J . Chem. Soc., Faraday Trans. I , 1986, 82, 13-34 Activity and Selectivity in Catalytic Reactions of Buta- 1,3-diene and But-1 -ene on Supported Vanadium Oxides Kenji Mori Kinu-ura Research Department, JGC Corp., Sunosaki-cho, Handa, Aichi 475, Japan Akira Miyamoto"? and Yuichi Murakami Department of Synthetic Chemistry, Faculty of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464, Japan The activity and selectivity in the oxidation of buta-1,3-diene, and oxidation and isomerization of but-1 -ene on unsupported and supported V,O, catalysts have been investigated in terms of the catalyst structure. The rate of oxidation is mainly determined by the number of surface V=O species on the catalyst for both buta-1,3-diene and but-1-ene. The roughness of the V,O, surface affected the activity for buta-173-diene, but not for but-I-ene oxidation. It was also found that TiO, support increases the activity of the surface V=O for but- 1-ene oxidation.The selectivity to maleic anhydride was determined by the number of V,O, layers on the support for both reactions. When the number of V,O, layers was 1 or 2, the selectivity was low, while it increased markedly with an increase in the number of V,O, layers to 5, and attained a constant value above 5 layers. Both V,Oj and support were active for the isomerization of but-1-ene to cis- and trans- but-2-ene. On V,O,, the cisltrans ratio was low, while it was as high as 3 for the A1,0, support. The rate and selectivity of the isomerization on supported catalysts were explained in terms of the structure of V,O, on the support.Difference in the structure-activity/selectivity correlation between oxidation and isomerization and that between but- 1 -ene oxidation and buta- 1,3-diene oxidation were also discussed. Supported metal oxide catalysts exhibit differing degrees of catalysis depending on the kind of support and on the composition of the cata1yst.lt2 However, the activity and selectivity on the supported metal-oxide catalyst have not been well clarified in terms of the structure of the metal oxide on the support. This seems to be due the lack of a well established method to determine the structure of such catalysts and, more especially, the number of active sites. For supported vanadium oxide catalysts, we have previously established the rectangular-pulse technique which allows the determination of the number of surface V=O species and the number of V205 layers on ~ u p p o r t .~ Furthermore, the structures of V205/Ti02 and V205/A1203 catalysts have been determined by various physico-chemical measurements together with the rectangular pulse techniq~e.~ By investigating the oxidation of benzene on well characterized vanadium oxide catalysts, the activity and selectivity in benzene oxidation have been revealed in terms of the structure of V205 on the s ~ p p o r t . ~ Since the structure-activity/selectivity correlation is expected to change greatly with the type of r e a c t i ~ n , ~ - ~ it seems interesting to investigate the correlation for various reactions. The purpose of this study is then to investigate the structure-activity/selectivity correlation for the reactions of buta- 173-diene and T Present address : Department of Hydrocarbon Chemistry, Faculty of Engmeering, Kyoto University, Yoshida Honmachi, Sakyo-ku, Kyoto 606, Japan.14 Supported Vanadium Oxide Catalysts Table 1.Physical and catalytic properties of unsupported V,O, catalysts for the oxidation of buta- 1 ,3-dienea ~~~ ~~~~~~ ~ L (W%E.T.) R, (RO/SB.E.T.) S(MA S(C0 SB.E.T. /pmd /pmol /pmol /pmol TF, +FR) +CO,) S(C0) catalyst /m2 g-l g-l m-, 8-I s-l m-, s-l /ks-l ( 7 3 (%I / w a I v,o,-u 5.4 22 4.1 12.2 2.26 555 60 40 1.35 V,O,-F 0.8 4 5.0 0.27 0.34 68 65 35 1.32 V,O,-RO 0.8 4 5.0 0.51 0.64 128 64 34 1.36 a SB. E.T., the B.E.T. (Brunauer-Emmett-Teller) surface area; L, the number of surface V=O species; R,, reaction rate; TFo, turnover frequency for the oxidation; S(MA + FR), selectivity to maleic anhydride and furan; S(CO+CO,), selectivity to CO and CO,; S(CO), selectivity to CO; S(CO,), selectivity to CO,.Reaction conditions: temperature = 622 K; partial pressure of buta-1,3-diene = 0.0074 atm; partial pressure of 0, = 0.397 atm. but- 1 -ene. These reactions were selected because (i) the partial oxidation product (maleic anhydride) is common to both reactants enabling us to discuss effects of the structure of the reactant molecule on the structure-activity/selectivity lo. l1 (ii) the isomerization of but-1-ene to cis- and trans-but-2-ene proceeds in addition to the oxidation and we can compare the structure-activity/selectivity correlation for the isomerization with that for the o ~ i d a t i o n , l ~ - ~ ~ and (iii) the reactions are important for the industrial production of maleic anhydride.Experimental Catalysts A V,O,-U catalyst was prepared by the thermal decomposition of NH4V0, in a stream of 0, at 773 K. A V,O,-F catalyst was prepared by fusing the V,O,-U catalyst at 1073 K for 18 h in air, followed by gradual cooling to room temperature. A V,O,-RO catalyst was prepared from the V,O,-F catalyst by the reduction-oxidation treatment, i.e. reduction in flowing H, at 673 K for 1 h followed by reoxidation in flowing 0, (20%) at 673 K for 1 h (this cycle was repeated 5 times). The number of surface V=O species (L) on the catalysts has been determined by using the rectangular-pulse technique3 and the results are shown in table 1 together with the results of the B.E.T.surface area (SB.E,.T.). The number of V,O, layers (N) for the catalysts was calculated from L using: N = 2/[J94(V,O,)I (1) where M(V,O,) is the molecular weight of V,O,. According to the results of X-ray diffraction, u.v.-visible spectra, i.r. spectra, X-ray photoelectron spectra and scanning electron micrographs of the catalysts,*> electronic properties of the catalysts do not change while the surface of V,O,-U or V,O,-RO is rougher than that of V,O,-F. TiO, (anatase) was prepared by hydrolysis of Ti(SO,), followed by calcination in air at 873 K, while Al,O, was commercially available (Sumitomo y-Al,03). B.E.T. surface areas of TiO, and Al,03 were 48 and 230m2 g-l, respectively.Vanadium oxides supported on carriers were prepared by impregnation of the carrier with an oxalic acid solution of NH,VO, followed by calcination at 773 K in a stream of 0, for 3 h. V,O,/TiO, and V,O,/Al,O, monolayer catalysts were prepared from V,O,/TiO, (10 mol % V,O,) and V,0,/Al,03 (25 mol % V,O,), respectively, in a manner similar to that described by Yoshida et aZ.15 The number of surface V=O species (L), the number of V,O, layers on the support (N), and the B.E.T. surface area of the supported catalysts have been determined4 and the results are shown in table 2, ML is the percentageK. Mori, A . Miyamoto and Y. Murakami 15 Table 2. Physical and catalytic properties of V,O,/TiO, and V,O,/AI,O, catalysts for the oxidation of buta-1 ,3-dienea ~- L R O S(MA S(C0 catalyst SB- E.T./pmol ML /pmol TF, +FR) +CO,) S(C0) (mol%V,O,) /m2 g-l ggl N (%) g-l sP1 /ks-l (2;) (%I /S(CO,) ~ _ _ 1 2 5 10 25 50 monolayer 2 5 10 25 35 50 monolayer 47 45 26 23 10 32 7.4 22 1 219 168 114 101 66 174 5.4 V,O,/TiO, 56 1-2 32 5.49 120 1-2 67 13.5 184 2-3 280 39.3 135 5-8 600 25.5 60 3WO 2800 21.5 31 5&60 6400 15.3 126 1 ~ 14.0 V2°5/A1203 3 1-2 11 0.35 77 1-2 27 4.03 355 1-3 67 18.2 405 2 4 220 28.2 249 5-15 660 15.8 365 3-7 330 24.5 20 1 - 1.18 22 504 12.2 v,o,-u 98 I12 21 3 189 358 492 111 117 52 51 70 67 64 59 555 32 41 60 60 56 61 41 9 28 51 56 57 55 27 60 68 59 40 40 44 39 59 91 72 49 44 43 45 73 40 -~ 0.89 0.97 1.22 1.35 1.44 1.44 1.03 0.54 0.57 0.75 1.20 1.16 1.25 0.55 1.35 N , number of V,O, layers on support; ML, percent theoretical monolayer of V,O, calculated from eqn (2); for the definition of S,.,.,., L, Ro, TF,, S(MA+FR), S(CO+CO,), S(C0) and S(CO,), see table 1.Reaction conditions: temperature = 622 K; partial pressure of buta-1,3-diene = 0.0074 atm; partial pressure of 0, = 0.397 atm. theoretical monolayer of V205,16 which is calculated from the V,O, content (x) and %.FA using: 100 (% 1 Nx4V,O,) ML = xM(V,O,) + (1 - x) M(support) SB. E.'V. where N is the Avogadro number, a(V,O,) is the area occupied by a V,O, unit (20.6 A,); and M(support) is molecular weight of the support (TiO, or A1203). Unless stated otherwise, the particle size of the catalyst was 28-48 mesh. Catalytic Activity Measurements Kinetic studies were carried out by using the continuous-flow reaction technique under the following conditions; total pressure = 1 atm (1 atm = 101.3 kPa), temperature = 58&666 K, partial pressure of 0, (Po) = M.467 atm, partial pressure of buta-1,3- diene = 0.0074 atm, partial pressure of but-1-ene = 0.0079 atm, and nitrogen was used as a balance gas.Reaction products identified were maleic anhydride (MA), furan(FR), CO, and CO, for the reaction of buta-1,3-diene, and MA, CO, CO,, cis-but-2-ene and trans-but-2-ene for the reaction of but-1-ene. The rate of formation of MA was determined by titration with 0.1 mol dm-3 NaOH after its collection in water. FR, CO, CO,, cis-but-2-ene and trans-but-2-ene were analysed by using gas chromatography. Particular care was paid to remove the heat of reaction and thus control the reactor16 Supported Vanadium Oxide Catalysts temperature (k 1 K).The glass reactor was heated to the reaction temperature by using a fluidized bed of sand, and the catalyst was diluted with a-Al,O,. Characterizations of Catalysts A catalyst in the steady-state reaction was rapidly cooled down to room temperature to measure the steady-state catalyst structure. 1.r. spectra of the catalyst were observed on a Jasco-EDR-3 1 emissionless i.r. diffuse-reflectance spectrometer using KBr as a diluent . l7 Results Effects of 0, Partial Pressure on Reaction Rates and Catalyst Structures for Buta-l,3-diene Oxidation In the oxidation of buta-l,3-diene on vanadium oxide catalysts, the following reactions were found to take place: C4H6 C4H203(MA) (3) C,H6 + C,H,O(FR) (4) C,H, + 4CO ( 5 ) C,H, + 4C0,.(6) From the stoichiometries of eqn (3)-(6), the rate of formation of each product [R(MA), R(FR), R(CO), R(CO,)] is defined as the rate of buta-l,3-diene converted to the product. It has been shown that MA is consecutively formed through FR,,.18 and this was confirmed in the present study. Taking this into account, the rate of formation partial oxidation products (MA and FR) [R(MA + FR)] or total oxidation products (CO and CO,) [R(CO+CO,)] is defined by eqn (7) and (8), respectively: R(MA+FR) = R(MA)+R(FR) (7) R(C0 + CO,) = R(C0) + R(C0,). (8) The total reaction rate (R,) and selectivites to partial oxidation products [S(MA + FR)], total oxidation products [S(CO + CO,)], CO[S(CO)] and CO,[S(CO,)] are given by eqn (9)-( 12), respectively : R, = R(MA + FR) + R(C0 + CO,) S(MA+FR) = R(MA+FR)/R, (9) S(C0 + CO,) = R(C0 + CO,)/K" (10) S(C0) = R(CO)/R, (1 1) S(C0,) = R(CO,)/R,.(12) Fig. 1 shows effects of Po on R,, R(MA+FR) and R(CO+CO,) for the V,O,-U catalyst at 622 K. The rates increased gradually with increasing Po to 0.3 atm, while they are almost constant when Po > 0.3 atm. In spite of the change of the reaction rates with Po, S(MA + FR), S(C0) and S(C0,) are independent of Po) as shown in fig. 2. The reaction rates decreased gradually after the stoppage of 0, supply and attained a negligible value 8 h after the reaction began. It was also confirmed that both the reaction rate and i.r. spectrum of the catalyst attained a steady state under this condition. Fig. 3 shows the i.r. spectra of the V,05-U catalyst in the steady-state reaction at variousK.Mori, A . Miyamoto and Y. Murakami 17 12 r 0 0.1 0,2 0 , 3 0,4 0,s Po/atm Fig. 1. Effect of Po on R,, R(MA + FR), and R(C0 + CO,) for buta- 1,3-diene oxidation on V,O,-U at 622K. 0, R,; 0, R(MA+FR); 0, R(CO+CO,). Partial presence of buta-1,3- diene = 0.0074 atm, W/F = 220 g s rnol-'. 0 0,l 0 , 2 0 , 3 0,4 0,5 Po/atm Fig. 2. Effect of Po on S(MA+FR), S(CO), and S(C0,) for buta-1,3-diene oxidation on V,O,-U at 622 K. 0, S(MA+ FR); 0, S(C0); A, S(C0,). Partial pressure ofbuta-173-diene = 0.0074 atm. partial pressures of 0,. With Po > 0.298 atm the catalyst in the steady-state reaction gave absorption bands at 1020 and 825 cm-l which are assigned to the stretching vibration of V=O species and the coupled vibration between V=O and V-0-V, respe~tive1y.l~~ 2o These absorptions gradually decrease as Po decreases below 0.298 atm, and new absorption bands at 990 and 910 cm-l are observed for the catalyst in the steady-state reaction when Po < 0.149 atm.These bands are assigned to lattice vibrations of V,0,.21 In order to quantify the change in the amount of V=O species with changing Po the relative absorbance at 1020 cm-I was calculated from the spectra in fig. 3 and the results are shown in fig. 4. As shown, the amount of V=O species increases almost linearly with Po to 0.298 atm and it is constant above this value.18 Supported Vanadium Oxide Catalysts wavenumberlcm-' Fig. 3. Infrared spectra of V,O,-U in the steady state of buta-1,3-diene oxidation at various partial pressures of 0,.Temperature = 622 K, partial pressure of buta-1,3-diene = 0.0074 atm. The numbers in parentheses represent the partial pressures of 0,. Activity and Selectivity in Buta-l,3-diene Oxidation under Conditions of Excess Oxygen Table 1 shows results of Ro, Ro/SB. E.T., S(MA + FR), S(C0 + CO,) and S(CO)/S(CO,) for unsupported vanadium oxide catalysts under excess oxygen conditions where the reaction rate was zeroth order with respect to Po and where the catalyst was confirmed to be in the highest oxidation state, i.e. V5+. In accordance with the decrease in SB,E.T., Ro for V,O,-F or V,O,-RO is much smaller than that for V20,-U. It should be noted that the specific activity (RO/SB.E.T.) for V,O,-U is much larger than that for V,O,-F and that for V,O,-RO is larger than that of V,O,-F.In contrast to the behaviour of the reaction rate, the selectivity, S(MA + FR), S(C0 + CO,) or S(CO)/S(CO,), does not change significantly with the kind of unsupported V,05. Table 2 shows results of Ro, S(MA+FR), S(CO+CO,) and S(CO)/S(CO,) for V,O,/TiO, and V,O,/Al,O3 catalysts with various V,O, contents. TiO, or Al,O, alone had a negligible activity for the buta-1,3-diene oxidation. The reaction rate (R,) for V,O,/TiO, increases markedly with an increase in V205 content from 0 to 5 mol % , passes a maximum at 5 mol % V,O,, and then decreases to the value of V,O,-U with further increase in V,O, content. The selectivity to MA and FR[S(MA+FR)] is low for V,O,/TiO, containing 1-2 mol % V,O,, while it is as high as 60% for catalysts with V,O, contents > 5 mol % .V2O,/A1,O3 with low V,05 content (2 mol % ) has a negligible activity.Ro increases markedly with increasing V,O, content from 5-25 mol % . R, forK . Mori, A . Miyamoto and Y. Murakami 19 0 0 , l 0.2 0 . 