Inorg. Chem. 2010, 49, 5611–5618 5611 DOI: 10.1021/ic100528n
Aerobic Oxidation of Lignin Models Using a Base Metal Vanadium Catalyst Susan K. Hanson,† R. Tom Baker,*,†,§ John C. Gordon,*,† Brian L. Scott,‡ and David L. Thorn*,† †
Chemistry Division and ‡Materials Physics Applications Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545. § Current address: Department of Chemistry, and Centre for Catalysis Research and Innovation, University of Ottawa, Ottawa, ON Canada K1N6N5
Received March 19, 2010
Dipicolinate vanadium(V) complexes oxidize lignin model complexes pinacol monomethyl ether (A), 2-phenoxyethanol (B), 1-phenyl-2-phenoxyethanol (C), and 1,2-diphenyl-2-methoxyethanol (D). With substrates having C-H bonds adjacent to the alcohol moiety (B-D), the C-H bond is broken in pyridine-d5 solvent, yielding 2-phenoxyacetaldehyde from B, 2-phenoxyacetophenone from C, and benzoin methyl ether from D. In DMSO-d6 solvent the reaction is slower, and both C-H and C-C bond cleavage products are observed for D. The vanadium(IV) products of these reactions have been identified and characterized. Catalytic oxidation of C and D has been demonstrated using air and (dipic)V(O)OiPr. For both substrates, the C-C bond between the alcohol and ether groups is broken in the catalytic oxidation. 1-Phenyl-2-phenoxyethanol is oxidized to a mixture of phenol, formic acid, benzoic acid, and 2-methoxyacetophenone. The products of oxidation of 1,2-diphenyl-2-methoxyethanol depend on the solvent; in DMSO benzaldehyde and methanol are the major products, while benzoic acid and methyl benzoate are the major products obtained in pyridine solvent. Phenyl substituents on the model complex facilitate the oxidation, with relative rates of oxidation D > C > B.
Introduction The development of alternatives to petroleum-based fuels and chemicals is becoming increasingly urgent because of concerns over climate change, growing world energy demand, and energy security issues. Biomass is the only renewable carbon feedstock available, and thus much recent effort has focused on developing technologies that convert biomass into chemicals and fuels.1,2 The majority of non food-derived biomass is in the form of lignocellulose, which is often not fully utilized because of difficulties associated with breaking down both lignin and cellulose.3 Recently, a number of methods have been reported to transform cellulose directly into more valuable materials such as glucose,4 sorbitol,5 *To whom correspondence should be addressed. E-mail: rbaker@ uottawa.ca (R.T.B.),
[email protected] (J.C.G.),
[email protected] (D.L.T.). (1) Dodds, D. R.; Gross, R. A. Science 2007, 318, 1250–1251. (2) Huber, G. W.; Iborra, S.; Corma, A. Chem. Rev. 2006, 106, 4044–4098. (3) Zhang, Y. H. P. J. Ind. Microbiol. Biotechnol. 2008, 35, 367–375. (4) Li, C.; Zhao, Z. K. Adv. Synth. Catal. 2007, 349, 1847–1850. (5) Fukuoka, A.; Dhepe, P. L. Angew. Chem., Int. Ed. 2006, 45, 5161– 5163. (6) Mascal, M.; Nikitin, E. B. Angew. Chem., Int. Ed. 2008, 47, 7924–7926. (7) Ji, N.; Zhang, T.; Zheng, M.; Wang, A.; Wang, H.; Wang, X.; Chen, J. G. Angew. Chem., Int. Ed. 2008, 47, 8510–8513. (8) Deng, H.; Lin, L.; Sun, Y.; Pang, C.; Zhuang, J.; Ouyang, P.; Li, Z.; Liu, S. Catal. Lett. 2008, 126, 106–111. (9) Kleinert, M.; Barth, T. Energy Fuels 2008, 22, 1371–1379. (10) Yan, N.; Zhao, C.; Dyson, P. J.; Wang, C.; Liu, L.; Kou, Y. ChemSusChem 2008, 1, 626–629. (11) Pu, Y.; Jiang, N.; Ragauskas, A. J. J. Wood. Chem. Technol. 2007, 27, 23–33.
