Copper-catalyzed cross-coupling interrupted by an opportunistic smiles rearrangement: An efficient domino approach to dibenzoxazepinones

Article abstract of DOI:10.1002/anie.201106786

Unexpected Smiles! An unusual and highly regioselective synthesis of dibenzoxazepinones by a domino sequence assisted by an unexpected Smiles rearrangement is reported. The process is effective on electronically differentiated phenols and shows a high tolerance to variation in the benzamide substituents. A plausible path for the reaction, supported by preliminary mechanistic data, is offered. Copyright

Full text of DOI:10.1002/anie.201106786

Angewandte
Chemie
DOI: 10.1002/anie.201106786
À
À
Domino C O/C N Bond Formation
Copper-Catalyzed Cross-Coupling Interrupted by an Opportunistic
Smiles Rearrangement: An Efficient Domino Approach to
Dibenzoxazepinones**
Matthew O. Kitching, Timothy E. Hurst, and Victor Snieckus*
À
À
The construction of C O and C N bonds by palladium- and
copper-mediated cross-coupling approaches has undergone
a renaissance in the last decade.[1] Although the utility of these
coupling approaches to effect a single-bond construction
event has now been established, the quest to improve
efficiency has driven the further exploration of one-pot
procedures for effecting multiple transformations, often by
a domino[2] approach. Success in this area has come from
impressive examples of ligand control,[3] finely tuned multi-
catalyst systems,[4] and exploitation of established multicom-
ponent reactions.[5] Based on our previous experience with
Ullmann[6] couplings of 2-halobenzamides,[7] combined with
the established Goldberg N-arylation methods,[1b,8] we envis-
aged that a copper-catalyzed one-pot annulation between
2-iodobenzamides 2 and 2-bromophenols 3 would allow
and incorporates the effective directed ortho metalation
strategy.[11] This allows for rapid construction, either singly or
in library format, of this class of tricyclics, which constitute
attractive synthetic targets due to their varied and extensive
biological activity.[12]
Considerable optimization (see the Supporting Informa-
tion) demonstrated that using CuI as catalyst, dibenzoyl-
methane (dbm) as ligand, K3PO4 as base, and DMF as solvent
is optimal for this process, giving a tricyclic product in 78%
yield (Scheme 2). A remarkable switch in selectivity was
À
À
access to dibenzoxazepinones 1 by selective C O and C N
bond forming events (Scheme 1).[9]
Scheme 1. Initial retrosynthetic analysis of dibenzoxazepinones 1.
Scheme 2. Initial study of copper-catalyzed annulations to dibenzoxa-
zepinones.
We herein report an unusual and highly selective copper-
initiated domino synthesis of dibenzoxazepinones that
observed when the apolar solvent toluene was used. This
produced the diaryl ether 6aa (69% yield), arising from an
Ullmann coupling, as the only product. Control experiments
performed without ligand showed that CuI alone was capable
of catalyzing the transformation, albeit in reduced and
irreproducible yield. Performing the reaction in the absence
of both CuI and ligand led to complete recovery of iodide 4a.
The structure of the tricyclic product was established by 2D
NMR spectroscopy and X-ray crystallographic analysis. To
our surprise, the expected Goldberg arylation product 7aa
had not been formed. Instead, the rearranged compound 8aa,
in which the oxygen and chlorine atoms share a meta- rather
than the expected para-relationship, was the sole product. We
reasoned that this product arose from interception of the
initially formed diaryl ether 6aa by a Smiles rearrangement,
À
includes the formal C N coupling of secondary benzamides
assisted by an unexpected Smiles rearrangement. Most
strategies to 1 are classical, multistep processes.[10] Our
efficient, modular, and convergent approach utilizes either
commercially available or readily prepared starting materials
[*] Dr. M. O. Kitching, Dr. T. E. Hurst, Prof.Dr. V. Snieckus
Department of Chemistry, Queen’s University
Kingston, Ontario, K7L 3N6 (Canada)
E-mail: snieckus@chem.queensu.ca
[**] We would like to thank Dr. O. B. Familoni for initial results, Dr. R.
Wang for X-ray structural analysis, and Dr. A. A. Neverov and Prof.
R. S. Brown for helpful mechanistic discussions. We thank the
NSERC Discovery Grant program for the support of our synthetic
programs.
À
instead of the expected C N Goldberg coupling process (see
below). Having established optimized conditions, we exam-
ined the reaction scope with respect to the amide
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201106786.