3 0,4 0,5 Po/atm Fig. 4. Amount of V=O in V,O,-U in the steady state of buta-1,3-diene oxidation at various pressures of 0,. V,O,/Al,O, (25 mol % V,O,) is considerably higher than Ro for V,O,/TiB, (5 rnol % V,O,). S(MA+ FR) for V,O,/A1,0, with low V20, content (2-5 mol "/o ) i s very low, while it increases markedly in the range from 5-25 mol:< and attains a constant value above 25 mol % . S(MA + FR) for the monolayer V,O,/TiO, and V,O,/Al,O, is low and total oxidation of buta- 1,3-diene to CO and CO, takes place preferentially.Although tables 1 and 2 show results for the reaction at 622 K, similar relationships were found to hold at any temperature examined (59G658 K). The apparent activation energy for R, was 20 kcal mol-1 for V,O,-U, 27 kcal mol-1 for V,O,-F, 25 kcal mol-l for V,O,-RO, 19-20 kcal mol-1 for V,O,/TiO, and 27-30 kcal mol-1 for V,O,/Al,O,. The selectivity to MA and FR decreased only sightly with increased temperature. Effects of P, on Rates and Catalyst Structures for But-1-ene Oxidation In the reaction of but-1-ene on vanadium oxide catalysts, the following reactions were found to take place: but-1 -ene (B) + 30, -+ C,H,O,(MA) + 3H,O (13) but- 1 -ene (B) + 40, -+ 4C0, + 4H,O (14) but-1-ene (B)+602 -+ 4C0,+4H20 (15) (16) (17) As with the case for buta-1,3-diene rate of formation of each product [R(MA), R(CO), R(CO,), R(c-B), or R(t-B)] is defined as the rate of but-1-ene converted to the product.Reactions (1 3)-( 15) are oxidation of but- 1 -ene, while reactions (1 6) and (1 7) are isomerization of the reactant. Rates of oxidation (R,) and isomerization (RI), total reaction rate (R), and selectivities to the oxidation products (S,) and isomerization product (S,) are then given by the following: but-1-ene (B) + cis-but-2-ene (c-B) but-I -ene (B) + trans-but-2-ene(t-B). R , = R(MA) + R(C0) + R(C0,) RI = R(c-B) + R(t-B) (18) (19)20 Supported Vanadium Oxide Catalysts 2,4 I 1 L 1 1 0 . 1 0.2 0 . 3 0,4 0 , 5 Po/atm Fig. 5. Effect of Po on R,, R(MA), R(C0) and R(C0,) for but-1-ene oxidation on V,O,-U at 648 K. 0, R,; 0, R(MA); A, R(C0); 0, R(C0,).Partial pressure of but-1-ene = 0.0079 atm, W/F = 5 13 g s rnolp1. R = Ro+RI (20) so = R,/R (21) S I = R,/R. (22) S(MA) = R(MA)/Ro (23) S(C0) = R(CO)/R, (24) S(C0,) = R(CO,)/R,. (25) S(c/t) = R(c-B)/R(t-B). (26) Selectivities to MA[S(MA)], CO[S(CO)] and CO,[S(CO,)] in the oxidation of but- 1 -ene are defined by: The cisltrans ratio [S(c/t)] in the isomerization of but-1-ene is similarly given by: Fig. 5 shows the effect of Po on R,, R(MA), R(C0) and R(C0,) for the V,O,-U catalyst at 648 K. The rates increase gradually with increasing Po to 0.15 atm, while they are almost constant at Po > 0.15 atm. In spite of the change the reaction rates with Po, S(MA), S(C0) and S(C0,) are independent of Po as shown in fig. 6. Fig. 7 shows the effect of Po on R,, R(c-B) and R(t-B) for V,O,-U at 648 K.Similar to the result for the oxidation (fig. 5), the rates increase gradually with increasing -Po to 0.15 and are almost constant at Po > 0.15 atm. It should be noted that the isomerization does not proceed in the absence of 0,, the same behaviour as that for the oxidation. Fig. 8 shows the results of S(c/t) and SI at various Po. When Po is high, trans-but-2-ene is preferentially formed, while the yield of cis-but-2-ene is slightly higher than that of trans-but-2-ene when Po is low. The value of S , is almost independent of Po except for the value at low Po where oxidation is slightly more favourable than isomerization. After the stoppage of 0, supply, the rates decreased gradually with increasing time and attained a negligible value at 8 h after the reaction began (similar to buta- 1,3-diene).It was confirmed that both the reaction rate and i.r. spectrum of the catalyst attainedK . Mori, A . Miyamoto and Y. Murakami 40 I-.------ 21 I 1 1 I I 0 0,1 0-2 0 , 3 0.4 0 . 5 Po/atm Fig. 6. Effect of Po on S(SM), S(C0) and S(C0,) for but-1-ene oxidation on V,O,-U at 648 K . 0, S(MA); A, S(C0); 0, S(C0,). Partial pressure of but-1-ene = 0.0079 atm. Po/atm Fig. 7. Effect of Po on R,, R(t-B) and R(c-B) for V,O,-U at 648 K. 0, R,, A, R(t-B); 0, R(c-B). Partial pressure of but-1-ene = 0.0079 atm, W/F = 513 g s mol-’. a steady state under this condition. Similar to the case for buta- I ,3-diene, the absorbance at 1020 cm-l was calculated from the i.r. spectra at various Po and the results are shown in fig.9. As shown, the amount of V=O species increases almost linearly with the increase in Po < 0.149 atm and it is constant above this value of Po. Activity and Selectivity in the Oxidation of But-l-ene under Excess Oxygen Conditions Table 3 shows results for Ro, RO/SR.E.T., S(MA), S(C0) and S(C0,) for unsupported vanadium oxide catalysts under excess oxygen conditions where the reaction rate was zeroth order with respect to Po and where the catalyst was confirmed to be in the highest oxidation state, i.e. V5+. In accordance with the decrease in SR.E.rr., R, for V,O,-F or22 Supported Vanadium Oxide Catalysts t 1 O a 4 1 I I I I 10 0 0,l 0,2 0 , 3 0,4 0.5 Po/atm Fig. 8. Effect of Po on S, and S(c/t) for V,O,-U. I I I 0,l 0,2 0 , 3 0,4 0,s 6 - I 0 Po/atm Fig.9. Amount of V=O in V,O,-U in the steady state of but-I-ene oxidation at various partial pressures of 0,. Table 3. Oxidation of but- 1 -ene on unsupported V,O, catalystsa _ ~ _ ~ _ _ _ _ _ _ -~ R, (ROISl3. E.T.) /pmol /pmol TF, S(MA) S(C0) S(C0,) catalyst g-l s-l mP2 s-' /ks-l (%I (%I (%> V,O,-U 0.91 0.17 43 46 31 23 V,O,-F 0.090 0.1 1 23 42 32 26 V,O,-RO 0.100 0.13 25 42 33 25 a R,, reaction rate for the oxidation; TF,, turnover frequency for the oxidation; S(MA), selectivity to maleic anhydride ; S(CO), selectivity to CO; S(CO,), selectivity to CO,. Reaction condi- tions: temperature = 622 K; partial pressure of but-1-ene = 0.00 79 atm; partial pressure of 0, = 0.397 atm.K . Mori, A . Miyamoto and Y. Murakami Table 4.Oxidation of but-1-ene on V,O,/TiO, and V,0,/A1,03 catalystsa R O /,urn01 TF, S(MA) S(C0) S(C0,) catalyst g-ls-l /ks-l (%) (%) (%) 1 2 5 10 25 50 mono 1 aye r 2 5 10 25 35 50 monolayer 5.38 11.2 17.7 13.1 5.64 2.11 11.3 0.12 1.23 9.94 17.0 15.0 10.7 0.84 0.94 V,O,/TiO, 96 24 93 32 96 40 97 45 94 47 68 48 90 33 40 