Synthesis of substituted indole from 2-aminobenzaldehyde through [1,2]-aryl shift

Article abstract of DOI:10.1021/jo1016713

A mild, efficient, and simple method for the synthesis of 3-ethoxycarbonylindoles has been developed. Addition of ethyl diazoacetate (EDA) to 2-aminobenzaldehydes cleanly affords the indole core. As opposed to other common approaches for the synthesis of indole, this method displays both excellent functional group tolerance and perfect regiochemical control. This allowed the synthesis of a variety of useful indole building blocks from 2-aminobenzaldehydes derived from readily available anthranilic acids.

Full text of DOI:10.1021/jo1016713

pubs.acs.org/joc
step, resulting in a poor regiochemical control for unsymme-
Synthesis of Substituted Indole from
2-Aminobenzaldehyde through [1,2]-Aryl Shift
trical ketones.4 For the Larock indole synthesis and its variants,
the necessity for transition metals and bases can limit the
substrate scope of the reaction.5
ꢀ
Patrick Levesque and Pierre-Andre Fournier*
An interesting alternative to the indole scaffold has been
reported by Pei and co-workers in which they generated
indoles directly from 1-(2-aminophenyl)-2-chloroethanone
and a Grignard reagent (Figure 1, top).6 Performing an
addition of an organometallic compound to this ketone
affords a benzylic alkoxide. The presence of a chlorine atom
at the R position promotes subsequent [1,2]-aryl migration,
generating 1-(2-aminophenyl)acetone. The indole core was
then rapidly formed following condensation of the aniline
onto the ketone. Although interesting, this strategy is limited
by the use of Grignard reagents and the availability of the
requisite starting R-chloroketone.
Merck Frosst Center for Therapeutic Research, 16711 Trans
Canada Highway, Kirkland, Qc, Canada, H9H 3L1
pierreandrefournier@gmail.com
Received August 25, 2010
Hossain and co-workers investigated a similar, but step-
wise approach to Pei’s indole synthesis.7,8 The requisite
benzylic alkoxide intermediate in this case was accessed via
acid-catalyzed addition of ethyl diazoacetate (EDA) to 2-nitro-
benzaldehyde (Figure 1, bottom). Subsequent [1,2]-aryl migra-
tion then furnished 2-(2-nitrophenyl)-3-oxopropanoate. The
desired indole core was obtained after reduction of the nitro
group and subsequent condensation of the aniline onto the
aldehyde. Although the use of Grignard reagents is success-
fully obviated, the main downside of this stepwise procedure
is that the hydrogenation step might be incompatible with
various functional groups on the indole core such as nitro and
bromine, groups that could allow for further functionalization.
Both Pei’s and Hossain’s strategy are similar in that they both
generate a benzylic alkoxide that undergoes [1,2]-aryl migration
via displacement of a suitable leaving group at the R-position.
On the other hand, the two strategies differ with regard to the
nucleophile used, and while one occurs in a basic media, the other
one is acid catalyzed. Here, we report a simple and straight-
forward addition of ethyl diazoacetate to readily available
A mild, efficient, and simple method for the synthesis of
3-ethoxycarbonylindoles has been developed. Addition
of ethyl diazoacetate (EDA) to 2-aminobenzaldehydes
cleanly affords the indole core. As opposed to other common
approaches for the synthesis of indole, this method dis-
plays both excellent functional group tolerance and per-
fect regiochemical control. This allowed the synthesis of a
variety of useful indole building blocks from 2-amino-
benzaldehydes derived from readily available anthranilic
acids.
The indole core represents one of the ubiquitous hetero-
cyclic scaffolds in natural products and pharmaceutical
compounds.1 The synthesis of this core has been mainly depen-
dent on two approaches: Fisher indole-type reactions and
variations of Larock indole synthesis.2,3 The most glaring
limitation of the Fisher indole-type reaction is the lack of stereo-
chemical control during the electrophilic aromatic substitution
(5) For a review on metal-catalyzed synthesis of indoles, see: (a) Zeni, G.;
Larock, R. C. Chem. Rev. 2004, 104, 2285–2210. (b) Cacchi, A.; Fabrizi, G.
Chem. Rev. 2005, 105, 2873–2920. (c) Alonso, F.; Beletskaya, I. P.; Yus, M.
