Search

Isotopic exchange between deuterium and cyclopentane and between pentane and ethane on the Rh/SiO 2 catalysts of the preceeding paper has been investigated after several different pretreatments. Although a more restricted number of catalysts was examined, as with 2,2-dimethylbutane, increased percentage exposed ( D h ) appears to lead to increased turnover frequency ( N t ). After the standard pretreatment, H 2 ,300°; He,450°, the rates in alkane(g) ⇌ alkane(ads) increase only slowly as D h increases, but the rates in monoadsorbed alkane ⇌ diadsorbed alkane decline relatively rapidly. As with 2,2-dimethylbutane, the pretreatment H 2 ,450° leads to the largest values of N t . He,450° following H 2 , 450° removes surface hydrogen atoms and leads to reconstruction to a surface of much lower activity for exchange. Cyclopentane exchanges much faster than pentane on these catalysts. It appears that adsorption and desorption of cyclopentane involve single-step processes to and from diadsorbed cyclopentane + 2H∗. Adsorption is faster than that with pentane because cyclopentane is essentially in the eclipsed conformation needed for this process. The rate of monoadsorbed ⇌ diadsorbed cyclopentane is similarly augmented. Adsorbate-surface strain reduces the rate constant for adsorption of 2,2-dimethylbutane and increases that for desorption. When coverages by the unhindered hydrocarbon are large, the relative rate of exchange of the hindered hydrocarbon vs that of the unhindered hydrocarbon will be larger than one would conclude merely from the relative rate constants of adsorption.

The rate of the reaction, H2+D2 → 2HD, is very rapid at −160 °C on the catalyst Cp′2Th-(CH3)2/dehydroxylated Al2O3, activated in H2+D2 and believed to contain supported Cp′2ThH+(Cp′ = η5-pentamethylcyclopentadienyl). Even at −195 °C, the turnover frequency is greater than 0.1 sec−1 at\(P_{H_2 + D_2 } = 1\) atm. Strong evidence is adduced that the exchange proceeds by a Rideal-Eley process, i.e. by a four-center transition state involving Th-H+D-D.

This contribution reports a detailed kinetic and mechanistic study of heterogeneous olefin hydrogenation by Cp{prime}{sub 2}Th(CH{sub 3}){sub 2} (Cp{prime} = {eta}{sup 5}-(CH{sub 3}){sub 5}C{sub 5}) adsorbed on dehydroxylated alumina (DA). For both propylene and isobutylene, the kinetics can be accommodated by a two-step sequence: (i) olefin insertion into Th-H (rate constant {nu}{sub 2}); (ii) hydrogenolysis of the resulting Th-alkyl bond (rate constant {nu}{sub 3}). Kinetic data can be fit to the rate law N{sub t} (turnover frequency) = {nu}{sub 2}{nu}{sub 2}P{sub H{sub 2}}/({nu}{sub 2} + {nu}{sub 3}P{sub H{sub 2}}), where the reaction is zero-order in olefin and where {nu}{sub 2}/{nu}{sub 3} = 0.55 bar{sup {minus}1} ({minus}63{degree}C) and 0.47 bar{sup {minus}1} ({degree}C) for propylene and isobutylene, respectively. Kinetic isotope measurements indicate N{sub t}(H{sub 2})/N{sub t}(D{sub 2}) = 1.5(1) for propylene hydrogenation at {minus}63{degree}C and 2.0(2) for isobutylene hydrogenation at 0{degree}C. In reaction with D{sub 2}, propylene yields exclusively 1,2-propane-d{sub 2}. Arrhenius activation energies for catalytic hydrogenation are 3.6(2) kcal mol{sup {minus}1} (propylene, {minus}63 to {minus}23{degree}C) and 5.3(2) kcal mol{sup {minus}1} (isobutylene, {minus}23 to +64{degree}C); for isobutylene, it could be determined that the activation enthalpies for olefin insertion and alkyl hydrogenolysis are comparable (6.2(10) and 4.2(1) kcal mol{sup {minus}1}, respectively).

Consideration des especes C m H n adsorbees sur des composes organometalliques, metaux du groupe VIII

Publisher Summary Most of the large number of acronyms added to BET between ICC 1 in 1956 and ICC 11 in 1996 represent new techniques for analyzing the products of reaction, for characterizing catalysts and for determining the structure of chemisorbed species. Use of these techniques has substantially altered the practice of catalysis. Only the elderly will remember the great improvement in the quality of life that resulted from the fulfillment of Emmett's prophecy, that gas chromatography appeared “destined to provide a rapid, accurate method for the analysis of complicated mixtures of products in a relatively simple, straightforward manner.” Adequate description of many catalysts will require a large number of bits of data as they are usually rather complicated materials rather than simple chemicals. Attempts at this were just beginning by ICC 1, but now, one expects authors to give specific surface areas and some details of the porosity of their catalysts. Automation of the former tedious point by point measurement of the N 2 adsorption isotherm has greatly facilitated this. Use of chemisorption to measure specific site numbers has become much more common, for example, CO for Cr 2 O 3 , and H 2 to measure H/M (usually equated to M S /M where M is an atom of a supported Group VIII metal). If the metal particles are roughly spherical, the H/M can be converted into an average particle size, but modern high resolution electron microscopy and wide angle x-ray scattering (WAXS) provide particle size distributions rather than mere averages. Further, in favorable cases, the first can provide atomic level resolution and lattice details and the second can diagnose the degree of strain in metallic particles larger than about 2 nm. However, most of these techniques require significant funding.

