- Email: [email protected]
Characterization and reactivity of vanadia–molybdena catalysts supported on γ-Al2O3
Applied Catalysis A: General 202 (2000) 133–139
Characterization and reactivity of vanadia–molybdena catalysts supported on ␥-Al2 O3 Komandur V.R. Chary∗ , Thallada Bhaskar, Jangam Jyothi Maheshwar, Kanaparthi Ramesh, Vattikonda Venkat Rao Catalysis Section, Indian Institute of Chemical Technology, Hyderabad 500 007, India Received 16 June 1999; received in revised form 5 January 2000; accepted 25 January 2000
Abstract The influence of molybdenum oxide on the dispersion of vanadium oxide supported on alumina was investigated. A series of V2 O5 –MoO3 catalysts with varying MoO3 content ranging from 1–5% (w/w) were prepared by impregnation of previously prepared 10% V2 O5 /␥-Al2 O3 with requisite amounts of ammonium molybdate solution. Dispersion of vanadia was determined by oxygen chemisorption at 195 K and was found to decrease with the increase of molybdena loading. The calcined catalyst samples were characterized by electron spin resonance (ESR) technique in their hydrogen reduced and unreduced conditions. The XRD results suggest formation of vanadium aluminate with the addition of MoO3 to V2 O5 /␥-Al2 O3 catalyst. The ESR spectra of hydrogen reduced catalysts show doublet structure at the low field parallel components, suggesting the presence of two chemically distinct V4+ on alumina species. It was found that the nature of V4+ is strongly influenced by the added molybdena. The catalytic properties were evaluated for the vapour-phase oxidation of methanol. The activity during oxidation reaction was found to decrease with the increase of molybdena loading. However, the selectivity of formaldehyde was found to increase with molybdena loading, indicating that the added MoO3 created additional sites for partial oxidation reaction. The presence of molybdenum oxide inhibited the interaction between vanadium and alumina, leading to poor dispersion of vanadia and lower catalytic activity during selective oxidation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Oxygen chemisorption; Electron spin resonance; Vanadia; Molybdena
1. Introduction Supported vanadium oxides have been widely investigated as they catalyse a number of reactions including selective oxidation, ammoxidation, as well as the selective catalytic reduction of NO by NH3 [1–17]. The catalytic properties of vanadia in selective oxidation reactions are strongly influenced by the method ∗ Corresponding author. Tel.: +91-40-7171510; fax: +91-40-7173757. E-mail address: [email protected] (K.V.R. Chary)
of preparation, nature of support and the type of promoter. It is generally believed that optimal catalytic activity and selectivity are achieved when one monolayer is dispersed on supported oxides. Traditionally, oxides of several elements such as Nb, Sb, P, K, Na, Cs, Rb, Mo, and W have been used as additives to supported vanadia catalysts in order to increase their activity and to improve their selectivity for oxidation reaction [18–20]. Molybdenum is frequently added as a promoter to vanadium-based catalysts for a number of selective oxidations [21–25]. For example, V2 O5 in combination with MoO3 was employed in selec-
0926-860X/00/$ – see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 0 0 ) 0 0 4 5 9 - 2
134
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
tive oxidation of benzene to maleic anhydride and o-xylene to phathalic anhydride [26]. Recently Dejoz et al. [27] reported the role of molybdenum in Mo doped V–Mg–O catalysts during oxidative dehydrogenation of n-butane and suggested that the incorporation of MoO3 favours oxydehydrogenated products more than undoped samples. The addition of molybdena improves selectivity of vanadia catalysts. The presence of molybdena in the catalyst system changes in phase composition of surface layers. Satsuma et al. [28] have reported the surface active sites of V2 O5 –MoO3 catalysts as determined by a NO-NH3 rectangular pulse method. One of the main factors controlling the efficiency of the supported vanadia catalysts is the local structure of vanadium oxide species stabilized on the support surface. Several spectroscopic techniques including 51 V solid state NMR [29–33], ESR [34–39], EXAFS [40,41] and ESCA [42,43] have been employed in the recent past to characterize supported vanadia systems in order to understand the nature of the supported vanadia species present on the surface. ESR offers a powerful and sensitive technique for investigating the oxidation states, surface and bulk co-ordination and the physical form of a transition metal oxide on a diamagnetic support [44]. ESR technique has been extensively employed to characterize supported vanadia systems in order to understand specific information about (i) the local structure surrounding V4+ , (ii) the differences in the microstructure arising from distinct nature of the supported oxide, (iii) the nature of paramagnetic species in the catalysts and (iv) the so-called carrier effect of the active component. Previous studies from our laboratory showed that V2 O5 forms a monolayer on the surface of ␥-Al2 O3 support at 10% w/w of V2 O5 . Evidence for the monolayer formation was also provided by the results of solid state 51 V NMR spectroscopy [32,33,48] and of radial electron distribution [45], X-ray diffraction and oxygen chemisorption [46] studies. In the present investigation, we report the influence of MoO3 on the dispersion of V2 O5 supported on ␥-Al2 O3 and its effect on the activity and selectivity of the catalysts towards partial oxidation of methanol to formaldehyde. We also report new complementary results of MoO3 on V2 O5 /␥-Al2 O3 catalysts by ESR spectroscopy. The purpose of this work has been to study the influence of
the presence of molybdenum oxide on the dispersion of vanadia and also on the catalytic properties during partial oxidation of methanol.
2. Experimental The catalysts preparation involves two steps. In the first step, 10% V2 O5 (w/w) supported on ␥-Al2 O3 (Harshaw Al-111-E) was prepared by wet impregnation of the support using an aqueous solution containing ammonium metavanadate. After impregnation, the sample was dried at 383 K and calcined in air at 773 K for 6 h. In the second step a series of V2 O5 –MoO3 catalysts with varying MoO3 content ranging from 1–5% (w/w) were prepared by wet impregnation of previously prepared and oven dried 10% V2 O5 /␥-Al2 O3 catalysts using stoichiometric amounts of an aqueous solution containing ammonium heptamolybdate. The samples were dried at 384 K for 16 h and further calcined at 773 K for 6 h in air. Oxygen chemisorption experiments were performed on pre-reduced catalysts using a static high vacuum adsorption apparatus, following the method of Parekh and Weller [47]. Briefly, the catalyst sample was reduced at 773 K in a continuous flow of hydrogen (40 ml/min) for 6 h and then evacuated at the same temperature for an hour prior to oxygen adsorption at 195 K. The amount of oxygen chemisorption was determined by the difference between two oxygen adsorption isotherms at 195 K. The details of experimental procedure are described elsewhere [46]. The ESR spectra of unreduced and hydrogen reduced samples were recorded on a Bruker ER 200D-SRC X-band spectrometer (with 100 kHz modulation) at 300 K. Prior to recording ESR spectra, the instrument was calibrated using DPPH as external standard. The reduced samples used in the ESR study were prepared in quartz tubes (25 cm long, 4 mm dia) which formed part of the catalyst reduction cell. The reduction was carried at 773 K for 2 h in a continuous flow (35 ml/min) of hydrogen. The set-up was subsequently evacuated at the reduction temperature for an hour, cooled to room temperature and finally evacuated for an hour at 1×10−6 Torr. The catalyst sample thus prepared was transferred to an ESR sample tube and sealed off under vacuum.