Chem. Rev. 2004, 104, 3079–3160. For recent variants of the Larock indole
synthesis, see: (d) Stuart, D. R.; Bertrand-Laperle, M.; Burgess, K. M. N.;
Fagnou, K. J. Am. Chem. Soc. 2008, 130, 16474–16475. (e) Leogane, O.;
Lebel, H. Angew. Chem., Int. Ed. 2008, 47, 350–352. (f) Wurtz, S.; Neumann,
J. J.; Droge, T.; Glorius, F. Angew. Chem., Int. Ed. 2008, 47, 7230–7233.
(g) Jia, Y.; Zhu, J. J. Org. Chem. 2006, 71, 7826–7834. (h) Nakamura, Y.;
Ukita, T. Org. Lett. 2002, 4, 2317–2320.
(1) (a) Somei, M.; Yamada, F. Nat. Prod. Rep. 2004, 21, 278–311. (b) Somei,
M.; Yamada, F. Nat. Prod. Rep. 2005, 22, 73–103. (c) Horton, D. A.; Bourne,
G. T.; Smythe, M. L. Chem. Rev. 2003, 103, 893–930. (d) Kawasaki, T.; Higuchi,
K. Nat. Prod. Rep. 2007, 24, 843–868. (e) Rahman, A.; Basha, A. Indole
Alkaloids; Harwood Academic Publishers: Amsterdam, The Netherlands,
1998; p 141. (f) Gupta, R. R. Heterocyclic Chemistry; Springer Publishing:
New York, 1999; Vol. 2, p 193.
(6) (a) Pei, T.; Chen, C.-Y.; Dormer, P. G.; Davies, I. W. Angew. Chem.,
Int. Ed. 2008, 47, 4231–4233. (b) Pei, T.; Tellers, D. M.; Streckfuss, E. C.;
Chen, C.-Y.; Davies, I. W. Tetrahedron 2009, 65, 3285–3291.
(7) (a) Islam, M. S.; Brennan, C.; Wang, Q.; Hossain, M. M. J. Org.
Chem. 2006, 71, 4675–4C677. (b) Dudley, M. E.; Morshed, Md. M.; Brennan,
C. L.; Islam, M. S.; Ahmad, M. S.; Atuu, M.-R.; Branstetter, B.; Hossain,
M. M. J. Org. Chem. 2004, 69, 7599–7608.
(2) For a general review on indole synthesis, see: (a) Humphrey, G. R.;
Kuethe, J. T. Chem. Rev. 2006, 106, 2875–2911. (b) Gribble, G. W. J. Chem.
Soc., Perkin Trans. 1 2000, 1045–1075.
(3) For some recent methods, see: (a) Du, Y.; Liu, R.; Linn, G.; Zhao, K.
Org. Lett. 2006, 8, 5919–5922. (b) Du, Y.; Chang, J.; Reiner, J.; Zhao, K.
J. Org. Chem. 2008, 73, 2007–2010. (c) Schlosser, M.; Ginanneschi, A.;
Leroux, F. Eur. J. Org. Chem. 2006, 2956–2969. (d) Kraus, G. A.; Guo, H.
Org. Lett. 2008, 10, 3061–3063.
(8) For leading references on acid-catalyzed reactions of substituted
diazomethanes and carbonyls, see: (a) Hashimoto, T.; Naganawa, Y.;
Maruoka, K. J. Am. Chem. Soc. 2008, 130, 2434–2435. (b) Hashimoto, T.;
Miyamoto, H.; Naganawa, Y.; Maruoka, K. J. Am. Chem. Soc. 2009, 131,
11280–11281. (c) Hashimoto, T.; Naganawa, Y.; Maruoka, K. J. Am. Chem.
Soc. 2009, 131, 6614–6617. (d) Moebius, D. C.; Kingsbury, J. S. J. Am. Chem.
Soc. 2009, 131, 878–879. (e) Wommack, A. J.; Moebius, D. C.; Travis, A. L.;
Kingsbury, J. S. Org. Lett. 2009, 11, 3202–3205. (f) Li, W.; Wang, J.; Hu, X.;
Shen, K.; Wang, W.; Chu, Y.; Lin, L.; Liu, X.; Feng, X. J. Am. Chem. Soc.
