C-C or C-O bond cleavage in a phenolic lignin model compound: Selectivity depends on vanadium catalyst

Article abstract of DOI:10.1002/anie.201107020

The aerobic oxidation of a phenolic lignin model compound with a vanadium catalyst results in the oxidative cleavage of the C-C bond between the aryl ring and the adjacent hydroxy-substituted carbon atom (see scheme). Labeling experiments indicate key mechanistic differences to a previously reported related C-O bond cleavage reaction. The selectivity in C-C versus C-O bond cleavage depends on the choice of the vanadium catalyst. Copyright

Full text of DOI:10.1002/anie.201107020

.
Angewandte
Communications
DOI: 10.1002/anie.201107020
Oxidative Cleavage
À
À
C C or C O Bond Cleavage in a Phenolic Lignin Model Compound:
Selectivity Depends on Vanadium Catalyst**
Susan K. Hanson,* Ruilian Wu, and Louis A. “Pete” Silks*
In light of diminishing fossil fuel reserves and increasing
demand for energy, nonfood biomass (lignocellulose) is an
attractive feedstock for the production of renewable chem-
icals and fuels.[1] Although it has traditionally been considered
an intractable by-product of the pulp and paper industries,
lignin is receiving increasing attention as a potential precursor
for valuable aromatic compounds.[2–4] Lignin is a polymer
composed of methoxy-substituted phenyl and phenolic sub-
units, and the complexity of its irregular structure poses
a significant challenge to its use as a renewable feedstock.[5,6]
However, a number of recent reports suggest that using
a homogeneous catalyst for reduction or oxidation of lignin
could be an effective way to depolymerize lignin in a selective
fashion.[2,7–9] In principle, the selectivity could be tuned by the
choice of the transition-metal catalyst and the ligand. For
À
Scheme 1. a) Non-oxidative C O bond cleavage reported by Son and
instance, Sergeev and Hartwig recently reported that a nickel
Toste.[12] b) Vanadium catalysts for the oxidation of lignin models.
carbene complex is an effective catalyst for hydrogenolysis of
aryl ether bonds in several lignin model compounds,[10] and
Bergman and co-workers found a ruthenium xantphos
À
catalyst that cleaves the C O bond in a lignin model
Herein, we present reactivity studies of a phenolic lignin
model compound 1b with catalysts 4 and 5. Remarkably, the
two vanadium catalysts show different selectivities for the
aerobic oxidation of 1b. Vanadium catalyst 5 mediates a new
compound by transfer hydrogenation.[11]
Catalytic aerobic oxidation is also appealing for the
production of fine chemicals from lignin, as this strategy
would not require added stoichiometric reagents, and would
afford water as the only by-product of oxidation.[7,8] Use of an
earth-abundant metal as a catalyst could provide additional
benefits in terms of availability and cost. Toward this
objective, Son and Toste recently reported a vanadium
À
type of reaction in which the C C bond between the aryl ring
and the adjacent alcohol group is broken to give 2,6-dimeth-
oxybenzoquinone (6) and acrolein derivative 7, and ketone
À
8b. In contrast, catalyst 4 affords C O bond cleavage
products 2b and 3, and ketone 8b. The significant difference
in selectivity between the two catalysts underscores the
potential of homogeneous catalysts for controlling the
selectivity in the aerobic oxidation of lignin.
À
catalyzed C O bond cleavage reaction in the lignin model
compound 1a, which afforded alkene 2a and 2-methoxyphe-
nol (3) as the major products (Scheme 1a).[12] Although the
overall reaction is redox-neutral, optimal activity was ach-
ieved by the use of catalyst 4 (10 mol%) under aerobic
conditions.[12] As an alternative, we have investigated vana-
We were interested in investigating the reactivity of
vanadium catalysts with the phenolic lignin model compound
1b, since prior work had focused on non-phenolic lignin
model compounds,[12,13] and it was envisioned that the
phenolic group could influence the course of the oxidation
reaction. Phenolic groups are abundant in lignin: spruce lignin
is estimated to have 15–30 free phenolic groups per 100
phenylpropanoid units.[14,15] The lignin model compound
[13C2]-1b was prepared as a 4:1 mixture of diastereomers
(erythro:threo).
