Synthesis of 3-benzazepines and azepino[4,5-b]heterocyclic ring systems via intramolecular Friedel-Crafts cyclization

Article abstract of DOI:10.1016/j.tetlet.2013.02.012

The intramolecular Friedel-Crafts cyclization was found to be effective for the synthesis of 3-benzazepines and azepino[4,5-b]heterocyclic ring systems using simple allylic bromides tethered to activated aromatic nuclei. Amongst the various Lewis acids screened, Bi(OTf)3 was found to be the most effective 'ecofriendly' catalyst for the cyclization.

Full text of DOI:10.1016/j.tetlet.2013.02.012

Tetrahedron Letters 54 (2013) 2018–2021
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of 3-benzazepines and azepino[4,5-b]heterocyclic ring systems via
intramolecular Friedel–Crafts cyclization
⇑
Robert B. Kargbo , Zohreh Sajjadi-Hashemi, Sujata Roy, Xiaomin Jin, R. Jason Herr
Medicinal Chemistry Department, Albany Molecular Research, Inc. (AMRI), PO Box 15098, 26 Corporate Circle, Albany, NY 12212-5098, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The intramolecular Friedel–Crafts cyclization was found to be effective for the synthesis of
3-benzazepines and azepino[4,5-b]heterocyclic ring systems using simple allylic bromides tethered to
activated aromatic nuclei. Amongst the various Lewis acids screened, Bi(OTf)3 was found to be the most
effective ‘ecofriendly’ catalyst for the cyclization.
Received 24 January 2013
Revised 31 January 2013
Accepted 4 February 2013
Available online 14 February 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Friedel–Crafts cyclization
3-Benzazepines
Allylic bromides
Activated aromatic nuclei
Alkaloids containing seven-membered ring core scaffolds are
found in many natural products and biologically active com-
pounds,1 and the substituted 3-benzazepine motif in particular
has been exploited as a scaffold for potential medicinal agents in
several therapeutic applications. Representative examples are
shown in Figure 1 that include azepino[4,5-b]indole FXR agonists
1,2 3-benzazepine dopamine agonist SCH23390 (2),3 the natural
product 3-benzazepine ( )-cephalotaxine (3),4 and azepino[4,
5-b]indole arboflorine (4).5 Consequently, significant synthetic
efforts have been devoted to the development of efficient protocols
for the preparation and direct functionalization of these fused cyc-
lic heteroaromatics. Many of these include intramolecular exten-
sions of Friedel–Crafts-type approaches6 and Tsuji–Trost/Heck
reactions,7 as well as radical cyclizations8 and thermal rearrange-
ment-promoted cyclizations,9 among others.10
The development of useful methods for the formation of nitro-
gen-containing seven-membered rings can constitute a greater
challenge than for smaller ring homologs due to entropic factors
in the ring closure, and increased ring strain in the corresponding
cyclic products.11 The efficiency of intramolecular Friedel–Crafts
(IMFC)-type reaction chemistry in particular often suffers from rel-
evant drawbacks such as the lack of regioselective control in the
cyclization step, and the typical need for high reaction tempera-
tures and stoichiometric amounts of Lewis acid catalysts, some of
which generate metal-containing waste streams of environmental
concern.6 Keeping all of these drawbacks in mind, we set out to
develop a thermally mild IMFC method to prepare 3-benzazepines
and azepino[4,5-b]heterocyclic rings that was both regioselective
and catalytic in an ecologically benign activating reagent.12
We initially envisioned the synthesis of azepino[4,5-b]hetereoa-
romatic systems through an IMFC reaction utilizing catalytic
amounts of an environmentally friendly metal. Inspired by the
reports of Cook and coworkers that demonstrated a remarkably
mild InCl3-catalyzed Friedel–Crafts cyclization utilizing both
electron-rich and electron-poor arene substrates,13 we started
O
O
Cl
H3C
H3C
N
N
CH3
N
HO
n
O
N
H
CO2iPr
F
1
2
FXR agonists
O
SCH23390
N
O
N
H
H
HO
N
H
O
4
H
N
CH3
H
OCH3
(-)-Cephalotaxine
3
Arboflorine
⇑
Corresponding author. Tel.: +1 518 512 2000; fax: +1 518 512 0000.