r 2010 American Chemical Society
5-(chloromethyl)furfural,6 and ethylene glycol.7 Less progress has been made with selective transformations of lignin,8-11 which is typically treated in paper and forest industries by kraft pulping or incineration.12,13 Currently, more than 40 million tons of lignin are produced annually worldwide, roughly 95% of which is burned for energy.14 In coming years, as technologies which employ fermentation or other selective conversions of cellulose and hemicellulose are implemented, production of lignin is anticipated to increase dramatically. The discovery of new methods to convert lignin directly to value-added products would allow for more efficient use of this natural resource. Lignin is a randomized polymer containing methoxylated phenoxy propanol units.15 A number of different linkages occur naturally; one of the most prevalent is the β-O-4 linkage shown in Figure 1,15 containing a C-C bond with 1,2hydroxy ether substituents. We envisioned that a selective oxidative cleavage of this carbon-carbon bond could be used to break lignin into smaller, more utilizable components, for example, aromatic product streams that could feed into existing industrial markets. Air (oxygen) would be the ideal oxidant for this transformation. While the oxidative C-C bond cleavage of 1,2-diols is well established for a variety of metals, including vanadium, iron, (12) Lora, J. H.; Glasser, W. G. J. Polym. Environ. 2002, 10, 39–48. (13) Chakar, F. S.; Ragauskas, A. J. Ind. Crop. Prod. 2004, 20, 131–141. (14) Kleinert, M.; Barth, T. Chem. Eng. Technol. 2008, 31, 736–745. (15) Reale, S.; di Tullio, A.; Spreti, N.; de Angelis, F. Mass Spec. Rev. 2004, 23, 87–126.
Published on Web 05/21/2010
pubs.acs.org/IC
5612 Inorganic Chemistry, Vol. 49, No. 12, 2010
Hanson et al.
Figure 1. β-O-4 linkage and lignin model complexes.
Figure 2. X-ray Structures of 3 and 5 (thermal ellipsoids at 50% probability, H atoms omitted for clarity). Selected bond lengths (A˚) for 3: V1-O1 = 1.568(2), V1-O7 = 1.775(2), V1-O6 = 2.547(2), V1-O4 = 1.929(2), V1-O2 = 1.949(2), V1-N1 = 2.060(3). For 5: V1-O1 = 1.576(2), V1-O6 = 1.784(2), V1-O7 = 2.438(2), V1-O2 = 1.925(2), V1-O4 = 1.931(2), V1-N1 = 2.053(2).
manganese, ruthenium, and polyoxometalate complexes,16-20 less is known about oxidative C-C bond cleavage of 1,2hydroxyethers.21 Permanganate and tribromide, both common reagents for C-C bond cleavage of vicinal diols, do not react with 1,2-hydroxyethers to break the carbon-carbon bond.22-25 Instead, oxidation of only the alcohol group (yielding 1,2-carbonyl-ethers) is observed, and different mechanistic pathways have been proposed for the reactions of (16) Riano, S.; Fernandez, D.; Fadini, L. Catal. Commun. 2008, 9, 1282– 1285. (17) Barroso, S.; Blay, G.; Fernandez, I.; Pedro, J. R.; Ruiz-Garcia, R.; Pardo, E.; Lloret, F.; Munoz, M. C. J. Mol. Catal. A. 2006, 243, 214–220. (18) Takezawa, E.; Sakaguchi, S.; Ishii, Y. Org. Lett. 1999, 1, 713–715. (19) Khenkin, A. M.; Neumann, R. J. Am. Chem. Soc. 2008, 130, 14474– 14476. (20) Felthouse, T. R. J. Am. Chem. Soc. 1987, 109, 7566–7568. (21) Baciocchi, E.; Bietti, M.; Putignani, L.; Steenken, S. J. Am. Chem. Soc. 1996, 118, 5952–5960. (22) Bhatia, I.; Banerji, K. K. J. Chem. Soc. Perkin Trans. II 1983, 1577–1580. (23) Goswami, G.; Kothari, S.; Banerji, K. K. Proc. Indian Acad. Sci.(Chem. Sci.) 2001, 113, 43–54. (24) Gosain, J.; Sharma, P. K. Proc. Indian Acad. Sci. (Chem. Sci.) 2003, 115, 135–145. (25) Bhatt, M.; Sharma, P. K.; Banerji, K. K. Indian J. Chem. 2002, 41B, 826–831. (26) For chromic acid, Pb(OAc)4, and Co(OAc)3, different rates of oxidation of glycols and their mono-ethers have been used to support a pathway for glycol C-C bond cleavage involving a chelating diolate complex. However, the products of the glycol mono-ether oxidations were not characterized. See: Trahanovsky, W. S.; Gilmore, J. R.; Heaton, P. C. J. Org. Chem. 1973, 38, 760–763. (27) Littler, J. S.; Waters, W. A. J. Chem. Soc. 1960, 2767–2772. (28) Morimoto, T.; Hirano, M. J. Chem. Soc., Perkin Trans. 2 1982, 1087– 1090. (29) Hintz, H. L.; Johnson, D. C. J. Org. Chem. 1967, 32, 556–564. (30) VO2þ has been reported to react with pinacol monomethyl ether, but the organic products of this reaction were not characterized: Jones, J. R.; Waters, W. A.; Littler, J. S. J. Chem. Soc. 1961, 630–633.
