Organic Letters
Letter
a
Scheme 1. This Work in Perspective and Mechanistic
Hypothesis
Table 1. Optimization of Reaction Conditions
b
c
entry
Lewis acid
mol % acid T/°C t/h conv (%) yield (%)
de
,
1
2
3
4
5
6
7
8
AcCl/MeOH
BF3·OEt2
Zn(OTf)2
Cu(OTf)2
Mg(OTf)2
Ca(OTf)2
B(C6F5)3
Sc(OTf)3
Sc(OTf)3
Sc(OTf)3
In(OTf)3
InCl3
150
100
5
5
5
5
10
5
5
5
0
6
6
100
100
100
0
0
0
100
70
100
50
100
100
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
84
d
d
rt
rt
rt
rt
rt
rt
rt
60
rt
rt
rt
24
24
24
24
3
24
24
24
8
66
78
f
9
10
11
12
43
89
96
5
5
6
a
Unless otherwise specified, all reactions were performed with 1.0
mmol of 10, 1.0 mmol of 11, and CH2Cl2 as the solvent in 4−5 mL.
c
not detected. Conversion of 10. Isolated yields of 12.
d
e
f
Acetophenone was formed. Solvent was THF (2 mL) Solvent
was dichloroethane
conversion within 24 h (entry 9), but the yield is inferior to
that of B(C6F5)3 (entry 7). Remarkably, indium(III) triflate
affords 89% yield of 12 in only 8 h (entry 11). Finally, the best
result is achieved by carrying the reaction out in the presence
of 5 mol % InCl3 in CH2Cl2 at room temperature. Under these
conditions, complete conversion is achieved in 6 h to give the
3-vinyl indole 12 as a single regioisomer in 96% isolated yield
(Table 1, entry 12).
As outlined in Scheme 1b, we pursued proof of concept in
the form of a direct C(sp2)-C(sp2) bond forming reaction
between vinyl azides with indoles to give vinyl indoles. To the
best of our knowledge, there are no examples of such a
transformation. Vinyl indoles are attractive intermediates for
the synthesis of alkaloids. Previously vinyl indoles have been
prepared from nitrimines and indoles.13 However, nitrimines
are not attractive starting materials as their preparation is very
low yielding (nitrimines were prepared in 12−42% yield, but
typically yields were less than 20%).13 Vinyl indoles have also
been prepared from ketones, but the conditions require strong
acid and high temperature, incompatible with many function-
alities.14
In order to identify an effective Lewis acid catalyst, (1-
azidovinyl)benzene 10 and 2-methyl-1H-indole 11 were
selected as suitable reactants for optimizing the reaction
conditions (Table 1). As expected, treatment of the 1:1
mixture of 10 and 11 with the Lewis acid BF3·OEt2 (Table 1,
entry 2) or a Brønsted acid (stoichiometric HCl formed in situ
from acetyl chloride and methanol, Table 1, entry 1) at 0 °C to
room temperature does not afford the desired 3-vinyl-indoles
12. However, acetophenone can be observed by TLC and
isolated after water workup.8a
A number of metal salt Lewis acids such as zinc(II) triflate,
copper(II) triflate, as well as magnesium(II) triflate and
calcium(II) triflate also fail to give any desired product (Table
1, entries 3−6). In contrast, even substoichiometric (10 mol
%) amounts of B(C6F5)3 leads to formation of vinyl indole
product 12 in 84% isolated yield (entry 7). The use of
lanthanide Lewis acids was also investigated. Scandium(III)
triflate catalyzes the reaction, but the reaction only reaches
70% conversion after 24 h (entry 8). In dichloroethane, the
reaction is even slower (entry 10). At an increased temperature
of 60 °C, scandium(III) triflate catalyzes the reaction to 100%
It was important to investigate whether the reaction
proceeds via hydrolysis of the (1-azidovinyl)benzene (10) to
the acetophenone observed in the early experiments (Table 1,
entries 1−3). Accordingly, a series of reactions of 10 with
varying stoichiometry of 2-methyl indole (11) were carried
out, and the reactions were monitored. Reaction in the absence
of catalyst returned the unreacted starting materials. Reaction
in the complete absence of 11 also affords no product, and the
vinyl azide 10 is recovered unchanged (Scheme 2a). Neither
acetophenone nor 1-phenyl-vinyl chloride was observed by IR
or NMR of the reaction mixture. Thus, the observed
acetophenone (Table 1, entries 1−3) is likely formed upon
addition of water at the end of the reaction. When the reaction
is carried out with 0.5 equiv of 11, the product 12 is isolated in
42% yield along with unreacted 10 (46%, Scheme 2b). Again,
no acetophenone is detected. Notably, acetophenone fails to
react with indole 10 under the standard reaction conditions
(Scheme 2c) and no product is observed. This makes the
possibility of the mechanism going through a ketone as an
intermediate highly unlikely.14 Additionally, signals attributable
to azide are observed by IR spectroscopy in the reaction
experiments support the mechanism outlined in Scheme 1b.
The reaction is readily scaled to gram scale. Thus, reaction
of (1-azidovinyl)benzene gives 12 in 96% yield at 100 mg scale
and a similar 91% yield at 2 g scale (Scheme 3). The scope of
the reaction is fairly wide (Scheme 3). Reaction of (1-
azidovinyl)benzene 13 (R1 = H) with 2-phenyl-indole leads to
B
Org. Lett. XXXX, XXX, XXX−XXX