E-mail address: robert.kargbo@amriglobal.com (R.B. Kargbo).
Figure 1. Examples of 3-benzazepine and azepino[4,5-b]heterocyclic bioactive
agents and natural products.
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.02.012

R. B. Kargbo et al. / Tetrahedron Letters 54 (2013) 2018–2021
2019
Table 1
equivalents of both Bi(OTf)3 and Hunig’s base, from which 6a was
obtained in 98% yield (entry 14).
Lewis acid catalyst screening reactions
Br
The other available bismuth(III) salts BiBr3 and Bi(NO3)3Á5H2O
were much less effective than Bi(OTf)3 (Table 1, entries 16 and
17) and the use of bismuth metal was completely ineffective (entry
18). It should also be noted that the reaction proceeded well in
non-coordinating solvents such as chloroform, toluene, 1,2-dichlo-
roethane, and hexane, but failed to provide significant product
amounts in solvents possessing donor heteroatoms such as aceto-
nitrile, THF, 1,4-dioxane, or methanol (see Supplementary data for
more details).
TsN
NTs
Lewis acid
4 Å mol. seives
CH2Cl2 (0.05 M)
rt, 16 h
OCH3
6a
(+/-)-
H3CO
OCH3
H3CO
5a
Entry
Lewis acid
Yield
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
InCl3 (20 mol %)
InCl3 (50 mol %)
InCl3 (100 mol %)
InBr3 (20 mol %)
In(0) (20 mol %)
Ag2OÁ4B2O3 (20 mol %)
AgOTf (20 mol %)
ZnCl2 (20 mol %)
FeCl3 (20 mol %)
TiCl4 (20 mol %)
SnCl4 (20 mol %)
Bi(OTf)3 (20 mol %)
Bi(OTf)3 (20 mol %)
Bi(OTf)3 (100 mol %)
BiCl3 (20 mol %)
0a
With promising reaction conditions in hand, we next explored a
limited substrate scope for the IMFC reaction with allylic bromide-
tethered precursors 5a–l (Table 2).17 Good to moderate yields were
obtained utilizing 20 mol % of Bi(OTf)3 with several substrates
bearing electron-rich aromatic moieties (entries 1 through 7). In
these cases where symmetrically-substituted aromatic nuclei (rel-
ative to the alkylamine functionality) were employed (e.g., 5a, 5d,
and 5e), single products were obtained (entries 1, 4, and 5). For
unsymmetrically-substituted aryl substrates (e.g., 5b, 5c, 5f, and
5g), regioisomeric mixtures of products were obtained, often in
preference for the products of the para-directing substitution of
the electron-donating moiety (para-products of entries 2, 3, 6,
and 7). In general, substrates with alkoxy-substituted groups gave
excellent yields (entries 1–4) compared to precursors with the
weaker electron-donating methyl functionality (entries 5–7),
which required increased catalyst loading to perform efficiently.
In comparison, the relatively electron-neutral precursor 5h was
less favorable for the cyclization reaction (entry 8), whereas the
electron-withdrawing effect of the 4-chlorophenyl substrate 5i
did not provide the corresponding 3-benzazepine adduct 6i, and
led only to trace amounts of product with a full loading of Lewis
acid (entry 9). Interestingly, the presence of a cyclopropyl group
in precursor 5j led to an improvement in yield versus parent sub-
strate 5h (entry 10 vs 8), possibly due to a favorable conformation
induced by sterics in the transition state, allowing better access to
the geometry required for the cyclization reaction.
We were pleased to find that heterocyclic precursors provided
excellent yields of azepino[4,5-b]heterocycles when conducted
below room temperature (entries 11 and 12). Both indole 5k and
benzothiophene 5l were shown to react at the C2 position to
provide the corresponding products in near-quantitative isolated
yields, and it is important to note that no protection of the indole
nitrogen atom was necessary (entry 11).
