Straightforward assembly of benzoxepines by means of a rhodium(III)- catalyzed C-H functionalization of o-vinylphenols

Article abstract of DOI:10.1021/ja410538w

Readily available o-vinylphenols undergo a formal (5 + 2) cycloaddition to alkynes when treated with catalytic amounts of [Cp*RhCl₂] ₂ and Cu(OAc)₂. The reaction, which involves the cleavage of the terminal C-H bond of the alkenyl moiety, generates highly valuable benzoxepine skeletons in a practical, versatile, and atom-economical manner. Using carbon monoxide instead of an alkyne as reaction partner leads to coumarin products which formally result from a (5 + 1) cycloaddition.

Full text of DOI:10.1021/ja410538w

Communication
pubs.acs.org/JACS
Straightforward Assembly of Benzoxepines by Means of a
Rhodium(III)-Catalyzed C−H Functionalization of o‑Vinylphenols
Andres Seoane, Noelia Casanova, Noelia Quinones, Jose L. Mascarenas,* and Moises Gulías*
́
́
́
̃
̃
́ ́ ́
Centro Singular de Investigacion en Química Bioloxica e Materiais Moleculares (CIQUS) and Departamento de Química Organica,
Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain
S
* Supporting Information
Our work started by identifying 2-hydroxystyrenes as readily
available substrates that might engage in rhodium-catalyzed
heteroannulations with alkynes via reactions involving a C−H
activation step. At the outset, the regiochemistry of the
potential annulation was quite unpredictable, as a priori there
are three different C−H positions susceptible to activation, and
therefore the reaction might lead to five-, six-, or seven-
membered rings. While the formation of benzofuranes from
phenols using Rh(III) catalysts has not been described,9
precedents in the annulation of naphthols with alkynes pointed
to the formation of chromene-type molecules (B) as a viable
outcome for the reaction (Figure 2).10
ABSTRACT: Readily available o-vinylphenols undergo a
formal (5 + 2) cycloaddition to alkynes when treated with
catalytic amounts of [Cp*RhCl2]2 and Cu(OAc)2. The
reaction, which involves the cleavage of the terminal C−H
bond of the alkenyl moiety, generates highly valuable
benzoxepine skeletons in a practical, versatile, and atom-
economical manner. Using carbon monoxide instead of an
alkyne as reaction partner leads to coumarin products
which formally result from a (5 + 1) cycloaddition.
etal-catalyzed cycloadditions involving the coordination
Mand activation of π-electrons have revolutionized the
way of making cyclic compounds.1 In recent years there have
been an increasing number of reports on a new type of metal-
catalyzed annulations that involve as a key step the activation of
C−H bonds.2 These reactions have provided for the easy
construction of a variety of rings, mainly five- and six-
membered heterocycles, through formal (3 + 2)3 or (4 + 2)4
cycloadditions. Remarkably, the assembly of larger rings by
means of related annulations remains to be developed.5
Herein we describe a new type of heteroannulation involving
a C−H activation process that allows the synthesis of
benzoxepines from extremely simple precursors in a formal
(5 + 2) cycloaddition reaction.6 The benzoxepine skeleton
forms the basic core of many molecules with pharmacological
importance such bauhinoxepin A, bulbophylol B or janoxepin
(see Figure 1),7 and therefore methods that allow their
assembly from readily available precursors are of major
interest.8
Figure 2. Different annulation options for o-vinylphenols using a
Rh(III) catalyst.
Remarkably, reaction of alkyne 2a with 2 equiv of 2-
vinylphenol (1a) in the presence of catalytic amounts of
[Cp*RhCl2]2 (Cp* = pentamethylcyclopentadienyl) and 2.1
equiv of Cu(OAc)2·H2O, in toluene at 100 °C, did not give the
benzofurane- or chromene-type of products but gave the
oxepine 3aa in 52% yield (Table 1, entry 1). Using [Cp*IrCl2]2
instead of the rhodium complex led to the majoritary recovery
of the starting material, while RuCl2(p-cymene)]2 induced the
decomposition of 1a. After screening several solvents, we found
that using acetonitrile instead of toluene leads to a considerable
increase in the yield up to 91%. Finally, we found that carrying
out the reaction under an air atmosphere (balloon) allowed to
decrease the amount of 1a and Cu(OAc)2 to 1.5 equiv and 0.5
equiv, respectively, without compromising the yield (97%, after
1h at 85 °C, entry 7). The loading of Cu(OAc)2 can be
decreased to 10%; however, the reaction is slower (entry 8).