2010, 132, 8532–8533.
(4) Robinson, B. The Fisher Indole Synthesis; John Wiley and Sons:
Chichester, UK, 1982.
DOI: 10.1021/jo1016713
r
Published on Web 09/13/2010
J. Org. Chem. 2010, 75, 7033–7036 7033
2010 American Chemical Society

Levesque and Fournier
JOCNote
not be separated from a mixture of oligomeric side products.
This was not surprising since the tendency of 2-aminoben-
zaldehyde to form polymers is well documented.9 To prevent
the formation of oligomers, we tried to protect the aniline
moiety with various groups. The reaction failed to produce
any indole product with both N-Ts and N-Boc substituents
(entries 2 and 3), but an isolated yield of 59% was obtained
when an N-Bn group was used (entry 4). The use of benzyl as
protecting groups for indoles is well documented and they
can be removed under basic conditions,10 acidic conditions,11 or
reductive conditions,12 allowing the selection of the depro-
tection conditions that best fit the substrate.
With a suitable protecting group identified, the reaction
was optimized further. For comparison purposes, all reac-
tions were stopped after 5 h and yields were determined by
HPLC. Performing the reaction at room temperature gave a
71% yield of the desired product, with only trace amounts of
starting benzaldehyde 1a and EDA remaining (entry 5).13
The yield is not improved further when the reaction was
allowed to stir overnight.
FIGURE 1. Pei’s and Hossain’s strategies for indole synthesis.
We then investigated the use of other acids to promote this
reaction. No product was observed when Ti(Oi-Pr)4 was
used as a catalyst, presumably because it is too weak a Lewis
acid (entry 6). When TiCl4 was employed, only 61% of 2a
was obtained although complete consumption of EDA and
the starting aldehyde 1a was observed (entry 7). Going to
SnCl4 led to a further decrease in yield (29%) and numerous
unknown side products were observed on the HPLC trace
(entry 8). Tetrafluoroboric acid did afford product 2a in a
67% yield, but full conversion of 1a was not achieved
although the EDA was fully consumed (entry 9). Lastly,
we tried the reaction with methylaluminium bis(2,6-di-tert-
butyl-4-methylphenoxide) (MAD),14 which furnished the
indole product in a 72% yield (entry 10). However, since it
provided no notable advantage to the simpler and cheaper
TABLE 1. Reaction Optimization for the Direct Synthesis of Indoles by
Lewis Acid Promoted Addition of Ethyl Diazoacetate to 2-Aminobenzal-
dehydesa
entry
R
Lewis acid
solvent
CH2Cl2
yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
H
Ts
BF3 OEt2
3
0b,c
0b
BF3 OEt2
BF3 OEt2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
DMF
MeCN
PhMe
CH2Cl2
ClCH2CH2Cl
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
3
Boc
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
0b
3
BF3 OEt2
BF3 OEt2
56b
71
0
3
3
Ti(Oi-Pr)4
TiCl4
SnCl4
61
29
67
72
0
BF3 OEt2, we decided to keep the latter for all subsequent
3
reactions.
HBF4
MADd
BF3 OEt2
We tried changing solvent to THF or DMF, but no pro-
duct was observed (entries 11 and 12). The reaction can be
done in other solvents like acetonitrile, toluene, chloroform,
and 1,2-dichloroethane (entries 13 to 16), but none proved to
be better than dichloromethane.
3
BF3 OEt2
BF3 OEt2
0
3
42
60
63
65
60e
66f
55g
77h
3
BF3 OEt2
BF3 OEt2
BF3 OEt2
3
3
3
We also tried to vary the amount of reactants employed
in the reaction. Both increasing and reducing the amount
of BF3 OEt2 used gave lower yields (entries 17 and 18). Low-
BF3 OEt2
BF3 OEt2
3
3
BF3 OEt2
BF3 OEt2
3
3
ering the amount of EDA to 2 equiv gave a lower yield of
3
aReactions were performed in 300 μL of solvent on a 0.06 mmol scale
and were stopped systematically after 5 h by dilution in acetonitrile to a
(9) (a) Opie, J. W.; Smith, L. I. Organic Synthesis; Wiley: New York, 1955;
Vol. III, pp 56-58. (b) Li, J.-F.; Zhao, Y.; Cai, M.-M.; Li, X.-F.; Li, J.-X.