These results clearly indicate that the cyclization requires an
electron rich aromatic ring to afford good efficiency to provide
the 3-benzazepine ring system, findings which are congruent with
the recent results of Inui and coworkers toward the synthesis of
2-benzazepines.18 On the basis of these findings, we suggest a reac-
tion mechanism as depicted in Figure 2, in which the coordination
of the bismuth metal to the allylic functionality can lead to two
distinct modes of activation. First, the Lewis acid may activate
0a
>98%b
Tracea
0a
40%c
23%c
Tracea
Tracea
Tracea
71%c
>98%b
33%b,d
>98%b,e
57%c
44%c
18%c
0a
BiBr3 (20 mol %)
Bi(NO3)3Á5H2O (20 mol %)
Bi(0) (20 mol %)
a
b
c
Starting material recovered.
Isolated yield.
NMR yield using 1,2,3-trimethoxybenzene as an internal standard.
Run without 4 Å molecular sieves.
Run without 4 Å molecular sieves, but 1 equiv of Hunig’s base added.
d
e
our investigation by employing this In(III) salt. As a test case, we
sought to induce the cyclization of allylic bromide 5a to vinyl-
substituted 3-benzazepine 6a (Table 1). Unfortunately, our first
attempts with catalytic InCl3 in dichloromethane at room temper-
ature failed to facilitate the reaction even with 50 mol % (entries 1
and 2), although a full equivalent was successful in affording 6a in
almost quantitative yield (entry 3). The use of InBr3 also failed to
induce cyclization at 20 mol %, as did indium metal itself (entries
4 and 5).
A survey to identify a more suitable alternative found that the
reaction was promoted to varying degrees by a few other Lewis
acids. Silver salts Ag2OÁ4B2O3 and AgOTf did afford low amounts
of product, with the balance of starting material 5a converted into
byproducts (Table 1, entries 6 and 7). Application of ZnCl2, FeCl3,
and TiCl4 to the reaction resulted in only trace amounts of product,
with full recovery of unreacted starting material (entries 8–10).
Although a catalytic amount of SnCl4 was effective in promoting
the reaction (entry 11), it was not selected for use due to its toxicity
issues.14 A number of other common Lewis acids also failed to pro-
mote the IMFC reaction (data not shown, see Supplementary data).
The utility of bismuth(III) salts has attracted recent attention
due to their low toxicity, low cost, and good stability and demon-
strated high catalytic efficiency, rendering them suitable Lewis
acids for green chemistry.15 Their use in Lewis acid-promoted
addition of aromatic nucleophiles to electron-deficient alkenes
made it a natural choice to facilitate our system. We were therefore
pleased to find that 20 mol % of Bi(OTf)3 was equally as effective as
100 mol % of InCl3 (Table 1, entries 12 vs 3), providing 6a from 5a
in almost quantitative isolated yield. It was interesting to note that
the efficacy drops off significantly in the absence of molecular
sieves, providing only 33% yield and substantial decomposition
byproducts (entry 13). In a simple experiment to probe whether
the molecular sieves perform the role of an acid scavenger,16 we
subjected 5a to conditions without molecular sieves, but with full
the allylic moiety by coordinating to both the olefin
p-system as
well as to the bromide to generate a neutral intermediate 7a or
continue to displace the halide leading to an ionic species 7b. Sup-
posing a pre-organization of the linear molecule into a boat-chair
type conformation, transition through either 7a or 7b allows elec-
trophilic addition placing the newly-formed vinylic substituent in
a pseudo-equatorial disposition. One could expect that electron-
donating substitution on the aromatic ring would better stabilize
the cationic cyclization adduct 8, which facilitates rearomitization
to generate the desired aryl-fused seven-membered ring product
6.19 This mechanism may also explain the detrimental effect of
donor atom-containing solvents on this IMFC reaction. Some
mechanistic investigations are in progress in order to better define
the mode of this unprecedented cyclization.

