Organic Letters
Letter
scope of application. Moreover, although it is highly site-
specific, directed C−H sulfonylation of arenes occurs mostly at
the position ortho to the directing group.2c,8 On the other
hand, one of the simplest and oldest strategies to substitute an
aromatic hydrogen atom by a sulfonyl group is probably
aromatic electrophilic substitution, akin to Friedel−Crafts
acylation (Scheme 2b).9 This strategy benefits from ease of
setup, a large variety of accessible catalysts, and a wide range of
commercially available reagents. Nevertheless, in most cases
Friedel−Crafts-type sulfonylation is applicable only to low-
weight benzene derivatives and suffers from poor regioselec-
tivity dictated by the innate orientation effect of the
substituents.
Table 1. Optimization of the Superacid-Promoted
Benzenesulfonylation of p-Methylacetanilide
b
We previously demonstrated10 that under superacid
conditions,11 aromatic amines (or amides)despite being
fully protonatedcan react with an electrophile provided that
the latter is further activated by the superacidic medium
(superelectrophilic activation).12 In particular, under these
conditions the orientation effect of the amine substituent is no
longer effective, and the site selectivity of the reaction is
dictated by the secondary substituent, leading to nonclassical
regioselectivity. Motivated by these previous results and by the
absence of general procedures for direct electrophilic
sulfonylation of aromatic amines,9 we aimed at developing a
complementary site-selective method that would allow the
direct functionalization of aniline and indole derivatives from
readily available sulfone precursors by taking advantage of
superelectrophilic activation (Scheme 2c).
We started our investigations with p-methylacetanilide as
model substrate and benzenesulfonic acid Aa as a cheap and
readily available sulfonyl source (Table 1). With neat HF/SbF5
(7/1) superacid, the desired product 1a was obtained after 2 h
at −20 °C in an encouraging 30% yield (Table 1, entry 1).
Delightfully, the substrate was selectively functionalized at the
meta position (with respect to the acetamido group),
confirming our initial hypothesis. The peculiar regioselectivity
of this transformation was also confirmed by X-ray analysis of
We next evaluated the impact of the acidity on the reactivity by
modulating the concentration of SbF5 (Table 1, entries 1−4)
and found that the optimal conditions were reached with a 2/1
HF/SbF5 mixture (73% yield; Table 1, entry 3). Increasing the
amount of benzenesulfonic acid to 2 equiv did not improve the
yield (Table 1, entry 5). While decreasing the temperature to
−40 °C completely inhibited the reaction (Table 1, entry 6),
performing the reaction at higher temperature led to a poor
yield and to the formation of the undesired regioisomer 1a′
(Table 1, entry 7). The formation of 1a′ can be explained by
an initial intramolecular rearrangement within the arenium ion
generated by the protonation of p-methylacetanilide to the
more stable m-methylacetanilide, which then reacts with
benzenesulfonic acid.13,14 No reaction occurred using the
weaker superacid TfOH (Table 1, entry 8), a result that
supports our initial hypothesis and the necessity to use
superelectrophilic activation to perform this transformation. As
expected, the use of excess of Lewis acids in dichloromethane
did not afford the desired product even at room temperature
with a prolonged reaction time (Table 1, entries 9−14). We
next evaluated the reactivity of p-methylacetanilide with other
sulfonyl precursors (Table 1, entries 15−23). Methyl sulfonate
Ba was found to be totally unreactive under these conditions
(Table 1, entry 15). Although good conversion of the starting
material was obtained with sulfonyl chloride Ca, 1a was
Yield [%]
a
entry
acid (v/v)
PhSO2Y (n)
T [°C]
1a
1a′
1
2
3
4
5
6
7
8
HF/SbF5 (7/1)
HF/SbF5 (3/1)
HF/SbF5 (2/1)
HF/SbF5 (1/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
TfOH
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (2.0)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Aa (1.2)
Ba (1.2)
Ca (1.2)
Da (1.2)
Ea (1.2)
Fa (1.2)
Ga (1.2)
Ha (1.2)
Ia (1.2)
Ja (1.2)
−20
−20
−20
−20
−20
−40
0
−20
20
20
20
20
20
20
−20
−20
−20
−20
−20
−20
−20
−20
30
58
73
49
72
0
15
0
0
0
0
0
0
0
0
10
12
0
0
68
3
86
72
0
0
0
0
0
0
45
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
c
d
e
9
SbF5
AlCl3
FeCl3
d
e
10
11
12
13
d
e
d
e
Cu(OTf)2
d
e
Bi(OTf)3
d
e
14
P2O5
15
16
17
18
19
20
21
22
23
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
HF/SbF5 (2/1)
f
−20
a
b
Used as the solvent unless stated otherwise. Yields of isolated
c
d
products. 58% yield after 1 h. The reaction time was extended to 20
e
f
h. 2 equiv in CH2Cl2. Concomitant formation of 3-chloro-4-
methylacetanilide in 50% yield.
obtained in a low 10% yield accompanied by the formation of
3-chloro-4-methylacetanilide in 50% yield (Table 1, entry 16).
In this case, reagent C is a source of chloride ions, which can
be oxidized in situ by SbV. The generated halenium ions then
react with the aromatic nucleophile to furnish the chlorinated
product, as previously observed.10b In order to avoid this
competitive process, sulfonyl fluoride Da was evaluated as the
reaction partner (Table 1, entry 17).15 However, the reaction
was very slow in this case, and 1a was isolated in only 12%
yield. We next turned our attention to sulfonyl reagents
derived from sulfonamide. While N,N-dimethylsulfonamide Ea
and N-acetylsulfonamide Fa were found to be unreactive
(Table 1, entries 18 and 19), using the symmetrical
sulfonimide Ga afforded 1a in a good 68% yield (Table 1,
entry 20). On the other hand, only traces of 1a were obtained
when the reaction was performed with imidazole derivative Ha
(Table 1, entry 21). Sulfonyl azide Ia and sulfonyl hydrazide Ja
were revealed to be also very efficient in this transformation,
broadening the panel of reagents with the ability to sulfonate
aromatic amines in superacid. Under these conditions, 1a was
4116
Org. Lett. 2021, 23, 4115−4120