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
135
A flow micro-reactor operating at atmospheric pressure was used for the catalytic oxidation of methanol to formaldehyde. About 0.25 g of the catalyst sample plugged with glass wool in a glass reactor was fed with methanol vapor at 448 K. The feed gas (air) was passed through two saturators containing methanol at a flow rate of 40 ml/min. The reaction products, mainly formaldehyde and dimethyl ether were analysed by an on-line gas chromatograph equipped with FID using a column containing 10% carbowax 20 M on chromosorb. Traces of carbon oxides were also formed during the reaction.
3. Results and discussion Results of oxygen chemisorption by various V2 O5 –MoO3 /␥-Al2 O3 catalysts are plotted as a function of MoO3 loading in Fig. 1. It is clear that oxygen uptake decreases linearly with the increase of MoO3 content in the catalyst. The dispersion of vanadia was calculated by the number of V-oxide units present in the sample and the number of ‘O’ chemisorbed, with the assumption that one ‘O’ atom corresponds to two V atoms as per the equation. V2 O5 + H2 → V2 O4 + H2 O V2 O4 + O → V2 O5
Fig. 1. Results of oxygen V2 O5 –MoO3 /␥-Al2 O3 catalysts.
chemisorption
by
various
Fig. 2. X-ray diffraction patterns of V2 O5 /␥-Al2 O3 MoO3 -V2 O5 /␥-Al2 O3 catalysts (a) 10% V2 O5 /␥-Al2 O3 (b) MoO3 -V2 O5 /␥-Al2 O3 (c) 2% MoO3 -V2 O5 /␥-Al2 O3 (d) MoO3 -V2 O5 /␥-Al2 O3 (e) 4% MoO3 -V2 O5 /␥-Al2 O3 (f) MoO3 -V2 O5 /␥-Al2 O3 .
and 1% 3% 5%
The dispersion of vanadia is found to decrease linearly with increase of MoO3 content (Fig. 1). The XRD patterns of calcined samples of V2 O5 /␥-Al2 O3 and V2 O5 –MoO3 /␥-Al2 O3 are shown in Fig. 2. As can be seen from the XRD pattern of 10% V2 O5 /␥-Al2 O3 (Fig. 2a) only ␥-Al2 O3 peaks were noticed and the XRD lines due to V2 O5 were absent. However, one cannot rule out the possibility of V2 O5 crystallites having sizes less than 4 nm, which is beyond the detection limit of our XRD technique. X-ray diffraction patterns of V2 O5 –MoO3 /␥-Al2 O3 samples with varying Mo loading are shown in Fig. 2b–f. The results indicate that XRD peaks characteristic of individual V2 O5 and MoO3 are absent in all the samples. However, a sharp peak is noticed at d=3.1 Å (Fig. 2b–f) and its intensity increases gradually with increase of Mo loading in the sample. The appearance of XRD peaks at d=3.1, 3.2, 2.4 Å in V2 O5 –MoO3 /␥-Al2 O3 catalysts is attributed to the formation of aluminium vanadate phase (JCPDS 19–56). Fig. 3 shows the representative ESR spectra of unreduced samples recorded at 300 K. The unpaired
136
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
Fig. 3. ESR spectra of unreduced V2 O5 /␥-Al2 O3 & MoO3 -V2 O5 /␥-Al2 O3 catalysts at 300 K (a) 10% V2 O5 /␥-Al2 O3 (b) 1% MoO3 and 10% V2 O5 on ␥-Al2 O3 (c) 3% MoO3 and 10% V2 O5 on ␥-Al2 O3 (d) 5% MoO3 and 10% V2 O5 on ␥-Al2 O3 .
electron associated with V4+ (3d1 ) interacts with nuclear magnetic moment of 51 V (I=7/2) to give rise to eight parallel and eight perpendicular components in the ESR spectrum. Fig. 3 suggests that the hyperfine structure of ESR spectrum does not change with the addition of MoO3 to V2 O5 /␥-Al2 O3 . The linear relationship between Hk (or H⊥ ) and nuclear quantum
Fig. 4. ESR spectra of hydrogen reduced V2 O5 /␥-Al2 O3 and MoO3 -V2 O5 /␥-Al2 O3 catalysts (a) 10% V2 O5 /␥-Al2 O3 (b) 1% MoO3 and 10% V2 O5 on ␥-Al2 O3 (c) 2% MoO3 and 10% V2 O5 on ␥-Al2 O3 (d) 3% MoO3 and 10% V2 O5 on ␥-Al2 O3 (e) 4% MoO3 and 10% V2 O5 on ␥-Al2 O3 (f) 4% MoO3 and 10% V2 O5 on ␥-Al2 O3 .