This chapter will discuss methods for the study of the mechanisms of heterogeneous catalytic reactions. The various examples of mechanistic studies have been chosen to illustrate techniques used in mechanistic studies and not to provide a general survey of the mechanisms which have been proposed for catalytic reactions. References have been chosen to be leading references rather than to indicate historical priority.

The activity of a series of supported Pt catalysts for the hydrogenation of benzene has been evaluated at 80 °C and 1 atm total pressure. Three sets of catalysts with a wide range of percent metal exposed were investigated: Pt/Al2O3, Pt/Al2O3-Cl and Pt/SiO2, with the Cl series obtained by exposure of Pt/Al2O3 to HCl or CCl4.The structure sensitivity of benzene hydrogenation on the chloride-free Pt/Al2O3 depended upon the reduction temperature employed in pretreatment. In general, the reaction was structure insensitive for high-temperature reduction (>450 °C), while for lower temperatures (100 and 300 °C) activity decreased with decreasing particle size. For Pt/Al2O3-Cl the reactivity behaviour was a function of chloride level. At low loadings (0.3 wt. %) the behaviour of these catalysts was similar to that of their chloride-free counterparts; however, a higher loading set (1 wt.%) showed a depression in activity over the entire particle size range except for very low percentage exposed metal. This is to be compared with the apparent augmentation of the activity of Pt/Al2O3 after HCl treatment reported by Bernard and Primet (J. Chem. Soc., Faraday Trans., 1990, 86, 567) for n-butane isomerization and that for methylcyclopropane hydrogenolysis (after either HCl or CCl4) reported by Ollendorff et al.(Appl. Catal., 1990, 62, 85), since these reactions are sensitive to acidity in a fashion that is not possible with benzene.

The activity and selectivity of a series of Pt/Al2O3 catalysts for the hydrogenolysis of methylcyclopropane and exchange between deuterium and cyclopentane have been determined. The series investigated included the range from 4.1 to 106 percentage exposed of platinum (Dh). Turnover frequencies, Nt, for methylcyclopropane hydrogenolysis at 0 °C determined after high-temperature pretreatment (H2, Tp > 350 °, 1) were independent of Dh as were the activation energies for i-butane and n-butane formation, 36.4 ± 2 and 42.3 ± 4 kJ mol−1, respectively. For cyclopentane-deuterium exchange at 81 °C the isotopic distribution patterns were clearly dependent upon Dh, particularly noted as a decrease in d 10 (d 2 -d 10 ) with increasing Dh. Activation energy for overall exchange was about 75 kJ mol−1 with some dependence on Dh for the individual distributions. Catalyst pretreatment conditions affect activity for both reactions. For methylcyclopropane a minimum is observed for Tp about 200 °C, with an inversion in ordering of Nt vs Dh at higher temperatures, as previously observed for Pt/SiO2. For cyclopentane, Nt decreases with increasing Dh at Tp = 450 °C. The method of catalyst preparation also showed some effect on Nt for methylcyclopropane, but not for cyclopentane. A summary discussion of the factors responsible for various pretreatment effects on Pt/SiO2 and Pt/Al2O3 is given for both reactions.

The reaction of M(eta/sup 5/-(CH/sub 3/)/sub 5/C/sub 5/)/sub 2/(CH/sub 3/)/sub 2/ and M(eta/sup 5/-(CH/sub 3/)/sub 5/C/sub 5/)/sub 2/(CD/sub 3/)/sub 2/ compounds, M = Th and U, with partially dehydroxylated and nearly completely dehydroxylated alumina yields methane via protolysis by surface OH, ring H atom abstraction, and elimination within the M(CH/sub 3/)/sub 2/ groups. The latter pathway is proposed on the basis of chemical evidence to result in alumina-stabilized actinide alkylidenes. 12 references, 2 tables.

Activation of Mo(CO)6/γ-Al2O3 at 100°C in flowing helium yields Mo(CO)3(ads.) which is modestly active in the metathesis of propylene at 53°C. Activation at higher temperatures results in a large increase in activity, the evolution of further CO and, by 280°C, of one molecule of hydrogen from surface OH groups. Activation beyond 280°C decreases the activity. Mo(CO)3(ads.) adsorbs one molecule from a pulse of O2 at 25°C to eject one molecule of CO and form Mo(CO)2O2(ads.), a species 15 times more active than Mo(CO)3(ads.). Interrelations are presented between these surface species and ones known molecularly.

The infrared spectra of molybdenum hexacarbonyl adsorbed on two types of γ-alumina have been determined after initial adsorption at 25°C and after activation in flowing helium at 100°C to produce Mo(CO)3ads. Nearly fully hydroxylated alumina (HA), partially dehydroxylated alumina (PDA, pretreated at 450–500°C), and nearly fully dehydroxylated alumina (DA, pretreated at 1000°C) were employed. Reaction between physisorbed Mo(CO)6 and the surface of alumina commences at 25°C or slightly above. It is proposed that surface O2− (σ-O−) or OH− (σ-OH) add to the carbon atom of one CO ligand with consequent labilization of other CO ligands for displacement by σ-O− or σ-OH. An infrared absorption band for Mo(CO)3ads was observed on HA, PDA, and DA at near 1580 cm−1 but only weakly with HA. On HA it is proposed that Mo(CO)3ads is primarily (σ-OH)3Mo(CO)3, and on DA, species like (σ-O)3Mo(CO)2(C00-σ). On PDA, (σ-O)3 would be replaced by (σ-O)2(σ-OH) or (σ-O)(σ-OH)2.