vicinal diols and their monoethers.22-28 In contrast, cerium(IV) has been shown to break the C-C bond of 2-methoxycyclohexanol to give adipaldehyde (1,6-hexane-dial) and methanol.29,30 N-iodosuccinimide is also an effective reagent for C-C bond cleavage of 1,2-hydroxy ethers (yielding carbonyl compounds and acetals),31 but no catalytic version of this reaction has previously been reported.32 The use of an earth-abundant (non-precious) metal catalyst and air as an oxidant would represent a significant advance in our ability to break carbon-carbon bonds in 1,2-hydroxyether compounds such as lignin. We now report that dipicolinate vanadium complexes oxidize the lignin model complexes pinacol monomethyl ether, 2-phenoxyethanol, 1-phenyl-2-phenoxyethanol, and 1,2-diphenyl-2-methoxyethanol.33 Both C-H and C-C cleavage modes have been observed in these substrates. Catalytic aerobic oxidative C-C bond cleavage has been demonstrated in 1-phenyl-2phenoxyethanol and 1,2-diphenyl-2-methoxyethanol. To the best of our knowledge, these are the first examples of catalytic aerobic C-C bond cleavage of 1,2-hydroxyether compounds. Results and Discussion Initial investigations aimed to gain insight into the stoichiometric reactivity of the 1,2-hydroxyether lignin models (31) McDonald, C. E.; Holcomb, H.; Leathers, T.; Ampadu-Nyarko, F.; Frommer, J. Tetrahedron Lett. 1990, 31, 6283–6286. (32) Electrolysis of 1,2-hydroxyethers and 1,2-diethers has been reported to give C-C bond cleavage: Shono, T.; Matsumura, Y.; Hashimoto, T.; Hibino, K.; Hamaguchi, H.; Aoki, T. J. Am. Chem. Soc. 1975, 97, 2546–2548. (33) A non-oxidative selective C-O bond cleavage of lignin model complexes catalyzed by vanadium was very recently reported. See: Son, S.; Toste, D. Angew. Chem., Int. Ed. 2010, 49, 3791-3794.
Article
with (dipic)VV. Reaction of (dipic)VV(O)OiPr (1a) or (dipic)VV(O)OEt (1b) with pinacol monomethyl ether (A), 2-phenoxyethanol (B), 1-phenyl-2-phenoxyethanol (C), or 1,2-diphenyl-2-methoxyethanol (D)34 in acetonitrile solvent yielded new vanadium(V) complexes where the alcoholether ligand was bound in a chelating fashion. Complexes (dipic)VV(O)(pinOMe) (2), (dipic)VV(O)(OPE) (3), (dipic)VV(O)(OPPE) (4) and (dipic)VV(O)(DPME) (5) (pinOMe = 2,3dimethyl-3-methoxy-2-butanoxide; OPE = 2-phenoxyethoxide; OPPE = 1-phenyl-2-phenoxyethoxide; DPME = 1,2-diphenyl-2-methoxyethoxide), were isolated in good yields (39-88%) and characterized by NMR and IR spectroscopy, elemental analysis, and X-ray crystallography (for 2, 3, and 5). The X-ray structures of 2 (Supporting Information, Figure S1), 3 (Figure 2), and 5 (Figure 2), display similar vanadium oxo bond distances of 1.573(2), 1.568(2), and 1.576(2) A˚, respectively. In each case the substrate binds in a chelating manner, similar to the previously reported vanadium(V) pinacolate complex (dipic)V(O)(pinOH).35 When pinacol monomethyl ether complex 2 was heated in pyr-d5 solution (6 h at 100 °C, or 3 weeks at 25 °C), the previously reported vanadium(IV) complex (dipic)VIV(O)(pyr)2 (6),35 acetone, 2-methoxypropene, and pinacol monomethyl ether were formed. Yields of the organic products were determined by integration of the 1H NMR spectrum against an internal standard ( p-xylene); acetone, 2-methoxypropene, and pinacol monomethyl ether were formed in 94%, 76%, and 92% of the theoretical maximum yields based on the oxidizing equivalents of vanadium consumed, suggesting the stoichiometry shown in eq 1.