0 16 7 28 28 42 41 41 43 43 47 42 0 43 46 V2°5/A1203 v,o,-u 37 34 36 33 32 32 34 34 33 25 32 32 29 31 31 39 34 24 22 21 20 33 66 60 47 27 25 24 69 23 23 a For the definition of R,, TF,, S(MA), S(CO), and S(CO,), see table 3. Reaction conditions: temperature = 622 K; partial pressure of but-1-ene = 0.0079 atm; partial pressure of 0, = 0.397 atm. V,O,-RO is much smaller than that for V,O,-U. However, note that the specific activity (Ro/SB.E.T.) is almost constant and that for V,O,-F is only sightly smaller than that for V,O,-U or V,O,-RO.The selectivity, S(MA), S(C0) or S(CO,), does not change significantly with the kind of unsupported V,O,. Table 4 shows results of R,, S(MA), S(C0,) for V,O,/TiO, and V205/A1,0, catalysts with various V,O, contents. TiO, or Al,O, alone had a negligible activity for the but- 1 -ene oxidation. R, for V,O,/TiO, increased markedly with an increase in V,O, content from 0-5 mol% , passed a maximum at 5 mol % V,O,, and then decreased to the value of V,O,-U with further increase in V,05 content. S(MA) is low for V,O,/TiO, containing 1 or 2 mol % V,O,, while it is as high as 40% for catalysts with > 5 mol V,O, . V205/A1,03 with low V,O, content (2 mol%) has only a negligible activity.R, increases markedly with increasing V,O, content from 5-25 mol % . R, for V,0,/A1,03 (25 mol % V,O,) is considerably higher than R, for V,O,/TiO, ( 5 mol % V,O,). S(MA) for V205/A1203 with low V,O, content (2-5molX) is very low, while it increases markedly in the range from 5-25 mol% and attains a constant value above 25 mol % . S(MA) for the monolayer V,O,/TiO, and V,O,/Al,O, is low and total oxidation of but- 1 -ene to CO and CO, takes place preferentially. Although tables 3 and 4 show results for the reaction at 622 K, similar relationships were found to hold at any temperature examined (580 - 666 K). The apparent activation energy for R, was 25-26 kcal mol-1 for unsupported V,O, catalysts, 22-23 kcal mol-1 for V,O,/TiO,, and 25-27 kcal mol-1 for V,0,/Al,03.24 Supported Vanadium Oxide Catalysts Table 5.Isomerization of but- 1 -ene on unsupported V,O, catalystsn v,o,-u 1.17 0.22 53 20 33 0.60 55 V,O,-RO 0.12 0.15 30 13 18 0.7 1 55 V205-F 0.10 0.13 25 10 15 0.67 52 a R,, rate of isomerization; TF,, turnover frequency for the isomerization ; TF(c-B), turnover frequency for the isomerization to cis-2-butene ; TF(t-B), turnover frequency for the isomerization to trans-but-2-ene; S,, selectivity to isomerization. Reaction conditions: temperature = 622 K; partial pressure of but-1-ene = 0.0079 atm; partial pressure of 0, = 0.397 atm. Table 6. Isomerization of but-1 -ene on V,O,/TiO, and V,O,/Al,O, catalystsn Oh I 2 5 10 25 50 monolayer Oh 2 5 10 25 35 50 monolayer 0.72 3.35 7.07 8.30 5.31 3.24 1.65 6.96 0.12 0.25 3.40 13.9 10.7 13.8 8.29 3.23 1.17 T 4 TF(c-B) /ks-l /ksP1 - 60 59 45 39 54 53 55 - 83 44 39 26 38 33 161 53 V,O,/TiO, 26 26 19 17 22 20 25 - V,O,/A1203 - 53 28 22 13 16 14 94 V,O,-U 20 - 0.57 34 0.78 33 0.78 26 0.74 22 0.76 32 0.70 33 0.62 31 0.80 - 3.00 30 1.78 16 1.70 18 1.22 13 0.95 22 0.70 19 0.72 68 1.39 100 38 39 32 29 36 44 38 100 67 73 58 38 48 43 79 33 0.60 55 ~~ ~ a For the definition of R,, TF, TF(c-B), TF(t-B), S(c/t) and S,, see table 5.Reaction conditions: temperature = 622 K; partial pressure of but-1-ene = 0.0079 atm; partial pressure of 0, = 0.397 atm. TiO, or A1,0, support alone. Activity and Selectivity in the Isomerization of But-l-ene under Excess Oxygen Conditions Table 5 shows results for R,, R,/SR. E.T., S(c/t) and S, for unsupported vanadium oxide catalysts under excess oxygen conditions.In line with the result for the oxidation (table 3), R, for V,O,-F or V,O,-RO is much smaller than that for V,O,-U. However, the specific activity (RI/SB.E.T.) is almost constant and that for V,O,-F is only slightlyK. Mori, A . Miyamoto and Y. Murakami 25 smaller than that for V205-U or V,O,-RO. S(c/t) and S , are almost constant and independent of the kind of unsupported V205. Table 6 shows results for R,, S(c/t) and S, for V,O,/TiO, and V,O,/Al,O, catalysts with various V,05 contents. In contrast to the behaviour in the oxidation, TiO, or A1,0, support alone exhibits a considerable activity for but- 1 -ene isomerization, although it is smaller than that for the supported V205 catalyst. R, for V,O,/TiO, increases markedly with an increase in V205 content from 0-5 m o l x , passes a maximum at 5 m o l x , and then decreases to the value of V20,-U with further increase in V,O, content.The S(c/t) for V,O,/TiO, is almost constant and independent of the V,05 content. V205/A1203 with low V205 content (2 mol%) shows low activity. R, increases markedly with increasing V,05 content from 5-10 mol:4. S(c/t) = 3 for the A1,0, support and is much higher for the unsupported V,05. As the V,O, content increases to 35 mol % , S(c/t) decreases gradually to the value for the unsupported V,05. S(c/t) for the monolayer V20,/Al,03 is also considerably higher than that of the unsupported V,03. Although tables 5 and 6 show results for the reaction at 622 K, similar relationships were found to hold at any temperature examined (580-666 K).The apparent activation energy for R, was 20-23 kcal mol-l for the unsupported V,O, catalysts, 21-23 kcal mol-1 for V,05/Ti02 and 24-26 kcal mol-l for V,05/Al,03. Discussion Active Oxygen Species for the Oxidation of Buta-1,3-diene and But-1-ene As shown in fig. 1, the rate of buta-1,3-diene oxidation increased with increasing Po up to 0.3 atm for V,O,-U and attained a constant value above this value of Po. The steady-state amount of the V=O species in the catalyst changed with Po similarly to the reaction rate (fig. 4). This suggests that the reaction proceeds by the reduction-oxidation mechanism (or Mars-van Karevelen mechanism)22 and that the active oxygen species for the buta-1,3-diene oxidation is surface V=O species, in accordance with the conclusion obtained by Akimoto et al.23 Adsorbed oxygen species are not responsible for this reaction, because no adsorbed oxygen species, such as O;, 0- or O;, were detected on the catalysts by either e.s.r.or t.p.d. measurements. The observed relationship between reaction rate and Po (fig. 1) can be explained in terms of the reduction-oxidation mechanism as follows: In the absence of 0, (Po = 0 atm), the catalyst is in the reduced state and no surface V=O species is present to oxidize buta-1,3-diene. As Po increases, reoxidation increases the oxidation state of the catalyst to increase the number of surface V=O species. This means that the reaction rate increases with increasing Po. In the presence of excess oxygen (P> 0.3 atm), the catalyst is in the highest oxidation state, i.e.V,O,. Therefore, the increase in Po does not lead to further increase in the oxidation state or the number of surface V=O species. Thus the reaction rate does not increase with Po under excess oxygen conditions. As shown in fig. 5 and 9, the behaviour of the reaction rate and the amount of V=O at various Po in but-1-ene oxidation are similar to those in buta-l,3-diene oxidation. Therefore, but-1 -ene oxidation also proceeds by the reduction-oxidation mechanism and the active oxygen species is surface V=O species. Structure Sensitivity of Buta-1,3-diene Oxidation The surface of V,05-F has been found to be significantly different from that of V,O,-U or V20,-R0.8~g Since the surface V=O species has been found to be the active oxygen species for buta- 1,3-diene oxidation, the turnover frequency (TF,) for this reaction can be defined by TF" = R,/L.