Angew. Chem. Int. Ed. 2012, 51, 2925 –2929
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2925

.
Angewandte
Communications
Table 1: Variation of benzamide ring substituents (4).
N-substituent, which demonstrated that simple alkyl, benzyl,
and 2,4-dimethoxybenzyl groups were equally well tolerated
(Scheme 3).
Entry
1
Substrate
Product
Yield [%]
50[a,b]
Scheme 3. Copper-catalyzed coupling–annulation of benzamides 9
with 2-bromo-4-chlorophenol (5a). [a] 78% yield was obtained when
the reaction was conducted using 10 g (36 mmol) of 9a.
2
3
4
5
6
80[a]
73[a]
85
With these encouraging results in hand, the coupling–
annulation of variously substituted N-ethyl-2-iodobenz-
amides 4 with 2-bromo-4-chlorophenol (5a) was undertaken
(Table 1). In terms of the effect of structural features,
introducing steric hindrance ortho to the amide proved
detrimental to the reaction, with the desired product 8ba
being isolated in only 14% yield after 6 h. However, increas-
ing the reaction time to 24 h led to a synthetically useful 50%
yield of 8ba (entry 1), along with small amounts of an
isomeric tricyclic compound, which was assigned as the
Goldberg coupled product 7ba on the basis of 2D NMR
spectroscopy (see the Supporting Information).[13] A hindered
3-methyl benzamide 4c was well tolerated, as were amides
bearing halogens (entries 3 and 4), an electron withdrawing
group (entry 5), and electron donating groups (entries 6 and
7). Significantly, the Smiles rearranged products were formed
in all cases, as shown by either X-ray crystallographic analysis
or 2D NMR spectroscopy.
86
86
7
74
To explore the scope of electronically and sterically
differentiated 2-bromophenols in the observed Smiles re-
arrangement (Table 2), the isomeric 2-bromocresols 5c and
5d were exposed to the standard reaction conditions, resulting
in the formation of rearranged products 8ac and 8ad, albeit in
low yield, with large amounts of the uncyclized biaryl ethers
also isolated. As in the previous case, simply extending the
reaction time to 24 h and 48 h, respectively, led to increased
yields (entries 2 and 3). 1-Bromo-2-naphthol also performed
well in the reaction to give the benzonaphthoxazepinone 8ae
(entry 4). However, sluggish reactivity was observed when
2-bromo-4-methoxyphenol 5 f was used. Upon extended
heating, 8af was obtained as the major product in 31%
yield (entry 5), along with small amounts of the isomeric
compound 8ag. This result suggests that electron-rich phenols
undergo rearrangement at a greatly reduced rate, allowing
Goldberg cyclization as a competing pathway. To confirm this
result, 8ag was independently prepared from 2-bromo-5-
methoxyphenol 5g under the standard conditions (entry 6),
furnishing material that was identical to the minor product
8ag of entry 5.
[a] 24 h reaction time. [b] The Goldberg coupled product 7ba was also
isolated in 12% yield.
was obtained (entry 7), the corresponding nitrile 8ai was
formed in poor yield (entry 8). Heterocycle-derived sub-
strates also appear to be excellent coupling partners, based on
the formation of pyridobenzoxazepinone 8aj in excellent
yield (entry 9). To further test the reaction scope, the one-pot
coupling–annulation of 2-bromo-4-chlorothiophenol (11) was
attempted. Unfortunately, the coupling of 4a with 11 proved
sluggish at 1208C. However, an equimolar mixture of both
Goldberg and rearranged products (12a and 12b), was
obtained after heating at 1508C for 48 h (entry 10).
Having established the scope of the dibenzoxazepinone
synthesis, we turned our attention to investigation of the
mechanism of this unusual process. With the requirement for
copper to initiate the process already established (see the
Supporting Information), the analogous single-step Ullmann
and Goldberg couplings were conducted. When a mixture of
N-ethyl toluamide and model phenol 5a were exposed to the
standard reaction conditions, no Goldberg coupling product
Electronically impoverished phenols gave more varied
results. While a high yield of 8ah bearing a fluoro substituent
2926 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2925 –2929

Angewandte
Chemie
Table 2: Variation of phenol ring substituents (5).
model amide 4a were combined under identical conditions
(Scheme 4), the Ullmann product 14 was obtained (58%
yield) along with the Smiles-rearranged tertiary salicylamide
15 (39% yield), whose structure was unambiguously con-
firmed by X-ray crystallography (see the Supporting Infor-
mation).
