J.G. Goodwin

The effect of alkali promotion on CO hydrogenation over Rh/TiO2

To develop a better understanding of the link between alkali promotion of hydrocarbon synthesis and that of oxygenate synthesis, CO hydrogenation over both unpromoted and promoted RhTiO2 was studied in a differential reactor at 250–300 °C, 1–10 atm, and COH2 = 2. The ability of the alkali species to promote the selectivity for oxygenated compounds increased in the order unpromoted Li > K > Cs. The promoters also showed a strong hydrogenation-suppression ability which may have caused the decrease in hydrocarbon selectivity and activity. The activities and selectivities of RhTiO2 and alkali-promoted RhTiO2 appear to be correlated with their catalytic abilities for hydrogenation, CO dissociation, and CO insertion. Results for unpromoted RhSiO2 are also presented for Comparison purposes.

The use of probe molecules in the study of CO hydrogenation over SiO2-supported Ni, Ru, Rh, and Pd

Abstract CO hydrogenation over Ni SiO 2 , Ru SiO 2 , Rh SiO 2 and Pd SiO 2 was studied by the addition of various probe molecules (C 2 H 4 , CH 3 CH 2 OH, and CH 3 CHO) to the reactant stream under synthesis conditions. The well-known differences among these catalysts in product formation (methane, higher hydrocarbons, C 2+ oxygenates and methanol) were shown to be due to differences in their activities to catalyze hydrogenation, hydrogenolysis, dehydrogenation, decarbonylation, CH x insertion, and CO insertion rather than just CO dissociation. Ni SiO 2 and Pd SiO 2 exhibited weak activities for the incorporation of these probe molecules into higher hydrocarbons and oxygenates. Rh SiO 2 , a good catalyst for the production of C 2 oxygenated compounds, showed strong activity for CO insertion and for the incorporation of ethylene and ethanol into C 3+ oxygenated compounds. Ru, a higher hydrocarbon synthesis catalyst, displayed fairly strong activities for the incorporation of ethanol and acetaldehyde into C 3+ hydrocarbons. Probing of the surface under reaction conditions by the addition of certain reacting molecules provides an excellent method for developing a better understanding of reaction mechanisms as well as of fundamental properties of catalysts under those conditions.

Surface structure dependence of reversible/weak H2 chemisorption on supported Ru

Abstract Earlier work on highly dispersed NaY-supported Ru ( d p = 0.9–1.6 nm ) showed that the reversible (weak) fraction of hydrogen chemisorption at 298 K is a function of average particle diameter, and it was suggested that this hydrogen is accommodated on lower-energy sites. In order to gain a more complete understanding of this weak Chemisorption, an in-depth investigation of the effect of particle size on reversible H 2 Chemisorption at 298 K was carried out using supported ruthenium catalysts with average metal particle diameters of 0.9–12.5 nm. Particular attention was paid to determining the reversibly chemisorbed hydrogen fraction under strictly the same conditions. It was found that this fraction exhibited a maximum of 30% of the total Chemisorption for an average Ru particle size of 1.6 nm. This fraction decreased to zero as the average Ru particle size increased above 2.5 nm. These findings enable us to suggest that loosely chemisorbed hydrogen at equilibrium on Ru may be accommodated on multiatomic sites (“ensembles”) of a similar type as the so-called B 5 sites described by R. van Hardeveld and F. Hartog ( Surf. Sci. 15 , 189, 1969). Causes for deviations from this functionality are discussed.