2020
R. B. Kargbo et al. / Tetrahedron Letters 54 (2013) 2018–2021
Table 2
Bi(OTf)3-catalyzed IMFC formation of 3-benzazepines and azepino[4,5-b]heterocyclic ring systems 6a
Lewis acid
NTs
R
4 Å mol. seives
R
NTs
CH2Cl2 (0.05 M)
rt, 16 h
Br
5a-l
(+/-)-6a-l
Entry
1
Product
Yieldb
Entry
Product
Yieldb
60%
(78%)
H3CO
H3C
NTs
NTs
NTs
2 : 1
98%
7
8
9
6a
H3C
H3CO
(+/-)-
ortho-6g
(+/-)-
para-6g
(+/-)-
75%
(98%)
30%
(55%)
O
O
NTs
NTs
NTs
2
3
O
9 : 1
(+/-)-ortho-6b
O
(+/-)-6h
(+/-)-para-6b
H3CO
72%
(95%)
0
(Trace)
NTs
NTs
NTs
Cl
H3CO
H3CO
H3CO
9 : 1
(+/-)-6i
(+/-)-par a-6c
(+/-)-ortho-6c
65%
(88%)
65%
(89%)
NTs
NTs
H3CO
4
5
6
10
11
12
(+/-)-6d
(+/-)-6j
50%
(77%)
NTs
NTs
98%c
98%c
H3C
N
H
(+/-)-6e
(+/-)-6k
62%
(82%)
H3C
NTs
NTs
NTs
S
H3C
H3C
(+/-)-6l
9 : 1
H3C
para-6f
(+/-)-
ortho-6f
(+/-)-
a
b
c
Bi(OTf)3 (20 mol %), 4 Å molecular sieves, CH2Cl2 (0.05 M), rt, 16 h.
Yield of isolated product; yield in parenthesis indicates reaction with 1 equiv of Bi(OTf)3.
Reaction carried out at 0 °C.
mechanistic investigations are in progress in order to better under-
stand the mode of cyclization, and extension to an asymmetric var-
iant to this transformation will be communicated in due course.
N
R
N
Ts
R
Ts
or
OTf
Br
Bi
Bi(OTf)3
Acknowledgment
OTf
Bi
Br
OTf
TfO
OTf
7a
7b
OTf
The authors wish to thank AMRI colleague Dr. Joel Walker for
many helpful discussions during the preparation of this
manuscript.
R
Br
N
Ts
NTs
Supplementary data
R
H
HBr
6
8
(+/-)-
Supplementary data associated with this article can be found, in
theonlineversion, at http://dx.doi.org/10.1016/j.tetlet.2013.02.012.
Figure 2. Proposed modes of cyclization.
References and notes
1. (a) Epstein, O. L.; Lee, S.; Cha, J. K. Angew. Chem., Int. Ed. 2006, 45, 4988; (b) Yet,
L. Chem. Rev. 2000, 100, 2963; (c) Kuroda, C.; Suzuki, H. Curr. Org. Chem. 2003, 7,
115; (d) Zou, Y.; Garayalde, D.; Wang, Q.; Nevado, C.; Goeke, A. Angew. Chem.,
Int. Ed. 2008, 47, 10110; (e) Justicia, J.; Oller-Lopez, J. L.; Campana, A. G.; Oltra, J.
E.; Cuerva, J. M.; Bunuel, E.; Cardenas, D. J. J. Am. Chem. Soc. 2005, 127, 14911;
In conclusion, we have defined a mild intramolecular Friedel–
Crafts cyclization (IMFC) catalyzed by an environmentally benign
Bi(III) salts for the preparation of 3-benzazepines products and
their corresponding azepino[4,5-b]heterocyclic congeners. Some

R. B. Kargbo et al. / Tetrahedron Letters 54 (2013) 2018–2021
2021
(f) Dzwiniel, T. L.; Stryker, J. M. J. Am. Chem. Soc. 2004, 126, 9184; (g) Soto, S.;
Vaz, E.; Dell’Aversana, C.; Alvarez, R.; Altucci, L.; de Lera, A. R. Org. Biomol. Chem.