With the optimized conditions in hand we investigated the
scope with regard to the alkyne component (Scheme 1).
Received: October 14, 2013
Figure 1. Representative compounds containing the oxepine core.
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja410538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
Table 1. Optimization of the Reaction
Scheme 2. Reaction with Phenols Equipped with Different
Substituents
a,b
1a
(equiv)
T
(°C)
yield
b
entry
catalyst
solvent
(%)
1
2
3
4
5
6
7
8
9
[Cp*RhCl2]2
[Cp*IrCl2]2
[RuCl2(p-cymene)]2
[Cp*RhCl2]2
[Cp*RhCl2]2
[Cp*RhCl2]2
[Cp*RhCl2]2
[Cp*RhCl2]2
none
2
toluene
toluene
toluene
DMF
100
100
100
100
100
85
52
0
2
2
traces
72
2
2
tAmylOH
CH3CN
CH3CN
CH3CN
CH3CN
83
2
91
c
1.5
1.5
1.5
85
97
87
0
d
85
85
a
0.33 mmol of 2a, 2 mL of solvent, 2.1 equiv of Cu(OAc)2·H2O.
b
c
Isolated yield of 3aa (based on 2a). 0.5 equiv Cu(OAc)2·H2O/air
d
balloon. 0.1 equiv of Cu(OAc)2·H2O/air balloon, 16 h.
a
ab
,
Reaction conditions: 0.33 mmol of 2, 0.50 mmol of 1, [Cp*RhCl2]2
(2.5 mol %), 0.5 equiv Cu(OAc)2·H2O, 2 mL of CH3CN at 85 °C, air
Scheme 1. Scope with Respect to the Alkyne Component
b
balloon, overnight. Isolated yield (based on 2).
phenolic substrates (1b−1l), when there are no commercial
sources, are easily assembled from the corresponding
salicylaldehydes through a Wittig reaction with a methylene-
phosphorous ylide.
We first investigated the influence of the substituent in the
position para to the hydroxyl group. As shown in Scheme 2, the
reaction works in substrates bearing substituents with either
electron-donating or electron-withdrawing properties, including
methoxy, methyl, phenyl, chloro, bromo, or ester groups, and
the oxepine products are obtained in good to excellent yields
(3ba−3ga, 60−89%) Moreover, substituents in the position
meta to the hydroxyl group such as methoxy, methyl, bromide,
and trifluoromethyl (phenols 1h−1k) or in ortho (1l), are also
tolerated, and the products 3ha−3la are isolated with good
yields. Finally, reaction of bromo-substituted vinyphenol (1f)
with the asymmetric alkyne 2g led to the corresponding
product 3fg in 68% yield as a single regioisomer.
a
Interestingly, the reaction does not proceed in substrates
with alkyl substituents at the terminal position of the alkene
such as (E)-2-(prop-1-en-1-yl)phenol, which decomposed
under the reaction conditions.
Reaction conditions: 0.33 mmol of 2, 0.50 mmol of 1a, [Cp*RhCl2]2
(2.5 mol %), 0.5 equiv Cu(OAc)2·H2O, 2 mL of CH3CN at 85 °C, air
b
balloon, overnight. Isolated yield based on 2.
Competition experiments revealed that the phenol 1g reacts
preferentially to 1a when mixed together with the alkyne 2a
under the standard reaction conditions (Scheme 3). However,
in separate experiments we found that 1a reacts faster than
electron-deficient substrates 1g of 1k.11 This divergence could
be explained in terms of an initial and irreversible formation of
a phenoxide−Rh complex, as this might be easier for the more
acidic substrate 1g; however, the subsequent C−H activation
step might be more favorable for the more electron-rich
substrate 1a.
