Scheme 1
.
sp3 C-H Activation of N-Alkyl-2-bromoaniline
Table 1. Optimization of Reaction Conditions
Derivatives
yield (%)a
entry
ligandb
base
additive
2
1
of N-alkyl-2-bromoaniline derivatives (Scheme 1) would
provide a straightforward access to the construction of the
indoline framework that is a common structural motif found
in many biologically active compounds including indoline
alkaloids.10,11 Herein, we report the palladium-catalyzed
cyclization of functionalized N-alkyl-2-bromoanilines via sp3
C-H activation, leading to various indolines, as well as the
2,3-dihydropyrrolo[3,2-b]pyridine derivative.12 This study
includes the first palladium-catalyzed sp3 C-H activation
of a simple alkyl group without the assistance of a pyridine
or quaternary carbon (R1 and/or R2 ) H, Scheme 1).
We first investigated the palladium-catalyzed cyclization
of N-protected 2-bromo-N-tert-butylaniline 1, since this
would have the benefit of a quaternary carbon as well as
nine reactive C-H bonds in the vicinity of the palladium
atom when forming the arylpalladium(II) intermediate.
Among various ligands examined (Table 1, entries 1-5),
1
2
3
4
5
Davephos
PPh3
XPhos
K2CO3
K2CO3
K2CO3
56
51
6
2
66
96
20
46
94
93
33
4
P(t-Bu)3·HBF4 K2CO3
PCy3·HBF4
PCy3·HBF4
PCy3·HBF4
PCy3·HBF4
PCy3·HBF4
PCy3·HBF4
K2CO3
6
7
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
t-BuCO2H >99 (98)
8c
9d
10
t-BuCO2H
t-BuCO2H
52
97
5
2
e
f
g
t-BuCO2H >99 (98)
a Yields based on 1H NMR (isolated yields in parentheses). b Structures
of DavePhos and XPhos are shown below. c DMA was used in place of
xylene. d Reaction time was 5 h. e Pd/ligand ) 2/4 mol %. f Pd/ligand )
3/6 mol %. g With 1.4 equiv of Cs2CO3.
13
PCy3·HBF4 was found to be the most effective for this
cyclization: treatment of aniline 1 with Pd(OAc)2 (5 mol %)
in the presence of PCy3·HBF4 (10 mol %) and K2CO3 at 140
°C gave the desired indoline 2 in 66% yield with 33%
recovery of the starting material (entry 5). When Cs2CO3
was used instead of K2CO3, the yield was improved to 96%
(entry 6). A more promising result was obtained by the
addition of pivalic acid (30 mol %) to the reaction mixture,
as reported by Fagnou and co-workers quite recently (entry
7).3d,9f,12 Polar solvents such as DMA were not suitable for
this cyclization (entry 8). Lowering the catalyst loading to 2
mol % slightly decreased the yield of 2 (97%, entry 9), while
a comparable result to entry 7 was obtained when using 3
mol % of Pd(OAc)2 and 1.4 equiv of Cs2CO3 (quant., entry
10).
Under the optimized conditions (Table 1, entry 10), we
examined the reaction of a variety of 2-halo-N-tert-butyla-
niline derivatives (Table 2). First, we investigated the
influence of the halogen atom on the cyclization and found
the chloride 3 and iodide 4 showed lower reactivity than the
bromide 1 to afford the indoline 2 in 47% and 84% respective
yields (entries 1 and 2).14 The reaction of the N-unprotected
substrate 5a led to the recovery of the starting material (93%)
without producing any detectable amounts of the desired
indoline 5b (entry 3).15 In contrast, N-trifluoroacetamide 6a
gave 87% yield of the corresponding indoline 6b. Next, the
reaction of bromoanilines 7a-13a bearing various substit-
uent(s) on the arene was investigated (entries 5-10). The
monomethyl and dimethyl-substituted anilines 7a and 8a
underwent smooth ring closure providing the indolines 7b
(9) (a) Dyder, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 103. (b)
Baudoin, O.; Herrbach, A.; Gue´ritte, F. Angew. Chem., Int. Ed. 2003, 42,
5736. (c) Barder, T. E.; Walker, S. D.; Martinelli, J. R.; Buchwald, S. L.
J. Am. Chem. Soc. 2005, 127, 4685. (d) Hitce, J.; Retailleau, P.; Baudoin,
O. Chem. Eur. J. 2007, 13, 792. (e) Giri, R.; Maugel, N.; Li, J.-J.; Wang,
D.-H.; Breazzano, S. P.; Saunders, L. B.; Yu, J.-Q. J. Am. Chem. Soc. 2007,
129, 3510. (f) Lafrance, M.; Gorelsky, S. I.; Fagnou, K. J. Am. Chem. Soc.
2007, 129, 14570
.
(10) (a) Llabres, J. M.; Viladomat, F.; Bastida, J.; Codina, C.; Rubiralta,
M. Phytochemistry 1986, 25, 2637. (b) Ghosal, S.; Rao, P. H.; Jaiswal,
D. K.; Kumar, Y.; Frahm, A. W. Phytochemistry 1981, 20, 2003. (c) Kam,
T.-S.; Subramaniam, G.; Lim, K.-H.; Choo, Y.-M. Tetrahedron Lett. 2004,
(14) Decreased yield in entry 2 can be attributed to catalyst poisoning
by the iodide anion, see for example: Campeau, L.-C.; Parisien, M.; Jean,
A.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 581.
(15) (a) The unsuccessful result with the N-unprotected substrate 5a
might be due to the formation of the four-membered azapalladacycle 18.
For related azapalladacycles, see: Sole´, D.; Serrano, O. Angew. Chem., Int.
Ed. 2007, 46, 7270. (b) Sole´, D.; Vallverdu´, L.; Solans, X.; Font-Bardia,
M.; Bonjoch, J. J. Am. Chem. Soc. 2003, 125, 1587. (c) See also: Gies,
A.-E.; Pfeffer, M.; Sirlin, C.; Spencer, J. Eur. J. Org. Chem, 1999, 8, 1957.
(d) Vicente, J.; Abad, J.-A.; Frankland, A. D.; Ram´ırez de Arellano, M. C.
Chem. Eur. J. 1999, 5, 3066.
45, 5995
.
(11) For synthesis of indolines, see: (a) Witulski, B.; Stengel, T. Angew.
Chem., Int. Ed. 1999, 38, 2426. (b) Thansandote, P.; Raemy, M.; Rudolph,
A.; Lautens, M Org. Lett. 2007, 9, 5255, and references therein
.
(12) During the course of our investigation and preparation of this
manuscript, a related Pd(0)-catalyzed sp3 C-H activation of 2-tert-butoxy-
bromobenzene in the presence of pivalic acid and its related substrates to
form 2,2-dialkyldihydrobenzofurans has appeared in the literature: see ref
9f. Mechanistic consideration as well as the effect of pivalic acid based on
DFT calculations was also reported.
(13) This phosphonium salt was used because it is more stable than the
corresponding free phosphine.
1760
Org. Lett., Vol. 10, No. 9, 2008