number (mI ) was utilized to deduce the values of gk , g⊥ , Ak and A⊥ [34]. These values are as follows: gk =1.92, g⊥ =2.00, Ak =199 G, A⊥ =88 G. Fig. 4 represents the ESR spectra of hydrogen-reduced samples of 10% V2 O5 /␥-Al2 O3 and various amounts of
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
MoO3 added to 10% V2 O5 /␥-Al2 O3 catalysts. The spectra also show the hyperfine splitting (hfs) due to 51 V (I=7/2) similar to that of unreduced catalyst samples and also in agreement with values reported in literature dealing with ESR studies of V2 O5 /␥-Al2 O3 catalysts [33–36,38,46]. The axially symmetric values and hfs constants remain practically constant and are given as follows, gk =1.94, g⊥ =2.00, Ak =180 G, A⊥ =80 G. It is interesting to note from Figs. 3 and 4 that no ESR signals due to Mo5+ ions have been observed in MoO3 added to V2 O5 /␥-Al2 O3 catalysts either in their reduced or unreduced conditions. One of the interesting features of the ESR spectra of hydrogen-reduced catalysts (Fig. 4) is the presence of doublet structure in the outermost low field parallel components (mI =7/2). These findings are in agreement with our earlier ESR studies on reduced V2 O5 /␥-Al2 O3 catalysts [34]. There, a similar doublet structure is observed at the low field parallel components of the ESR spectra at loadings above 10% V2 O5 . With increasing MoO3 concentration in the catalyst the doublet structure is well resolved and merges with the main peak (Fig. 4d). The doublet structure in the low field parallel components suggests the presence of two chemically distinct V4+ species. The area under the ESR absorption curve is a measure of V4+ concentration and, thus, the ratio of areas (AM/AS, where AM is the area of main peak and AS is the area under satellite peak) represents the variation in the relative concentrations of two chemically distinct V4+ ions in the catalyst. The results of Fig. 4 show that the area of the satellite peak increases upon the addition of MoO3 . However, the area of the main peak decreases with MoO3 loading. The gk and Ak values for the second set of peaks (satellite) are as follows: gk =1.91, Ak =178 G. A comparison of the gk values for the satellite peaks with those of main peaks suggests that they are quite different for the two species, suggesting that the ligand field around the two V4+ species would also be different. Our gk and Ak values resemble those from previous ESR studies on V2 O5 /␥-Al2 O3 catalysts [34] and those of Inomata et al. [35] and Sharma et al. [36], suggesting that the satellite peaks represent V4+ ions formed on the catalyst surface. The reducibility of supported vanadia catalysts depends on the nature of the supported oxide and on the vanadia loading. Recent findings of Koranne et al. [51] using the TPR of V2 O5 /␥-Al2 O3 , catalysts have
137
shown that the average oxidation state of vanadium after reduction at 1173 K was consistent with the stoichiometry V5+ → V4+ . Nag and Massooth [43] and Haber [52] et al. have also hypothesized that the V4+ oxidation state is stabilized on Al2 O3 probably due to removal of an oxygen from the V–O–V linkage in a dimeric surface vanadia species. Another feature of the samples reduced at 773 K is that, in the spectra for V2 O5 /␥-Al2 O3 and MoO3 –V2 O5 /␥-Al2 O3 samples, the low field parallel components of the ESR spectra show the doublet structure, suggesting the presence of two chemically distinct V4+ species. The present ESR results of V2 O5 –MoO3 /␥-Al2 O3 catalysts are in agreement with our earlier ESR work of V2 O5 /␥-Al2 O3 catalysts [34], where a doublet structure appeared in the samples exceeding monolayer V2 O5 concentration on alumina. Many authors have reported the reduction behavior of supported MoO3 catalysts. Seyedmonir and Howe [53] used reduction temperatures of 673 and 773 K on silica-supported samples containing polymolybdate species and molybdenum trioxide. They concluded that at 673 K reduction of polymolybdate occurred with the formation of Mo5+ species, while at 773 K reduction of MoO3 to Mo4+ species occurred. Thus, in the present study we also cannot rule out the possibility of over reduction of Mo6+ to Mo4+ in V2 O5 –MoO3 /␥-Al2 O3 catalyst which is ESR inactive. The percentage conversion of methanol during partial oxidation is plotted as a function of MoO3 content in Fig. 5. The conversion of methanol decreases with MoO3 loading on V2 O5 /␥-Al2 O3 . This further supports the decrease of dispersion of vanadia with addition of MoO3 . The selectivities of formaldehyde (due to partial oxidation) and dimethyl ether (due to dehydration) during methanol oxidation reaction at 448 K by various catalysts are plotted as a function of MoO3 content in Fig. 6. The selectivity of formaldehyde is found to increase linearly with MoO3 content in the catalyst. However, the selectivity for dimethyl ether decreases linearly as a function of MoO3 . Under the similar experimental conditions V2 O5 /␥-Al2 O3 , catalysts have shown both dehydration and partial oxidation functionalities. However, with the increase of MoO3 loading in the catalyst, the selectivity of formaldehyde increases. The selectivity of MoO3 –V2 O5 /␥-Al2 O3 catalysts towards formaldehyde is higher than that of the corresponding
138
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
Fig. 5. The percentage of conversion of methanol during partial oxidation as a function of MoO3 content.
V2 O5 /Al2 O3 catalysts. It is instructive to consider what can possibly happen to the molybdena added to the V2 O5 /␥-Al2 O3 monolayer catalysts. Three possibilities can be conceived-(i) part of the added molybdena can interact with hydroxyl groups
Fig. 6. The selectivities of formaldehyde and dimethyl ether as a function of MoO3 content.
on alumina left after vanadia impregnation, because it is well known that molybdena interacts with the hydroxyl groups on alumina [49,50], (ii) the molybdena may cover part of the monolayer vanadia, (iii) the molybdena might interact with vanadia to form a surface compound which is difficult to reduce. Now let us consider the influence of the above possibilities on oxygen uptake. According to the first possibility, the oxygen chemisorption should increase, as it is known that the reduced molybdena on aluminium oxide chemisorb oxygen [47]. It is experimentally found that oxygen chemisorption decreases with the addition of molybdena. Therefore, the first possibility seems to be ruled out if the molybdena reduces as does the vanadia. The second possibility predicts a more or less constant oxygen chemisorption with the increase of molybdena, if the molybdena covering V2 O4 units chemisorb oxygen otherwise a decrease is visualized. The third possibility indicates a decrease in oxygen chemisorption, since the surface compound formation is expected to decrease the dispersion of both the components. In reality, all of the three possibilities may simultaneously occur to different extents. The decrease in oxygen chemisorption indicates that the third possibility is the most important one. However, one cannot rule out the possibility that a small fraction of molybdena interacts with the hydroxyl groups of bare alumina surfaces left out on the monolayer V2 O5 /␥-Al2 O3 . Because of this phenomena, the increase in oxygen chemisorption is expected to be small. Therefore, the net decrease in oxygen chemisorption led us to believe that the third possibility predominates. The increase of formaldehyde selectivity with MoO3 content might be due to the promotion by Mo species. The decrease in dehydration activity with the addition of molybdena might be a consequence of decrease in acidity of the support as a result of elimination of hydroxyl groups. The XRD results of MoO3 –V2 O3 /␥-Al2 O3 samples suggest the formation of vanadium-aluminate. Thus, the selectivity of dimethyl ether over V2 O5 /␥-Al2 O3 catalysts indicates the acidic property of V2 O5 /␥-Al2 O3 surface and the decrease of selectivity of DME with MoO3 addition is due to formation of surface aluminate. The addition of molybdena to V2 O5 /␥-Al2 O3 catalyst considerably increases the selectivity to formaldehyde during the partial oxidation of methanol.
