Effect of vanadium dispersion and of support properties on the catalytic activity of V-containing mesoporous silicas

Introduction SBA-15 and MCF mesoporous silicas that are characterized by large and uniform pores size, thick walls and high surface area, represent very interesting supports for catalytically active species (e.g. vanadium) allowing a large concentration of accessible, isolated and well defined active sites to be obtained [1]. On the other hand, vanadium-based systems are effective in many oxidation reactions, which can be kinetically modeled through a MvK mechanism. In this work, two series of highly dispersed vanadium-containing systems (namely V-SBA-15 and V-MCF) were prepared by direct synthesis and tested in the catalytic decomposition of dichloromethane, the most stable chlorinated-alkane. Physico-chemical and catalytic properties of both V-SBA-15 and V-MCF systems were compared with those of V-containing mesoporous samples prepared by wet impregnation and of a non-porous V-containing silica sample. Experimental part V-SBA-15 materials with different V-loadings (0.7-4.8 wt.%) were prepared by direct synthesis as reported in the literature [2]. Likewise, two V-MCF samples (2.0-4.5 wt.% vanadium) were prepared by a new direct synthesis (currently under publication). For comparison, two catalysts were synthesized by impregnation of SBA-15 and MCF supports (referred to as V/SBA(i) and V/MCF(i), respectively). A non-porous sample with 3 wt.% vanadium (denoted as V-Si) was obtained by flame pyrolysis method, according to Ref. [3]. Samples were characterized by means of powders XRD, FESEM and TEM microscopies, N2 sorption isotherms at -196 °C, H2-TPR, EPR, 51 V MAS NMR, DR UV-vis, micro-Raman and FT-IR spectroscopies. The decomposition of dichloromethane was used as probe reaction for catalytic oxidation of Cl-VOCs. Catalytic measurements were carried out in a fixed-bed reactor under both aerobic and anaerobic conditions at different reaction temperatures (200-500 °C). Results and discussion The samples prepared by direct synthesis showed higher specific surface areas (up to 820 m2g-1 for V-SBA-15 and 925 m2g-1 for V-MCF) as compared to impregnated ones (530 m2g-1 for V/SBA-15(i) and 645 m2g-1 for V/MCF(i)), in which a decrease occurred of both surface area and porous volume. Micro-Raman spectroscopy of dehydrated V-SBA-15 and V-MCF samples showed V=O stretching modes of isolated V species in tetrahedral coordination (1035 cm-1), whereas typical bands of polymeric VOx species and micro-crystalline V2O5 were observed with V-impregnated samples. Likewise, TEM images revealed the presence of VOx aggregates only at the surface of impregnated samples. DR UV-vis spectra of dehydrated V-SBA-15 systems showed the occurrence of V5+ species in tetrahedral coordination (absorption band at about 260 nm), whereas upon rehydration such species were converted to octahedral ones by coordination of water molecules (absorption band at about 380 nm). By contrast, a higher amount of octahedral coordinated V species was observed with impregnated samples under dehydrated conditions. Indeed, IR spectroscopy showed that more abundant and acidic sites were obtained by impregnation. H2-TPR analysis showed a much easier V reducibility in both V-SBA-15 and V-MCF samples, due to better V dispersion reached by direct synthesis. Higher catalytic activity for the dichloromethane decomposition was achieved over V-SBA-15 and V-MCF samples directly synthesized with respect to impregnated samples, as consequence of their much higher V dispersion (Fig. 1). Furthermore, better dichloromethane conversions were obtained over V-SBA-15 as compared to V-MCF systems (a similar trend was observed with impregnated samples). These findings seem to be related to the different mesoporous systems under study: molecules should be more retained within pores of SBA-15 framework as compared to ultra-large pores of three-dimensional MCF systems. On the other hand, a non-porous sample (V-Si), with V species well dispersed and incorporated into the silica framework, exhibited lower activity than mesoporous samples, confirming the important role of mesoporosity in total oxidation reactions. [1] J.M. Thomas, J.C. Hernandez-Garrido, R. Raja, R.G. Bell, Phys. Chem. Phys. 11 (2009) 2799. [2] M. Piumetti, B. Bonelli, M. Armandi, L. Gaberova, S. Casale, P. Massiani, E. Garrone, Micropor. Mesopor. Mater. (2010) doi:10.1016/j.micromeso.2010.04.011 [3] I. Rossetti, L. Fabbrini, N. Ballarini, F. Cavani, A. Cericola, B. Bonelli, M. Piumetti, E. Garrone, H. Dyrbeck, E. A. Blekkan,L. Forni, J. Catal. 256 (2008) 45