2012, 10, 2101; (h) Enders, D.; Lenzen, A.; Raabe, G. Angew. Chem., Int. Ed. 2005,
44, 3766. and references therein.
13. (a) Hayashi, R.; Cook, G. R. Org. Lett. 2007, 9, 1311; (b) Kaneko, M.; Hayashi, R.;
Cook, G. R. Tetrahedron Lett. 2007, 48, 7085; (c) Cook, G. R.; Hayashi, R. Org. Lett.
2006, 8, 1045.
14. (a) Scott, L. E.; Orvig, C. Chem. Rev. 2009, 109, 4885; (b) Gallagher, W. P.;
Terstiege, I.; Maleczka, R. E., Jr. J. Am. Chem. Soc. 2001, 123, 3194.
2. (a) Flatt, B.; Martin, R.; Wang, T.-L.; Mahaney, P.; Murphy, B.; Gu, X.-H.; Foster,
P.; Li, J.; Pircher, P.; Petrowski, M.; Schulman, I.; Westin, S.; Wrobel, J.; Yan, G.;
Bischoff, E.; Daige, C.; Mohan, R. J. Med. Chem. 2009, 52, 904; (b) Lundquist, J. T.,
IV; Harnish, D. C.; Kim, C. Y.; Mehlmann, J. F.; Unwalla, R. J.; Phipps, K. M.;
Crawley, M. L.; Commons, T.; Green, D. M.; Xu, W.; Hum, W.-T.; Eta, J. E.;
Feingold, I.; Patel, V.; Evans, M. J.; Lai, K.; Borges-Marcucci, L.; Mahaney, P. E.;
Wrobel, J. E. Bioorg. Med. Chem. Lett. 2010, 53, 1774.
3. Wu, W.-L.; Burnett, D. A.; Spring, R.; Greenlee, W. J.; Smith, M.; Favreau, L.;
Fawzi, A.; Zhang, H.; Lachowicz, J. E. J. Med. Chem. 2005, 48, 680.
4. (a) Hajer, A.; Bastien, N. Nat. Prod. Rep. 2012, 29, 845; (b) Li, W. Z.; Duo, W.;
Zhuang, C. Org. Lett. 2011, 13, 3538; (c) Hiroshi, M.; Yuta, N.; Takahiro, H.;
Wiwied, E.; Aty, W.; Kanami, M.; Setsuko, S.; Yusuke, H. Heterocycles 2010, 81,
441.
5. (a) Yu, D.; Hui-Ying, H.; Hui, L.; Yuan-Ping, R.; Pei-Qiang, H. Synlett 2011, 565;
(b) Kuan-Hon, L.; Osamu, H.; Kanki, K.; Takashi, K.; Masahiko, H.; Toh-Seok, K. J.
Nat. Prod. 2007, 70, 1302; (c) Kuan-Hon, L.; Toh-Seok, K. Org. Lett. 2006, 8, 1733.
6. (a) Bandini, M.; Tragni, M.; Umani-Ronchi, A. Adv. Synth. Catal. 2009, 351, 2521;
(b) Clark, R. D.; Weinhardt, K. K.; Berger, W. J.; Fisher, L. E.; Brown, C. M.;
MacKinnon, A. C.; Kilpatrick, A. T.; Spedding, M. J. Med. Chem. 1990, 33, 633.
7. (a) Tietze, L. F.; Thede, K.; Sannicolò, F. Chem. Commun. 2000, 7, 583; (b) Tietze,
L. F.; Schimpf, R. Angew. Chem., Int. Ed. 1994, 33, 1089; (c) Stewart, S. G.; Heath,
C. H.; Ghisalberti, E. L. Eur. J. Org. Chem. 2009, 1934.
15. (a) Leonard, N. M.; Wieland, L. C.; Mohan, R. S. Tetrahedron 2002, 58, 8373; (b)
Briand, G. G.; Burford, N. Chem. Rev. 1999, 99, 2601; (c) Ge, R. G.; Sun, H. Z. Acc.