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
4. Conclusions The addition of molybdena to V2 O5 /␥-Al2 O3 catalyst is found to decrease the dispersion of vanadia supported on alumina. Molybdena addition to V2 O5 catalysts lowers the catalytic activity during partial oxidation of methanol. However, it favours the selectivity improvement during selective oxidation. The lower catalytic performance of MoO3 –V2 O5 /␥-Al2 O3 catalysts is attributed to the inhibition of interaction between V2 O5 and ␥-Al2 O3 surface. Acknowledgements T. Bhaskar thanks University Grants Commission (UGC), New Delhi, for the award of Senior Research Fellowship. K. Ramesh thanks Director, IICT, for providing a Research Trainee position. References [1] D.J. Hucknall, Selective Oxidation of Hydrocarbons, Academic Press, New York, 1974. [2] G.C. Bond, S. Flamerz Tahir, Appl. Catal. 71, 1991, 1 (and references therein). [3] G.C. Bond, Appl. Catal. A: Gen. 157 (1997) 91. [4] G. Deo, I.E. Wachs, J. Haber, Crit. Rev. Surf. Chem. 4 (1994) 1. [5] I.E. Wachs, B.M. Weckhuysen, Appl. Catal. 157 (1997) 91. [6] K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl. Catal. 74 (1991) 1. [7] K.V.R. Chary, G. Kishan, T. Bhaskar, J. Phys. Chem. 102 (1998) 6792. [8] R.Y. Saleh, I.E. Wachs, S.S. Chan, C.C. Chersich, J. Catal. 98 (1986) 102. [9] C.R. Dias, M.F. Portela, G.C. Bond, J. Catal. 157 (1995) 344. [10] M.S. Wainwright, N.R. Foster, Catal. Rev. Sci. Eng. 19 (1979) 211. [11] F. Cavalli, F. Cavani, I. Maneti, F. Trifiro, Catal. Today 1 (1987) 245. [12] M. Sanati, A. Anderson, J. Mol. Catal. 59 (1990) 233. [13] K.V.R. Chary, G. Kishan, T. Bhaskar, K.V. Narayana, J. Chem. Res. (S). 1998, 314. [14] A. Andersson, S.T. Lundin, J. Catal. 65 (1980) 9. [15] A. Andersson, J. Catal. 69 (1981) 465. [16] H. Bosch, J.J. Janssen, Catal. Today 2 (1988) 369. [17] H. Schneider, S. Tschudin, M. Schneider, A. Wokaun, A. Baiker, J. Catal. 147 (1994) 5. [18] G. Deo, I.E. Wachs, J. Catal. 146 (1994) 335. [19] J. Zhu, B. Robenstrof, S.L.T. Andersson, J. Chem. Soc. Faraday Trans. 85 (1989) 3645. [20] A. van Hengstum, J. Pranger, J.G. van Ommen, P.J. Gellings, Appl. Catal. 11 (1984) 317.
139
[21] A. Satsuma, F. Okada, A. Hattori, A. Miyamoto, T. Hattori, Y. Murakami, Appl. Catal. 72 (1991) 295. [22] H.K. Matralis, Ch. Papadopoulou, Ch. Kordulis, A. Aguilar Elguezabal, V. Cortes Corberan, Appl. Catal. A Gen. 126 (1995) 365. [23] K.V.R. Chary, V. Venkat Rao, G. Muralidhar, P. Kanta Rao, Catal. Lett. 7 (1990) 389. [24] M. Najbar, J. Chem. Soc. Faraday Trans. 82 (1986) 1673. [25] Z.X. Liu, Y.Q. LI, S.X. Qi, K. Xie, N.J. Wu, Q.X. Bao, Appl. Catal. 56 (1989) 207. [26] A. Bielanski, M. Najbar, Appl. Catal. A: Gen. 157 (1997) 223. [27] A. Dejoz, J.M. Lopez Nieto, F. Marquez, M.I. Vazquez, App. Catal. A: Gen 180 (1989) 83. [28] A. Satsuma, A. Hattori, K. Mizutani, A. Furuta, A. Miyamoto, T. Hattori, Y. Murakami, J. Phys. Chem. 93 (1989) 1484. [29] B. Taouk, M. Guelton, J. Grimblot, J.P. Bonnelle, J. Phys. Chem. 92 (1988) 6700. [30] L.R. Le Coustumer, B. Taouk, M. Le Meur, E. Payen, M. Guelton, J. Grimblot, J. Phys. Chem. 92 (1988) 1230. [31] H. Eckert, I.E. Wachs, J. Phys. Chem. 93 (1989) 6796. [32] K.V.R. Chary, V. Venkat Rao, V.M. Mastikhin, J. Chem. Soc. Chem. Commun. 1988, 202. [33] K.V.R. Chary, G. Kishan, J. Phys. Chem. 99 (1995) 14424. [34] K.V.R. Chary, B.M. Reddy, N.K. Nag, V.S. Subrahmanyam, C.S. Sunandana, J. Phys. Chem. 88 (1984) 2622. [35] M. Inomata, K. Mori, A. Miyamoto, Y. Murakami, J. Phys. Chem. 87 (1983) 761. [36] V.K. Sharma, A. Wokaun, A. Baiker, J. Phys. Chem. 90 (1986) 2715. [37] N.K. Nag, K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl. Catal. 31 (1987) 73. [38] M. Akito, M. Usami, E. Echigoya, Bull. Chem. Soc. Japan 51 (1978) 2195. [39] K. Dyrek, M. Labanowska, J. Catal. 96 (1986) 32. [40] R. Kozlowski, R.F. Pettifer, J.M. Thomas, J. Phys. Chem. 87 (1983) 5172. [41] T. Tanaka, H. Ymashita, R. Tsuchitani, T. Funabiki, S. Yoshida, J. Chem. Soc. Faraday Trans. 84 (1988) 2987. [42] G.C. Bond, J.P. Zurita, S. Flamerz, Appl. Catal. 27 (1986) 353. [43] N.K. Nag, E.E. Massoth, J. Catal. 124 (1990) 127. [44] V.B. Kazanskii, in: Proceedings of the 6th International Congress on Catalysis, London, 1976, p. 50. [45] G. Bergeret, P. Gallezot, K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl.Catal. 40 (1988) 191. [46] K.V.R. Chary, B.M. Reddy, N.K. Nag, B.R. Rao, V.S. Subrahmanyam, Appl. Catal. 9 (1984) 225. [47] B.S. Parekh, S.W. Weller, J. Catal. 47 (1977) 100. [48] M. Najbar, S. Niziol, J. Solid State Chem. 26 (1978) 339. [49] F.E. Massoth, G. Muralidhar, J. Shabtai, J. Catal. 85 (1984) 53. [50] N. Yamagata, Y. Owada, S. Okazaki, K. Tanabe, J. Catal. 47 (1977) 358. [51] M.M. Koranne, J.G. Goodwin Jr., G. Marcelin, J. Catal. 148 (1994) 369. [52] J. Haber, A. Kozlowska, R. Kozlowska, J. Catal. 102 (1986) 52. [53] S.R. Seyedmonir, R.F. Howe, J. Catal. 110 (1988) 216.
