677
2006, 110, 677-679 Published on Web 12/18/2005
Chemoselective Alkane Oxidation by Superoxo-Vanadium(V) in Vanadosilicate Molecular Sieves Vasudev N. Shetti, M. Jansi Rani, D. Srinivas,* and Paul Ratnasamy National Chemical Laboratory, Pune 411 008, India ReceiVed: NoVember 11, 2005; In Final Form: December 9, 2005
Electron paramagnetic resonance (EPR) spectroscopy of reactive superoxo-vanadium(V) species in vanadosilicate molecular sieves (microporous VS-1 and mesoporous V-MCM-41) generated on contact with H2O2, tert-butyl hydroperoxide (TBHP), or (H2 + O2) is reported for the first time. By suitable choice of the silicate structure, solvent, and oxidant, we could control the vanadium-(O2-•) bond (i.e., the V-O bond) covalency, the mode of O-O cleavage (in the superoxo species), and, therefore, chemoselectivity in the oxidation of n-hexane: Oxidation by TBHP over V-MCM-41, for example, yielded 27.2% of (n-hexanol + n-hexanal + n-hexanoic acid), among the highest chemoselectivities for oxidation of the terminal -CH3 in a linear paraffin reported to date. Over these vanadosilicates, oxidation of the primary C-H bond occurs only via a homolytic O-O bond cleavage; the secondary C-H bond oxidations may proceed via both the homo- and heterolytic O-O cleavage mechanisms.
Vanadosilicate molecular sieves exhibit efficient catalytic activity in selective oxidation of hydrocarbons.1 Interestingly, they catalyze the terminal, primary C-H bond oxidation of n-alkanes, which their titanium analogues cannot. Vanadosilicates also exhibit higher efficiency in the photocatalytic decomposition of NO.2 The type of oxygen species generated and the mode of O-O bond cleavage in the reactive oxygen intermediate influence the catalytic activity and product selectivity.1 While there have been a large number of studies on the reactive oxygen species generated over titanosilicates using a range of spectroscopic techniques,3 similar studies on vanadosilicates are scarce.4 We report here, for the first time, in situ electron paramagnetic resonance (EPR) spectroscopic evidence for the formation of reactive superoxo-vanadium(V) intermediates in microporous VS-1 and mesoporous V-MCM-41 molecular sieves. In the case of metal complexes in homogeneous solutions, the superoxo-vanadium species were found to be short-lived.5 Stable species were, however, observed on silicasupported V2O5 on reduction in hydrogen followed by oxygen dosing.6 In the present study, we have observed that the superoxo species, generated in situ in the EPR cells on contacting VS-1 or V-MCM-41 with H2O2 or tert-butyl hydroperoxide (TBHP), are structurally different from those on silica-supported V2O5. VS-1 and V-MCM-41 were prepared and characterized according to known procedures1,7 (see sections S1 and S2 of the Supporting Information). In the calcined state, VS-1 and V-MCM-41 are EPR silent, indicating that vanadium is in the +5 oxidation state. When contacted with 30% aqueous H2O2, the samples became paramagnetic and showed EPR spectra typical of a superoxide radical8 (hereafter referred to as type A species; Figure 1(i)) with a well-resolved eight-line splitting pattern in both the gzz and gyy regions. These superhyperfine * Corresponding author. E-mail:
[email protected]. Phone: 91-0202590 2018. Fax: 91-020-2589 3761.
10.1021/jp0565296 CCC: $33.50
Figure 1. EPR spectra of superoxo-vanadium(V) species in vanadosilicates generated by contacting with (i) aq H2O2 (30%) and (ii) aq TBHP (70%). The vanadium hyperfine features are indicated. The spectra were recorded at 80 K and 9.45 GHz.