À
dium complexes as catalysts for oxidative C C bond cleavage
of lignin models,[13] and recently found that the 8-quinolinate
complex 5 catalyzes the aerobic oxidation of benzylic and
allylic alcohols when a basic additive is used.[13c]
[*] Dr. S. K. Hanson, Dr. R. Wu, Dr. L. A. “Pete” Silks
Chemistry and Bioscience Divisions, Los Alamos National Labo-
ratory
The reactivity of [13C2]-1b was then tested with the 8-
quinolinate vanadium complex 5. Unexpectedly, the reaction
of [13C2]-1b with 5 (1 equiv) in [D5]-pyridine under air at 808C
MS J582, Los Alamos 87545 (USA)
E-mail: skhanson@lanl.gov
[**] This work was supported by Los Alamos National Laboratory LDRD
(20100160ER). We would also like to thank the NSF Center for
Enabling New Technologies through Catalysis.
À
afforded coproducts derived from cleavage of the C(alkyl)
C(phenyl) bond, 2,6-dimethoxybenzoquinone (6, 40%) and
acrolein derivative [13C2]-7 (38%), as well as ketone [13C2]-8b
(20%; Scheme 2).[16] The 1H NMR spectrum (CDCl3) of
acrolein derivative [13C2]-7 shows a characteristic resonance
Supporting information for this article (including synthetic proce-
dures, experimental details, and NMR spectra of reaction products)
is available on the WWW under http://dx.doi.org/10.1002/anie.
201107020.
2
for the aldehyde proton at 9.46 ppm (d, JCÀH = 32 Hz), and
3410
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3410 –3413

Angewandte
Chemie
13
À
À
Scheme 2. Oxidation of phenolic lignin model compound [ C2]-1b with vanadium catalysts 4 (C O bond cleavage) and 5 (C C bond cleavage).
the 13C NMR spectrum of [13C2]-7 shows resonances at 158.4
and 107.4 ppm (1JCÀC = 80 Hz), which are consistent with the
chemical shifts of an enol ether.
consumed and an intractable mixture of products was
obtained when an independently synthesized sample of
ketone 8b was heated with vanadium catalyst 5 (1 equiv)
under air in [D5]-pyridine (48 h, 808C). Likewise, attempts to
isolate acrolein product 7 showed that it was not stable when
stored as a solid at room temperature for several weeks. MS
analysis of the catalytic reaction mixture showed the presence
of several higher molecular weight species, which potentially
originated from oligomerization of reaction intermediates or
unstable products. Competing polymerization has previously
been reported in the electrochemical oxidation of several
phenolic lignin model compounds.[24]
The formation of acrolein derivative 7 was surprising, as
À
the cleavage of the C(alkyl) C(phenyl) bond is a rather
unusual reaction. Vanadium complexes have been reported to
[17]
À
oxidatively break the C C bonds between vicinal diols, a-
hydroxyketones,[18] and a-hydroxyethers,[13a] as well as medi-
ate the decarboxylation of 1,2- and 1,3-hydroxycarboxylic
acids,[19] but vanadium-mediated cleavage of a benzylic C C
À
bond has not been reported. In a closely related reaction,
À
cobalt–salen complexes have been shown to break C C bonds
À
in several phenolic compounds, including isoeugenol, 3-
The unusual cleavage reaction of the C(alkyl) C(phenyl)
methoxy-4-hydroxyacetophenone, and 1b.[20–22] For cobalt,
bond observed in phenolic lignin model [13C2]-1b with
vanadium complex 5 differed significantly from the non-
À
the C C bond cleavage is proposed to proceed through
À
a superoxocobalt intermediate, which likely abstracts a hydro-
gen atom from the substrate to generate a cobalt-bound
oxygen-centered radical.[23,23]
oxidative C O bond cleavage in non-phenolic lignin model
1a reported for Tosteꢀs catalyst 4.[12] Consequently, we
wondered if the incorporation of the phenolic functional
group into the lignin model was responsible for the difference
in reactivity. To gain more insight into the influence of the
phenolic group, we tested the reaction of phenolic lignin
model [13C2]-1b with Tosteꢀs catalyst 4. Surprisingly, when
[13C2]-1b was heated with 4 (1 equiv) under air (808C, 48 h) in
Encouraged by the observation of this new reaction mode,
we tested the catalytic aerobic oxidation of [13C2]-1b by using
catalyst 5. Phenolic lignin model [13C2]-1b was heated under
air with 5 (10 mol%) in [D5]-pyridine at 808C. Complete
consumption of [13C2]-1b was observed after 48 hours at
808C; the reaction afforded 6 (30%), [13C2]-7 (34%), and
CD3CN, complete conversion of [13C2]-1b was observed and
13
13
À
À
[ C2]-8b (27%), thus demonstrating that the C(alkyl)
a mixture of C O bond cleavage coproducts [ C2]-2b (28%)
C(phenyl) bond could be broken catalytically. The C C
bond cleavage reaction also occurred in ethyl acetate, [D8]-
THF, or 2-methyltetrahydrofuran (2-MeTHF) by using
and 3, and ketone [13C2]-8b (43%) was obtained. Formation
of acrolein product [13C2]-7 was not observed, thus indicating
that the reaction selectivity was dependent on the specific
vanadium catalyst, and not merely changing because of the
incorporation of the phenolic group into the substrate. A
similar product distribution ([13C2]-2b, 30%; [13C2]-8b, 41%)
was obtained when the reaction of [13C2]-1b with 4 was carried
out in [D5]-pyridine (808C, 48 h), thus suggesting that the
observed selectivity was not significantly affected by the
presence of the basic pyridine. The catalytic oxidation of
À
a
combination of catalyst
5
(10 mol%) and NEt3
(10 mol%). When the reaction was carried out in ethyl
acetate (48 h, 808C), formation of 6 (44%), [13C2]-7 (45%),
and [13C2]-8b (14%) was observed. Comparable yields of 6
(26%), [13C2]-7 (45%), and [13C2]-8b (11%) were obtained in
2-MeTHF.
