J. Luo et al. / Tetrahedron Letters 55 (2014) 1552–1556
1553
R2
H
R1
R3
R2
N3
Cat. Pd, base
R2 = H
Cat. Pd, base
1
R1
R1
N
R3
N
R3
R2 ≠ H
+
4
3
OCO2Me
2d
Scheme 1. Pd-catalyzed synthesis of 3,4-diallylquinolines and 3-allylquinolines.
the product 3a was obtained in 69% isolated yield (entry 11). The
decrease in reaction temperature to 80 °C gave the product 3a in
60% isolated yield (entry 20). Decreasing the amounts of base and
catalyst resulted in lower yields (entries 21 and 22).
Results and discussion
Initially, we selected 1-azido-2-(3-phenylprop-2-ynyl) benzene
1a as a model substrate. The results are summarized in Table 1.
When allyl acetate 2a was used, the reaction proceeded smoothly
and gave the desired 3,4-diallylquinoline 3a in 39% isolated yield
(entry 1). Allyl chloride 2b produced the desired product 3a in
low yield (entry 2). No desired product was obtained with allyl tri-
butyltin 2c (entry 3). The best result was obtained with allyl
methyl carbonate 2d, giving the product 3a in 47% isolated yield
together with unidentified products (entry 4). Also, the use of 2a
together with 2d produced 3a in lower yield (entry 5). Thus, allyl
methyl carbonate was utilized as an allyl source for further optimi-
zation of reaction conditions.
Our research was focused on the optimization of palladium cat-
alysts, solvents, bases, and temperatures, and the results are sum-
marized in Table 2. No desired product 3a was obtained in the
absence of palladium catalyst (entry 1). A low yield (26%) was ob-
tained when the reaction was carried out in the presence of
Pd(PPh3)4 but in the absence of base (entry 5), indicating the impor-
tance of the combined use of both palladium catalyst and base. Next
we tested two palladium catalysts, Pd2(dba)3ÁCHCl3 and Pd(PPh3)4
(entries 2–4 and 6), the latter gave much better result. The use of
ligands, XantPhos, and S-Phos, gave 8% and 35% yields in the pres-
ence of Pd2(dba)3ÁCHCl3, respectively (entries 3 and 4). Among the
various bases we investigated (entries 6–13), the use of K3PO4
and NaOAc afforded 3a in high yields (entries 10 and 11). The
screening of various solvents, such as 1,4-dioxane, AcOEt, benzene,
CH3CN, CH3CN + H2O, THF, DMF, revealed that the choice of sol-
vents played an important role for the formation of 3a (entries
11, 14–19). DMF was found to be the most suitable solvent, and
With optimized conditions in hand, we carried out the reactions
between various 1-azido-2-(2-propynyl) benzenes 1 and allyl
methyl carbonate 2d, and the results are summarized in Table 3.
The substrate 1b, having a methyl at the para-position of the aro-
matic ring, afforded the corresponding cyclized product 3b in 66%
isolate yield through method A (entry 1). The substrate 1c, bearing
fluorine atoms at 3,5-positions on the aromatic ring, afforded the
desired product 3c in 55% yield (entry 2). Furthermore, the sub-
strates 1d, having a cyclohexyl group at the alkyne terminus, gave
the expected products 3d in moderate yields via the methods A
and B (entries 3 and 4, respectively); here, a mixture of unidenti-
fied by-products was formed, but they were easily separated from
the desired quinoline by column chromatography. It is noteworthy
that the method B gave a little higher yield than the method A,
although it took a longer reaction time. The substrates 1e and 1f,
having a methyl at the meta-position and a chloro group at para-
position of the aromatic ring, afforded products 3e and 3f in 60%
and 64% isolate yield through method A (entries 5 and 6). The sub-
strates 1g, 1h and 1i, in which the aromatic ring was substituted
with bromo and chloro groups, afforded the corresponding prod-
ucts 3g, 3h and 3i in moderate yields, respectively (entries 7–9,
10). The substrate 1j afforded the desired product 3j in 45% yield
with method A and 58% yield with method B (entries 11 and 12).
Other substituted allyl carbonates such as crotyl ethyl carbonate
2e and ethyl 2-methylallyl carbonate 2f, instead of allyl methyl
carbonate 2d, were also investigated and gave only trace amounts
of the products, 2-phenylquinoline as main product10 was formed
(yield 60% for 2e, 67% for 2f).
Table 1
Optimization of the allyl sourcesa
Pd(PPh3)4 (10 mol%),
K3PO4 (5 eq)
+
OCO2Et
2e
Ph
N3
1a
DMF, 100 o
C
Pd(PPh3)4 (10 mol%),
K2CO3 (5 eq)
X
N
Ph
+
Ph
DMF, 100 o C, 24 h
3k (trace)
N3
N
Ph
1a
2
3a
Pd(PPh3)4 (10 mol%),
K3PO4 (5 eq)
Entry
Allyl sources
3a, Yieldb (%)
OCO2Et
+
1
2
3
44 (39)c
OAc
Cl
Ph
N3
1a
DMF, 100 o
C
23
0
N
Ph
2f
3m
SnBu3
(trace)
52 (47)c
OCO2Me
2a + 2d
4
5
21d
a
All the reactions were carried out using 0.05 mmol of 1a and 5 equiv of 2 in the
It is noteworthy that the reaction of 1n, having the OAc group at
R2 (at the benzylic position), proceeded smoothly and gave the
3-allylquinoline 4 in 67% isolated yield (Scheme 2). On the basis
of the results above, it is clearly indicated that R2 group plays a
key role in the selective synthesis of allylated quinolines.
presence of Pd(PPh3)4 (10 mol %), and K2CO3 (5 equiv), in 1 mL DMF at 100 °C for
24 h.
b
1H NMR yield was determined by using p-xylene as an internal standard.
Isolated yield is shown in parentheses.
2.5 equiv 2a and 2.5 equiv 2d were used respectively.
c
d