features reveal that the superoxide radical is coordinated to a vanadium center, forming a 1:1 metal-superoxo complex (V(O2-•)).5,6 Vanadium superhyperfine features in the gxx region are not resolved. The rhombic g-anisotropy indicates that the O2-• ion is coordinated in an end-on fashion.8 The nature of the silicate structure has a marked effect on the EPR spin Hamiltonian parameters of the superoxovanadium species (gzz ) 2.0203, gyy ) 2.0104, gxx ) 2.0012, Azz ) 7.54 G, Ayy ) 4.60 G, Axx ) 1.1 G (VS-1/H2O2); gzz ) 2.0210, gyy ) 2.0102, gxx ) 2.0012, Azz ) 6.90 G, Ayy ) 4.53 G, Axx ) 1.1 G (V-MCM-41/H2O2)). The Azz parameter is lower for the species generated over V-MCM-41 than those generated over VS-1. The relative intensity of the signal is higher in the case of V-MCM-41 than in VS-1 (Figure 1(i)). Solvents have a marked effect on the vanadium superhyperfine coupling © 2006 American Chemical Society
678 J. Phys. Chem. B, Vol. 110, No. 2, 2006
Letters
SCHEME 1: Superoxo-Vanadium(V) Species Generated in Vanadosilicate Molecular Sieves Contacted with (H2 + O2), H2O2, and TBHP
constants. The Azz value for VS-1 samples decreased in different solvents: water (7.54 G) > acetone (7.37 G) > methanol (7.0 G) > acetonitrile (6.8 G). The superoxide species generated with (H2 + O2) on amorphous V2O5-SiO2 showed resolved hyperfine features even in the gxx region (gzz ) 2.023, gyy ) 2.011, gxx ) 2.004, Azz ) 9.7 G, Ayy ) 6.8 G, Axx ) 5.9 G),6 significantly different from those generated on VS-1 and V-MCM-41 (see section S3 of the Supporting Information). When the vanadosilicates were contacted with 70% aqueous TBHP instead of H2O2, two different types of superoxovanadium species (hereafter referred to as type B and C species) were generated (Figure 1(ii)). Unlike the type A species generated with H2O2, the type B and C species generated with TBHP did not show the characteristic vanadium superhyperfine features (compare the spectra in Figure 1(ii) with those in Figure 1(i)). Only the type B superoxide species were generated on VS-1 (gzz ) 2.0322, gyy ) 2.0098, gxx ) 2.0035). An additional superoxide species (C) was also generated on V-MCM-41 (gzz ) 2.0330, gyy ) 2.010, gxx ) 2.0043 (species B); g| ) 2.0235, g⊥ ) 2.0080 (species C)) (Figure 1(ii)). While species B was characterized by rhombic g-anisotropy corresponding to an endon coordinated superoxide ion, species C (generated only on V-MCM-41) was characterized by axial g-anisotropy corresponding to a side-on coordinated superoxide species.8 Species B and C differed in their gzz/| parameter. The gzz/| parameter of the species generated using TBHP (types B and C) was higher than those generated from H2O2 (type A). However, all of these values are well within the range for a superoxide ion coordinated to a cation with a +5 oxidation state.8 TBHP in an aqueous medium (70%) and in a nonaqueous medium (50%) gave rise to similar types of superoxide species. Scheme 1 summarizes the different types of superoxo-vanadium(V) species generated in vanadosilicates. When a known quantity of allyl alcohol or n-hexane was added, the superoxide signals decreased in intensity with time (Figure 2). The signals disappeared after 20 min in the case of allyl alcohol, and a longer time was needed for complete disappearance with n-hexane. Hence, the superoxo species generated on vanadosilicates on contact with H2O2 or TBHP are consumed by reaction with allyl alcohol or n-hexane to give the oxidation products. The vanadium superhyperfine coupling in the A-type species generated using H2O2 or (H2 + O2) arises due to a partial delocalization of electron density from the orbitals of superoxo oxygen to the vanadium orbitals. Such delocalization of electron density occurs only if the vanadium-(O2-•) bond (i.e., the V-O bond) in vanado-superoxo species is covalent. The B- and C-type superoxo-vanadium species generated using TBHP did not show vanadium superhyperfine features (see Figure 1(ii)), indicating that the unpaired electron is localized on the superoxo oxygen orbitals. Hence, the V-O bond of B- and C-type species is ionic. Similar relations between superhyperfine coupling constants and the nature of bonding were observed also in the
Figure 2. Reactivity of superoxo-vanadium(V) species in allyl alcohol and n-hexane oxidations. The reactions in the presence of substrates were conducted in EPR cells for different periods of time at 323-333 K, the samples were quenched to 80 K, and spectra (ν ) 9.45 GHz) were recorded.
TABLE 1: Oxidation of n-Hexane over Vanadosilicatesa vanadosilicateb
oxidant
solvent
VS-1 VS-1 VS-1 V-MCM-41 V-MCM-41
H2O2 H2O2 H2O2 H2O2 TBHP
(CH3)2CO CH3OH CH3CN CH3CN CH3CN
C-H oxidation selectivity (%) n-hexane TOFc conv. (%) (h-1) primary secondary 4.2 6.0 8.1 54.0 21.3
2 3 5 94 37
6.8 9.2 10.5 20.9 27.2
93.2 90.8 89.5 79.1 72.8
a n-Hexane ) 2.5 g; n-hexane/oxidant ) 3 (mol); solvent ) 12.5 mL; VS-1 ) 0.4 g; V-MCM-41 ) 0.15 g; temperature ) 373 K; time ) 8 h. b Si/V output molar ratio (AAS): VS-1, 320; V-MCM-41, 350. c Turnover frequency (TOF) ) moles of n-hexane converted per mole of vanadium per hour.