Three aspects of the surface chemistry and catalytic activity of organometallic complexes on alumina are covered with particular attention to the effect of the surface concentration of OH− present after various degrees of dehydroxylation. (1) The formation of Lewis acid-Lewis base pair sites during dehydroxylation of alumina. (2) The effect of the degree of dehydroxylation upon the course of the reaction of Mo(CO)6 with the surface of alumina during heating in He and in H2. Variously activated Mo(CO) 6 Al 2 O 3 are catalysts for the hydrogenation of propylene, the exchange of alkanes with D2, the hydrogenolysis of alkanes, the hydrogenolysis of cyclopropane, the hydrogenation of CO, and the metathesis of olefins. Metallic Mo(0) Al 2 O 3 is among the most active catalysts known for the first three reactions. (3) Organoactinide complexes on alumina. The complexes (Cp′)2Th(CH3)2, (Cp′)2U(CH3)2, and (Cp′)2ThH2 (where Cp′ is pentamethylcyclopentadienyl) lead to very active catalysts for the hydrogenation of propylene at −63 °C and the polymerization of ethylene at 25 °C. The surface chemistry of the reactions of the complexes with the surface of alumina during activation has been partially elucidated. High catalytic activity is favored by nearly complete dehydroxylation of the alumina support.

A set of platinum catalysts on γ-alumina was prepared by impregnation with Pt(NH3)2(NO2)2 aq and Pt(NH3)4(NO3)2 aq. Calcination at 200–400 °C followed by reduction at 300–500 °C gave percentages exposed of platinum as measured by hydrogen chemisorption at 25 °C (Dh) ∼- 100%. Direct reduction of incompletely dried material led to lower values of Dh. The ratio D t D h (Dt is the percentage exposed by hydrogen-oxygen titration at 25 °C) was unity for Dh 90%, not all oxygen adsorbed at 25 °C reacted with pulses of hydrogen at 25 °C. The ratio O/Pts increased with increasing temperature of adsorption of oxygen and at large enough O/Pts, reaction with hydrogen at 25 °C became incomplete for all catalysts. The value of O/Pts at which reaction became incomplete was structure sensitive and decreased with increasing Dh, from about 1.15 forDh = 36% to 0.5 for Dh = 94%. At smaller Dh, O/Pts became about 2 after exposure to oxygen at 300 °C; at Dh near 100%, O/Pt was about 1.15. Properties of Pt/SiO2 and Pt/Al2O3 for these reactions are compared.

Cyclopentene and cyclooctene have been hydrogenated in a batch reactor on supported platinum catalysts at temperatures such that over 99.99% of the olefin was in the adsorbed state and primarily on the support. Reaction temperatures were −75 °C for cyclopentene and −22 °C for cyclooctene and P(H2) was usually 100 Torr. At coverages of one molecule of cyclopentene per nm2 on Pt/(wide pore silica gel), the rate was zero order in olefin up to a conversion of about 70% following which the rate declined. The rate was about one-half order in hydrogen. Pt/alumina (PHF, American Cyanamid) and Pt/silica gel gave similar results. Transport of olefin from the surface of the support to crystallites of platinum occurred primarily by surface migration rather than through the vapor phase. Significant contribution to hydrogenation by hydrogen spillover from platinum to silica can be excluded. Small particles of platinum on silica gel gave somewhat larger values of turnover frequency, Nt, than did larger particles, at least after the pretreatment O2,300 °;H2,300 °;He,450 °; cool in He. Nt for cyclopentene was roughly equal to that for propylene in conventional vapor phase hydrogenation extrapolated to −75 dgC.

The preparation of Mo(0)/Al{sub 2}O{sub 3} by Mo(CO){sub 6}/DA;H{sub 3},300-500{degree};H{sub 2},650-950{degree} (Mo(CO){sub 6} deposited on highly dehydroxylated alumina, heated in flowing He to 300-500{degree}C and then in H{sub 2} to 650-950{degree}C) and by MoO{sub 3}/Al{sub 2}O{sub 3};H{sub 2},950{degree} give essentially equivalent results with respect to activity and selectivity in the hydrogenolyses of cyclopropane (CP) and methaylcyclopropane (MCP) at 100{degree}C and of propane at 250{degree}C. Thus, one can prepared metallic Mo on alumina by the easier and more flexible preparation from MoO{sub 3}/Al{sub 2}O{sub 3}. MCP undergoes single hydrogenolysis to isobutane and butane and double hydrogenolysis to (a) CH{sub 4} + C{sub 3}H{sub 8} and to (b) 2C{sub 2}H{sub 6}. The product ratio, butane/isobutane, is greater than unity which is uncommon on metals. After on to several pulses in a pulse reactor, catalyst samples were heated to 950{degree}C in flowing H{sub 2}. The amount of liberated methane gave the amount of carbonaceous deposit. With one to seven pulses of CP or MCP in H{sub 2}, (C atoms in deposit)/(C atoms in product) was 0.03-0.04 but only 0.007 with propane. Activities usually declined with pulse number, but selectivities changed little. The Mo(0)/Al{sub 2}O{sub 3} catalyst is initially metallic molybdenum and metallic molybdenummore » is catalytically active.« less

Experiments employing extended X-ray absorption fine structure and wide-angle X-ray diffraction, at the Cornell University High Energy Synchrotron Source, have shown that Pd catalysts of high percentage metal exposed stored in air consist primarily of crystalline PdO. Reduction of these catalysts in hydrogen results in the formation of palladium, whereas similar treatment of catalysts of low percentage metal exposed leads to palladium hydride.