Symmetrical alkynes bearing electron-rich or electron-deficient
aryl substituents (2b and 2c) led to the expected products 3ab
and 3ac in good yields (85% and 71%). Dialkyl-substituted
alkynes, such as 2d, 2e, and 2f, also participate in the process,
although the yields are slightly lower (52−65%).
Interestingly, with unsymmetrical aryl-alkyl alkynes the
reaction takes place with high regioselectivity, leading to
products in which the phenyl group is in the carbon tethered to
the oxygen group of the product. Thus, alkyne 2g afforded the
product 3ag in excellent yield and regioselectivity (14:1).
Similar results were obtained with an alkynylester (3ah, 99%
yield, 8:1 regioselectivity) or with an alkyne bearing a hydroxy
group like 2i (65% yield, 11:1).
Intermolecular competition experiments between 1a and the
deuterated analogue 1a-d2, demonstrated a kinetic isotope
effect (kH/kD = 2.7), which suggests that the C−H bond
cleavage is involved in the rate-limiting step.
The reaction is compatible with a wide variety of substituents
in the aryl group of the vinylphenol (Scheme 2). The required
Of mechanistic relevance, treatment of allylphenol 4 with
diphenylacetylene under the standard conditions led to
B
dx.doi.org/10.1021/ja410538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
the re-aromatized intermediate (II). From intermediate II the
process should involve alkyne coordination followed by
migratory insertion to give intermediate III, which evolves
through reductive elimination to the final product and a Rh(I)
species which is then reoxidized to enter a new catalytic cycle.
Consistent with the formation of rhodacycle II, we found
that treatment of the o-alkenyl phenols with carbon monoxide
(balloon pressure) under the reaction conditions produces
highly appealing coumarin products (6) in very good yields
(Scheme 6).13
Scheme 3. Competition Experiments
a
Scheme 6. Synthesis of Coumarins
recovery of a majority of the starting materials (Scheme 4),
which suggests that the conjugation of the vinyl moiety to the
aryl group is critical for a successful outcome.
Scheme 4. Reaction of 2-Allylphenol
a
Reaction conditions: 0.50 mmol of 1a, [Cp*RhCl2]2 (2.5 mol %), 1.2
equiv Cu(OAc)2·H2O, 2 mL of CH3CN at 85 °C, carbon monoxide
balloon, overnight.
Although a precise reaction mechanism cannot be definitively
established, a proposal consistent with the current data is
outlined in Scheme 5. The process most probably starts with
the phenolic substrate 1 replacing one of the acetates of the
catalyst to give intermediate I.12 This complex might evolve to
the rhodacycle II either through a typical concerted metal-
ation−deprotonation step (CMD) or by an intramolecular
electrophilic attack of the conjugated alkene to the electrophilic
rhodium (I) followed by a base-assisted deprotonation to yield
In summary, we have developed the first example of a metal-
catalyzed (5 + 2) cycloaddition formally involving a C−H
activation process. The method provides a fast, efficient, and
practical route to benzoxepines using commercial or readily
available o-vinylphenols and alkynes as starting materials.
Replacement of the alkyne component by carbon monoxide
allows assembly of coumarin derivatives in a straightforward
manner. Further mechanistic studies are currently underway
and will be reported in due course.
Scheme 5. Proposed Mechanistic Cycle
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and characterization for new com-
pounds. This material is available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
joseluis.mascarenas@usc.es
moises.gulias@usc.es
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are thankful for the financial support provided by the
Spanish Grants (SAF2010-20822-C02 and CSD2007-00006
Consolider Ingenio 2010), the ERDF and the Xunta de Galicia
Grants GRC2010/12, GR2013-041, INCITE09 209084PR and
EM2013/036. M.G thanks Xunta de Galicia for a Parga Pondal
contract.
C
dx.doi.org/10.1021/ja410538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(10) (a) Morimoto, K.; Hirano, K.; Satoh, T.; Miura, M. J. Org. Chem.
2011, 76, 9548−9551. (b) Mochida, S.; Shimizu, M.; Hirano, K.;
Satoh, T.; Miura, M. Chem. Asian J. 2010, 5, 847−851. For a related
activation of internal C−H bonds of alkenes see: (c) Hu, J.; Hirao, H.;
Li, Y.; Zhou, J. Angew. Chem., Int. Ed. 2013, 52, 8676−8680.