(27)26 Supported Vanadium Oxide Catalysts Values of TF, at 622 K were calculated from the results of R, and L for various catalysts and the results are shown in table 1. It is evident that TF, changes significantly with the type of catalyst; TF, (V,O,-U) % TFo (V,O,-RO) > TF, (V,O,-F). This behaviour is in contrast to those in the benzene oxidation,, and indicates that buta-1,3-diene oxidation on V,O, catalysts is a structure-sensitive reaction. Fusion of a solid would generally lead to a smooth surface with a decreased number of surface defects (e.g. steps, kinks or vacancies), while severe redox treatment of a solid with few surface defects would tend to increase their number.Furthermore, no impurity peaks were observed in the X.P.S. spectrum of V,O,-U, V,O,-F or V,O,-RO. Since the surface V=O species has been shown to be the active oxygen species, the surface V=O species at the surface defects are considered to be much more active than that in the smooth (010) plane. Structure Sensitivity of But-1-ene Oxidation Values of TF, were calculated from the results of R, and L for various catalysts aiid results are also shown in tables 3 and 4. TF, changes only slightly with the kind of unsupported catalysts, a behaviour in contrast to that in buta-l,3-diene oxidation. Thus, the surface V=O species at the surface defects are considered to have almost the same activity as that in the smooth (010) plane in this case.Activity of Supported Catalysts for Buta-l,3-diene Oxidation In general, the activity of a supported catalyst is determined by two factors; (i) the number of active sites and (ii) the specific activity of the active site, i.e. the turnover frequency. The separation of these two factors is indispensable for detailed understanding the role of support in a given reaction. As shown in table 2, R, for V,O,/Al,O, and V,O,/TiO, catalysts is larger than that for V,O,-U, indicating the promoting effect of A1,0, or TiO, support. The number of surface V=O species ( L ) for V,O,/Al,O, and V,O,/TiO, catalysts is much larger than that for the unsupported V,05, while the turnover frequency (TF) for the supported catalysts is smaller than that for the V,O,-U.This means that the effect of A1,0, or TiO, is to increase the number of surface V=O species, but the specific activity of the surface V=O species is not increased by the support. Fig. 10 shows the relationship between TF and the number of V,O, layers ( N ) for V,O,/TiO, and V,O,/Al,O,. TF for V,O,/TiO, decreases monotonically with decreasing number of V,O, layers on TiO,. This indicates that the retarding effect of TiO, on the specific activity of the surface V=O species becomes greater as the number of V,O, layers decreases. According to Vejux and C~urtine,~* there is a remarkable fit of the crystallographic patterns between the (010) face of V,O, and the TiO, surface. It is therefore considered that a smooth V,O, surface with few defects is formed for the V,O,/TiO, catalysts having a low concentration of V,O,, and that the number of surface defects increases with the content of V,O,.The activity of V=O at the surface defect is much higher than that in the smooth (010) face hence the significant reduction in TF for V,O,/TiO, compared with that for V,O,-U and the increase in 7 7 with increasing content of V,O, in V,O,/TiO,. Selectivity in Buta-1,3-diene Oxidation Fig. 11 shows relationship between the selectivity to partial oxidation products [S(MA + FR)] and the conversion of buta- 1,3-diene, which were obtained from the results under excess oxygen conditions at various temperatures. S(MA + FR) is independent of the conversion for any catalyst. This indicates that consecutive oxidation of the partial oxidation product to CO or CO, was negligible under the present experimentalK.Mori, A . Miyamoto and Y. Murakami 27 1 2 10 190 503 no. of Vz05 layers (N> Fig. 10. Relationshp between the turnover frequency (TF,) for buta- 1,3-diene oxidation and the number of V,O, layers on support. 0, V,O,/TiO,; A, V,O,/Al,O,; a, V,O,-U. I I 0 10 20 30 40 5 conversion (%) Fig. 11. Relationship between the conversion and selectivity for buta-l,3-diene oxidation on unsupported V,O,, V,O,/TiO,, and V,O,/Al,O, catalysts. A, V,O,-U ; A, V,O,-F ; A, V,O,-RO. 0, 0, @, (>, 8, 0, are V,O,/TiO, with V,O, contents 1, 2, 5 , 10, 25, 50 m o l x , and the monolayer catalyst, respectively. 0, o>, ($, 0, 0 , 0, + are V,O,/Al,O, with V,O, contents 2, 5. 10, 25, 35, 50 mol % , and the monolayer catalyst, respectively.Reaction temperature = 590- 658 K, partial pressure of buta-1,3-diene = 0.0074 atm, partial pressure of 0, = 0.397 atm. conditions. In other words, the difference in selectivity among the catalysts is not brought about by the consecutive oxidation of the partial oxidation product. As shown in fig. 4, the oxidation state of the V,O,-U catalyst changes greatly with Po. Under excess oxygen conditions, the catalyst is kept in the highest oxidation state (V5+) while it is reduced as Po decreases. As shown in fig. 2, S(MA + FR), S(C0,) and S(C0) do not change with Po. This means that the selectivity in buta- 1,3-diene oxidation is independent of the oxidation state of the catalyst, at least under the present experimental conditions (conversion of buta- 1,3-diene < 40 % , consecutive oxidation of the partial oxidation product to CO and CO, is negligible and the catalyst is in its steady state at a given level of Po).Table 1 shows that the selectivity is almost independent of the kind of unsupported V,O, catalysts, in contrast to the behaviour of the activity of the catalyst. This indicates that the selectivity is not affected by a change in the surface structure. The selectivity under excess 0, conditions changes greatly with the catalysts (table 2). Fig. 12 shows plots of the selectivity to partial oxidation products [S(MA + FR)] against the number28 60- 2o Supported Vanadium Oxide Catalysts 9 0 0 y.-.-. 