Entry
1
Substrate
Product
Yield [%]
75[a]
2
3
37[a]
Scheme 4. Single-step Ullmann coupling control experiment.
74[b]
Although Smiles rearrangements of diaryl ethers bearing
secondary amides have been previously reported,[14] the
systems investigated typically involve arenes bearing multiple
electron-withdrawing groups (for example, NO2, CN, F) or
strongly electron-deficient heteroaromatic rings, such as
pyridine.[15] In stark contrast, our reaction progresses
smoothly in the presence of both electron-neutral and
electron-rich phenols. Perhaps most surprisingly, these sub-
strates often outperform their electron-poor counterparts,
which are less-tolerated. Our results constitute, to the best of
our knowledge, the first report on the generality of the Smiles
rearrangement of electron-rich diaryl ethers bearing secon-
dary amides. Furthermore, in all but three cases examined,
exclusive regioselectivity for the Smiles rearrangement prod-
uct is observed.[16,17]
To further support the intermediacy of the Ullmann
product 6, and to investigate the Smiles rearrangement in
more depth, we examined the progress of the reaction from
isolated diaryl ether 6aa (Scheme 5). As expected, exposure
of 6aa to the optimized, copper-catalyzed conditions afforded
the Smiles rearranged product 8aa in quantitative yield,
providing further evidence for 6aa as an intermediate in the
reaction pathway. We reasoned that, as with the initiation of
4
67
5
6
7
8
9
31[c,d]
70[a]
72[b]
32[a]
95
41 (12a)[e]
38 (12b)[e]
10
[a] 24 h reaction time. [b] 48 h reaction time. [c] 72 h reaction time.
[d] The Goldberg coupled product 8ag was also isolated in 7% yield.
[e] 1508C for 48 h.
was detected; rather, unreacted amide was recovered (98%
Scheme 5. Proposed Ullmann–Smiles-cyclization reaction pathway to
yield) instead. In contrast, when 4-chlorophenol (13) and
dibenzoxazepinones.
Angew. Chem. Int. Ed. 2012, 51, 2925 –2929
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2927

.
Angewandte
Communications
this domino process, the final ring closure of 16 to produce
8aa resulted from a copper-catalyzed Ullmann coupling
between the newly generated phenol and the residual aryl
bromide. To confirm this assumption, the reaction of 6aa was
also conducted both without copper and ligand (base only)
and under purely thermal conditions (no copper, ligand, or
base). Although no reaction was observed in the latter case,
we were surprised to find that the dibenzoxazepinone product
8aa was obtained in quantitative yield under base-mediated
copper-free conditions,[18,19] demonstrating that copper is not
a prerequisite for the ring-closing etherification.[20] Based on
these observations, we propose a mechanism involving
a copper-catalyzed Ullmann coupling, followed by a base-
mediated Smiles rearrangement and final ring-closure
(Scheme 5).
As a final point of mechanistic significance, our inability
to isolate or detect a Smiles rearrangement product from any
of our reaction mixtures gave us pause for concern. To
demonstrate the transient generation of rearrangement
product 19 under the reaction conditions, we investigated
the possibility of trapping 19 in situ to confirm its presence in
the reaction mixture (Scheme 6). Towards this aim, amide 17
was treated with phenol 5g under our standard reaction
conditions. To our delight, product 20 was obtained in 73%
yield, thus demonstrating the intermediacy of 19 in the
reaction pathway.
of dibenzoxazepinones, with significant advantages over
established multi-step processes. Further work to expand
the scope of this method, to extend it to an analogous
coupling–annulation process with 2-iodobenzenesulfon-
amides, and to study its mechanism in detail will be reported
in due course.
Received: September 24, 2011
Published online: February 6, 2012
À
À
Keywords: C N bond formation · C O bond formation ·
copper · domino reactions · Smiles rearrangement
.
[1] For palladium amination, see: a) D. S. Surry, S. L. Buchwald,
Angew. Chem. 2008, 120, 6438 – 6461; Angew. Chem. Int. Ed.
2008, 47, 6338 – 6361; for copper, see: b) K. Kunz, U. Scholz, D.
Ganzer, Synlett 2003, 2428 – 2439; c) S. V. Ley, A. W. Thomas,
Angew. Chem. 2003, 115, 5558 – 5607; Angew. Chem. Int. Ed.
2003, 42, 5400 – 5449.
[2] L. F. Tietze, U. Beifuss, Angew. Chem. 1993, 105, 137 – 170;
Angew. Chem. Int. Ed. Engl. 1993, 32, 131 – 163.