case of cobalt-dioxygen complexes.9 It is known from the studies on titanosilicates3 and also vanadosilicates1 that the coordination number of vanadium increases on contact with H2O2 or TBHP from 4 to 5 or 6; V(OSi)3(OR)(O2-•)-type oxo species have been proposed to form.1 Here, R ) H and tertbutyl in the case of H2O2 and TBHP, respectively. The electron donating tert-butoxy group makes the vanadium center in the oxo-vanadium species rich in electron density, and hence, the delocalization of electron density from superoxo oxygen to vanadium is not favored. As a consequence, the superoxovanadium species generated using TBHP possibly does not show superhyperfine coupling features in the EPR spectrum. A covalent V-O bond, wherein part of the electron density on oxygen is delocalized onto the orbitals of V, leads to a lower amount of electron density on the superoxide ion. Since these electrons are being removed from the antibonding orbital of O2-•, this strengthens the O-O bond, thereby facilitating its heterolytic cleavage. An ionic V-O bond, on the contrary, weakens the O-O bond and leads to a homolytic O-O cleavage. Our EPR spectroscopic studies reveal that the covalency of the V-O bond of superoxo-vanadium species in VS-1 is greater than that in V-MCM-41. In different solvents, it decreases in the following order: acetone > methanol > acetonitrile. Hence, it follows that the probability of a heterolytic O-O cleavage in oxidations over VS-1 should be greater than those on V-MCM-41 and in the above solvents it should also decrease in the above order. In the oxidation of n-hexane, the selectivity for the secondary versus primary C-H bond oxidation is more over VS-1 than V-MCM-41 (Table 1) and this selectivity in the above solvents decreases in the following
Letters order: acetone > methanol > acetonitrile1c (see section S4 of the Supporting Information). With different oxidants, the primary C-H bond oxidation is more favored with TBHP than with H2O2 (Table 1). These parallel variations of chemoselective primary C-H bond oxidation and the tendency for a homolytic O-O cleavage indicate that the primary C-H bond oxidation in n-alkanes is favored, probably by a homolytic cleavage of the O-O bond in the oxidant. Oxidation of secondary C-H bonds can proceed by both the homo- and heterolytic O-O cleavage mechanisms. In conclusion, we provide, for the first time, EPR spectroscopic evidence for participation of oxygen radical ions in selective oxidations over vanadosilicate molecular sieves (Figure 2). In addition, our results indicate that the type of cleavage of the O-O bond in the oxidant (H2O2 or TBHP) influences the chemoselectivity of oxidation: A heterolytic cleavage leads to oxidation of secondary C-H bonds, while the primary C-H bonds (in paraffins) are preferentially oxidized by a homolytic cleavage of the O-O bond. Acknowledgment. One of the authors (V.N.S.) thanks the Council of Scientific and Industrial Research (CSIR), New Delhi, for the award of a Senior Research Fellowship. Supporting Information Available: Catalyst preparation, characterization, and reactivity. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) (a) Hari Prasad Rao, P. R.; Ramaswamy, A. V.; Ratnasamy, P. J. Catal. 1992, 137, 225-231. (b) Hari Prasad Rao, P. R.; Belhekar, A. A.; Hedge, S. G.; Ramaswamy, A. V.; Ratnasamy, P. J. Catal. 1993, 141, 595603. (c) Hari Prasad Rao, P. R.; Ramaswamy, A. V.; Ratnasamy, P. J. Catal. 1993, 141, 604-611. (d) Ramesh Reddy, K.; Ramaswamy, A. V.; Ratnasamy, P. J. Catal. 1993, 143, 275-285. (e) Ramaswamy, A. V.; Sivasanker, S. Catal. Lett. 1993, 22, 239. (f) Ramaswamy, A. V.; Sivasanker, S.; Ratnasamy, P. Microporous Mater. 1994, 2, 451-458. (g) Kumar, P.; Kumar, R.; Pandey, B. Synlett 1995, 289-298. (h) Mal, N. K.; Ramaswamy, A. V. Appl. Catal., A 1996, 143, 75-85. (i) Singh, A. P.; Selvam, T. Appl.