0 / 0 4 4QcI b I &I Fig. 12. Relationship between the selectivity for buta-1,3-diene oxidation and the number of V,O, layers on the support.0, V,O,/TiO,; A, V,O,/Al,O,; 0 , V,O,-U. of V,O, layers on support (N). When N = 1 or 2, S(MA + FR) is low, while it increases markedly with an increase in N to 5, and attains a constant value above 5 layers. It is interesting to note that the relationship between S(MA + FR) and N is quite similar for both V,O,/TiO, and V,O,/Al,O, catalysts, while the structures of the V,O,/TiO, catalysts differ significantly from those of V,O,/Al,O, catalysts.4 This indicates that the number of V,O, layers is an important factor for determining the selectivities in the oxidation of buta-l,3-diene; V,O, layers are necessary for the selective oxidation of buta-l,3-diene to FR or MA. The structure-selectivity correlation for buta- 1,3-diene oxidation is similar to that for benzene oxidation., The roughness of the catalyst surface does not affect the selectivity for both reactions.The number of V,05 layers is an important factor determining the selectivity for both reactions: the monolayer catalyst exhibits low selectivity to the partial oxidation product and V,O, layers are necessary for the selective oxidation. Although further studies are necessary to clarify molecular mechanism for the correlation between S(MA+FR) and N , a discussion similar to that described for benzene oxidation may be applicable to buta- 1,3-diene oxidation. As described above, the reaction proceeds by the reduction-oxidation mechanism and the surface V=O species has the active role, It has also been shown that there are Brarnsted acid sites adjacent to the surface V=O species on the supported vanadium oxide cataly~t.~.l9 These suggest that the buta-1,3-diene molecule is adsorbed and activated on the Brmsted acid site and that the reaction is initiated by the nucleophilic attack of the oxygen atom of a surface V=O species to the adsorbed buta-l,3-diene molecule to form an intermediate. Judging from the stoichiometry of the reaction (fig. 3 and 4), subsequent introduction of oxygen atoms to the intermediate species is necessary to complete the reaction. Since the reaction proceeds by the reduction-oxidation mechanism, these oxygen atoms are not directly supplied from gaseous 0,, but supplied by the oxygen of the catalyst. It is well known that the oxygen of V,O, can migrate from the bulk to surface.When the V,O, layers on support are sufficiently thick, the oxygens in the V20, layers can be supplied to oxidize the intermediate species. As for a catalyst with monolayer V,O, or with very thin V20, layers on support, oxygen cannot be supplied from the bulk, but is supplied by the migration of surface oxygen. This change in the mode of oxidation of the intermediate species with the number of V20, layers may provide one of the reasons for the correlation between S(MA+ FR) and N . The change in S(CO)/S(CO,) with the catalyst (table 2) may also be explained by the change in the mode of oxidation of the intermediate species. According to Bond et al.25 the active phase in buta-1,3-diene oxidation is the binaryK.Mori, A . Miyamoto and Y. Murakami 29 oxide Vo.04Tio.9602. The presence of binary oxide was not confirmed for the present catalyst, probably because the calcination temperature was considerably lower than that used by Bond et al. Since the electronic state of vanadium is greatly modified by the formation of the binary oxide, the structure-activity/selectivity correlation for such catalysts may provide an interesting subject for future investigation. Activity of Supported Catalysts for But-1-ene Oxidation Ro for V,O,/Al,O, is greater than that for the unsupported V,O, (table 4), indicating the promoting effect of the A1,0, support. The number of surface V=O species on V,O,/A1,0, is much greater than that on unsupported V,O,, while TFo for V,O,/Al,O, is not greater than that for unsupported V,O,.This means that the promoting effect of Al,O, is to increase the number of active sites. The rate for V,O,/TiO, is also much larger than that for unsupported V,O,. Both L and TFo are increased by supporting V,O, on TiO,. This means that the promoting effect of the TiO, support is caused by two factors: an increase in the number of surface V=O species and an increase in the activity of these species. It is well known that oxygen evolution from V,O, in V,O,/TiO, (anatase) occurs at a temperature much lower than that in unsupported V,0,.661 26 Thus it is believed that the V=O species on V,O,/TiO, are catalytically more active those on unsupported V,O,. The result shown in table 4 provides experimental evidence for the validity of this inference.Selectivity in But-1-ene Oxidation Fig. 13 shows some examples of relationships between the selectivity in but-1-ene oxidation [S(MA) and S(CO)] and the conversion of but-1 -ene, which were obtained from the results under excess oxygen conditions at various temperatures. Either S(MA) or S(C0) is independent of the conversion for any catalyst. This indicates that consecutive oxidation of MA to CO and CO, (or CO to CO,) is negligible under the present experimental conditions. In other words, the difference in the selectivity among catalysts is brought about by the difference in the catalyst structure, but not by consecutive oxidation. The oxidation state of the V,O,-U catalyst changes greatly with Po (fig. 9). Under excess oxygen conditions, the catalyst is kept in the highest oxidation state (V5+) while it is reduced as Po decreases.As shown in fig. 6, S(MA), S(C0,) and S(C0) do not change with Po. This means that the selectivity in but-1-ene oxidation is independent of the oxidation state of the catalyst, at least under the present experimental conditions (consecutive oxidation is negligible and the catalyst is in its steady-state at a given level Table 3 shows that the selectivity is almost independent of the kind of unsupported V,O, catalysts, indicating that it is not affected by a change in the surface roughness. On the other hand, the selectivity under excess 0, condition (table 4) changes greatly with the catalysts. Fig. 14 shows the selectivity to maleic anhydride [S(MA)] against the number of V,O, layers on support.When N = 1 or 2, S(MA) is low, while it increases markedly with an increase in N to 5, and attains a constant value above 5 layers. It is interesting to note that the relationship between S(MA) and N is almost common to both V,O,/TiO, and V,O,/Al,O, catalysts, while the structures of the V,O,/TiO, catalysts differ significantly from those of V,O,/Al,O, cataly~ts.~ This indicates that the number of V,O, layers is an important factor for determining the selectivities in the oxidation of but-1-ene; V,O, layers are necessary for the selective oxidation of but-1-ene to MA. of Po).30 Supported Vanadium Oxide Catalysts Q 20 I 0 20 40 60 conversion (70) 0 Fig. 13. Relationship between the conversion and selectivity for but- 1 -ene oxidation on unsupported V,O,, V,O,/TiO, and V,O,/Al,O, catalysts. A, V,O,-U; A, V,O,-F; A, V,O,-RO.0 , 0 , 0 , 0, 9. + are V,O,/TiO, with V,O, contents 1, 2, 5, 10, 25 and 50 m o l x , respectively. 0, a, @, 0, 0, are V,O,/A1,0, with V,O, contents 2, 5 , 10, 25, 35 and 50 m o l x , respectively. Reaction temperature = 580-666 K, partial pressure of but-1-ene = 0.0079 atm, partial pressure of 0, = 0.397 atm. k , , . ,,,., , , , , , , * , , , , 0 1 2 10 100 500 no. of V , 0 5 layers (N) Fig. 14. Relationship between the selectivity for but-1-ene oxidation and the number of V,O, layers on support. 