Characterization and reactivity of vanadia–molybdena catalysts supported on ␥-Al2 O3 Komandur V.R. Chary∗ , Thallada Bhaskar, Jangam Jyothi Maheshwar, Kanaparthi Ramesh, Vattikonda Venkat Rao Catalysis Section, Indian Institute of Chemical Technology, Hyderabad 500 007, India Received 16 June 1999; received in revised form 5 January 2000; accepted 25 January 2000
Abstract The influence of molybdenum oxide on the dispersion of vanadium oxide supported on alumina was investigated. A series of V2 O5 –MoO3 catalysts with varying MoO3 content ranging from 1–5% (w/w) were prepared by impregnation of previously prepared 10% V2 O5 /␥-Al2 O3 with requisite amounts of ammonium molybdate solution. Dispersion of vanadia was determined by oxygen chemisorption at 195 K and was found to decrease with the increase of molybdena loading. The calcined catalyst samples were characterized by electron spin resonance (ESR) technique in their hydrogen reduced and unreduced conditions. The XRD results suggest formation of vanadium aluminate with the addition of MoO3 to V2 O5 /␥-Al2 O3 catalyst. The ESR spectra of hydrogen reduced catalysts show doublet structure at the low field parallel components, suggesting the presence of two chemically distinct V4+ on alumina species. It was found that the nature of V4+ is strongly influenced by the added molybdena. The catalytic properties were evaluated for the vapour-phase oxidation of methanol. The activity during oxidation reaction was found to decrease with the increase of molybdena loading. However, the selectivity of formaldehyde was found to increase with molybdena loading, indicating that the added MoO3 created additional sites for partial oxidation reaction. The presence of molybdenum oxide inhibited the interaction between vanadium and alumina, leading to poor dispersion of vanadia and lower catalytic activity during selective oxidation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Oxygen chemisorption; Electron spin resonance; Vanadia; Molybdena
1. Introduction Supported vanadium oxides have been widely investigated as they catalyse a number of reactions including selective oxidation, ammoxidation, as well as the selective catalytic reduction of NO by NH3 [1–17]. The catalytic properties of vanadia in selective oxidation reactions are strongly influenced by the method ∗ Corresponding author. Tel.: +91-40-7171510; fax: +91-40-7173757. E-mail address: [email protected] (K.V.R. Chary)
of preparation, nature of support and the type of promoter. It is generally believed that optimal catalytic activity and selectivity are achieved when one monolayer is dispersed on supported oxides. Traditionally, oxides of several elements such as Nb, Sb, P, K, Na, Cs, Rb, Mo, and W have been used as additives to supported vanadia catalysts in order to increase their activity and to improve their selectivity for oxidation reaction [18–20]. Molybdenum is frequently added as a promoter to vanadium-based catalysts for a number of selective oxidations [21–25]. For example, V2 O5 in combination with MoO3 was employed in selec-
0926-860X/00/$ – see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 0 0 ) 0 0 4 5 9 - 2
134
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
tive oxidation of benzene to maleic anhydride and o-xylene to phathalic anhydride [26]. Recently Dejoz et al. [27] reported the role of molybdenum in Mo doped V–Mg–O catalysts during oxidative dehydrogenation of n-butane and suggested that the incorporation of MoO3 favours oxydehydrogenated products more than undoped samples. The addition of molybdena improves selectivity of vanadia catalysts. The presence of molybdena in the catalyst system changes in phase composition of surface layers. Satsuma et al. [28] have reported the surface active sites of V2 O5 –MoO3 catalysts as determined by a NO-NH3 rectangular pulse method. One of the main factors controlling the efficiency of the supported vanadia catalysts is the local structure of vanadium oxide species stabilized on the support surface. Several spectroscopic techniques including 51 V solid state NMR [29–33], ESR [34–39], EXAFS [40,41] and ESCA [42,43] have been employed in the recent past to characterize supported vanadia systems in order to understand the nature of the supported vanadia species present on the surface. ESR offers a powerful and sensitive technique for investigating the oxidation states, surface and bulk co-ordination and the physical form of a transition metal oxide on a diamagnetic support [44]. ESR technique has been extensively employed to characterize supported vanadia systems in order to understand specific information about (i) the local structure surrounding V4+ , (ii) the differences in the microstructure arising from distinct nature of the supported oxide, (iii) the nature of paramagnetic species in the catalysts and (iv) the so-called carrier effect of the active component. Previous studies from our laboratory showed that V2 O5 forms a monolayer on the surface of ␥-Al2 O3 support at 10% w/w of V2 O5 . Evidence for the monolayer formation was also provided by the results of solid state 51 V NMR spectroscopy [32,33,48] and of radial electron distribution [45], X-ray diffraction and oxygen chemisorption [46] studies. In the present investigation, we report the influence of MoO3 on the dispersion of V2 O5 supported on ␥-Al2 O3 and its effect on the activity and selectivity of the catalysts towards partial oxidation of methanol to formaldehyde. We also report new complementary results of MoO3 on V2 O5 /␥-Al2 O3 catalysts by ESR spectroscopy. The purpose of this work has been to study the influence of
the presence of molybdenum oxide on the dispersion of vanadia and also on the catalytic properties during partial oxidation of methanol.
2. Experimental The catalysts preparation involves two steps. In the first step, 10% V2 O5 (w/w) supported on ␥-Al2 O3 (Harshaw Al-111-E) was prepared by wet impregnation of the support using an aqueous solution containing ammonium metavanadate. After impregnation, the sample was dried at 383 K and calcined in air at 773 K for 6 h. In the second step a series of V2 O5 –MoO3 catalysts with varying MoO3 content ranging from 1–5% (w/w) were prepared by wet impregnation of previously prepared and oven dried 10% V2 O5 /␥-Al2 O3 catalysts using stoichiometric amounts of an aqueous solution containing ammonium heptamolybdate. The samples were dried at 384 K for 16 h and further calcined at 773 K for 6 h in air. Oxygen chemisorption experiments were performed on pre-reduced catalysts using a static high vacuum adsorption apparatus, following the method of Parekh and Weller [47]. Briefly, the catalyst sample was reduced at 773 K in a continuous flow of hydrogen (40 ml/min) for 6 h and then evacuated at the same temperature for an hour prior to oxygen adsorption at 195 K. The amount of oxygen chemisorption was determined by the difference between two oxygen adsorption isotherms at 195 K. The details of experimental procedure are described elsewhere [46]. The ESR spectra of unreduced and hydrogen reduced samples were recorded on a Bruker ER 200D-SRC X-band spectrometer (with 100 kHz modulation) at 300 K. Prior to recording ESR spectra, the instrument was calibrated using DPPH as external standard. The reduced samples used in the ESR study were prepared in quartz tubes (25 cm long, 4 mm dia) which formed part of the catalyst reduction cell. The reduction was carried at 773 K for 2 h in a continuous flow (35 ml/min) of hydrogen. The set-up was subsequently evacuated at the reduction temperature for an hour, cooled to room temperature and finally evacuated for an hour at 1×10−6 Torr. The catalyst sample thus prepared was transferred to an ESR sample tube and sealed off under vacuum.