Some of the H 2 adsorbed on chromia (microcrystalline α-chromia, amorphous chromia, and Cr 2 O 3 /SiO 2 ) at temperatures below −130°C is not removed by flowing helium at the temperature of adsorption, T . Such adsorbed hydrogen, H ur ( T ), increases monotonically as T is lowered from −130 to −205°C. H irr (−196), CO irr (−78), and activity for H 2 + D 2 = 2HD at −130°C increase in parallel as the temperature of activation in helium of chromia hydrogel is increased from 300 to 500°C. The hydrogel was pretreated with H 2 at 300°C. CO irr (−78) completely poisons H irr (−196). Pulses of D 2 in −196°C completely and rapidly displace H 2 (g) from H irr (−196) without formation of HD. At −163°C substantial HD is formed. The weak chemisorption of H 2 at lower temperatures appears to be a molecular polarization adsorption at Cr 3+ ( cus ) O 2− ( cus ) ion pairs. Displacement by D 2 may involve adsorption to form H 2 ∗ D 2 followed by preferential release of H 2 . An isotope effect of about 50 favors adsorption of D 2 vs H 2 at −196°C. the isotopic exchange at temperatures somewhat above −196°C appears to involve rearrangement of H 2 ∗ D 2 to HD∗HD perhaps via H − ∗(HD 2 ) + .

The initial reaction of HFe/sub 4/(CH)(CO)/sub 12/ with partially dehydroxylated alumina produces (HFe/sub 4/(C)(CO)/sub 12/)/sup -/, which is bound to the surface. Evolution of CO occurs slowly by extensive loss of CO from a small number of cluster molecules. This process leads to the coexistence of intact (HFe/sub 4/(C)(CO)/sub 12/)/sup -/ plus an iron or iron carbide like species. In keeping with this interpretation, the supported material displays activities for the reduction of Co and the hydrogenation of benzene and product distributions that are typical of iron metal.

Cyclopentene, di-t-butylacetylene, 3-hexyne, cis- and trans-di-t-butylethylene, bicyclo[2,2,1] heptene, and bicyclo[2,2,2]octene were hydrogenated with H2 and with D2 in a liquid-phase, batch reactor at 20 °C on PdSiO2andPtSiO2. The effect of the presence of unsaturated hydrocarbon on the reaction, H2 + D2 → 2HD, was also investigated. In hydrogenation of the acetylenes, trans-ene and alkane were formed via α,α,β-triadsorbed alkane and this path was promoted by relief of intraadsorbate strain in the adsorbed acetylene during hydrogenation of di-t-butylacetylene. Hydrogenations of the cis- and trans-di-t-butylethylene proceeded by different mechanisms because of intraadsorbate strain in the adsorbed cis-ene. The ratio of isomerization to hydrogenation in the reaction of the cis-olefin with H2 was very much larger than usual because of the large adsorbate-surface strain in the adsorbed trans-ene. Adsorbed hydrogen was in preequilibrium on neither catalyst and adsorbed olefin was not in preequilibrium on PtSiO2 and probably not on PdSiO2 except perhaps for trans-di-t-butylethylene. The effect of the two kinds of steric strain on rates of hydrogenation is analyzed.

Results of studies of the interaction of hydrogen and oxygen and of the activity and selectivity for hydrogenolysis of methylcyclopropane and for deuterium exchange with cyclopentane are given for a series of Pd SiO 2 catalysts previously described (Nandi, R. K., Pitchai, R., Wong, S. S., Cohen, J. B., Burwell, R. L., Jr., and Butt, J. B., J. Catal.70, 298, 1981; Nandi, R. K., Georgopoulos, P., Cohen, J. B., Butt, J. B., Burwell, R. L., Jr., and Bilderback, D. H., J. Catal.77, 421 (1982)). A comparison with prior studies of a similar series of Pt SiO 2 (Otero-Schipper, P. H., Wachter, W. A., Butt, J. B., Burwell, R. L., Jr., and Cohen, J. B., J. Catal.53, 414, 1978) is also given. Exchange patterns of the two series with cyclopentane are similar, although some differences in the details of D6–10 suggest that reaction via α,α-diadsorbed cyclopentane is relatively more important on Pt than on Pd. The turnover frequencies for i-butane formation in the hydrogenolysis reaction are six to eight times higher for Pt SiO 2 than for Pd SiO 2 , while n-butane formation on the two is about the same. As a result, the selectivities, iso/normal, are much lower on Pd. The patterns of structure sensitivity for this reaction on the two series also differ considerably, with an apparent maximum being observed for Pd SiO 2 for a catalyst of 65.5% exposed. Finally, the dependence of the nature of hydrogenolysis structure-sensitivity on conditions of pretreatment, very pronounced on Pt SiO 2 , is less on Pd SiO 2 .

Oxidation of diphenylpicrylhydrazine (DPPH.H) to diphenylpicrylhydrazyl (DPPH) is effected by mesoporous chromias with excess oxygen but not by their microporous equivalents, the pores of which are too small to permit entrance and egress of DPPH.H and DPPH. Of the iodometric oxidizing capacity of the macroporous chromias, 30–40% forms DPPH, 10% remains unused and the remainder is consumed in oxidizing DPPH to diamagnetic products. The reaction is nearly complete in one day at 25 °C. Addition of oxygen at this point leads to a substantial rate of catalytic oxidation which continues for days. The catalytic site is probably Cr(V) which is difficultly reducible because of some special structural feature. Mesoporous chromias devoid of excess oxygen upon which oxygen has been adsorbed at −78 °C oxidize DPPH.H to DPPH but they provide little catalytic oxidation. CrO3 on wide-pore silica also oxidizes DPPH.H to DPPH but its catalytic activity is small and dies rapidly. The lower catalytic activity is probably associated with the much higher dispersion of the CrO3SiO2. Studies by EPR of CrO3SiO2 gave results consistent with Cr(V) being the catalytic site. In the reduction of Cr(VI) by DPPH2, Cr(V) is an intermediate which reaches a maximum concentration of at least 9% and then declines.