In contrast to compound 2, all the compounds having secondary C-H bonds on the coordinated alcohol moiety underwent oxidative C-H bond cleavage under stoichiometric conditions in pyridine solution. Heating 2-phenoxyethoxide complex 3 in pyr-d5 solution (10 min at 100 °C) afforded 6, 2-phenoxyacetaldehyde (78%), and 2-phenoxyethanol (98%), suggesting the stoichiometry shown in eq 2. The formation of 2-phenoxyacetaldehyde was confirmed by comparison with an authentic sample prepared according to a literature procedure.36
When dissolved in pyridine-d5 solution, 1-phenyl-2-phenoxyethoxide complex 4 reacted rapidly at room temperature (10 min) to form 6 and a 1:1 ratio of 2-phenoxyacetophenone (34) Prepared as an 85:15 mixture of u (R,S þ S,R):l (S,S þ R,R) diastereomers by the acid-catalyzed ring opening of trans-stilbene oxide in CH3OH (see Experimental Section for details). (35) Hanson, S. K.; Baker, R. T.; Gordon, J. C.; Scott, B. L.; Sutton, A. D.; Thorn, D. L. J. Am. Chem. Soc. 2009, 131, 428–429. (36) Speranza, G.; Mueller, B.; Orlandi, M.; Morelli, C. F.; Manitto, P.; Schink, B. Helv. Chim. Acta 2003, 86, 2629–2636.
Inorganic Chemistry, Vol. 49, No. 12, 2010
5613
and 1-phenyl-2-phenoxyethanol (eq 2). 2-Phenoxyacetophenone was prepared independently according to a published procedure37 and spiked into the reaction mixture, confirming the identity of this product. Complex 5 also underwent C-H bond cleavage when dissolved in pyr-d5 at room temperature, reacting completely within 10 min to form 6 and a 1:1 ratio of benzoin methyl ether (2-methoxy-2-phenylacetophenone) and 1,2-diphenyl-2-methoxyethanol (eq 3)
However, in the absence of pyridine, conversion of 5 was significantly slower, and in addition to C-H bond cleavage, C-C bond cleavage was observed even for compounds having secondary C-H bonds on the coordinated alcohol moiety. Heating a DMSO-d6 solution of complex 5 (30 min at 100 °C) resulted in a mixture of C-H and C-C bond cleavage products (eq 4). The organic products of this reaction consisted of benzoin methyl ether (30%), coproducts methanol and benzaldehyde (66%), and 1,2-diphenyl2-methoxyethanol (100%). The green vanadium product of this reaction was the new vanadium(IV) complex (dipic)VIV(O)(DMSO)2 (7). Complex 7 showed a VdO stretch in the IR spectrum at 948 cm-1 and was characterized by X-ray crystallography (Figure 3). The VdO bond length in 7 (1.605(2) A˚) is similar to that observed in 6 (1.611(2) A˚).35 Complex 7 could be independently prepared by the thermolysis of the pinacolate complex (dipic)V(O)(Hpin)35 in DMSO solvent and was isolated in 83% yield.
The mechanism of the C-C cleavage reaction depicted in eq 4 is currently under investigation, but one possible pathway involves initial reaction of 5 to form benzaldehyde and a methoxybenzyl radical. Subsequent reaction of the methoxybenzyl radical with water and a second equivalent of 5 would form methanol, a second equivalent of benzaldehyde, and release 1,2-diphenyl-2-methoxyethanol.38 Similar radical pathways have been proposed for the cleavage reactions of (37) Bu, X.; Jing, H.; Wang, L.; Chang, T.; Jin, L.; Liang, Y. J. Mol. Cat. A. 2006, 259, 121–124. (38) Attempts to test this hypothesis by the addition of radical traps to the reaction were inconclusive. Complex 5 reacted directly with BHT (2,6-ditert-butyl-4-methylphenol) and PPh3 to give products that were not further characterized. When 9,10-dihydroanthracene was added to the reaction mixture, only trace (
