Michael R. Gelsthorpe

Hydrogenolysis of alkanes. Part 4. —Hydrogenolysis of propane, n-butane and isobutane over Pt/ Al2O3 and Pt–Re/Al2O3 catalysts

The hydrogenolysis of propane, n-butane and isobutane on Pt/Al2O3(0.3 and 0.6 % Pt) and on Pt–Re/Al2O3(0.3 % Pt, 0.3 % Re) has been investigated by a thermal cycling technique that permits evaluation of initial rates of deactivation and of consequential changes in product distributions. Less deactivation is found with propane than with the butanes; with n-butane on Pt/Al2O3 catalysts, two types of sites are apparent, a one of which, having a higher probability of giving central C—C bond fission, is selectively deactivated. However, the relative chances of desorption and of further hydrogenolysis of the adsorbed intermediates are the same at both types of site.With the Pt–Re/Al2O3 catalyst, further bond-breaking in the adsorbed species formed from all three alkanes is relatively more favoured than with the Pt/Al2O3 catalysts, and with n-butane the chance of central bond-fission is enhanced. No changes in product distribution result from initial deactivation, which is comparable to that found with Pt/Al2O3 catalysts. Isomerisation of n-butane is slight (> 3 %) on all three catalysts.Multiply adsorbed species are suggested as possible intermediates, and differences in product selectivities are attributed chiefly to variations in electron density at the active site.

A study of EUROPT-1 by temperature-programmed reduction

Reduction of EUROPT-1 in the ‘as-received’ state begins at 200 K, attains a maximum rate at about 330 K and is complete by 423 K: after correction for the overlap between hydrogen uptake and desorption, and for hydrogen chemisorption, the hydrogen consumed in the reduction gives the initial composition of the catalyst as PtO0.9. Re-oxidation does not occur immediately on exposure to air at room temperature, but slowly, over a period of days, and gives initially another oxide phase more easily reducible than that present in the ‘as-received’ material. Oxidation of the reduced catalyst begins at 183 K and is complete at 470 K, but the product of this accelerated oxidation is in part more reactive than that formed by oxidation at room temperature. The results are consistent with the concept that ‘as-received’ EUROPT-1 contains a disordered Pt oxide phase.

Sustained hysteresis in a catalysed reaction: Hydrogenolysis of alkanes on Re/Al2O3 catalysts

Alkane hydrogenolysis on Re/Al2O3 catalysts has been studied by a thermal cycling procedure. Because deactivation occurs during heating and is then reversed at higher temperatures, rates are much faster during cooling than during heating. Rates during subsequent thermal cycles are reproducible. Re is 104 times more active than Pt for hydrogenolysis.

Modified Aluminophosphate Microporous Solids

Abstract A combination of physical techniques, including FTIR, XRD, gas adsorption and thermal methods, have been used to characterise two modified ALPO-5 phases. The first of these, designated as ZAPO-5, was prepared by the incorporation of zinc at the crystallisation step, whereas the second, Si-ALPO-5, resulted from interaction with SiCl 4. The latter phase was found to possess mesopores as well as micropores. The surface acidity of both these solids and of pristine ALPO-5 was investigated.

Modified aluminophosphate molecular sieves: preparation and characterisation

Abstract Crystalline microporous aluminium phosphate has been prepared in a variety of ways. The influence of the preparative conditions affecting the solid state properties of the resultant solids were studied. Quencg cooling of the reaction mixture led to a change of crystal morphology, but not in structure. Incorporation of zinc also affected the crystal morphology, but the pore structure was essentially unaltered for both systems.

Hydrogenolysis of propane, n-butane and isobutane on rhenium/alumina catalysts

The hydrogenolysis of n-butane and isobutane and of propane has been studied on Re/Al2O3 catalysts containing 1.5, 2.2 and 2.75% Re; they were reduced at 763 K. The reactor temperature was raised stepwise from its initial value (330–360 K in later experiments) to 600 K, analyses being performed at 10 or 15 K intervals, the process being repeated as the temperature was returned to its starting value. This thermal cycle was repeated at least once in every case.With n-butane, the rate initially obeyed the Arrhenius law (the α-phase), but deactivation set in at about 400 K and the rate passed through a maximum. It subsequently renewed its upward trend (the β-phase), but it appeared that the surface was cleansed at high temperature, as the rate was much faster for most of the cooling stage (γ-phase). In the case of n-butane, the rates at each point of the cycle were quite reproducible in the second and subsequent cycles; the rates for isobutane and for propane showed little change with cycle number, although with propane deactivation started at low temperature. Rates, activation energies and product selectivities were closely reproduced, especially in the γ-phase. This sustained hysteresis of catalytic rate is explained by an alternate coking and cleansing of the surface of the Re particles.

Promotion and Selective Poisoning of Supported Metal Catalysts

The term ‘promoter’ is generally applied to substances added to metallic catalysts to improve their performance. There is no distinction in principle between promoters that enhance selectivity or lifetime, and selective poisons that eliminate unwanted reactions. This concept is illustrated by reference to the effect of (i) Re on Pt/Al 2 O 3 , (ii) transition metal oxide additives to Ru/SiO 2 , and (iii) high-temperature reduction of Ru/TiO 2 , using alkane hydrogenolysis and isomerisation as test reactions. The results are interpreted in terms of a decreasing size of the active metal ensemble, and, with Pt-Re, the creation of new active centres.

Induction of mesoporosity in ALPO-5. Treatment with silicon tetrachloride

Reaction of microporous AlPO4(ALPO-5) with SiCl4 vapour in a nitrogen atmosphere at 870 K results in the partial isomorphous substitution of P predominantly by Si. The product has fewer but stronger acid sites than the parent phase, whilst crystallinity is retained. Isothermal adsorption studies with H2O and N2 probe molecules indicates that some adsorption sites are lost via reaction with SiCl4, and there is evidence for the development of mesoporosity in addition to the micropores already present. It is argued that reaction occurs preferentially at the intercrystallite regions in the aggregates, and at intracrystalline defects, i.e. defects in the pore structure. Comparisons with SAPO-5 are also made.

The effect of carbonaceous deposits on product selectivity in alkane hydrogenolysis

The hydrogenolysis of propane and of n-butane on Pt/Al2O3 and Pt-Re/Al2O3 catalysts has been studied using a stepwise temperature programme which reveals clearly the effects of carbonaceous deposits on both rate and product selectivities. On Pt/Al2O3 catalysts, deposits formed during the reaction of n-butane suppress C2H6 formation and enhance C3H8 formation, due solely to a decrease in the probability of central C-C bond fission. On Pt-Re/Al2O3, such deposits have little or no effect on product selectivities.

The efficient removal of organic templating molecules from aluminophosphate molecular sieves

Treatment with methanolic HCl leads to the complete removal of templating molecules from the title solid; calcination in air, a process normally used in the case of zeolites, leads to deposition of carbonaceous material in the present system.