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
135
A flow micro-reactor operating at atmospheric pressure was used for the catalytic oxidation of methanol to formaldehyde. About 0.25 g of the catalyst sample plugged with glass wool in a glass reactor was fed with methanol vapor at 448 K. The feed gas (air) was passed through two saturators containing methanol at a flow rate of 40 ml/min. The reaction products, mainly formaldehyde and dimethyl ether were analysed by an on-line gas chromatograph equipped with FID using a column containing 10% carbowax 20 M on chromosorb. Traces of carbon oxides were also formed during the reaction.
3. Results and discussion Results of oxygen chemisorption by various V2 O5 –MoO3 /␥-Al2 O3 catalysts are plotted as a function of MoO3 loading in Fig. 1. It is clear that oxygen uptake decreases linearly with the increase of MoO3 content in the catalyst. The dispersion of vanadia was calculated by the number of V-oxide units present in the sample and the number of ‘O’ chemisorbed, with the assumption that one ‘O’ atom corresponds to two V atoms as per the equation. V2 O5 + H2 → V2 O4 + H2 O V2 O4 + O → V2 O5
Fig. 1. Results of oxygen V2 O5 –MoO3 /␥-Al2 O3 catalysts.
chemisorption
by
various
Fig. 2. X-ray diffraction patterns of V2 O5 /␥-Al2 O3 MoO3 -V2 O5 /␥-Al2 O3 catalysts (a) 10% V2 O5 /␥-Al2 O3 (b) MoO3 -V2 O5 /␥-Al2 O3 (c) 2% MoO3 -V2 O5 /␥-Al2 O3 (d) MoO3 -V2 O5 /␥-Al2 O3 (e) 4% MoO3 -V2 O5 /␥-Al2 O3 (f) MoO3 -V2 O5 /␥-Al2 O3 .
and 1% 3% 5%
The dispersion of vanadia is found to decrease linearly with increase of MoO3 content (Fig. 1). The XRD patterns of calcined samples of V2 O5 /␥-Al2 O3 and V2 O5 –MoO3 /␥-Al2 O3 are shown in Fig. 2. As can be seen from the XRD pattern of 10% V2 O5 /␥-Al2 O3 (Fig. 2a) only ␥-Al2 O3 peaks were noticed and the XRD lines due to V2 O5 were absent. However, one cannot rule out the possibility of V2 O5 crystallites having sizes less than 4 nm, which is beyond the detection limit of our XRD technique. X-ray diffraction patterns of V2 O5 –MoO3 /␥-Al2 O3 samples with varying Mo loading are shown in Fig. 2b–f. The results indicate that XRD peaks characteristic of individual V2 O5 and MoO3 are absent in all the samples. However, a sharp peak is noticed at d=3.1 Å (Fig. 2b–f) and its intensity increases gradually with increase of Mo loading in the sample. The appearance of XRD peaks at d=3.1, 3.2, 2.4 Å in V2 O5 –MoO3 /␥-Al2 O3 catalysts is attributed to the formation of aluminium vanadate phase (JCPDS 19–56). Fig. 3 shows the representative ESR spectra of unreduced samples recorded at 300 K. The unpaired
136
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
Fig. 3. ESR spectra of unreduced V2 O5 /␥-Al2 O3 & MoO3 -V2 O5 /␥-Al2 O3 catalysts at 300 K (a) 10% V2 O5 /␥-Al2 O3 (b) 1% MoO3 and 10% V2 O5 on ␥-Al2 O3 (c) 3% MoO3 and 10% V2 O5 on ␥-Al2 O3 (d) 5% MoO3 and 10% V2 O5 on ␥-Al2 O3 .
electron associated with V4+ (3d1 ) interacts with nuclear magnetic moment of 51 V (I=7/2) to give rise to eight parallel and eight perpendicular components in the ESR spectrum. Fig. 3 suggests that the hyperfine structure of ESR spectrum does not change with the addition of MoO3 to V2 O5 /␥-Al2 O3 . The linear relationship between Hk (or H⊥ ) and nuclear quantum
Fig. 4. ESR spectra of hydrogen reduced V2 O5 /␥-Al2 O3 and MoO3 -V2 O5 /␥-Al2 O3 catalysts (a) 10% V2 O5 /␥-Al2 O3 (b) 1% MoO3 and 10% V2 O5 on ␥-Al2 O3 (c) 2% MoO3 and 10% V2 O5 on ␥-Al2 O3 (d) 3% MoO3 and 10% V2 O5 on ␥-Al2 O3 (e) 4% MoO3 and 10% V2 O5 on ␥-Al2 O3 (f) 4% MoO3 and 10% V2 O5 on ␥-Al2 O3 .