[3] As a selected example, see: B. P. Fors, S. L. Buchwald, J. Am.
Chem. Soc. 2010, 132, 15914 – 15917.
[4] As a selected example, see: J. Panteleev, L. Zhang, M. Lautens,
Angew. Chem. 2011, 123, 9255 – 9258; Angew. Chem. Int. Ed.
2011, 50, 9089 – 9092.
[5] M. DꢀSouza, T. J. J. Mꢁller, Chem. Soc. Rev. 2007, 36, 1095 –
1108.
[6] F. Monnier, M. Taillefer, Angew. Chem. 2009, 121, 7088 – 7105;
Angew. Chem. Int. Ed. 2009, 48, 6954 – 6971.
[7] A. V. Kalinin, J. F. Bower, P. Riebel, V. Snieckus, J. Org. Chem.
1999, 64, 2986 – 2987.
[8] E. Racine, F. Monnier, J.-P. Vors, M. Taillefer, Org. Lett. 2011, 13,
2818 – 2821.
[9] For example, see: a) D. Maiti, S. L. Buchwald, J. Am. Chem. Soc.
2009, 131, 17423 – 17429; b) A. Shafir, P. A. Lichtor, S. L.
Buchwald, J. Am. Chem. Soc. 2007, 129, 3490 – 3491; c) G. E.
Job, S. L. Buchwald, Org. Lett. 2002, 4, 3703 – 3706; d) F.
Monnier, M. Taillefer, Angew. Chem. 2008, 120, 3140 – 3143;
Angew. Chem. Int. Ed. 2008, 47, 3096 – 3099.
[10] By intramolecular Friedel–Crafts acylation, see: a) Y. Nagai, A.
Irie, H. Nakamura, K. Hino, H. Uno, H. Nishimura, J. Med.
Chem. 1982, 25, 1065 – 1070; b) B. S. Wagh, B. P. Patil, M. S. Jam,
S. S. Harak, S. B. Wagh, Heterocycl. Commun. 2007, 13, 165 –
172; by oxidation of dibenzo(b,f)(1,4)-oxazepines, see: c) K.
Brewster, R. A. Chittenden, J. M. Harrison, T. D. Inch, C.
Brown, J. Chem. Soc. Perkin Trans. 1 1976, 1291 – 1296; by
Beckmann rearrangement, see: d) K. Nagarajan, C. L. Kulkarni,
A. Venkateswarlu, Indian J. Chem. 1974, 12, 247 – 251; e) A. T.
Troshchenko, T. P. Lobanova, Zh. Org. Khim. 1967, 3, 501 – 503
[J. Org. Chem. USSR (Engl. Transl.) 1967, 3, 480 – 482]; by
carbonylative coupling, see: f) Q. Yang, H. Cao, A. Robertson,
H. Alper, J. Org. Chem. 2010, 75, 6297 – 6299; by Ugi four-
component reactions, see: g) J. Wu, Y. Jiang, W.-M. Dai, Synlett
2009, 1162 – 1166; by SNAr and amide formation, see: h) I. G.
Abramov, A. V. Smirnov, L. S. Kalandadze, V. N. Sakharov, V. V.
Plakhtinskii, Heterocycles 2003, 60, 1611 – 1614; i) A. V. Samet,
V. N. Marshalkin, K. A. Kislyi, N. B. Chernysheva, Y. A. Stre-
lenko, V. V. Semenov, J. Org. Chem. 2005, 70, 9371 – 9376;
j) R. A. Bunce, J. E. Schammerhorn, J. Heterocycl. Chem. 2006,
43, 1031 – 1035; k) H. J. M. Gijsen, D. Berthelot, M. Zaja, B.
Brꢂne, I. Geuens, M. Mercken, J. Med. Chem. 2010, 53, 7011 –
7020; for solid phase synthesis, see: l) X. Ouyang, N. Tamayo,
A. S. Kiselyov, Tetrahedron 1999, 55, 2827 – 2834; m) N. D. Hone,
J. I. Salter, J. C. Reader, Tetrahedron Lett. 2003, 44, 8169 – 8172.
Scheme 6. In situ trapping of Smiles rearrangement intermediate 19.