1
Although the H NMR spectrum ([D5]-pyridine) of the
catalytic reaction mixture did not show distinct signals that
could be attributed to unidentified products, the 13C NMR
spectrum showed two additional broad signals at 85.8 and
62.0 ppm, and examination of the mass balance for the overall
reaction indicated that not all material was accounted for. We
postulated that the low mass balance observed in the reaction
of 1b could be due to instability of the products, which could
polymerize or further react under the experimental condi-
tions. Consistent with this idea, 40% of the material was
[13C2]-1b by Tosteꢀs complex 4 (10 mol%) in [D5]-pyridine
13
À
also afforded C O bond cleavage products [ C2]-2b (32%)
and 3, and [13C2]-8b (44%).[25] The 13C NMR spectrum
(CDCl3) of compound [13C2]-2b showed characteristic reso-
nances at 132.2 and 129.7 ppm (1JCÀC = 69 Hz).
In an effort to increase the yields of monomeric products,
a mixture of catalysts 4 (5 mol%) and 5 (5 mol%), and NEt3
(10 mol%) was evaluated for the oxidation of [13C2]-1b in
ethyl acetate. Complete consumption of [13C2]-1b was
Angew. Chem. Int. Ed. 2012, 51, 3410 –3413
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3411

.
Angewandte
Communications
observed after heating the mixture at 808C for 48 hours, and
À
À
a mixture of C O (2b (24%), 3) and C C bond cleavage
products (5 (36%), 6 (30%)), and ketone 8b (14%) was
obtained. [13C1]-formic acid (4%) was also detected by 1H and
13C NMR spectroscopy. Although the yields of monomeric
products were slightly increased by using a combination of
both catalysts, the mass balance (ca. 74% for the 13C-labeled
portion of the molecule) and the presence of a brown
precipitate suggested that some undesired side reactions still
occurred.
Intrigued by the selectivity difference between the two
vanadium catalysts, we aimed to gain insight into the reaction
mechanisms that might underlie the observed selectivities.
Scheme 4. Oxidation of non-phenolic lignin model compound [13C2]-1c
with vanadium catalyst 5.
À
Previously, a one-electron pathway was proposed for the C O
bond cleavage reaction reported by Toste, in which the
benzylic hydrogen is abstracted to generate a vanadium-
bound ketyl radical intermediate.[12] To further examine the
role of the benzylic C H bond in the C O and C C bond
cleavage reactions, the lignin model compound [13C2,D1]-1b
was prepared with a deuterium incorporated into the benzylic
position.
observed and ketone [13C2]-8c (55%) and dehydrated ketone
À
À
À
13
À
[ C2]-9 (10%) were obtained. No C C bond cleavage
products were observed, thus indicating that the phenolic
À
group plays an important role in the C(alkyl) C(phenyl)
Compared to the reaction of [13C2]-1b, the reaction of
[13C2,D1]-1b with Tosteꢀs catalyst 4 (1 equiv) in CD3CN
(808C, 48 h) was found to be qualitatively slower. When the
reactions of complex 4 with [13C2]-1b and [13C2,D1]-1b were
monitored side-by-side, only 56% of [13C2,D1]-1b had reacted
after 18 hours, as compared to 83% of [13C2]-1b. A conversion
of 87% was observed for [13C2,D1]-1b after 72 hours at 808C,
and 2b (14%), 3, and 8b (35%) were obtained. Dimethox-
ybenzoquinone (6) was also a minor product of the reaction
(8%). The slower reaction rate observed for [13C2,D1]-1b and
the increase in the yield of 8b relative to 2b implicate
bond cleavage pathway. C O bond cleavage product [ C2]-2c
13
À
was also not detected in the reaction mixture, thus suggesting
that the formation of 2 and 8 proceeds by distinct reaction
pathways.