0, V,O,/TiO,; A, V,O,/Al,O,; 0, V,O,-U. Active Sites for Isomerization Fig. 15 shows some examples of relationships between S(c/t) and the conversion of but-1-ene, which were obtained from the results under excess oxygen conditions at various temperatures.S(c/t) is independent of the conversion for any catalyst, indicating that consecutive isomerization of cis-but-2-ene to trans-but-2-ene or of trans-but-2-ene to cis-but-2-ene was negligible under the present experimental conditions. According to the results of previous investigations on the isomerization ofK. Mori, A . Miyamoto and Y. Murakami 31 I 1 1 60 0,21 0 20 40 1 conversion (%) 1 Fig. 15. Relationship between the conversion and S ( c / t ) for but-1-ene isomerization on unsupported V,O,, V,O,/TiO,, and V,O,/Al,O, catalysts. A, V,O,-U; A, V,O,-F; A, V,O,-RO. 0, 0 , 0 , 0 , 8, 4 are V,O,/TiO, with V,O, contents 1, 2, 5, 10, 25 and 50 mol % , respectively. 0, (3, 0, 8, 8, are V,O,/Al,O, with V,O, contents 2, 5, 10, 25, 35 and 50 mol%, respectively. Reaction temperature = 580-666 K, partial pressure of but-1-ene = 0.0079 atm, partial pressure of 0, = 0.397 atm.but-1-ene,l2-l4 the value of S(c/t) is a sensitive reflection of the reaction mechanism: if the reaction proceeds on the Brsnsted acid site, the s-butyl cation is formed as an intermediate to give 0.5 < S(c/t) < 1 .O. If a n-ally1 anion intermediate is formed by the interaction of but-1-ene with metal oxides such as MgO, ZnO, La,O, and A1,0,, S(c/t) > 2. As shown in table 5, S(c/t) for the unsupported V205 is in the range 0.6-0.7, suggesting that Brsnsted-acid site plays an important role as an active site for the isomerization of but-1-ene. This is consistent with the results of characterization of the unsupported V,05 because i.r.spectra of NH, and pyridine adsorbed on the unsupported V205 indicate the presence of the Brsnsted-acid site on the catalyst, but not of the Lewis-acid ~ i t e . ~ , ~ ~ As shown in table 6, Al,O, support alone gives a value of S(c/t) as high as 3. This is also consistent with the presence of the Lewis-acid site and the absence of the Brsnsted-acid site as revealed by i.r. ~pectra.~ The value of S(c/t) for TiO, support alone (0.67) is also consistent with the results of i.r. spectra of NH, and pyridine on TiO,, since the presence of the Brsnsted-acid site has been confirmed in addition to the Lewis-acid site.4 S(c/t) for V,O5/Al@, decreases gradually with the increase in V,O, content from 0 to 35 mol % .This can be explained in terms of the structure of the V205/Al,0, catalyst as follows: when the V,05 content is 25 mol % or lower, the A1,0, surface is not completely covered by v205,4 therefore the isomerization can proceed on both A120, and V205 surfaces, leading to a value of S(c/t) between 3.0 (A1203 support alone) and 0.7 (V,O, alone). When the V,O, content is 35 mol % or higher, the whole Al,O, surface is covered with V205 layers and S(c/t) for the catalyst is close to the value for the unsupported V,O,. As for the monolayer V,O,/Al,O, catalyst, a considerable portion of A120, is exposed to the catalyst ~urface.~ This explains high value of S(c/t) for the monolayer V,05/A1,0, catalyst. S(c/t) for the V205/Ti02 catalyst is almost constant and close to the value for the unsupported catalyst, indicating that the Brsnsted acid site is important for the isomerization of but-1-ene on a V205/Ti02 catalyst.2 F A R 132 Supported Vanadium Oxide Catalysts As shown in fig. 7, the rate of but-1-ene isomerization increases with increasing Po to 0.1 5 atm for V,O,-U and attains aconstant value above this value of Po. The steady-state amount of the V5+=0 species in the catalyst changes with Po similarly to the reaction rate (fig. 9). Since the Brsnsted-acid site is the active site for the isomerization of but-1-ene on the unsupported vanadium oxide catalyst, this suggests that the surface V=O species play an essential role in promoting the reaction on the Brsnsted-acid site. According to quantum-chemical calculations of vanadium oxide clusters with various oxidation states,28 the V5+=0 species effectively increases the positive charge on the Brsnsted-acid site adjacent to the V5+=0 species, while its reduction leads to the decrease in the positive charge of the Brsnsted-acid site.This suggests that the promoting effect of the surface V5+=0 species on the isomerization is brought about by the increase in the strength of the Brsnsted-acid site. Difference in the Catalytic Behaviour between Oxidation and Isomerization Judging from the stoichiometries of the oxidation [eqn (1 3)-( 15)] and isomerization [eqn (1 6-1 7)], the structure-activity/selectivity correlation for the oxidation is expected to be different from that for the isomerization. If we take into account the effect of Al,O, or TiO, support on the isomerization, however, the catalytic behaviour in the isomerization is apparently very similar to that in the oxidation : R , and R, change similarly with the treatment of the unsupported V,O, catalyst (tables 3 and 5 ) and with Po (fig.5 and 7). This can be seen quantitatively in the value of S, (percentage of the isomerization rate in the total reaction rate). ST is almost constant and independent of the kind of unsupported catalyst (table 5 ) or Po (fig. 8). Similarly, S , for supported catalysts (except for the one with low V,O, content) does not differ significantly from that for the unsupported V,O, (table 6). The surface V=O species is the active oxygen species for the oxidation of but-1-ene. As discussed above, surface V=O also plays an essential role in the isomerization of but-1-ene by increasing the strength of the Brsnsted-acid site which is the active site for isomerization.This explains the apparent similarity in the catalytic behaviour between the isomerization and oxidation. A closer inspection of the turnover frequency for the isomerization (TF,) (table 6) indicates that TiO, support does not increase TF, for unsupported V,O,, in contrast to the behaviour in oxidation (table 4). This can also be explained in terms of the above mentioned difference in the role of surface V=O between both reactions. Difference in the Structure-Activity/Selectivity Correlation between But-1 -ene Oxidation and Buta-1,3-diene Oxidation According to Ai et the oxidation of but-1-ene to MA proceeds by the following mechanism : but-1-ene buta-1,3-diene - furan - MA co, co2 1 co, coz co, co, Although no buta-1,3-diene or furan was detected as a reaction product, the present results are consistent with this mechanism.This is because the turnover frequency for but-1 -me oxidation is much smaller than that for buta-1,3-diene oxidation and because the relationship between S(MA) and N (fig. 14) is similar to that for buta-1,3-diene oxidation. It should however be noted that the structure-activity correlation for but- 1 -ene oxidation is much different from that of the buta-1,3-diene oxidation. In contrast to theK. Mori, A . Miyamoto and Y. Murakami 33 behaviour in but-l-ene oxidation, the