive catalysis are discussed.     

A Mechanistic Study of Sulfur Poisoning of the Water-Gas-Shift Reaction over Pd/Ceria

The effect of sulfur on the water-gas-shift (WGS) activity of Pd/ceria catalysts has been studied using steady-state rate measurements, pulse-reactor studies, and FTIR. After exposing Pd/ceria to SO₂ at 673 K in an oxidizing environment, the WGS rates dropped to a value close to that observed on Pd/alumina. Both pulse-reactor and FTIR measurements showed that cerium sulfates can be readily reduced by CO and re-oxidized by O₂ at 723 K; however, unlike reduced ceria, the Ce₂O₂S formed by reduction of the sulfates cannot be re-oxidized by H₂O or CO₂. The implications of these measurements for understanding the oxygen-storage capacity (OSC) of three-way catalysts are discussed.     

Kinetic Study of Metal-Catalyzed Reaction of Solid Calcium Oxide with Gas Mixtures of Methane and Water Vapor to Form Hydrogen

With noble metal catalysts (Pd, Pt, Rh, Ir) present, hydrogen is formed by the interaction of solid calcium oxide with gas mixtures of methane and water vapor, according to CaO + CH₄ + 2H₂O CaCO₃ + 4H₂. Among the metals, Ir and Rh are so active that the reaction takes place at temperatures as low as 600 K. Rate data obtained with these metals show a nearly first order with respect to CH₄ pressure, while a negative order with respect to H₂O vapor pressure. The apparent activation energies are 171 and 217 kJ/mol for the Ir- and Rh-catalyzed reactions, respectively. On the other hand, Ni does not catalyze the reaction below 733 K, probably due to its strong interaction with H₂O vapor.     

FTIR study of the rearrangement of adsorbed CO species on Al2O3-supported rhodium catalysts

The adsorption of CO at low temperatures (130–293 K) has been investigated on Rh/Al₂O₃ catalysts of low (0.001–1 wt%) Rh loadings by means of Fourier transform infrared spectroscopy. The surface structure of Rh produced at different reduction temperatures (573 and 1173 K) was shock-cooled to 130 K, where the addition of CO caused the appearance of the band due to bridge-bonded CO ((Rh⁰)₂–CO) on all samples. The appearance of the bands due to gem-dicarbonyl (Rh⁺(CO)₂) and linearly bonded CO (Rh_x–CO) depended on the Rh content and the reduction temperature of the catalysts. The positions and the integrated absorbances of the symmetric and asymmetric stretchings of the Rh⁺(CO)₂ changed with temperature. On the basis of the above findings the rearrangement of the adsorbed CO species (indirectly that of surface Rh) is discussed. This revised version was published online in July 2006 with corrections to the Cover Date.     

Methanation on K+-modified Pt/SiO2: the impact of reaction conditions on the effective role of the promotor

This paper reports on the first study that the authors know of of the effect of alkali promotion of Pt on methanation. Methanation was investigated on a 4.5 wt% Pt/SiO₂ catalyst promoted with different amounts of K⁺ (K⁺/Pt=0, 0.1, and 0.2) for two different temperature ranges (503-552 K and 573-665 K). The methanation rate was 10-70% lower on the promoted catalysts for reaction temperatures of 573 to 665 K. In this temperature range, the relative decrease in rate upon promotion was a function of K⁺ loading and did not vary with temperature, , or time-on-stream. In addition, there was no significant effect of K⁺-promotion on activation energy (ca. 29 kcal/mol) or methanation reaction orders with respect to CO and H₂ (-0.1-0.0 and 0.4-0.6, respectively). However, there was a decrease in the number of methane-destined surface intermediates upon promotion as determined by steady-state isotopic transient kinetic analysis (SSITKA). All these observations lead to the conclusion that, in this higher reaction temperature range, K⁺ acts mainly as a site-blocking agent for methanation on Pt and does not change the reaction rate of the limiting step, probably hydrogenation. Between 503 and 552 K, the activation energy and reaction orders with respect to H₂ and CO were also not affected by K⁺. However, the catalyst with a K⁺/Pt ratio of 0.1 showed the highest methanation activity. In this lower temperature range and for all the catalysts, the apparent activation energies were also found to be lower, 18 vs. 29 kcal/mol, compared to those at higher temperatures. The reaction order with respect to CO was higher (0.2-0.3) in comparison with what was observed in the higher temperature range (ca. -0.1-0.0). These results suggest, that, in the low temperature range and for low loadings of K⁺, K⁺ affects the rate-determining step resulting in a rate increase greater than the decrease due to the blockage effect. Thus, K⁺ serves as a rate promoter at low reaction temperatures while its only effective function is site blockage at higher temperatures. This revised version was published online in July 2006 with corrections to the Cover Date.     