Modern techniques for analysis of X-ray diffraction profiles have been applied to several of the platinum-silica gel catalysts described in Part I to determine average platinum crystallite size, percentage exposed, crystallite shape, size distribution, lattice parameter, residual stresses and strains, presence or absence of faulting, and mean-square amplitude of vibration. In addition the surface area of the gel was determined with small-angle scattering (SAS). There is good agreement between the percentage exposed of platinum measured by gas adsorption in Part I and the results reported here, calculated from crystallite sizes, indicating the crystallite size is the true platinum particle size. Crystallites to sizes as small as ≈25 A have been examined. Furthermore, the crystallites are equiaxed in shape (and definitely not cuboidal) and are strain and defect free, except in the case where the average size is near the pore size of the gel. Size distributions are sharper when the catalyst preparation is by impregnation rather than by ion exchange. The size distributions indicate that there is coalescence of some surface species during preparation rather than coarsening (Ostwald ripening). The meansquare amplitude of vibration of platinum increases by ≈30% as the particles decrease in size from ≈100 to ≈25 A. There is no change in the lattice parameter greater than ≈0.1%. The gel surface areas determined by SAS are in agreement with those determined in Part I by physisorbtion of nitrogen.

Chromia hydrogel precipitated by the urea method and dried in air at 25 °C ( 25 ) has little or no excess oxygen beyond Cr 2 O 3 . Dried at 85 °C ( 85 ), it has about 125 meq/mole Cr and 135 has 400–500. Removal of the excess oxygen by hydrogen at 300 °C gives an amorphous, microporous material designated A-Cr(III). Heating 25 to 300 °C in vacuo or in flowing helium gives a nearly identical A-Cr(III) and both can be heated to nearly 500 °C in vacuo or in helium without crystallization. The excess oxygen of 85 and 135 is unavailable to I − aq in heterogeneous iodometry. Activation in vacuo to progressively higher temperatures develops availability to I − aq and, after activation to about 350 °C, almost all excess oxygen of 135 reacts with I − aq. Activation to still higher temperatures leads to crystallization to mesoporous α-Cr 2 O 3 (probably by the fluxing action of Cr(VI)) with loss of about 70% of the area, 85 crystallizes at somewhat higher temperatures. By 450 °C, both have about 2.8 atoms of Cr(VI) per nm 2 . At 300 °C, hydrogen converts them to XCr(III), the crystalline analog of ACr(III). Oxygen adsorbed on ACr(III) at−78 °C (probably as Cr(IV)O 2 post− ) is completely available to I − aq, behaves as a 4- e oxidant and is reduced by hydrogen at 140 °C. Adsorption of oxygen at 100 °C is more complicated. Chromia with adsorbed oxygen crystallizes upon heating in vacuo at 400 °C.

Results of a detailed, quantitative investigation are reported of surface chemistry and catalysis involving selected organoactinides and partially dehydroxylated (PDA) or dehydroxylated (DA) alumina supports. For the complexes Cp'/sub 2/M(CH/sub 3/)/sub 2/ and Cp'/sub 2/M(CD/sub 3/)/sub 2/ (Cp' = eta/sup 5/ - (CH/sub 3/)/sub 5/; M = Th, U), methane-evolving surface reaction pathways are identified as M-CH/sub 3/ protonolysis via surface OH (especially on PDA), Cp' H atom abstraction, and intramolecular elimination of methane within M(CH/sub 3/)/sub 2/ units. This latter process is proposed on the basis of methylene transfer to acetone and some olefin metathesis activity to result in Al/sup 3 +/-stabilized alkylidenes. Hydogenolysis studies indicate that ca. 25% of the Cp'/sub 2/M(CH/sub 3/)/sub 2//DA surface M-CH/sub 3/ groups are removable as methane; reduction of methyl chloride to methane confirms the presence of surface M-H groups produced by hydrogenolysis. The Cp'/sub 2/M(CH/sub 3/)/sub 2//DA complexes are active catalysts for propylene hydrogenation following a variety of pretreatment conditions, with N/sub t/ approx. = 0.5 s/sup -1/ in a flow reactor at -63/sup 0/C (about 10 times more active than typical Pt/SiO/sub 2/ catalysts under the same conditions). M = Th and U are comparable in hydrogenation activity, and CO poisoningmore » experiments indicate that ca. 3% of the adsorbed molecules are catalytically active. Cp'/sub 2/M(CH/sub 3/)/sub 2/ complexes on PDA and silica gel are considerably less active catalysts. The Cp'/sub 2/M(CH/sub 3/)/sub 2//DA systems are also active catalysts for ethylene polymerization and weakly active for butene isomerization. Experiments with Cp'/sub 2/Th(CH/sub 2/C(CH/sub 3/)/sub 3/)/sub 2/ and (Cp'/sub 2/Th(..mu..-H)H)/sub 2/ on DA reveal activity for propylene hydrogenation comparable to the Cp'/sub 2/M(CH/sub 3/)/sub 2/ systems. In contrast, more coordinatively saturated Cp/sub 3/UCH/sub 3/ and Cp/sub 3/Th(n-C/sub 4/H/sub 9/) (Cp = eta/sup 5/-C/sub 5/H/sub 5/) are far less active, while Cp'Th(CH/sub 2/C/sub 6/H/sub 5/)/sub 3/ is far more active (N/sub t/ approx. = 10 s/sup -1/).« less