(11) NMR analysis of the reaction using an internal standard
indicated that, while 3aa was formed in 45% yield after 10 min of
reaction, 3ga was formed in 28% yield. The formation of the
trifluoromethylaryl derivative 3ka is also much slower than that of 3aa
(see SI for more details).
(12) The formation of RhCp*(OAc)2 as the active species has been
proposed for similar reactions, see for instance: (a) Hyster, T. K.;
Rovis, T. J. Am. Chem. Soc. 2010, 132, 10565−10569. (b) Xu, L.; Zhu,
Q.; Huang, G.; Cheng, B.; Xia, Y. J. Org. Chem. 2011, 77, 3017−3024.
(13) For related metal-catalyzed carbonylation methods, see:
(a) Ferguson, J.; Zeng, F.; Alper, H. Org. Lett. 2012, 14, 5602−
5605. (b) Inamoto, K.; Kadokawa, J.; Kondo, Y. Org. Lett. 2013, 15,
3962−3965. (c) Sasano, K.; Takaya, J.; Iwasawa, N. J. Am. Chem. Soc.
2013, 135, 10954−10957. (d) Luo, S.; Luo, F.-X.; Zhang, X.-S.; Shi,
Z.-J. Angew. Chem., Int. Ed. 2013, 52, 10598−10601. (e) Jiang, G.-J.;
Fu, X.-F.; Li, Q.; Yu, Z.-X. Org. Lett. 2012, 14, 692−695.
REFERENCES
■
(1) For selected reviews of metal-catalyzed cycloadditions see:
(a) Lautens, M.; Klute, W.; Tam, W. Chem. Rev. 1996, 96, 49−92.
(b) Schore, N. E. Chem. Rev. 1988, 88, 1081−1119. (c) Gulias, M.;
Lopez, F.; Mascarenas, J. L. Pure Appl. Chem. 2011, 83, 495−506.
(d) Lopez, F.; Mascarenas, J. L. Beilstein J. Org. Chem. 2011, 7, 1075−
1094. (e) Inglesby, P. A.; Evans, P. A. Chem. Soc. Rev. 2010, 39, 2791−
2805. (f) Yu, Z.-X.; Wang, Y.; Wang, Y. Chem. Asian J. 2010, 5, 1072−
1088.
(2) For selected recent reviews of metal-catalyzed C−H function-
alizations, see: (a) Wencel-Delord, J.; Glorius, F. Nat. Chem. 2013, 5,
369−375. (b) Engle, K. M.; Mei, T.-S.; Wasa, M.; Yu, J.-Q. Acc. Chem.
Res. 2012, 45, 788−802. (c) Yeung, C. S.; Dong, V. M. Chem. Rev.
2011, 111, 1215−1292. (d) Wencel-Delord, J.; Droege, T.; Liu, F.;
Glorius, F. Chem. Soc. Rev. 2011, 40, 4740−4761. (e) Sun, C.-L.; Li, B.-
J.; Shi, Z.-J. Chem. Commun. 2010, 46, 677−685. (f) Fagnou, K. Top.
Curr. Chem. 2010, 292, 35−56. (g) Daugulis, O. Top. Curr. Chem.
2010, 292, 57−84. (h) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010,
110, 1147−1169. (i) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem.
Rev. 2010, 110, 624−655. For a review of rhodium-catalyzed oxidative
annulation of aromatic substrates with alkynes or alkenes, see:
(j) Satoh, T.; Miura, M. Chem.Eur. J. 2010, 16, 11212−11222.
(3) For selected examples, see: (a) Stuart, D. R.; Bertrand-Laperle,
M.; Burgess, K. M. N.; Fagnou, K. J. Am. Chem. Soc. 2008, 130,
16474−16475. (b) Shi, Z.; Zhang, C.; Li, S.; Pan, D.; Ding, S.; Cui, Y.;
Jiao, N. Angew. Chem., Int. Ed. 2009, 48, 4572−4576. (c) Rakshit, S.;
Patureau, F. W.; Glorius, F. J. Am. Chem. Soc. 2010, 132, 9585−9587.