Acetate formation and explosive decomposition during ethanol oxidation on Rh

The adsorption and reaction of ethanol with the Rh(110) surface has been studied using a thermal molecular beam system and temperature programmed desorption. On the clean surface, ethanol shows a very simple dehydrogenation, producing hydrogen in the gas phase, adsorbed CO (which is desorbed by heating to 550 K) and carbon. Since in alcohol synthesis reactions it is likely that the surface will be partially oxidised, the reaction with predosed oxygen was also investigated. The reaction pathway then becomes much more complex. The main changes are (i) CH₄ and H₂O evolution during adsorption, and (ii)Acetate formation by oxygen insertion in the molecule. The acetate shows very unusual decomposition kinetics — a surface explosion with a very narrow peak-yielding CO₂ and H₂ in the gas phase and adsorbed C. The acetate is always seen on Rh catalysts which are selective for alcohol synthesis from CO and H₂, and it is proposed that oxidic promoters such as vanadia may act to stabilise this intermediate.     

Elucidation of CO2 Formation Mechanism in CO + NO Reaction on Pd(111) and Pd(110) Surfaces Using IR Chemiluminescence Method

The infrared (IR) chemiluminescence technique was applied to steady-state CO oxidation by NO on Pd(111) and Pd(110). From a comparison of IR emission spectra of CO₂ between the CO + NO and CO + O₂ reactions, it was found that the vibrational energy states of CO₂ in the CO + NO reaction were similar to those in the CO + O₂ reaction. This indicates that the reaction path of CO₂ formation in CO + NO is the same as that in CO + O₂, although the vibrational states are very dependent on the surface structure.     

Effect of metal oxide additives on the CO hydrogenation to methanol over Rh/SiO2 and Pd/SiO2

Catalysts were prepared from ultra pure SiO₂, Pd and Rh nitrates and chlorides, and by doping with Al, Fe, Na, K or Ca nitrate. The activities and selectivities of the Pd and Rh catalysts were investigated at 553 K, H₂/CO=2 or 3 and 2.5 or 4 MPa respectively. Additives had a strong influence on the catalytic properties. The doping with alkali and alkaline earth oxides led to a strong suppression of the CO dissociation. Particularly basic additives, such as Ca, had a strong promoting effect on the methanol production. This may confirm that the formation of methanol occurs through formate intermediates.     

Oxidative dehydrogenation of ethane to ethylene over alumina-supported vanadium oxide catalysts: Relationship between molecular structures and chemical reactivity