Nickel tetracarbonyl physisorbed from helium carrier onto partially dehydroxylated γ-alumina at 23–50°C rapidly reacts with surface sites to form a yellow complex, probably adsorbed nickel tricarbonyl. Carbon monoxide rapidly reverses this reaction. In flowing helium, the yellow complex loses further carbon monoxide to form a gray species in which CO Ni is about 1.4. Heating to 100°C reduces the ratio to 0.4. The gray species are small clusters of metallic nickel covered with carbon monoxide. Heating to about 450°C liberates additionally about 0.33 molecules of hydrogen, 0.1 of carbon monoxide, and 0.1 of carbon dioxide. This oxidation of nickel(0) by surface hydroxyl groups occurs even in the presence of hydrogen. Heating nickel carbonyl on highly dehydroxylated alumina liberates no hydrogen.

If Mo(CO) 6 /(dehydroxylated alumina, γ− + δ-Al 2 O 3 ) is activated in helium to 500 °C, clusters of Mo about 4 nm in diameter are formed. They are contaminated with some retained carbon which is essentially completely removed as methane by hydrogen at 650 °C. The average oxidation number of the molybdenum is about +0.3 and it is suggested that the clusters consist of Mo 2+ plus Mo(0) in which the charge is delocalized and balanced by AlO − groups at the surface. The clusters are not identical with metallic molybdenum since they do not chemisorb nitrogen. Materials so prepared exhibit a turnover frequency for formation of methane at 300 °C of about 0.04 sec −1 per surface atom of molybdenum in flowing hydrogen plus carbon monoxide (mole ratio = 5). Smaller amounts of ethane and propane are also formed. Activation at 300–500 °C in helium or hydrogen leads to similar turnover frequencies and selectivities. Activation of Mo(CO) 6 /(alumina dehydroxylated at 475 °C) at 300 °C in hydrogen or helium results in a material in which Mo is ionic and highly dispersed. Turnover frequencies for formation of methane at 300 °C are about 0.001 sec −1 per atom of molybdenum and selectivities are about the same as for the catalyst on dehydroxylated alumina. Activation in hydrogen results in formation of about 1.4 molecules of methane per molecule of Mo(CO) 6 and the resulting water, in essence, further oxidizes the molybdenum beyond the Mo 2+ formed initially by reaction of Mo(0) with surface hydroxyl groups, Mo + 2AlOH 8 → Mo 2+ + H 2 + 2AlO 8 post− .

The hydrogenolyses in a pulse reactor of propane, butane, isobutane, pentane, isopentane, cyclopentane, and neohexane occur at roughly equal rates on Mo(0)Al2O3, but those of ethane and neopentane are slower and that of cyclopropane is several times faster. With large enough alkanes, single hydrogenolysis was accompanied by double and triple hydrogenolysis as initial reactions, but the C2(ads) fragment underwent further cleavage rather slowly. Cleavage of both terminal and medial bonds was observed, but that of terminal bonds predominated. The activity of Mo(0)Al2O3 is roughly equal to that of IrAl2O3, but divergence among the rates of hydrogenolysis of various alkanes is smaller on Mo(0)Al2O3. Since butenes gave rates of hydrogenolysis at 300 °C about equal to that of butane, a 1,2-diadsorbed alkane is not likely to be an intermediate in hydrogenolysis. The rate of hydrogenolysis of propane and butane and the degree of multiple hydrogenolysis of cyclopropane increased rapidly as the average ON (oxidation number) of the molybdenum decreased in the series, ON ≅ +4, ON ≅ +2, and ON = 0, but the rate of hydrogenolysis of cyclopropane changed little. The hydrogenolysis of alkanes is likely to involve two or more simultaneous mechanisms occurring at significant rates.

Pulsed hydrogenation of ethylene at -31/sup 0/C and the hydrogenolysis of cyclopropane at 0/sup 0/C in hydrogen or helium carrier gas on silica-supported platinum showed that the conversion was catalyzed by the platinum surface and that carbonaceous residues, which accumulated at a rate of approx. 0.05Vertical Bar3< of each pulse, were inactive. The residues were removed by reaction with hydrogen at 50/sup 0/-100/sup 0/C. The data were compatible with a mechanism in which the platinum surface is almost covered by a monolayer of hydrocarbon during the reaction and in which the adsorbed material reacts sequentially with two adsorbed hydrogen atoms and desorbs as alkane.

Isotopic exchange between deuterium and 2,2-dimethylbutane (neohexane) has been investigated on a series of Pt SiO 2 (percentage exposed of metal, Dh = 6.3–81%) and Pd SiO 2 (Dh = 13.8–96.7%) catalysts at temperatures in the vicinity of 100 °C. Catalysts were investigated after three different pretreatments: standard: O2,300 °; H2,300 °; He,450 ° and H2,450 °: O2,300 °; H2,450 ° and H2, 100 °:O2,300 °; H2,100 °. One might expect exchange into the methylene group to be subject to substantial steric hindrance from the adjacent t-butyl group and the degree of steric hindrance to be less at edge atoms than at atoms in (100) and (111) faces, i.e., relatively less on small crystallites than on larger ones. In general, the effect of Dh upon the relative yield of ethyl-d5 in neohexane did not accord with this expectation. A possible explanation is that the bounding faces of the crystallites are defective and contain many vacancies which generate sites about equivalent to edge sites in the degree of hindrance. If so, usual tests of structure sensitivity could be seriously affected.