J. Phys. Chem. B, Vol. 110, No. 2, 2006 679 Catal., A 1996, 143, 111-124. (j) Singh, A. P.; Selvam, T. J. Mol. Catal. A: Chem. 1996, 113, 489-497. (k) Kannan, S.; Sen, T.; Sivasanker, S. J. Catal. 1997, 170, 304-310. (l) Laha, S. C.; Kumar, R. Microporous Mesoporous Mater. 2002, 53, 163-177. (m) Shylesh, S.; Singh, A. P. J. Catal. 2004, 228, 333-346. (n) Shylesh, S.; Singh, A. P. J. Catal. 2005, 233, 359-371. (o) Trukhan, N. N.; Derevyankin. A. Y.; Shmakov, A. N.; Paukshtis, E. A.; Kholdeeva, O. A.; Romannikov, V. N. Microporous Mesoporous Mater. 2001, 44-45, 603-608. (p) Laha, S. C.; Kumar, R. Microporous Mesoporous Mater. 2002, 53, 163-177. (q) Shiraishi, Y.; Naito, T.; Hirai, T. Ind. Eng. Chem. Res. 2003, 42, 6034-6039. (2) (a) Anpo, M.; Zhang, S. G.; Higashimoto, S.; Matsuoka, M.; Yamashita, H. J. Phys. Chem. B 1999, 103, 9295-9301. (b) Higashimoto, S.; Matsuoka, M.; Zhang, S. G.; Yamashita, H.; Kitao, O.; Hidaka, H.; Anpo, M. Microporous Mesoporous Mater. 2001, 48, 329-335. (3) (a) Ratnasamy, P.; Srinivas, D.; Kno¨zinger, K. AdV. Catal. 2004, 48, 1-169. (b) Thomas, J. M.; Sankar, G. Acc. Chem. Res. 2001, 34, 571. (c) Sankar, G.; Thomas, J. M.; Catlow, C. R. A.; Barker, C. M.; Gleeson, D.; Kaltsoyannis, N. J. Phys. Chem. B 2001, 105, 9028. (d) Lin, W.; Frei, H. J. Am. Chem. Soc. 2002, 124, 9292. (e) Bordiga, S.; Damin, A.; Bonino, F.; Ricchiardi, G.; Lamberti, C.; Zecchiana, A. Angew. Chem., Int. Ed. 2002, 41, 4734. (f) Bordiga, S.; Damin, A.; Bonino, F.; Ricchiardi, G.; Zecchiana, A.; Tagliapietra, R.; Lamberti, C. Phys. Chem. Chem. Phys. 2003, 5, 4390. (g) Srinivas, D.; Manikandan, P.; Laha, S. C.; Kumar, R.; Ratnasamy, P. J. Catal. 2003, 217, 160. (h) Shetti, V. N.; Manikandan, P.; Srinivas, D.; Ratnasamy, P. J. Catal. 2003, 216, 461. (4) (a) Tatsumi, T.; Watanabe, Y.; Himasawa, Y.; Tsuchiya, J. Res. Chem. Intermed. 1998, 24, 529-540. (b) Gasymov, A. M.; Przheval’skaya, L. K.; Shvets, V. A.; Kazanski, V. B. Kinet. Katal. 1984, 25, 358-362. (5) (a) Kelm, H.; Kruger, H.-J. Angew. Chem. 2001, 113, 2406-2410. (b) Thompson, R. C. Inorg. Chem. 1982, 21, 859-863. (c) Wilshire, J. P.; Sawyer, D. T. J. Am. Chem. Soc. 1978, 100, 3972-3973. (d) Brooks, H. B.; Sicilio, F. Inorg. Chem. 1971, 10, 2530-2534. (6) (a) Shvets, V. A.; Sarichev, M. E.; Kasansky, V. B. J. Catal. 1968, 11, 378-379. (b) Shvets, V. A.; Kazansky, V. B. J. Catal. 1972, 25, 123130. (c) Nikisha, V. V.; Shelimov, B. N.; Shvets, V. A.; Griva, A. P.; Kazansky, V. B. J. Catal. 1973, 28, 230-235. (d) Shelimov, B. N.; Naccache, C.; Che, M. J. Catal. 1975, 37, 279-286. (e) Shelimov, B. N.; Che, M. J. Catal. 1978, 51, 143-149. (7) (a) Prakash, A. M.; Kevan, L. J. Phys. Chem. B 2000, 104, 68606868. (b) Chatterjee, M.; Iwasaki, T.; Hayashi, H.; Onodera, Y.; Ebina, T.; Nagase, T. Chem. Mater. 1999, 11, 1368-1375. (8) (a) Anpo, M.; Che, M.; Fubini, B.; Garrone, E.; Giamello, E.; Paganini, M. C. Top. Catal. 1999, 8, 189. (b) Che, M.; Tench, A. J. AdV. Catal. 1982, 31, 77. (9) Jones, R. D.; Summerville, D. A.; Basolo, F. Chem. ReV. 1979, 79, 139.