The influence of vanadium oxide loading in the supported VO_x/Al₂O₃ catalyst system upon the dehydrated surface vanadia molecular structure, surface acidic properties, reduction characteristics and the catalytic oxidative dehydrogenation (ODH) of ethane to ethylene was investigated. Characterization of the supported VO_x/Al₂O₃ catalysts by XPS surface analysis and Raman spectroscopy revealed that vanadia was highly dispersed on the Al₂O₃ support as a two-dimensional surface VO_x overlayer with monolayer surface coverage corresponding to 9 V/nm². Furthermore, Raman revealed that the extent of polymerization of surface VO_x species increases with surface vanadia coverage in the sub-monolayer region. Pyridine chemisorption-IR studies revealed that the number of surface Brønsted acid sites increases with increasing surface VO_x coverage and parallels the extent of polymerization in the sub-monolayer region. The reducibility of the surface VO_x species was monitored by both H₂-TPR and in situ Raman spectroscopy and also revealed that the reducibility of the surface VO_x species increases with surface VO_x coverage and parallels the extent of polymerization in the sub-monolayer region. The fraction of monomeric and polymeric surface VO_x species has been quantitatively calculated by a novel UV–Vis DRS method. The overall ethane ODH TOF value, however, is constant with surface vanadia coverage in the sub-monolayer region. The constant ethane TOF reveals that both isolated and polymeric surface VO_x species possess essentially the same TOF value for ethane activation. The reducibility and Brønsted acidity of the surface VO_x species, however, do affect the ethylene selectivity. The highest selectivity to ethylene was obtained at a surface vanadia density of 2.2 V/nm², which corresponds to a little more than 0.25 monolayer coverage. Below 2.2 V/nm², exposed Al support cations are responsible for converting ethylene to CO. Above 2.2 V/nm², the enhanced reducibility and surface Brønsted acidity appear to decrease the ethylene selectivity, which may also be due to higher conversion levels. Above monolayer coverage, crystalline V₂O₅ nanoparticles are also present and do not contribute to ethane activation, but are responsible for unselective conversion of ethylene to CO. The crystalline V₂O₅ nanoparticles also react with the Al₂O₃ support at elevated temperatures via a solid-state reaction to form crystalline AlVO₄, which suppresses ethylene combustion of the crystalline V₂O₅ nanoparticles. The molecular structure–chemical characteristics of the surface VO_x species demonstrate that neither the terminal VO nor bridging VOV bonds influence the chemical properties of the supported VO_x/Al₂O₃ catalysts, and that the bridging VOAl bond represents the catalytic active site for ethane activation.     

Understanding the correlation between calcination temperature and performance in low-temperature methanation over Ni-Zr/Al2O3 catalysts

Low-temperature methanation of CO in the continuous stirred tank reactor (CSTR) over Zr doped Ni/Al₂O₃ catalyst calcined at different temperatures (673, 723, and 823 K) was investigated. XRD, TPR, XPS, ICP, SEM, and S-TPR techniques were employed to characterize the fresh and spent catalysts. Based on the characterization results, it was found that low-temperature (673 K) calcination could effectively prohibit the formation of NiAl₂O₄ spinel, thereby resulting in more reducible NiO particles, which were the essential precursor of methanation active sites over the catalyst surface. Thus, the highest CO conversion of 93.6% was achieved over the 25N3ZA-673 catalyst. In addition, the deactivation rate of 25N3ZA-673 was relatively slow in comparison to 25N3ZA-823 due to the presence of more reducible NiO. The formed nickel carbonyl species (Ni[CO]_x), which quickly decomposed at a higher reaction temperature, was closely related to the catalyst deactivation. Therefore, 25N3ZA-673 possessed much better stability at 593 K than that at 553 K.     

Rhodium-catalyzed partial oxidation of methane to CO and H2. Transient studies on its mechanism

The reaction of methane with surface oxygen as well as the interaction of methane/oxygen mixtures with a Rh(1 wt%)/-Al₂O₃ catalyst was studied by applying the temporal-analysisof-product (TAP) reactor. The product distribution was strongly affected by the degree of surface reduction. CO₂ is formed as a primary product via a redox mechanism with the participation of surface oxygen. The dehydrogenation of methane yielding carbon deposits on the surface occurs on reduced surface sites. The formation of CO proceeds with high selectivity (up to 96%)at 1013 K via fast reaction of surface carbon species with CO₂.     

Fabrication of an Effusive Molecular Beam Instrument for Surface Reaction Kinetics—CO Oxidation and NO Reduction on Pd(111) Surfaces

A simple molecular beam instrument (MBI) was fabricated for measuring the fundamental parameters in catalysis such as, sticking coefficient, transient and steady state kinetics and reaction mechanism of gas/vapor phase reactions on metal surfaces. Important aspects of MBI fabrication are given in detail. Nitric oxide (NO) decomposition and NO reduction with carbon monoxide (CO) on Pd(111) surfaces were studied. Interesting results were observed for the above reactions and they support the efficiency of the MBI to derive the fundamental parameters of adsorption and catalysis. Sustenance of CO oxidation at 400 K is dependent mostly on the absence of CO-poisoning; apparently, CO + O recombination is the rate determining step ≤400 K. NO adsorption measurements on Pd(111) surface clearly indicating a typical precursor kinetics. Displacement of the chemisorbed CO by NO on Pd(111) surfaces was observed directly with NO + CO beams in the transient kinetics. It is also relatively easy to identify the rate-determining step directly from the MBI data and the same was demonstrated for the above reactions.