Eur. J. Med. Chem. 2009, 44, 2796–2806.
(10) (a) Haddach, A. A.; Kelleman, A.; Deaton-Rewolinski, M. V.
Tetrahedron Lett. 2002, 43, 399–402. (b) Yu, H.; Yu, Z. Angew. Chem., Int.
Ed. 2009, 48, 2929–2933. (c) Sanz, R.; Castroviejo, M. P.; Guilarte, V.; Perez,
A.; Fananas, F. J. J. Org. Chem. 2007, 72, 5113–5118.
(11) (a) Murakami, Y.; Watanabe, T.; Kobayashi, A.; Yokoyama, Y.
Synthesis 1984, 738–740. (b) Bennasar, M.-L.; Vidal, B.; Bosch, J. J. Org.
Chem. 1996, 61, 1916–1917. (c) Bennasar, M.-L.; Vidal, B.; Bosch, J. J. Org.
Chem. 1997, 62, 3597–3609. (d) Malapel-Andrieu, B.; Merour, J.-Y. Tetra-
hedron 1998, 54, 11079–11094. (e) Stearman, C. J.; Wilson, M.; Padwa, A.
J. Org. Chem. 2009, 74, 3491–3499.
(12) (a) Elhalem, E.; Bailey, B. N.; Docampo, R.; Ujvary, I.; Szajnman,
S. H.; Rodriguez, J. B. J. Med. Chem. 2002, 45, 3984–3999. (b) Pedras,
M. S. C.; Zheng, Q.-A.; Gadagi, R. A. Chem Commun. 2007, 368–370.
(13) Performing the reaction at a lower temperature or at various con-
centrations had no effect on reaction yield.
b
total volume of 50 mL. Yields were determined by HPLC. Reactions
performed on 100 mg of benzaldehyde and isolated yield after flash
chromatography are reported. Reagents were mixed at -78 °C and warmed
to rt overnight. cOligomeric mixture, see text for details. dMethylalumi-
nium bis(2,6-di-tert-butyl-4-methylphenoxide). e0.5 equiv of Lewis acid
was used. f1.5 equiv of Lewis acid was used. g2 equiv of EDA was used.
h5 equiv of EDA was used.
2-aminobenzaldehydes to afford 3-ethoxycarbonylindoles
directly under completely chemoselective conditions.
Initially, when 3 equiv of EDA was combined with 1 equiv
of BF3 OEt2 and 1 equiv of 2-aminobenzaldehyde at -78 °C, a
3
small amount of indole was observed after the reaction mix-
ture was allowed to warm to room temperature overnight
(Table 1, entry 1). Unfortunately, the desired product could
(14) Maruoka, K.; Concepcion, A. B.; Yamamoto, H. J. Org. Chem.
1994, 59, 4725–4726.
7034 J. Org. Chem. Vol. 75, No. 20, 2010

Levesque and Fournier
JOCNote
consumption of 1a (entry 20). We thus established that the
TABLE 2. Synthesis of Various Substituted Indolesa
optimal conditions were to use 5.0 equiv of EDA and 1.0
equiv of BF3 OEt2 at room temperature, at a concentration
of 0.2 M in CH2Cl2.
3
We repeated the reaction on 1a under optimized condi-
tions and 2a was isolated in a 74% yield (Table 2, entry 1).
This represents an 18% improvement over the isolated yield
obtained with unoptimized conditions (Table 1, entry 4). We
then synthesized multiple aldehydes 1 bearing different
substituents to see how it impacts reactivity.15 Not surpris-
ingly, the reaction tolerated various N-alkyl groups on the
aniline including PMB and methyl groups. Both indoles 2b
and 2c were prepared from their corresponding 2-aminoben-
zaldehyde 1b and 1c in 77% and 74% yield, respectively
(entries 2 and 3). The reaction gives higher yields when the
aromatic ring is electron deficient (entries 4-8). The method
is also tolerant of bromine and iodine substituents (entries 5
and 6), both of which could be subsequently functionalized
by using cross-coupling reactions for the generation of more
diversified intermediates. It is important to note that all the
reactions with electron-neutral and deficient benzaldehydes
1 were very rapid and completed within 30 min, except for
the case of 1d where the benzaldehyde is ortho disubstituted
(60 min reaction time).