Key differences emerge in the selectivity of vanadium
À
catalysts 4 and 5. While Tosteꢀs catalyst 4 breaks the C O
bond in lignin model 1b through a pathway that involves
À
cleavage of the benzylic C H bond, quinolinate complex 5
À
breaks the C(alkyl) C(phenyl) bond through a pathway that
À
does not involve the benzylic C H bond. A phenoxy group is
required for the C C bond cleavage reaction, which may
proceed through a phenoxy radical intermediate, as has been
À
À
cleavage of the benzylic C H bond duringthe formation of
both 2b and 8b.
À
proposed for related cobalt-mediated C C bond cleavage
A change in product distribution was also observed upon
reactions.[21–24]
oxidation of [13C2,D1]-1b with quinolinate catalyst
5
Overall, the remarkably different selectivities observed
for the two vanadium catalysts show the complexity of
vanadium-mediated oxidations and support the viability of
homogeneous catalysts for controlling selectivity in the
aerobic oxidation of lignin. Efforts to explore vanadium-
mediated oxidation of other types of lignin substructures are
currently under way.
(1 equiv). When the reaction of [13C2,D1]-1b with 5 was
carried out under air in [D5]-pyridine (808C, 48 h; Scheme 3),
the products consisted of acrolein derivative [13C2,D1]-7
Received: October 4, 2011
Revised: December 22, 2011
Published online: January 20, 2012
Scheme 3. Oxidation of [13C2,D1]-1b with vanadium catalyst 5.
Keywords: cleavage reactions · isotopic labeling · lignin ·
oxidation · vanadium
.
(57%) and 2,6-dimethoxybenzoquinone 6 (54%), with less
than 3% formation of [13C2]-8b. Both the retention of the
deuterium label in the product [13C2,D1]-7 and the increased
yield of 7 relative to 8b suggest that cleavage of the benzylic
[1] a) A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411 –
2502; b) D. R. Dodds, R. A. Gross, Science 2007, 318, 1250 –
1251.
[2] a) J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius, B. M. Weck-
huysen, Chem. Rev. 2010, 110, 3552 – 3599; b) S. R. Collinson, W.
Thielemans, Coord. Chem. Rev. 2010, 254, 1854 – 1870.
[3] a) C. Crestini, M. Crucianelli, M. Orlandi, R. Saladino, Catal.
Today 2010, 156, 8 – 22.
À
C H bond is not involved in the formation of 7.
With subsequent experiments, we aimed to establish
whether the phenolic group is necessary for the oxidative
À
C C bond cleavage reaction. When non-phenolic lignin
model [13C2]-1c was heated under air with 5 (10 mol%) in
[D5]-pyridine (808C, 48 h; Scheme 4), 65% conversion was
3412
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3410 –3413

Angewandte
Chemie
[4] a) M. Kleinert, T. Barth, Energy Fuels 2008, 22, 1371 – 1379;
b) K. Barta, T. D. Matson, M. L. Fettig, S. L. Scott, A. V. Iretskii,
P. C. Ford, Green Chem. 2010, 12, 1640 – 1647; c) N. Yan, C.
Zhao, P. J. Dyson, C. Wang, L.-T. Liu, Y. Kou, ChemSusChem
2008, 1, 626 – 629.
[14] U. Tuor, H. Wariishi, H. E. Schoemaker, M. H. Gold, Biochem-
istry 1992, 31, 4986 – 4995.
[15] E. Alder, Wood Sci. Technol. 1977, 11, 169 – 218.
[16] Yields were determined by integration of the 1H NMR spectrum
against an internal standard.
[5] S. Reale, A. Di Tullio, N. Spreti, F. De Angelis, Mass Spectrom.
Rev. 2004, 23, 87 – 126.
[6] J. Ralph, K. Lundquist, G. Brunow, F. Lu, H. Kim, P. F. Schatz,
J. M. Marita, R. D. Hatfield, S. A. Ralph, J. H. Christensen, W.
Boerjan, Phytochem. Rev. 2004, 3, 29 – 60.