The hydrogenation of propylene and the hydrogenolysis of cyclopropane has been studied on Mo(CO) 6 Al 2 O 3 activated in flowing He or H2. The γ-alumina was either partially dehydroxylated in flowing He at 475 °C (PDA) or nearly completely dehydroxylated at 950 °C (DA). In a pulse reactor with hydrogen carrier, Mo(CO) 3 PDA converted propylene both to ethylene + 2-butene (metathesis) and to propane (hydrogenation). Activation in He or in H2 to 300 °C (average ON ∼- +2 and +3, respectively) greatly enhanced the activity of Mo(CO) 3 PDA for hydrogenation which had to be measured in a flow reactor at −46 °C. Mo(CO) 6 DA activated to 300–500 °C in He or in H2 (average ON near zero) was an even more active catalyst and heating DA;He,300 ° in H2 to 500 °C led to a still more active catalyst. Hydrogenation was approximately zero order in propylene and first order in hydrogen. The same materials catalyzed the hydrogenolysis of cyclopropane at 0–100 °C. DA;He,300 ° and PDA;He,300 ° were the most active catalysts. Both single and double hydrogenolysis occurred as initial reactions to form propane and methane + ethane, respectively. Double hydrogenolysis was favored by molybdenum in lower ON.

Clean chromias of smaller and larger degrees of dehydroxylation, i.e., smaller and larger surface concentrations in Cr3+(cus), were prepared by activation in flowing helium to 315°C (microporous, amorphous) and to 520°C mesoporous, microcrystalline). These chromias were then exposed to argon, hydrogen, methane, or nitrogen at −160 to −78°C, flushed with helium, and the residual amount of adsorbate was determined after various times. From these data and from displacement experiments in which pulses of one gas were passed over chromia on which another gas was adsorbed, the relative strengths of adsorption at Cr3+(cus) are Ne ⪡ Ar ⪡ H2 ⪡ CH4

Pt catalysts with a wide range of percentage metal exposed have been examined by wide-angle and small-angle X-ray scattering, extended X-ray absorption fine structure (EXAFS), and X-ray fluorescence. The wide-angle studies were carried out in a special cell that permitted treatment in a gas environment at moderate temperatures. For high percentage metal exposed, the wide-angle patterns (as well as all EXAFS spectra) were obtained with X rays from the storage ring at Cornell University. After exposure to air, a catalyst with 63.5% Pt exposed consists of particles of crystalline Pt 3 O 4 with a core of Pt, whereas for 81% exposed, the catalyst is nearly fully oxidized. There is no indication of any difference between the PtPt distance in these catalysts and that in bulk Pt, nor is there any difference for 40% exposed or less (for which the exposed catalyst is Pt metal), even after reductions in H 2 . There is little contact area between the catalyst and the support. For the lowest percentage metal exposed (6.3%), reductions in hydrogen do not change the microstrains within particles, but there is reduction in particle size, which is greater the higher the reduction temperature. The activity of this catalyst decreases in the same manner. The mean-square amplitude of vibration increases with the temperature of H 2 reduction for several catalysts, as does the catalytic activity.

Simple techniques for studies of X-ray diffraction from PdSiO2 catalysts during catalytic reactions are reported here. The catalytic activity of PdSiO2 of low percentage metal exposed for methylcyclopropane hydrogenolysis at 0 °C is less for catalyst cooled from 450 °C in H2 than for that cooled in He. This effect appears to result from differing amounts of hydride formation. Ease of formation of hydride decreases with decreasing Pd particle size. Exposing the catalysts with Dh = 13.8 and 29.3% to hydrogen at ~25 °C results in nearly complete conversion of particles of Pd to hydride. Purging with helium (even at 0 °C) reconverts the hydride to palladium, but this reconversion exhibits an induction period; hydride formation is more rapid. Passing hydrogen plus methylcyclopropane results in the conversion into hydride of a substantial fraction of catalyst particles originally present as palladium. The palladium hydride particles of catalyst produced by cooling from 450 °C in H2 remain as PdH0.7 during the hydrogenolysis reaction. The lattice parameter of palladium particles is indistinguishable from that of bulk palladium, at least for particle diameters ≥45 A.

The adsorption of hydrogen, carbon monoxide, and oxygen and X-ray line profile analysis have been investigated on a series of Rh SiO 2 catalysts made by ion exchange with Rh(NH3)5(H2O)3+ and by impregnation with Rh4(CO)12 and Rh6(CO)16. The percentage exposed (Dh) of the series of catalysts as measured by hydrogen chemisorption at 25 °C varied from 11 to 108%. Ease of reduction of the oxidized catalysts decreased as Dh increased and reduction in hydrogen was incomplete in hydrogen at 25°C except for the catalysts of smallest Dh. The rates of hydrogenolysis of methylcyclopropane (MCP) at 0 °C decreased with increasing Dh by a factor of about 9 over the range of Dh of the series of catalysts for catalysts pretreated 02, 300 °, 0.5;H2,300 °,1;He, 450 °, 1 where the last number in each triad was the time of exposure in hours. Omission of He,450 °, 1 led to decreased rates and augmented structure sensitivity, but H2,25 °, 1 without He,450 °, 1 led to inverted structure sensitivity, presumably because of residual oxygen on rhodium. The activation energy for hydrogenolysis was about 37 kJ mol−. The ratio of isobutane to butane in the products was about 20 for Dh = 11%, declined rapidly with increasing Dh and became about 3.5 for Dh > 50%. Pretreatment conditions (temperature of pretreatment by H2, presence or not of terminal He,450 °, 1) had little effect on the ratio.