number (mI ) was utilized to deduce the values of gk , g⊥ , Ak and A⊥ [34]. These values are as follows: gk =1.92, g⊥ =2.00, Ak =199 G, A⊥ =88 G. Fig. 4 represents the ESR spectra of hydrogen-reduced samples of 10% V2 O5 /␥-Al2 O3 and various amounts of
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
MoO3 added to 10% V2 O5 /␥-Al2 O3 catalysts. The spectra also show the hyperfine splitting (hfs) due to 51 V (I=7/2) similar to that of unreduced catalyst samples and also in agreement with values reported in literature dealing with ESR studies of V2 O5 /␥-Al2 O3 catalysts [33–36,38,46]. The axially symmetric values and hfs constants remain practically constant and are given as follows, gk =1.94, g⊥ =2.00, Ak =180 G, A⊥ =80 G. It is interesting to note from Figs. 3 and 4 that no ESR signals due to Mo5+ ions have been observed in MoO3 added to V2 O5 /␥-Al2 O3 catalysts either in their reduced or unreduced conditions. One of the interesting features of the ESR spectra of hydrogen-reduced catalysts (Fig. 4) is the presence of doublet structure in the outermost low field parallel components (mI =7/2). These findings are in agreement with our earlier ESR studies on reduced V2 O5 /␥-Al2 O3 catalysts [34]. There, a similar doublet structure is observed at the low field parallel components of the ESR spectra at loadings above 10% V2 O5 . With increasing MoO3 concentration in the catalyst the doublet structure is well resolved and merges with the main peak (Fig. 4d). The doublet structure in the low field parallel components suggests the presence of two chemically distinct V4+ species. The area under the ESR absorption curve is a measure of V4+ concentration and, thus, the ratio of areas (AM/AS, where AM is the area of main peak and AS is the area under satellite peak) represents the variation in the relative concentrations of two chemically distinct V4+ ions in the catalyst. The results of Fig. 4 show that the area of the satellite peak increases upon the addition of MoO3 . However, the area of the main peak decreases with MoO3 loading. The gk and Ak values for the second set of peaks (satellite) are as follows: gk =1.91, Ak =178 G. A comparison of the gk values for the satellite peaks with those of main peaks suggests that they are quite different for the two species, suggesting that the ligand field around the two V4+ species would also be different. Our gk and Ak values resemble those from previous ESR studies on V2 O5 /␥-Al2 O3 catalysts [34] and those of Inomata et al. [35] and Sharma et al. [36], suggesting that the satellite peaks represent V4+ ions formed on the catalyst surface. The reducibility of supported vanadia catalysts depends on the nature of the supported oxide and on the vanadia loading. Recent findings of Koranne et al. [51] using the TPR of V2 O5 /␥-Al2 O3 , catalysts have
137
shown that the average oxidation state of vanadium after reduction at 1173 K was consistent with the stoichiometry V5+ → V4+ . Nag and Massooth [43] and Haber [52] et al. have also hypothesized that the V4+ oxidation state is stabilized on Al2 O3 probably due to removal of an oxygen from the V–O–V linkage in a dimeric surface vanadia species. Another feature of the samples reduced at 773 K is that, in the spectra for V2 O5 /␥-Al2 O3 and MoO3 –V2 O5 /␥-Al2 O3 samples, the low field parallel components of the ESR spectra show the doublet structure, suggesting the presence of two chemically distinct V4+ species. The present ESR results of V2 O5 –MoO3 /␥-Al2 O3 catalysts are in agreement with our earlier ESR work of V2 O5 /␥-Al2 O3 catalysts [34], where a doublet structure appeared in the samples exceeding monolayer V2 O5 concentration on alumina. Many authors have reported the reduction behavior of supported MoO3 catalysts. Seyedmonir and Howe [53] used reduction temperatures of 673 and 773 K on silica-supported samples containing polymolybdate species and molybdenum trioxide. They concluded that at 673 K reduction of polymolybdate occurred with the formation of Mo5+ species, while at 773 K reduction of MoO3 to Mo4+ species occurred. Thus, in the present study we also cannot rule out the possibility of over reduction of Mo6+ to Mo4+ in V2 O5 –MoO3 /␥-Al2 O3 catalyst which is ESR inactive. The percentage conversion of methanol during partial oxidation is plotted as a function of MoO3 content in Fig. 5. The conversion of methanol decreases with MoO3 loading on V2 O5 /␥-Al2 O3 . This further supports the decrease of dispersion of vanadia with addition of MoO3 . The selectivities of formaldehyde (due to partial oxidation) and dimethyl ether (due to dehydration) during methanol oxidation reaction at 448 K by various catalysts are plotted as a function of MoO3 content in Fig. 6. The selectivity of formaldehyde is found to increase linearly with MoO3 content in the catalyst. However, the selectivity for dimethyl ether decreases linearly as a function of MoO3 . Under the similar experimental conditions V2 O5 /␥-Al2 O3 , catalysts have shown both dehydration and partial oxidation functionalities. However, with the increase of MoO3 loading in the catalyst, the selectivity of formaldehyde increases. The selectivity of MoO3 –V2 O5 /␥-Al2 O3 catalysts towards formaldehyde is higher than that of the corresponding
138
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
Fig. 5. The percentage of conversion of methanol during partial oxidation as a function of MoO3 content.
V2 O5 /Al2 O3 catalysts. It is instructive to consider what can possibly happen to the molybdena added to the V2 O5 /␥-Al2 O3 monolayer catalysts. Three possibilities can be conceived-(i) part of the added molybdena can interact with hydroxyl groups
Fig. 6. The selectivities of formaldehyde and dimethyl ether as a function of MoO3 content.
on alumina left after vanadia impregnation, because it is well known that molybdena interacts with the hydroxyl groups on alumina [49,50], (ii) the molybdena may cover part of the monolayer vanadia, (iii) the molybdena might interact with vanadia to form a surface compound which is difficult to reduce. Now let us consider the influence of the above possibilities on oxygen uptake. According to the first possibility, the oxygen chemisorption should increase, as it is known that the reduced molybdena on aluminium oxide chemisorb oxygen [47]. It is experimentally found that oxygen chemisorption decreases with the addition of molybdena. Therefore, the first possibility seems to be ruled out if the molybdena reduces as does the vanadia. The second possibility predicts a more or less constant oxygen chemisorption with the increase of molybdena, if the molybdena covering V2 O4 units chemisorb oxygen otherwise a decrease is visualized. The third possibility indicates a decrease in oxygen chemisorption, since the surface compound formation is expected to decrease the dispersion of both the components. In reality, all of the three possibilities may simultaneously occur to different extents. The decrease in oxygen chemisorption indicates that the third possibility is the most important one. However, one cannot rule out the possibility that a small fraction of molybdena interacts with the hydroxyl groups of bare alumina surfaces left out on the monolayer V2 O5 /␥-Al2 O3 . Because of this phenomena, the increase in oxygen chemisorption is expected to be small. Therefore, the net decrease in oxygen chemisorption led us to believe that the third possibility predominates. The increase of formaldehyde selectivity with MoO3 content might be due to the promotion by Mo species. The decrease in dehydration activity with the addition of molybdena might be a consequence of decrease in acidity of the support as a result of elimination of hydroxyl groups. The XRD results of MoO3 –V2 O3 /␥-Al2 O3 samples suggest the formation of vanadium-aluminate. Thus, the selectivity of dimethyl ether over V2 O5 /␥-Al2 O3 catalysts indicates the acidic property of V2 O5 /␥-Al2 O3 surface and the decrease of selectivity of DME with MoO3 addition is due to formation of surface aluminate. The addition of molybdena to V2 O5 /␥-Al2 O3 catalyst considerably increases the selectivity to formaldehyde during the partial oxidation of methanol.