Chem. Res. 2007, 40, 267; (d) Yang, N.; Sun, H. Coord. Chem. Rev. 2007, 251,
2354; (e) Suzuki, H.; Matano, Y. Organobismuth Chemistry; Elsevier:
Amsterdam, 2001; (f) Varala, R.; Alam, M. M.; Adapa, S. R. Synlett 2003, 67;
(g) Le Roux, C.; Dubac, J. Synlett 2002, 181; (h) Loh, T.-P.; Chua, G.-L. Advances
in Organic Synthesis In Atta-ur-Rahman, Ed.; Bentham Science Publishers: The
Netherlands, 2005; p 173. and references therein.
16. It has been shown that organic reaction mixtures containing 4 Å molecular
sieves (Type 4A sodium aluminosilicates) are typically basic in nature, with
measured pH of slurry solutions in the pH 9–11 range. For more information,
see: (a) Davis, M. E. Ind. Eng. Chem. Res. 1991, 30, 1675 and references therein;
(b) Herr, R. J.; Meckler, H.; Scuderi, Jr., F. Org. Process Res. Dev. 2000, 4, 43; (c)
Sigma Aldrich Technical Information Bulletin AL-143: Mineral Adsorbents,
Filter Agents, and Drying Agents (Accessed 1/14/13).
17. General experimental procedure: Bi(OTf)3 (0.05 mmol) and 4 Å molecular sieves
(200 mg) were placed in a small screw-cap scintillation vial equipped with a
magnetic stir bar. Allylic halide (0.5 mmol) and dichloromethane (1 mL) were
added and the mixture was stirred at room temperature for 16 h. The mixture
was filtered through a pad of Celite, eluting with ethyl acetate, and the solvent
was removed under reduced pressure. The residue was purified by column
chromatography on silica gel to afford analytically pure cyclized product. See
Supplementary data for details concerning preparation of precursors.
18. So, M.;Kotake, T.; Matsuura, K.; Inui, M.;Kamimura, A. J. Org. Chem. 2012, 77, 4017.
19. Alternatively, the concept of Lewis acid-assisted Brønsted acid (LBA) may not
be entirely ruled out as a mechanism. See: (a) Kanno, O.; Kuriyama, W.; Wang,
Z. J.; Toste, F. D. Angew. Chem., Int. Ed. 2011, 50, 9919; (b) Yamamoto, H.;
Futatsugi, K. Angew. Chem., Int. Ed. 2005, 44, 1924; A mechanism proceeding via
proton transfer likewise cannot be discounted. See: (c) Rosenfeld, D. C.;
Shekhar, S.; Takemiya, A.; Utsunomiya, M.; Hartwig, J. F. Org. Lett. 2006, 8,
4179; (d) Lemière, G.; Cacciuttolo, B.; Belhassen, E.; Duñach, E. Org. Lett. 2012,
14, 2750.
8. Cid, M. M.; Dominguez, D.; Castedo, L.; Vazquez-Lopez, E. M. Tetrahedron 1999,
55, 5599.
9. Eckelbarger, J. D.; Wilmot, J. T.; Gin, D. Y. J. Am. Chem. Soc. 2006, 128, 10370.
10. (a) Wada, Y.; Kaga, H.; Uchiito, S.; Kumazawa, E.; Tomiki, M.; Onozaki, Y.;
Kurono, N.; Tokuda, M.; Ohkuma, T.; Orito, K. J. Org. Chem. 2007, 72, 7301; (b)
Wirt, U.; Schepmann, D.; Wünsch, B. Eur. J. Org. Chem. 2007, 462.
11. Noyori, R.; Hayakawa, Y. Org. React. 1983, 29, 163.
12. For a general review of Friedel–Crafts reactions, see: Olah, G. A.; Krishnamurti,
R.; Prakash, G. K. S. In Friedel–Crafts Alkylations in Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 3, p
293.