(d) Stuart, D. R.; Alsabeh, P.; Kuhn, M.; Fagnou, K. J. Am. Chem. Soc.
2010, 132, 18326−18339. (e) Huestis, M. P.; Chan, L.; Stuart, D. R.;
Fagnou, K. Angew. Chem., Int. Ed. 2011, 50, 1338−1341. (f) Chen, J.;
Pang, Q.; Sun, Y.; Li, X. J. Org. Chem. 2011, 76, 3523−3526.
(4) For selected examples, see: (a) Guimond, N.; Gouliaras, C.;
Fagnou, K. J. Am. Chem. Soc. 2010, 132, 6908−6909. (b) Ackermann,
L.; Lygin, A. V.; Hofmann, N. Angew. Chem., Int. Ed. 2011, 50, 6379−
6382. (c) Neely, J. M.; Rovis, T. J. Am. Chem. Soc. 2013, 135, 66−69.
(d) Hyster, T. K.; Knorr, L.; Ward, T. R.; Rovis, T. Science 2012, 338,
500−503. (e) Ye, B.; Cramer, N. Science 2012, 338, 504−506. (f) Xu,
X.; Liu, Y.; Park, C.-M. Angew. Chem., Int. Ed. 2012, 51, 9372−9376.
(g) Tan, X.; Liu, B.; Li, X.; Li, B.; Xu, S.; Song, H.; Wang, B. J. Am.
Chem. Soc. 2012, 134, 16163−16166. (h) Quinones, N.; Seoane, A.;
Garcia-Fandino, R.; Mascarenas, J. L.; Gulias, M. Chem. Sci. 2013, 4,
2874−2879. (i) Luo, C.-Z.; Gandeepan, P.; Cheng, C.-H. Chem.
Commun. 2013, 49, 8528−8530. (j) Huckins, J. R.; Bercot, E. A.; Thiel,
O. R.; Hwang, T.-L.; Bio, M. M J. Am. Chem. Soc. 2013, 135, 14492−
14495. (k) Dooley, J. D.; Reddy Chidipudi, S.; Lam, H. W. J. Am.
Chem. Soc. 2013, 135, 10829−10836.
(5) Formal (4 + 3) annulations based on C-H activation have been
recently reported: (a) Cui, S.; Zhang, Y.; Wu, Q. Chem. Sci. 2013, 4,
3421−3426. (b) Cui, S.; Zhang, Y.; Wang, D.; Wu, Q. Chem. Sci 2013,
4, 3912−3916. (c) Shi, Z.; Grohmann, C.; Glorius, F. Angew. Chem.,
Int. Ed. 2013, 52, 5393−5397.
(6) For representative intermolecular (5 + 2) annulations leading to
azepine skeletons, see: (a) Wender, P. A.; Pedersen, T. M.; Scanio, M.
J. C. J. Am. Chem. Soc. 2002, 124, 15154−15155. (b) Zhou, M.-B.;
Song, R.-J.; Wang, C.-Y.; Li, J.-H. Angew. Chem., Int. Ed. 2013, 52,
10805−10808.
(7) (a) Narita, K.; Nakamura, K.; Abe, Y.; Katoh, T. Eur. J. Org.
Chem. 2011, 4985−4988 and references therein. (b) Sprogoe, K.;
Manniche, S.; Larsen, T. O.; Christophersen, C. Tetrahedron 2005, 61,
8718−8721.
(8) For a review on the synthesis of oxepines including benzoxepines,
see: Snyder, N. L.; Haines, H. M.; Peczuh, M. W. Tetrahedron 2006,
62, 9301−9320.
(9) For Pd(II)-catalyzed annulation of phenols and alkynes leading to
benzofuranes see: (a) Kuram, M. R.; Bhanuchandra, M.; Sahoo, A. K.
Angew. Chem., Int. Ed. 2013, 52, 4607−4612. For a Rh(III)-catalyzed
synthesis of benzofuranes from N-Phenoxyacetamides see: (b) Liu, G.;
Shen, Y.; Zhou, Z.; Lu, X. Angew. Chem., Int. Ed. 2013, 52, 6033−6037.
D
dx.doi.org/10.1021/ja410538w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX