During the past few years, transition-metal-catalyzed
direct benzylic CꢀH functionalization of 2-alkylazaarenes
has attracted much attention.7 These reactions were sup-
posed to be initiated by the formation of metal enamide
species through the electrophilic metalation of 2-alkyla-
zaarenes. In the course of our ongoing efforts devoted
toward studying transition-metal-catalyzed CꢀH func-
tionalization,8 we discovered that, by virtue of copper
as an inexpensive catalyst, indolizine derivatives were able
to be accessed directly by annulation of 2-alkylazaarenes
and cinnamic acids featuring CꢀH olefination and
decarboxylative amination processes. This methodology
addresses many current limitations for the synthesis of
indolizines, furnishing a diverse collection of valuable C-2
arylated indolizines using readily and commercially avail-
able reagents.9
the base were employed (Table 1, entry 22). When the
reaction was conducted at a lower temperature, it pro-
ceeded with a lower yield, and a higher reaction tempera-
ture did not help increase the yield (Table 1, entry 23 and
Supporting Information). Further attempts to substitute
DMF with other solvents failed to yield better results (see
Supporting Information).
Table 1. Reaction Optimizationa
Our initial attempt started with the reaction of 2-ethyl-
pyridine and cinnamic acid at 140 °C under a N2 atmo-
sphere. Screening of the copper catalysts (Table 1, entries
1ꢀ6) revealed that Cu(OAc)2 was optimal to give the
annulation product 3a in 45% yield (Table 1, entry 6). In
contrast to other copper-catalyzed CꢀH activation
processes,10 the reaction failed to work under oxidative
reaction conditions (Table 1, entry 7 and Supporting
Information). Surprisingly, the product was isolated in
39% yield when copper bronze was employed as the
catalyst (Table 1, entry 8), and the reaction failed again
when the atmosphere was switched from N2 toO2 (Table 1,
entry 9). It was found that employing 2 equiv LiOAc as an
additive elevated the yield to51% (Table 1, entry 10), while
other additives such as NaOAc, KOAc, and LiCl were less
effective (Table 1, entries 11ꢀ13). To improve the yield, we
exploited active metal powders in the Cu(OAc)2 system to
generate in situ Cu(0) species (Table 1, entries 14ꢀ17).
Gratifyingly, adding 0.5 equiv of nickel powder increased
the yield of the product to 62% (Table 1, entry 17). No
reaction was observed in the absence of copper catalyst
(Table 1, entry 18), and the ligand (1,10-phenanthroline)
was required to obtain a higher yield (Table 1, entry 19 and
Supporting Information). It was found that, by extending
the reaction time to 36 h, the yield of the product reached
the highest yield (66%) (Table 1, entry 20). A 3 equiv
amount of2-ethylpyridinesubstratewas required toget the
best yield, and the yield decreased to 31% when 1 equiv of
2-ethylpyridine as the substrate and 2 equiv of pyridine as
time
(h)
yield
(%)b
entry
catalyst
CuI
additive
1
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
36
48
36
36
trace
11
28
23
0
2
CuCl
3
Cu(OTf)2
CuCl2
4
5
CuBr2
6
Cu(OAc)2
Cu(OAc)2
copper bronze
copper bronze
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
ꢀ
45
0
7c
8
39
0
9c
10
11
12
13
14
15
16
17
18
19d
20
21
22e
23f
LiOAc
51
29
trace
46
38
44
49
62
0
NaOAc
KOAc
LiCl
Fe powder þ LiOAc
Al powder þ LiOAc
Zn powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Ni powder þ LiOAc
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
Cu(OAc)2
35
66
65
31
58
a Reaction conditions: 2-ethylpyridine (1.5 mmol), cinnamic acid
(0.5 mmol), catalyst (0.1 mmol), 1,10-phenanthroline (0.1 mmol), additive
(0.25 mmol for reductive metal powder and 1 mmol for the other), DMF
(N,N-dimethylformamide, 1 mL), 140 °C, N2. b Isolated yield. c Under
O2. d In the absence of 1,10-phenanthroline. e 0.5 mmol of 2-ethylpyr-
idine and 1.0 mmol of pyridine were employed in the reaction. f The
reaction was performed at 160 °C.
(7) (a) Qian, B.; Guo, S.; Shao, J.; Zhu, Q.; Yang, L.; Xia, C.; Huang,
H. J. Am. Chem. Soc. 2010, 132, 3650. (b) Rueping, M.; Tolstoluzhsky,
N. Org. Lett. 2011, 13, 1095. (c) Song, G.; Su, Y.; Gong, X.; Han, K.; Li,
X. Org. Lett. 2011, 13, 1968. (d) Burton, P. M.; Morris, J. A. Org. Lett.
2010, 12, 5359. (e) Jiang, H.; Chen, H.; Wang, A.; Liu, X. Chem.
Commun. 2010, 46, 7259.
(8) (a) Yang, Y.; Cheng, K.; Zhang, Y. Org. Lett. 2009, 11, 5606. (b)
Yang, Y.; Chen, L.; Zhang, Z.; Zhang, Y. Org. Lett. 2011, 13, 1342.
(9) Direct C-2 arylation of indolizines is difficult to realize because
the CꢀH arylation of indolizines occurs on the most electron-rich C-3
position; see: (a) Park, C.-H.; Ryabova, V.; Seregin, I. V.; Sromek,
A. W.; Gevorgyan, V. Org. Lett. 2004, 6, 1159. For the synthesis of C-2
arylated indolizines through palladium-catalyzed approaches, see:
(b) Chernyak, D.; Skontos, C.; Gevorgyan, V. Org. Lett. 2010, 12,
3242. (c) Chernyak, D.; Gevorgyan, V. Org. Lett. 2010, 12, 5558.
(10) For a recent review on copper-catalyzed CꢀH functionalizatioin
under oxidative reaction conditions, see: Wendlandt, A. E.; Suess,
A. M.; Stahl, S. S. Angew. Chem., Int. Ed. 2011, 50, 11062.
With the optimal reaction conditions in hand, we next
examined the scope of substituted 2-alkylazaarenes in this
annulation process as shown in Scheme 1. Various pyr-
idines with C-2 functional groups such as CH2CN,
CH2COOEt, CH2Ph, and n-C5H11 were compatible with
the reaction conditions, resulting in the formation of the
desired products in moderate yields (3bꢀ3e). When 2-pico-
line participated in this reaction, the corresponding prod-
uct 3f was isolated in 45% yield. 2,6-Lutidine and 2,3-
lutidine were transferred via this reaction to give the
expected indolizines with methyl substituents on C-5 and
C-8 positions respectively (3g and 3h). It is noteworthy that
958
Org. Lett., Vol. 14, No. 4, 2012