The interaction of Mo(CO) 6 with γ-alumina has been studied in the temperature range of −17 to 1000 °C. It is likely that most of the dominant surface species formed during the decomposition of the complex have been identified. Activation in flowing helium at 100 °C of Mo(CO) 6 on partially hydroxylated alumina results in the formation of Mo(CO) 3 ads which is 100% dispersed and probably molecularly dispersed. Above 200 °C, the remaining CO of Mo(CO) 3 ads is liberated but detectable amounts of Mo(CO) 2 ads and Mo(CO)ads do not accumulate. It is likely that “(σ-O − ) 2 Mo” is the dominant species after activation near 270 °C. After heating to 500 °C, the average oxidation number of the decarbonylated Mo is about 5.6. The oxidation of the originally zero-valent Mo occurs due to reaction with surface hydroxyl groups of the alumina. Mo(VI) also results from heating Mo(CO) 2 O 2 ads and “Mo(CO)(O 2 ) 2 ads” [both formed by reaction of Mo(CO) 3 ads with O 2 ] to 500 °C. Alumina pretreated at 950 °C behaves as a quite different support. Due to the absence of hydroxyl groups, heating Mo(CO) 6 on this material leads to the formation of molybdenum mostly in oxidation number zero. Since Mo(0) supported on alumina is difficult to prepare by other techniques, dehydroxylated alumina may provide a useful synthesis for this type of catalyst.

The activities and selectivities of the series of Pt SiO 2 catalysts described in Parts I [Uchijima, T., Hermann, J. M., Inoue, Y., Burwell, R. L., Jr., Butt, J. B., and Cohen, J. B., J. Catal. 50, 111 (1977)] and II [Sashital, S., Cohen, J. B., Burwell, R. L., Jr., and Butt, J. B., J. Catal. 50, 111 (1977)] have been determined for the hydrogenation of cyclopropane, methylcyclopropane, and propene. All three reactions are found to be mildly structure sensitive under the conditions of investigation, with variations in the turnover numbers by a factor of about; 3 over the range of metal exposure. Activation energies were essentially the same for the three reactions and in all cases were independent of the percentage of Pt exposed. On this basis, a simple geometric correlation is shown to provide a reasonable explanation of the data, except where significant strain exists within the metal crystallites. It is also shown that an excellent correlation of activity vs metal vibrational amplitude is obtained for these catalysts.

Metallic molybdenum on dehydroxylated alumina (prepared by impregnating Mo(CO)6 onto alumina which had been dehydroxylated in flowing helium at 950 °C and then treating the Mo(CO)6Al2O3 by He,300–500 °;H2,650–950 °) is one of the more active catalysts for the hydrogenolysis of propane. Its turnover frequency per surface Mo atom at 250 °C is ~0.2 sec−1 at P = l atm and C3H8H2 = 0.038. Both single and double hydrogenolyses occur simultaneously, the latter having the larger apparent activation energy. MoAl2O3 catalysts with higher oxidation numbers of molybdenum are less active. Mo(0)Al2O3 is easily poisoned for hydrogenolysis by O2, CO, H2O, and even to some extent by N2. Initial activity is restored by H2 at ~950 °C. MoO3/hydroxylated alumina is reduced by H2 at 650 °C to an average ON (oxidation number) of Mo of +2 and by 800–950 °C to an ON of zero. The catalytic activities of these materials for the hydrogenolysis of propane at 200–300 °C, the hydrogenation of propylene at −46 °C, the metathesis of propylene at 65 °C, and the hydrogenolysis of cyclopropane at 0–100 °C are very nearly the same as those of Mo2+Al2O3 and Mo(O)Al2O3 made from Mo(CO)6Al2O3.

Isotopic exchange between cyclopentane and deuterium, mostly at 81/sup 0/C, has been investigated on a set of 10 Pt/SiO/sub 2/ catalysts whose percentages exposed ranged from 6.3 to 81%. The standard pretreatment, O/sub 2/, 300/sup 0/C; H/sub 2/, 300/sup 0/C; He 450/sup 0/C with cooling in helium, leads to turnover numbers for formation of exchanged cyclopentanes, which decline by a factor of less than 2 as the percentage exposed increases from 6.3 to 63.5% and then at 81% rise to about the value for 6.3%. Selectivity to various exchanged species varies with percentage exposed and, in particular, D/sub 6//D/sub 10/ (the ratio of initial yields of d/sub 6/ and d/sub 10/) decreases substantially as the percentage exposed increases. This suggests that exchange of the hydrogen atoms on both sides of the cyclopentane ring proceeds mainly via formation of ..cap alpha..,..cap alpha..-diadsorbed cyclopentane on catalysts with a smaller percentage exposed and by rollover on catalysts with a large percentage exposed. Other pretreatment conditions such as O/sub 2/, 300/sup 0/C; H/sub 2/, 100/sup 0/C and O/sub 2/, 300/sup 0/C; H/sub 2/, 450/sup 0/C with cooling in hydrogen give larger structure sensitivites for activity and different selectivities. Catalysts deactivate rather rapidly but withmore » little change in selectivity. Deactivation is accompanied by the formation of carbonaceous residues. C/Pt/sub s/ becomes approximately 0.5 and the activity declines 10% in about 10 min on stream. Ninety percent or more of the carbon in the residue occurs in a cyclopentane structure because 90% of the carbon is removed as cyclopentane by hydrogen at 100/sup 0/C. Repetition of the standard pretreatment regenerates the initial activity and selectivity.« less