STUDIES ON RUTHENIUM CATALYSTS .3. EFFECTS OF TYPE OF SUPPORT, METHOD OF PREPARATION AND CONDITIONS OF USE OF RU/TIO2 CATALYSTS ON THEIR BEHAVIOR IN NORMAL-BUTANE HYDROGENOLYSIS

Progress in developing theoretical models to describe the behaviour of supported metal catalysts requires experimental measurements of the highest reliability on structure-sensitive reactions, with a full record of the catalysts' history (preparation, pretreatment etc.). The study of Ru/TiO2 catalysts derived from RuCl3, and subjected to both low- and high-temperature reductions, constitutes a rigorous test of experimental control. Depending on the precise schedule and sequence of the pretreatments, Ru particles may either (a) be contaminated by Cl-, (b) suffer from either a partial or complete SMSI or (c) be relatively clean. Each state shows a characteristic range of activities and selectivities for the test reaction (n-butane hydrogenolysis); isothermal (433 K) rates vary by more than 10(5), and contaminated surfaces give high ethane selectivities (> 1.3). The variability of Ru/TiO2 catalysts containing 0.1-5% Ru prepared on a small scale by various workers over a period of time is assessed with respect to these kinetic parameters. Rates vary considerably, especially after high-temperature reductions (HTR); selectivities are less variable. Oxidation and low-temperature (433 K) reduction after HTR gives more uniform products. Differences in rate are not accounted for by variations in dispersion, because H-2 chemisorption does not accurately count the number of active sites, especially when the surface is contaminated. Repeat preparations by a single operator with the starting materials can however afford closely-reproducible products. Varying the acidity of the RuCl3 solution used for impregnation or the crystal structure of the support is without major effect; calcining the RUCl3/TiO2 precursor before the first HTR has a marked influence, as the Cl- concentration is thereby lowered. One catalyst made by ion exchange using [Ru(NH3)6]3+ was almost inactive for hydrogenolysis, but showed activity for dehydrogenation and isomerisation. Catalysts made by impregnation lose up to 60% of their initial activity and thereafter are perfectly stable; deactivation causes only small changes in product selectivities.