K.V.R. Chary et al. / Applied Catalysis A: General 202 (2000) 133–139
4. Conclusions The addition of molybdena to V2 O5 /␥-Al2 O3 catalyst is found to decrease the dispersion of vanadia supported on alumina. Molybdena addition to V2 O5 catalysts lowers the catalytic activity during partial oxidation of methanol. However, it favours the selectivity improvement during selective oxidation. The lower catalytic performance of MoO3 –V2 O5 /␥-Al2 O3 catalysts is attributed to the inhibition of interaction between V2 O5 and ␥-Al2 O3 surface. Acknowledgements T. Bhaskar thanks University Grants Commission (UGC), New Delhi, for the award of Senior Research Fellowship. K. Ramesh thanks Director, IICT, for providing a Research Trainee position. References [1] D.J. Hucknall, Selective Oxidation of Hydrocarbons, Academic Press, New York, 1974. [2] G.C. Bond, S. Flamerz Tahir, Appl. Catal. 71, 1991, 1 (and references therein). [3] G.C. Bond, Appl. Catal. A: Gen. 157 (1997) 91. [4] G. Deo, I.E. Wachs, J. Haber, Crit. Rev. Surf. Chem. 4 (1994) 1. [5] I.E. Wachs, B.M. Weckhuysen, Appl. Catal. 157 (1997) 91. [6] K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl. Catal. 74 (1991) 1. [7] K.V.R. Chary, G. Kishan, T. Bhaskar, J. Phys. Chem. 102 (1998) 6792. [8] R.Y. Saleh, I.E. Wachs, S.S. Chan, C.C. Chersich, J. Catal. 98 (1986) 102. [9] C.R. Dias, M.F. Portela, G.C. Bond, J. Catal. 157 (1995) 344. [10] M.S. Wainwright, N.R. Foster, Catal. Rev. Sci. Eng. 19 (1979) 211. [11] F. Cavalli, F. Cavani, I. Maneti, F. Trifiro, Catal. Today 1 (1987) 245. [12] M. Sanati, A. Anderson, J. Mol. Catal. 59 (1990) 233. [13] K.V.R. Chary, G. Kishan, T. Bhaskar, K.V. Narayana, J. Chem. Res. (S). 1998, 314. [14] A. Andersson, S.T. Lundin, J. Catal. 65 (1980) 9. [15] A. Andersson, J. Catal. 69 (1981) 465. [16] H. Bosch, J.J. Janssen, Catal. Today 2 (1988) 369. [17] H. Schneider, S. Tschudin, M. Schneider, A. Wokaun, A. Baiker, J. Catal. 147 (1994) 5. [18] G. Deo, I.E. Wachs, J. Catal. 146 (1994) 335. [19] J. Zhu, B. Robenstrof, S.L.T. Andersson, J. Chem. Soc. Faraday Trans. 85 (1989) 3645. [20] A. van Hengstum, J. Pranger, J.G. van Ommen, P.J. Gellings, Appl. Catal. 11 (1984) 317.
139
[21] A. Satsuma, F. Okada, A. Hattori, A. Miyamoto, T. Hattori, Y. Murakami, Appl. Catal. 72 (1991) 295. [22] H.K. Matralis, Ch. Papadopoulou, Ch. Kordulis, A. Aguilar Elguezabal, V. Cortes Corberan, Appl. Catal. A Gen. 126 (1995) 365. [23] K.V.R. Chary, V. Venkat Rao, G. Muralidhar, P. Kanta Rao, Catal. Lett. 7 (1990) 389. [24] M. Najbar, J. Chem. Soc. Faraday Trans. 82 (1986) 1673. [25] Z.X. Liu, Y.Q. LI, S.X. Qi, K. Xie, N.J. Wu, Q.X. Bao, Appl. Catal. 56 (1989) 207. [26] A. Bielanski, M. Najbar, Appl. Catal. A: Gen. 157 (1997) 223. [27] A. Dejoz, J.M. Lopez Nieto, F. Marquez, M.I. Vazquez, App. Catal. A: Gen 180 (1989) 83. [28] A. Satsuma, A. Hattori, K. Mizutani, A. Furuta, A. Miyamoto, T. Hattori, Y. Murakami, J. Phys. Chem. 93 (1989) 1484. [29] B. Taouk, M. Guelton, J. Grimblot, J.P. Bonnelle, J. Phys. Chem. 92 (1988) 6700. [30] L.R. Le Coustumer, B. Taouk, M. Le Meur, E. Payen, M. Guelton, J. Grimblot, J. Phys. Chem. 92 (1988) 1230. [31] H. Eckert, I.E. Wachs, J. Phys. Chem. 93 (1989) 6796. [32] K.V.R. Chary, V. Venkat Rao, V.M. Mastikhin, J. Chem. Soc. Chem. Commun. 1988, 202. [33] K.V.R. Chary, G. Kishan, J. Phys. Chem. 99 (1995) 14424. [34] K.V.R. Chary, B.M. Reddy, N.K. Nag, V.S. Subrahmanyam, C.S. Sunandana, J. Phys. Chem. 88 (1984) 2622. [35] M. Inomata, K. Mori, A. Miyamoto, Y. Murakami, J. Phys. Chem. 87 (1983) 761. [36] V.K. Sharma, A. Wokaun, A. Baiker, J. Phys. Chem. 90 (1986) 2715. [37] N.K. Nag, K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl. Catal. 31 (1987) 73. [38] M. Akito, M. Usami, E. Echigoya, Bull. Chem. Soc. Japan 51 (1978) 2195. [39] K. Dyrek, M. Labanowska, J. Catal. 96 (1986) 32. [40] R. Kozlowski, R.F. Pettifer, J.M. Thomas, J. Phys. Chem. 87 (1983) 5172. [41] T. Tanaka, H. Ymashita, R. Tsuchitani, T. Funabiki, S. Yoshida, J. Chem. Soc. Faraday Trans. 84 (1988) 2987. [42] G.C. Bond, J.P. Zurita, S. Flamerz, Appl. Catal. 27 (1986) 353. [43] N.K. Nag, E.E. Massoth, J. Catal. 124 (1990) 127. [44] V.B. Kazanskii, in: Proceedings of the 6th International Congress on Catalysis, London, 1976, p. 50. [45] G. Bergeret, P. Gallezot, K.V.R. Chary, B. Rama Rao, V.S. Subrahmanyam, Appl.Catal. 40 (1988) 191. [46] K.V.R. Chary, B.M. Reddy, N.K. Nag, B.R. Rao, V.S. Subrahmanyam, Appl. Catal. 9 (1984) 225. [47] B.S. Parekh, S.W. Weller, J. Catal. 47 (1977) 100. [48] M. Najbar, S. Niziol, J. Solid State Chem. 26 (1978) 339. [49] F.E. Massoth, G. Muralidhar, J. Shabtai, J. Catal. 85 (1984) 53. [50] N. Yamagata, Y. Owada, S. Okazaki, K. Tanabe, J. Catal. 47 (1977) 358. [51] M.M. Koranne, J.G. Goodwin Jr., G. Marcelin, J. Catal. 148 (1994) 369. [52] J. Haber, A. Kozlowska, R. Kozlowska, J. Catal. 102 (1986) 52. [53] S.R. Seyedmonir, R.F. Howe, J. Catal. 110 (1988) 216.