Catalysis Letters 64 (2000) 191–196 191 Dynamic states of V-oxide species: reducibility and performance for methane oxidation on V 2 O 5 /SiO 2 catalysts as a function of coverage M.A. Ba˜ nares a,* , J.H. Cardoso b , F. Agull´ o-Rueda c , J.M. Correa-Bueno b and J.L.G. Fierro a a Instituto de Cat´ alisis y Petroleoqu´ ımica, CSIC, Campus UAM-Cantoblanco, E-28049 Madrid, Spain E-mail: mbanares@icp.csic.es b Departamento de Engenharia Qu´ ımica, UFSCAR, S˜ ao Carlos, S˜ ao Paulo, Brazil c Instituto de Ciencia de Materiales, CSIC, Campus UAM-Cantoblanco, E-28049 Madrid, Spain Received 16 July 1999; accepted 26 November 1999 Temperature-programmed in situ Raman spectroscopy is used to understand the effect of surface vanadia coverage on the structure, reducibility and performance for the oxidation of methane on V 2 O 5 /SiO 2 catalysts. The vanadia coverage on silica has no effect on its structure below its dispersion-limit loading (“monolayer” coverage); however, the interactions among surface vanadia species under reducing conditions become increasingly important. This interaction appears to operate through the sharing of oxygen sites facilitating the reduction, but it does not alter the total reducibility. The probability for this interaction to take place increases with vanadium oxide surface coverage. It is therefore expected that under reaction conditions the catalyst with higher vanadia coverage would have a greater capacity to release oxygen. This would increase the activity per vanadium site. Keywords: supported vanadia, surface interaction, coverage effect, reducibility, methane oxidation, temperature-programmed Raman, TOF 1. Introduction The direct conversion of methane to more valuable C 1 oxygenates is a very appealing route for the conversion of the large reserves of natural gas (>90% CH 4 ) [1,2]. Much work in this direction has been done in recent years, al- though the results reported so far have not always been encouraging [3,4]. It has been shown that the rate of de- struction of HCHO on the catalyst surface is higher than in the gas phase [5]. We have recently reported that product distribution depends on the BET area of the support [6] and on the type of the redox oxide [7]. However, from a molecular point of view, the role of the coverage of a sup- ported redox oxide in activity and selectivity is not fully understood. The TOF values for vanadia in a V 2 O 5 /SiO 2 series are affected by the coverage of vanadia on silica in the sub-“monolayer” region [8]. This is an unexpected re- sult since the nature of the surface vanadium oxide species on silica is not affected by their surface coverage below its dispersion limit loading [9]. This work aims to explain the contradiction observed where the same sites afford different TOF numbers for methane conversion. In situ TP-Raman characterization provides valuable information about the behavior of sur- face vanadia sites at different coverage values under differ- ent conditions (catalytic reaction, TPR, TPO). * To whom correspondence should be addressed. 2. Experimental Commercial non-porous silica (Aerosil 200, BET area 174 m 2 /g, and particle size ca. 14 nm) was used as the starting material. Silica samples with lower BET areas were prepared by calcining several aliquots of the start- ing silica at temperatures in the 1273–1423 K range for a period between 3 and 9 h. The carriers were labeled “Sy”, where y refers to the BET area of the support. Sy supports were impregnated with an aqueous solution of ammonium metavanadate (Aldrich) and hydrogen peroxide in a rotary evaporator at 343 K. The vanadium oxide concentration was close to 0.3 and 0.8 V atoms/nm 2 of the silica support, which is below monolayer capacity of vanadia on silica (ca. 1 V atom/nm 2 ). The impregnates were dried at 383 K and calcined in two steps: 623 K for 2 h and 923 K for 5 h. After calcination, they were sieved to a particle size from 0.125 to 0.250 mm diameter. The catalysts are re- ferred to as “xVy”; in this case x denotes the coverage of vanadium, expressed in vanadium atoms per nm 2 of the silica support. The vanadium concentration of each sample was deter- mined by atomic absorption spectrophotometry in a Perkin– Elmer 3030 instrument. The BET area of the catalysts was calculated by the BET method from the N 2 adsorption– desorption isotherms recorded at 77 K in a Micromeritics ASAP 2000 apparatus. XP spectra were obtained using an ESCALAB 200R spectrometer fitted with a hemispherical electron analyzer and a Mg anode X-ray exciting source J.C. Baltzer AG, Science Publishers