Performing the reaction with a more electron-rich benzal-
dehyde 1 gave lower yields (entry 9) and the reaction time was
also much longer (2 h). The presence of meta and para alkyl
groups had little influence on the reaction yields (entries 11
and 12) although ortho substitution did lead to a lower yield
(entry 10).
The presence of another fused aromatic ring was tolerated
since the reaction of the naphthalene derivative 1m furnished
2m in 80% yield (entry 13). The method is also suitable for
the preparation of 7-azaindole derivative 2n, which was
obtained in 52% yield from the corresponding aldehyde
1n. The lower yield in this case can again be attributed to
the lower reactivity observed with electron-rich benzalde-
hydes. Finally, we tried the reaction with the acetophenone
derivative 1o, which furnished the desired 2,3-disubstituted
indole 2o in a still respectable 46% yield. The reaction time
required was much longer for this substrate.16
In conclusion, we developed reaction conditions for the
direct transformation of N-alkyl-2-aminobenzaldehyde to
3-ethoxycarbonylindoles. The reaction proceeds in moderate
to good yields from readily available 2-(alkylamino)benzalde-
hyde. A wide variety of 4-, 5-, and 6-substituted indole cores
can be synthesized under very mild conditions with this pro-
cedure. Further studies to extend the scope of this transfor-
mation to include disubstituted diazo as well as acetophe-
none derivatives are currently underway.
Experimental Section
General Procedure for the Synthesis of Indoles: Ethyl 1-Ben-
zylindole-3-carboxylate (2a). To a solution of 2-(benzylamino)-
benzaldehyde (1a) (100 mg, 0.47 mmol) and EDA (243 μL, 2.18
mmol) in DCM (2.4 mL) at 0 °C was added BF3 OEt2 (60 μL,
0.47 mmol) dropwise. The ice bath was removed upon completion
3
aReaction conditions: 1 (1 equiv) and EDA (5 equiv) in CH2Cl2 [0.2
b
M], add BF3 OEt2 (1 equiv) at 0 °C, stirred at rt. Isolated yield after
3
column chromatography. cSee text for details.
(15) Benzaldehydes 1 were readily prepared on gram scale from the
corresponding anthranilic acids. See the Supporting Information for details.
(16) Further investigations revealed that the desired addition/1,2-aryl
shift step occurred quite rapidly (less than 2 h), but the subsequent con-
densation was very long.
55% (entry 19). On the other hand, increasing the amount of
EDA to 5.0 equiv improved the yield to 77% and allowed full
J. Org. Chem. Vol. 75, No. 20, 2010 7035

Levesque and Fournier
JOCNote
of the addition and the solution was stirred at room temperature
for 30 min. The reaction mixture was poured into a saturated
aqueous solution of NaHCO3 and it was extracted (3ꢀ) with
DCM. The combined organic layers were dried over MgSO4,
filtered, and concentrated under reduced pressure to yield an
orange oil. The crude residue was purified by flash chromatog-
raphy (10:1 hexanes:EtOAc) to afford 97 mg (74%) of com-
pound 2a as a white solid. 1H NMR (400 MHz, CDCl3) δ 8.26
(d, J=7.64 Hz, 1 H), 7.90 (s, 1 H), 7.37-7.24 (m, 6 H), 7.19 (d,
J=6.76 Hz, 2 H), 5.37 (s, 2 H), 4.53-4.36 (m, 2 H), 1.47 (t, J=
7.10 Hz, 3 H). 13C NMR (101 MHz, CDCl3) δ 164.7, 136.4,
135.6, 134.2, 128.6, 127.7, 126.6, 126.4, 122.5, 121.6, 121.4, 109.9,
107.5, 59.3, 50.3, 14.2. HRMS (ESI) m/z calcd for C18H18NO2
[M þ H]þ 280.1338, found 280.1344.
Acknowledgment. The author would like to thank Merck
Frosst for financial support.
Supporting Information Available: Procedural and spectral
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
7036 J. Org. Chem. Vol. 75, No. 20, 2010