In summary, we have uncovered an unusual and highly
regioselective copper-initiated domino reaction which pro-
vides
a general route to dibenzoxazepinones 8 from
2-iodobenzamides 4 and 2-bromophenols 5. As supported
by detailed mechanistic studies, the reactions proceed first by
a copper-catalyzed Ullmann coupling, followed by a base-
mediated Smiles rearrangement and final ring-closing process
(Scheme 5). Our initial observation of a Smiles reaction was
unexpected owing to the absence of strong electron with-
drawing groups.[14] It therefore provides a caveat to those in
À
À
the currently active area of C O and C N bond coupling
studies.
This method provides an efficient, scalable (see Scheme 3,
footnote a), and modular approach for the rapid construction
2928 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2925 –2929

Angewandte
Chemie
[11] a) J. Clayden, Organolithiums: Selectivity in Synthesis, Perga-
mon, Elsevier, 2002, pp. 21 – 90; b) T. Macklin, V. Snieckus in
Handbook of CH Transformations (Ed.: G. Dyker), WileyVCH,
New York, 2005, pp. 106 – 119; c) C. G. Hartung, V. Snieckus in
Modern Arene Chemistry (Ed.: D. Astruc) WileyVCH, New
York, 2002, pp. 330 – 367.
[12] For HIV-1 RT inhibition, see: S. Vilar, L. Santan, E. Uriarte, J.
Med. Chem. 2006, 49, 1118 – 1124; for CNS agents, see: J. F. F.
Liegeois, F. A. Rogister, J. Bruhwyler, J. Damas, T. P. Nguyen,
M. O. Inarejos, E. M. G. Chleide, M. G. A. Mercier, J. E.
Delarge, J. Med. Chem. 1994, 37, 519 – 525; for p38 MAP
kinase inhibition see: A. Dorn, V. Schattel, S. Laufer, Bioorg.
Med. Chem. Lett. 2010, 20, 3074 – 3077; as non-carbohydrate
inhibitors of E. coli ribosomal A site, see: S. P. Maddaford, M.
Motamed, K. B. Turner, M. S. K. Choi, J. Ramnauth, S. Rakhit,
R. R. Hudgins, D. Fabris, P. E. Johnson, Bioorg. Med. Chem.
Lett. 2004, 14, 5987 – 5990; for anti-inflammatory activity, see:
J. K. Chakrabarti, T. A. Hicks, Eur. J. Med. Chem. 1987, 22, 161 –
163.
[15] During the preparation of this manuscript a report of the
synthesis of dibenzoxazepinones also utilizing a Smiles rear-
rangement solely on electron deficient substrates appeared; see:
Y. Liu, C. Chu, A. Huang, C. Zhan, Y. Ma, C. Ma, ACS Comb.
Sci. 2011, 13, 547 – 553.
[16] J. R. Proudfoot, U. R. Patel, S. J. Campbell, J. Org. Chem. 1993,
58, 6996 – 7000.
[17] In the two examples using phenolic substrates, the Smiles
product remains the dominant product (ratios of isolated
products 8ba/7ba 4.2:1, 8af/8ag 4.4:1). Only in the case of
dibenzothiazepinones 12a and 12b is no selectivity observed.
[18] To rule out the possibility of metal contamination, 8aa was also
prepared from N-ethyl-2-fluoro-4-methylbenzamide under
copper-free conditions. See the Supporting Information for
details.
[19] For metal contamination that significantly affects the results of
catalytic reactions, see: a) R. K. Arvela, N. E. Leadbeater, M. S.
Sangi, V. A. Williams, P. Granados, R. D. Singer, J. Org. Chem.
2005, 70, 161 – 168; b) S. L. Buchwald, C. Bolm, Angew. Chem.
2009, 121, 5694 – 5695; Angew. Chem. Int. Ed. 2009, 48, 5586 –
5587.
[13] J. F. Bunnett, T. Okamoto, J. Am. Chem. Soc. 1956, 78, 5363 –
5367.
[14] a) V. Sapegin, V. N. Sakharov, L. S. Kalandadze, A. V. Smirnov,
T. A. Khristolyubova, V. V. Plakhtinskii, A. V. Ivashchenko,
Mendeleev Commun. 2008, 18, 281 – 283; b) O. Reinaud, P.
Capdevielle, M. Maumy, J. Chem. Soc. Perkin Trans. 1 1991,
2129 – 2134.
[20] It is not possible to rule out catalysis by homeopathic quantities
of copper or other metals present in commercially available
bases (see ref. [19]). However, the use of the same batch of
K3PO4 did not catalyze either the overall transformation or the
initial Ullmann reaction (see the Supporting Information).
Angew. Chem. Int. Ed. 2012, 51, 2925 –2929
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2929