Similar to our oxidative double cyclization,4 the cascade
cyclization of 1a did not produce any undesired monocy-
clization products. The reaction afforded a mixture of triple
cyclization products 2a and 3a (in 1.5:1 ratio) in 54% yield.
Further screening of different ligands was proven to be
fruitful: using quinoline in place of pyridine as ligand
improved both product yield and ratio (entry 2 vs 1).
However, other quinoline derivatives did not result in
satisfactory reaction conversions (entries 3-6). Interestingly,
isoquinoline was found to be the best ligand for the cy-
clization of 1a, furnishing 2a as the major product (in 5.6:1
ratio) in 76% combined yield (entry 7).
Scheme 1. Top Route: Heck-Type Cyclization Involves
Oxidative Addition to Aryl/Vinyl Halides or Triflates
(Pre-activated Carbon Nucleophiles), thus Requiring Pd(0)
Catalysts. Bottom Route: Oxidative Cyclization Is Initiated
through Nucleophilic Attacks on Pd(II)-olefin Complexes, and
External Oxidants Are Required To Regenerate Pd(II) from
Pd(0) Species
A comparison of basicities of these N-heteroaromatics
indicates a correlation with the reaction conversion and
product ratio.7 Therefore, we hypothesized that the N-ligand
basicities play a crucial role in suppressing olefin isomer-
ization.
With quinoline and isoquinoline as ligands, we then
evaluated the scope of the cascade cyclization, and the results
are summarized in Table 2. Substrates 1a-c, which bear
different para substituents, underwent triple cyclization
smoothly to furnish 2a-c as the sole diastereomers in good
yields (entries 1-6). Notably, the triple cyclization reactions
were found to be highly diastereoselective (dr > 24:1).
Compared with reactions under the Pd(OAc)2/isoquinoline
system, the cyclization reactions of 1a-c using the Pd(OAc)2/
quinoline system required similar reaction time to reach
complete conversion. However, the Pd(OAc)2/isoquinoline
system resulted in better yields of products 2a-c by
lowering the yields of olefin-isomerized product 3. For
instance, the cyclization of 1c with the Pd(OAc)2/quinoline
system furnished product 2c in 34% yield and olefin-
isomerized product 3c in 29% yield (entry 6). When the
ligand was changed to isoquinoline, 1c could be converted
to 2c in 55% yield by suppressing the formation of olefin-
isomerized product 3c (5% yield; entry 5). On the other
hand, the cyclization efficiency of 1a-c did not correlate
strongly with their electronic properties, which differs
from the case of our recently reported double cyclization
reactions involving unsaturated anilides. The aforemen-
tioned system exhibited a significant electronic depen-
dence, allowing electron-withdrawing anilides to cyclize
faster than did their electron-donating counterparts.4 With
the Pd(OAc)2/isoquinoline system, substrate 1d, which
possesses a homologated olefinic tethered chain, cyclized
in a diastereoselective manner to afford 2d as the major
product (45% yield; entry 7) and trace amounts of olefin-
isomerized products (2% combined yield). The yield of
undesired monocyclization products, even in the absence
of tandem relays, which are usually required in polycy-
clization reactions. To the best of our knowledge, no
successful examples of palladium-catalyzed oxidative
cascade cyclization have been reported for the formation
of more than two bonds.5 Herein we report palladium-
catalyzed, highly diastereoselective oxidative cascade
cyclization of unsaturated anilides, leading to the forma-
tion of three bonds in one step.
We first investigated the cyclization of 1a under a
Pd(OAc)2/pyridine catalyst system (Table 1, entry 1).6
Table 1. Screening of N-Ligands for the Pd-Catalyzed Oxidative
Cascade Cyclizationa
convnb
(%)
% yield
product ratio
entry
ligand
pyridine
(2a + 3a)b
(2a:3a)b
(4) Yip, K.-T.; Yang, M.; Law, K.-L.; Zhu, N.-Y.; Yang, D. J. Am.
Chem. Soc. 2006, 128, 3130–3131.
1
2
3
4
5
6
7
73
86
43
42
65
52
89
56
70
39
26
24
39
76
1.5:1
4.2:1
1:8.3
1:3.2
1:6.3
1:1.4
5.6:1
quinoline
(5) For examples of palladium-catalyzed oxidative cyclizations involving
the formation of more than one bond, see: (a) Alexanian, E. J.; Lee, C.;
Sorensen, E. J. J. Am. Chem. Soc. 2005, 127, 7690–7691. (b) Tietze, L. F.;
Sommer, K. M.; Zinngrebe, J.; Stecker, F. Angew. Chem., Int. Ed. 2005,
44, 257–259. (c) Liu, C.; Widenhoefer, R. A. J. Am. Chem. Soc. 2004,
126, 10250–10251. (d) Arai, M. A.; Kuraishi, M.; Arai, T.; Sasai, H. J. Am.
Chem. Soc. 2001, 123, 2907–2908. (e) Andersson, P. G.; Ba¨ckvall, J.-E.
J. Am. Chem. Soc. 1992, 114, 8696–8698. (f) Tamaru, Y.; Hojo, M.;
Higashimura, H.; Yoshida, Z.-I. J. Am. Chem. Soc. 1988, 110, 3994–4002.
(6) For screening of other reaction parameters, including the effects of
solvent, temperature, and oxidant, see the Supporting Information.
(7) For details, see the Supporting Information.
6-methoxyquinoline
8-methylquinoline
acridine
quinoxaline
isoquinoline
a All reactions were performed at 70 °C using 1a (0.2 mmol), ligand
(40 mol %), and Pd(OAc)2 (10 mol %) in toluene (2 mL) under O2 (1 atm).
b
1
Determined by H NMR.
1912
Org. Lett., Vol. 11, No. 9, 2009