activity for buta-l,3-diene oxidation is very sensi- tive to the change in surface structure of V,O,; TFo for V,O,-F is much smaller than that for V,O,-U or V,O,-RO.Furthermore, TiO, support does not increase TF, in buta- 1,3-diene oxidation. Judging from eqn (28) the role of catalyst in the oxidation of but- 1 -ene to buta-173-diene is to abstract two hydrogen atoms from but-l-ene. On the other hand, the role in the oxidation of buta-1,3-diene to furan involves (i) donation of an oxygen atom, (ii) abstraction of two hydrogen atoms, and (iii) transformation of the geometry of conjugated dienes from s-trans to s-cis structures: CH-CH - II II CH C H , \ / CH2 \\ CH- CH \\ CH2 0 We have previously demonstrated for the oxidation of H, and CO on vanadium oxide catalysts that role of catalyst in a reaction significantly affects the structure sensitivity of the reaction.29 The role of the surface V=O in the H, oxidation (a structure-insensitive reaction) is to dissociate H-H bonds and form V-OH species, while that in the CO oxidation (a structure-sensitive reaction) is to donate its oxygen to a CO molecule to form a CO, molecule.Thus, the abovementioned differences in the role of the catalyst would explain the difference in the structure-activity correlation between but- 1 -ene oxidation and buta- 1,3-diene oxidation This work was partially supported by a Grant-in-Aid for Scientific Research (no. 59470097) and for Encouragement of Young Scientists (no. 59750655) from the Ministry of Education, Science and Culture, Japan. References 1 2 3 4 5 6 7 8 9 10 (a) C. F. Cullis and D. J. Hucknall, Catalysis (The Chemical Society, London, 1982), vol.5, p. 273; (6) M. S. Wainwright and N. R. Foster, Catal. Rec., 1979, 19, 21 1 ; (c) D. B. Dadyburjoir, S. S. Jewur and E. Ruckenstein, Catal. Rev., 1979, 19, 293; ( d ) A. Bielanski and J. Haber, Catal. Reti., 1979. 19, 1 ; (e) J. Haber, Proc. 8th Int. Congr. Catal. (West Berlin, 1984), 1-85 and references therein. D. J. Hucknall, Selective Oxidation of Hydrocarbons (Academic Press, London, 1974), p. 75 and references therein. (a) A. Miyamoto, Y. Yamazaki, M. Inomata and Y. Murakami, J. Phj,s. Chem., 1981, 85, 2366; (h) M. Inomata, A. Miyamoto and Y. Murakami, J . Phys. Chem., 1981, 85, 2372. (u) M. Inomata, K. Mori, A. Miyamoto, T. Ui and Y. Murakami, J. Phys. Chem., 1983, 87, 754; (b) M. Inomata, K. Mori, A. Miyamoto and Y.Murakami, J. Phys. Chem., 1983,87,761; (c) Y. Murakami, M. Inomata, K. Mori, T. Ui, K. Suzuki, A. Miyamoto and T. Hattori, Preparation of Catalysts III, ed. G. Poncelet, P. Grange and P. A. Jacobs (Elsevier, Amsterdam, 1983), p. 531. (a) K. Mori, M. Inomata, A. Miyamoto and Y. Murakami, J. Chem. Soc., Faraday Trans. I , 1984,80, 2655; (b) K. Mori, M. Inomata, A. Miyamoto and Y. Murakami, J. Phys. Chem., 1983, 87, 4560. (a) G. C. Bond and K. Briickman, Faraday Discus3 Chem. Soc., 1981, 72, 235; (h) D. J. Cole, C. F. Cullis and D. J. Hucknall, J. Chem. Soc., Faraday Trans. 1, 1976, 72, 2185; (c) A. J . van Hengstum, J. G. van Ommen, H. Bosch and P. J. Gellings, Proc. 8th Int. Congr. Cutal. (West Berlin, 1984), vol. IV, p. 297; ( d ) R. Koziolowski, R.F. Pettifier and J. M. Thomas, J . Phys. Chem., 1983, 87, 5175; (e) T. Ono, Y. Nakagawa, H. Miyata and Y. Kubokawa, Bull. Chem. Soc. Jpn, 1984, 57, 1205; (e) I. E. Wachs, S. S. Chan, C. C. Chersich and Y. Saleh, Catalysis on the Energy Scene, ed. S. Kaliaguine and A. Mahay (Elsevier, Amsterdam, 1984), p. 275; ( f ) R. D. Srivastava and A. B. Stiles, J. Catal., 1982, 77, 192. A. Miyamoto, K. Mori, M. Inomata and Y. Murakami, Proc. 8th Int. Congr. Catal. (West Berlin, 1984), W-285 and references therein. K. Mori, A. Miyamoto and Y. Murakami, J . Phys. Chem., 1984,88, 2735. K. Mori, A. Miyamoto and Y. Murakami, J . Phys. Chem., 1984, 88, 2741. (a) G. Centi, I. Manenti, A. Riva and F. Trifiro, Appl. Catul., 1984, 9, 177; (b) L. Morselli, F. Trifiro and L. Urban, J. Catal., 1982, 75, 112; (c) F.Cavani, G . Centi and F. Trifiro, Ind. Eng. Chem., Prod. ,.a L-L34 Supported Vanadium Oxide Catalysts Res. Dev., 1983,22,570; ( d ) W. E. Slinkard and P. B. Degrrot, J . Catal., 1981,68,423; (e) R. L. Varma and D. N. Saraf, J . Catal., 1978,55,361; (f) M. Nakamura, K. Kawai and Y . Fujiwara, J . Catal., 1974, 34,345; (g) T. Seiyama, K. Nita, T. Maehara, N. Yamazoe and Y. Takita, J. Catal., 1977,49, 164, and references therein. 11 (a) M. Ali, K. Harada, and S . Suzuki, Kogyo Kaguku Zassi, 1970, 73, 524; (6) M. Ai, J . Catal., 1981, 67, 110; (c) M. Ai, Bull. Chem. SOC. Jpn, 1970, 43, 3490; ( d ) M. Ai, Bull. Chem. SOC. Jpn, 1977, 50, 355; (e) M. Ai, J. Catal., 1975,40,318; (f) M. Ai, T. Niikuni and S. Suzuki, Kogyo Kagaku Zassi, 1970, 73, 165, and references therein. 12 J. W. Hightower and W. K. Hall, Chem. Eng. Prog., Symp. Ser., 1967, 63, 122, and references therein. 13 K. Tanabe, Solid Acids and Bases (Kodansha, Tokyo and Academic Press, London, 1970), chap. 5, and references therein. 14 (a) J. Goldwasser and W. K. Hall, J . Catal., 1981,71,53; (6) M. P. Rosynek, J. S. Fox and J. L. Jensen, J . Catal., 1981, 71, 194; (c) H. Forster and R. Seeleman, J . Chem. SOC., Faraday Trans. I , 1978, 74, 1435; ( d ) E. A. Irvine and D. Taylor, J. Chem. SOC., Furaday Trans. I, 1978, 74, 1590; (e) H. Hattori, K. Maruyama and K. Tanabe, Bull. Chem. Soc. Jpn, 1977,50,2187; v) S . Tsuchiya, M. Mikami and H. Imamura, Bull. Chem. SOC. Jpn, 1984, 57, 571, and references therein. 15 S. Yoshida, T. Iguchi, S. Ishida and K. Tarama, Bull. Chem. SOC. Jpn, 1972, 45, 476. 16 F. Roozeboom, T. Fransen, P. Mars and P. J. Gellings, 2. Anog. Allg. Chem., 1979, 25, 449. 17 M. Niwa, T. Hattori, M. Takahashi, K. Shirai, M. Watanabe and Y. Murakami, Anal. Chem., 1979, 51, 46. 18 (a) M. Ai, K. Harada and S . Suzuki, Kogyo Kagaku Zassi, 1970, 73, 524; (b) M. Ai, J . Catal., 1981, 67, 110; (c) F. Cavani, G. Centi, I. Manenti, A. Riva and F. Trifiro, Znd. Eng. Chem., Prod. Res. Dev., 1983,22, 565; ( d ) F. Cavani, G. Centi and F. Trifiro, Znd, Eng. Chem., Prod. Res. Dev., 1983, 22, 570 and references therein. 19 M. Inomata, A. Miyamoto and Y . Murakami, J. Catal., 1980, 62, 140. 20 K. Tarama, S. Yoshida, S. Ishida and H. Kakioka, Bull. Chem. Soc. Jpn, 1969, 41, 2840. 21 L. D. Frederickson and D. M. Hansen, Anal. Chem., 1963, 35, 818. 22 P. Mars and D. W. van Krevelen, Chem. Eng. Sci., 1951, 3, 41. 23 M. Akimoto, M. Usami, and E. Echigoya, Bull. Chem. SOC. Jpn, 1978, 51, 2195. 24 A. Vejux and P. Courtine, J. Solid State Chem., 1978, 23, 93. 25 G. C. Bond, A. J. Sarkany and G. D. Parfitt, J . Catal., 1979, 57, 176. 26 (a) A. Vejux and P. Courtine, J. Solid State Chem., 1978, 23, 93; (b) G. C. Bond, A. J. Sarkany arid 27 M. Inomata, A. Miyamoto and Y. Murakami, J . Catal., 1980, 62, 140. 28 A. Miyamoto, M. Inomata, A. Hattori, T. Ui and Y. Murakami, J. Mol. Catal., 1982, 16, 315. 29 K. Mori, M. Miura, A. Miyamoto and Y . Murakami, J . Phys. Chem., 1984, 88, 5232. G. D. Parfitt, J. Catal., 1979, 57, 176. Paper 4/ 1995; Received 22nd Noaember, 1984