Reaction analysis of potassium promotion of Ru-catalyzed CO hydrogenation using steady-state isotopic transients

Abstract The effects of alkali on catalyst activity and deactivation during CO hydrogenation have been studied in the past mostly based on an available surface-metal atoms approach. However, such an approach cannot easily distinguish to what extent the modifier brings about changes in surface concentrations of reaction intermediates or affects site activity during steady-state reaction. Steady-state isotopic transient analysis (SSITKA) with carbon tracing was used to decouple the effects of potassium on the methane-producing sites during steady-state CO hydrogenation over Ru/SiO 2 catalysts having modifier loadings of up to (K/Ru) atom = 0.2. The SSITKA results indicate that, during steady-state CO hydrogenation, carbidic carbon evolved into methane via a high-reactivity (C 1α ) and a low-reactivity (C 1β ) trajectory. With increasing amounts of K + the average “true” intrinsic turnover frequency ( k ) of both of these carbidic pools decreased, as did their steady-state surface abundance. Relative to C 1β , the C 1α pool was affected to a slightly greater extent, both in terms of its reactivity and abundance. It is likely that potassium was able to strengthen the carbon-metal interaction which made hydrogenation of the carbon adlayer more difficult, resulting in smaller methane-destined pools of active surface carbon. With time-on-stream, deactivation by deposition of inactive carbon did not significantly affect the product distribution or the methane rate constant at the prevailing K + doping levels; instead, deactivation was due to a loss in the steady-state abundance of carbon-containing surface intermediates exiting as methane. Implications of the role of potassium during steady-state CO hydrogenation in influencing the active metal surface, the carbidic adlayer, and the latters transformation into unreactive carbon are addressed.

The effect of chlorine modification of silica-supported Ru on its CO hydrogenation properties

Abstract An investigation of CO hydrogenation on Cl-modified supported Ru catalysts has been carried out using both steady-state Fischer-Tropsch synthesis and steady-state isotopic transient kinetic analysis of methanation. The steady-state reaction results illustrate how the presence of chlorine acts to decrease catalytic activity and to enhance the selectivity of methane formation even though it is present on the catalyst only during the initial stages of the reaction. The deactivation results for F-T as well as the isotopic transient results suggest that structural rearrangements induced by the presence of chlorine, rather than selective site blocking or electronic interactions, may be the primary mechanism of chlorine modification of the catalytic properties of supported ruthenium for CO hydrogenation. Isotopic transients indicated that the decrease in methanation activity with increasing initial Cl concentration was a function of a decrease in the number of reactive surface intermediates (or sites) and not of a change in site activity.

Effect of dealumination on the catalytic activity of acid zeolites for the gas phase synthesis of MTBE

Abstract The gas phase MTBE synthesis reaction was studied on a number of zeolite catalysts, in order to obtain a better understanding of the impact of acidity on zeolite activity for MTBE formation away from thermodynamic equilibrium limitations. The catalysts investigated included a series of dealuminated HY zeolites with different acid properties. In addition, H-ZSM-5, an amorphous silica-alumina, and Amberlyst-15 resin were investigated for comparison. An increase in acidity of the HY zeolites produced by dealumination was found to result in a significant enhancement in the intrinsic activity for MTBE formation. The ratio of extra-lattice to lattice Al appears to be an important parameter for determining the catalytic behavior of zeolites for this reaction.

Structure sensitivity of reactions between cyclopropane and hydrogen on supported ruthenium catalysts

Reaction (1) is called “hydrogenation,” reaction (2) “selective hydrocracking,” and reaction (3) “nonselective hydrocracking.” On Pd, It-, and Pt only reaction (1) is observed, while Co, Rh, and OS are active for reactions (1) and (2), and Fe, Ni, and Ru are able to catalyze all three reactions (I-5). The structure sensitivity of the three cyclopropane reactions on Ru catalysts has not yet been investigated in great detail. In previous work it was discovered that the nonselective hydrocracking reaction occurred only on catalysts containing relatively large Ru particles (I). On highly dispersed Ru catalysts, the nonselective hydrocracking reaction did not take place at all, and instead an increase in activity for reactions (1) and (2) with increasing Ru dispersion was noted (4). These findings indicated an apparent structure sensitivity of all three reactions of cyclopropane, and prompted us to investigate this phenomenon more closely. In the present study, the dispersion of Ru and the nature of the support are varied systematically in order to evaluate their relative contributions to the activity and selectivity in the three cyclopropane reactions.

