J. Cheng et al. / Tetrahedron Letters 55 (2014) 4044–4046
4045
Table 1
and environmentally-friendly iodocyclization process of 2-tosyla-
Effect of catalysts and conditions on the yield of 4a
mino-phenylprop-1-yn-3ols with 2.0 equiv molecular iodine (I2)
in MeOH at 60 °C. In fact, this electrophilic addition–cyclization
tandem synthetic strategy could trace back to the synthesis of
isoquinolines.10 Larock10a,b and Wu10c,d utilized iminoalkynes and
2-alkynylbenzaldehyde oximes, respectively as substrates reacted
with electrophilic reagents to prepare 2-halogen isoquinolines
under metal-free conditions. Whatever the construction of 3-halo-
gen quinolines or 2-halogen isoquinolines, a common point of
these reported tandem reactions is the use of the stronger electro-
philic reagents, including I2, Br2, ICl, NBS, and NIS.
However, aqueous halogen acid has not been yet employed suc-
cessfully as halogen source of the 3-position of quinoline. Besides,
we noticed that cyclization of 1-(2-aminoaryl)-2-yn-1-ols could
occur in two manners,11 5-exo-dig resulting in an indole derivative
and 6-endo-dig producing a quinoline derivative catalyzed by
Pd(II) or Cu(II). Herein, we report a novel and regioselective proto-
col for the synthesis of 3-chloride or 3-bromide quinoline deriva-
tives using aqueous halogen acid as halogen source catalyzed by
Cu(II). To the best of our knowledge, this methodology has not
been previously reported.
Entry
Catalysta
Temperature (°C)
Timeb (h)
Yieldc (%)
1
2
3
4
5
6
7
8
9
—
60
60
60
60
60
60
60
25
60
80
60
60
12
8
n.r.d
n.r.
n.r.
Trace
25
33
87
70
93
BF3ÁEt2O
ZnCl2
SnCl4
CuSO4
Cu(OAc)2
CuCl
CuCl2
CuCl2
CuCl2
CuCl2
CuCl2
12
10
12
12
6
24
3
2
10
11
12
76
3
3
86e
0f
a
b
c
d
e
f
Besides amount of BF3ÁEt2O 50 mol %, others used 5 mol %.
Monitored by thin layer chromatography.
Isolated yield.
No reaction.
Toluene as solvent.
Using aq KCl instead of aq HCl as halogen source.
entries 8–11) could catalyze the synthesis of 4a efficiently, and
finally the optimized condition (Table 1, entry 9) was acquired
after the study of the influence of temperature and time on the
yield of 4a. Meanwhile, we investigated the effect of temperature
on the yield of 4a and found that the higher the temperature, the
more the amount of 5a (Table 1, entry 10). And the product was
only 5a when using saturated solution of KCl instead of concen-
trated hydrochloric acid as halogen source (Table 1, entry 12),
which indicated that acid condition could promote the tandem
reaction. Besides, copper bromide or copper iodide was not used
because of the need to maintain the purity of halogen source in
the tandem reaction.
After separated by silica gel column chromatography, the struc-
ture of 4a was established by 1H, 13C NMR, HMQC, and ESI-HRMS
spectra. Firstly, HRMS spectra indicated that the accurate molecu-
lar weight matched the formula of 4a. Secondly, compared with
NMR spectra of 5a, it was distinct that 4a reduced an aromatic pro-
ton and increased a quaternary carbon atom in NMR spectra. There
were three aromatic protons in low field of 1H NMR, the signal of
H-7 showed dual double peaks (J5,7 = 2.4 Hz, J8,7 = 9.2 Hz) at dH
7.61 and the signal of H-5 and H-8 all displayed double peaks at
dH 8.14 and 7.88, which respectively correlated with the signal at
dC 130 (C-7), 123 (C-5), and 131 (C-8) in HMQC spectra. Finally,
the signal of cyclopropyl appeared obviously at dH 1.16 (2H), 1.27
(2H), 2.82 (1H) and dC 11 (2C), 14 (1C) in high field of NMR.
To further demonstrate the scope and flexibility of the present
optimized conditions, different o-trifluoroacetyl anilines were then
explored.14 As illustrated in Table 2 3-chloride quinolines were
obtained efficiently with high regioselectivity (Table 2, entries
1–11 yield 80–93%) under the optimized conditions, and the elec-
tronic nature of the R1 group did not play a key role in this annu-
lation process (R1 = Cl, F, Me or OMe). However, the synthesis of
3-bromide quinolines was more difficult due to a bigger steric hin-
drance of bromide atom. When the neighboring R2 group was
cyclopropyl, the yield of tandem product 4 was lower (Table 2,
entries 12–15, yield 50–64%); when the R2 group was the bigger
group such as tert-butyl or phenyl, but to our surprise, the major
product was not 3-halogen quinoline 4 but only 3-unsubstituted
quinoline 5 (Table 2, entries 16–21).
Initially, our original goal was to synthesize 6-chloro-2-cyclo-
propyl-4-(trifluoromethyl)quinoline (5a), but it was unexpected
to detect that the major product was converted into 3-chloride
quinoline (4a) when we used concentrated hydrochloric acid to
quench the excess Grignard regent in the process of one-pot reac-
tion (Scheme 2). And then, to test the viability of the regioselectiv-
ity and the suitability of this one-pot system for preparing 4a, we
carried out the above experiment quantificationally again. We
treated propargylic alcohol 3a, which was the reactive residue of
o-trifluoroacetyl aniline 1a and 1.2 equiv (cyclopropylethy-
nyl)magnesium chloride 2a without any isolation, with 1.5 equiv
concentrated hydrochloric acid and catalytic amount of CuCl2
(3–5 mol %) undergoing a cyclization–halogenation tandem reac-
tion to obtain the target molecule 4a characterized by 1H, 13C,
HMQC NMR, and HRMS.
The following is the optimal reaction conditions of the above
synthesis of 4a. Firstly, 1a was easily obtained in 3 steps from
p-chloroaniline with 75% yield,12 and then reacted with alkynyl
Grignard regent 2a in THF at 0 °C.13 Secondly, in the presence of
1.5 equiv concentrated hydrochloric acid, the effect of catalysts
and conditions on the yield of 4a was investigated in the one-pot
system (Table 1). In preliminary optimized experiments, the reac-
tion did not happen under the condition of catalyst-free or using
boron trifluoride, zinc chloride, or tin tetrachloride as catalyst
(Table 1, entries 1–4). The target molecule 4a was monitored until
using Cu salt as catalyst, but the yield of 4a was low and the major
product was by-product 5a occurred through cyclization rather
than tandem reaction when using copper acetate (Table 1, entry
5) or copper sulfate (Table 1, entry 6) as catalyst. To our delight,
cuprous chloride (Table 1, entry 7) and copper chloride (Table 1,
O
F3C
OH
Cl
Cl
MgCl
Cl
2a
CF3
NH2
NH2
3a
THF
CF3
1a
CF3
N
5
8
Based on the above observations and the reported results of the
tandem reaction,9,11 a plausible reaction mechanism was proposed
as showed in Scheme 3. After Grignard addition reaction, propargy-
lic alcohol 3 and copper chloride transformed into complex I. And
then acetylenic bond that was activated by Cu(II) reacted an
electrophilic addition with HCl. It was worthwhile to notice that
Cl
Cl
CuCl2
aq. HCl
+
7
N
5a (by-product)
4a
Scheme 2. One-pot synthesis of 4a.
Cheng, Jie