[17] a) J. S. Littler, A. I. Mallet, W. A. Waters, J. Chem. Soc. 1960,
2761 – 2766; b) M. Kirihara, K. Yoshida, T. Noguchi, S. Naito, N.
Matsumoto, Y. Ema, M. Torii, Y. Ishizuka, I. Souta, Tetrahedron
Lett. 2010, 51, 3619 – 3622.
[18] M. Kirihara, S. Takizawa, T. Momose, J. Chem. Soc. Perkin
Trans. 1 1998, 7 – 8.
[19] a) J. R. Jones, W. A. Waters, J. A. Littler, J. Chem. Soc. 1961,
630 – 633; b) I. K. Meier, J. Schwartz, J. Am. Chem. Soc. 1989,
111, 3069 – 3070.
[20] R. S. Drago, B. B. Corden, C. W. Barnes, J. Am. Chem. Soc. 1986,
108, 2453 – 2454.
[21] C. Canevali, M. Orlandi, L. Pardi, B. Rindone, R. Scotti, J. Sipila,
F. Morazzoni, J. Chem. Soc. Dalton Trans. 2002, 3007 – 3014.
[22] a) E. Bolzacchini, C. Canevali, F. Morazzoni, M. Orlandi, B.
Rindone, R. Scotti, J. Chem. Soc. Dalton Trans. 1997, 4695 –
4699; b) E. Bolzacchini, L. B. Chiavetto, C. Canevali, F. Mor-
azzoni, M. Orlandi, B. Rindone, J. Mol. Catal. A 1996, 112, 347 –
351.
[7] a) W. Partenheimer, Adv. Synth. Catal. 2009, 351, 456 – 466; b) R.
DiCosimo, H. C. Szabo, J. Org. Chem. 1988, 53, 1673 – 1679.
[8] a) J. Zakzeski, A. L. Jongerius, B. M. Weckhuysen, Green Chem.
2010, 12, 1225 – 1236; b) K. Stꢁrk, N. Taccardi, A. Bosmann, P.
Wasserscheid, ChemSusChem 2010, 3, 719 – 723.
[9] For representative examples, see: a) M. Nagy, K. David, G. J. P.
Britovsek, A. J. Ragauskas, Holzforschung 2009, 63, 513 – 520;
b) T. Q. Hu, B. R. James, C. L. Lee, J. Pulp Pap. Sci. 1997, 23,
J200 – J205; c) R. Patt, H. J. Mielisch, J. Odermatt, K. Wieghardt,
T. Weyermuller, US patent, US2001/0025695, 2001; d) Y. Cui,
C. L. Chen, J. S. Gratzl, R. Patt, J. Mol. Catal. A 1999, 144, 411 –
417; e) T. Tzedakis, Y. Benzada, M. Comtat, Ind. Eng. Chem.
Res. 2001, 40, 3435 – 3444.
[10] A. G. Sergeev, J. F. Hartwig, Science 2011, 332, 439 – 443.
[11] J. M. Nichols, L. M. Bishop, R. G. Bergman, J. A. Ellman, J. Am.
Chem. Soc. 2010, 132, 12554 – 12555.
[12] S. Son, F. D. Toste, Angew. Chem. 2010, 122, 3879 – 3882; Angew.
Chem. Int. Ed. 2010, 49, 3791 – 3794.
[13] a) S. K. Hanson, R. T. Baker, J. C. Gordon, B. L. Scott, D. L.
Thorn, Inorg. Chem. 2010, 49, 5611 – 5618; b) B. Sedai, C. Dꢂaz-
Urrutia, R. T. Baker, R. Wu, L. A. Silks, S. K. Hanson, ACS
Catal. 2011, 1, 794 – 804; c) S. K. Hanson, R. Wu, L. A. Silks, Org.
Lett. 2011, 13, 1908 – 1911.
[23] A. Zombeck, R. S. Drago, B. B. Corden, J. H. Gaul, J. Am.
Chem. Soc. 1981, 103, 7580 – 7585.
[24] V. L. Pardini, C. Z. Smith, J. H. P. Utley, R. R. Vargas, H.
Viertler, J. Org. Chem. 1991, 56, 7305 – 7313.
[25] Ketone 8b was found to be unstable under the reaction
conditions. When a sample of independently prepared 8b was
heated under air in CD3CN (808C, 48 h) with 4 (1 equiv),
approximately 44% conversion was observed, and an intractable
mixture of products was obtained.
Angew. Chem. Int. Ed. 2012, 51, 3410 –3413
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3413