Potassium dispersion on silica-supported ruthenium catalysts

Abstract Alkali modifiers are known to be quite effective at improving catalyst activity or selectivity for several metal-catalyzed reactions of industrial importance. Yet it is still difficult to address the location and distribution of alkali species in most catalysts. This paper reports on an investigation of the potassium dispersion in a series of 3 wt% Ru/SiO 2 catalysts sequentially doped with potassium nitrate up to (K/Ru)awm = 0.2 followed by rereduction. This series was evaluated extensively using gas volumetric hydrogen chemisorption and the structure-sensitive ethane hydrogenolysis reaction. Hydrogen chemisorption results indicate that the alkali was apparently atomically dispersed on the ruthenium surface. The added potassium species interfered with hydrogen chemisorption on a one-to-one atomic basis. Potassium addition resulted in a decrease in the apparent activation energy and an increase in the apparent hydrogen reaction order for ethane hydrogenolysis. Using the statistical poisoning model of Martin ( Catal. Reu.-Sci. Eng. 30 , 519 (1988)) which assumes that the metal surface is uniform for adatom adsorption, the apparent ensemble required for the reaction was estimated to be made up of 12 ± 3 adjacent exposed surface ruthenium atoms. Using an extension of Martins model, this structure-sensitive reaction also revealed that at the higher potassium levels the alkali dispersion became nonuniform. This nonuniform dispersion is suggested to be due to a preference of the dopant for certain metal sites. Because of this nonuniform dispersion, the “true” reaction ensemble size is suggested to be less than 12.

Use of linear modeling in steady-state isotopic-transient kinetic analysis of surface-catalyzed reactions: Application to plug-flow reactors

Abstract Determination of kinetic parameters from steady-state isotopic transient kinetic analysis (SSITKA) requires modeling of the reactor system and the catalyst surface. In this work, a novel application of linear-modeling methods was developed for SSITKA, using transfer functions for non-differential-length catalyst beds in non-ideal plug-flow reactors (PFRs), which are often employed in SSITKA. Various linear relationhips between the catalyst surface and the gas phase were derived - including linear convolution, which provides a new rigorous method for generating calculated isotopic transient responses from catalyst-surface and reaction-system models. Incorporation of linear convolution - which avoids a priori gas-phase behavior correction which is problematic when using non-ideal PFRs - in parametric and nonparametric kinetic analyses provides for increased accuracy in the determination of kinetic parameters in SSITKA. The linear modeling techniques developed were applied to a PFR transient-response model for an irreversible reaction occurring in a non-differential-length catalyst bed. It was determined that the dependency upon the catalyst-bed length in PFRs is unimportant if gradientless conditions - differential conversion of reactant and irreversible reaction - are maintained in a catalyst bed of non-differential length. The results illustrate how the linear modeling techniques developed can be used with SSITKA to test assumed catalyst-surface reaction models.

The role of the zeolite in the hydrogenolysis of C2 and C3 hydrocarbons on RuNaY catalysts

Abstract The catalytic properties for the hydrogenolysis of ethane, propane and cyclopropane of a series of highly dispersed RuNaY catalysts have been investigated. These catalysts have activities and selectivities for ethane and propane hydrogenolysis similar to other supported ruthenium catalysts. However, the activity of the RuNaY for cyclopropane hydrogenolysis is much higher than that of Ru on conventional oxide supports, while the selectivities remain in a range expected for well-dispersed ruthenium. The increase in activity for the RuNaY catalysts is due mainly to the presence of highly dispersed Ru particles made possible by the zeolite support. A destabilization of the cyclopropane ring by the electrostatic field of the zeolite, however, does not seem to contribute significantly to the observed rate increase. It appears that the ring opening of cyclopropane and the hydrogenolysis of cyclopropane to ethane and methane have a common intermediate, the formation of which is rate determining for both reactions. The discovery that on Ru the ring opening of cyclopropane is structure sensitive is surprising since this reaction is generally considered as a classic example for structure insensitivity.