Surface organometallic chemistry on metals: Preparation of bimetallic catalysts by controlled hydrogenolysis of Sn(n-C₄H₉)₄ on a Ni/SiO₂ catalyst

Selective hydrogenolysis of Sn(n-C₄H₉)₄ on a Ni/SiO2 catalyst has been carried out at various temperatures and at various coverages of the metallic surface by the organotin(IV) compound. The surface reaction was followed by chemical analysis of the gaseous and surface products, electron microscopy, infrared spectroscopy, magnetic measurements, and EXAFS measurements. At room temperature, and in the absence of metallic particles, Sn(n-C₄H₉)₄ is simply physisorbed on the silica surface and can be easily extracted by n-heptane. In the presence of metallic Ni, and provided that the amount of Sn introduced represents less than ca. a monolayer of the metallic surface, hydrogenolysis of Sn(n-C₄H₉)₄ occurs exclusively on the Ni particles, as evidenced by STEM-EDAX measurements, infrared spectroscopy, and EXAFS measurements. Analysis of the gaseous products evolved at increasing temperatures, infrared measurements, and EXAFS measurements indicate a stepwise cleavage of Sn-alkyl bonds. At temperatures lower than 323 K, there is formation of relatively stable surface organometallic fragments which can be formulated as Ni_s[Sn(n-C₄H₉)x]_y in which "Ni-_s" represents surface nickel atoms (x = 2, 3). The stoichiometry of the surface reaction leading to stable surface organometallic fragments depends both on the surface coverage (y) of the nickel particle by the organometallic compound and on the temperature of the hydrogenolysis. For the sample obtained after hydrogenolysis at 323 K for 10 h and for a Sn/Ni_s ratio of 0.5, an average of 2.5 butyl groups remain on the surface. The EXAFS experiments (at the Sn K-edge) carried out on such a sample show that adsorbed Sn is surrounded on average by 2.5 light backscatterers (very probably C type) and by 1.3 heavy backscatterers (most probably Ni). Various possible structures have been proposed, taking into account the results of the analytical and EXAFS data, as well as the magnetic and chemisorption measurements. For reaction temperatures greater than 373 K, all butyl groups are hydrogenolysed mainly as butane. STEM-EDAX experiments show that, provided that the amount of Sn introduced is lower than a monolayer of the Ni_s, the signal of Sn is always associated with that of Ni, indicating the lack of migration of tin to silica. Electron microscopy indicates that the average metallic particle size on the surface increases by a value of about 1.5 nm. For a Sn/Ni_s ratio of 0.5, EXAFS studies show that each "naked" tin atom is surrounded by 4 Ni atoms of less, suggesting that tin atoms have not migrated into the nickel particles. Magnetic measurements suggest that each tin atom has then the same magnetic effect as 4.5 chemisorbed hydrogen atoms.