Organic Letters
Letter
Scheme 1. Context of This Work
The evaluation of different solvents (entries 2−4) showed
that 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) was optimal for
the transformation. Among different iron(II) salts that were
investigated (entries 2, 5−7), iron sulfate afforded the product
in the highest yield. Interestingly, the yield of the reaction
could be further increased by improving the leaving group
ability of the sulfonyl group. Employing NsONHMe·TfOH
(
2c) as an aminating reagent afforded the desired product 3a
in 67% yield (entry 8). The air- and moisture-stable reagent 2c,
which has not been reported previously, can be readily
accessed in three steps from commercially available materials
in 51% overall yield and is stable for several months if stored at
−
20 °C. The reaction can be performed under air and does not
require anhydrous solvents. Interestingly, when the Boc-
protected precursor of the aminating reagent is deprotected
in situ by the addition of triflic acid, the desired product can be
obtained in 21% yield (entry 10).
Having the optimized conditions in hand, we set out to
investigate the generality of the reaction on a range of arenes
(
Scheme 2). Both electron-neutral and electron-rich arenes
could undergo N-methylamination in moderate to good yields.
Unfortunately, electron-poor arenes proved to be unsuitable
for the reported process. The more electron-poor the arene
substrate was, the lower was the conversion and thus the yield
of the amination reaction. However, on systems with an
additional electron-donating group, even strongly electron-
withdrawing groups such as ester (3g) and nitro (3h) groups
were tolerated. (Pseudo)halides that are known to undergo
oxidative addition with late transition metals remained
untouched (3i−k, 3m−p). Arenes bearing aliphatic halides
1
6
aminochlorination reaction. We thus questioned whether
these reagents could enable the direct synthesis of unprotected
secondary anilines from arenes via innate C−H amination.
Initial investigations using the previously reported pivalate
reagent (2a), however, resulted in low yields of the desired
product 3a (Table 1, entry 1). Given that tuning the leaving
group ability of the reagent can be essential to achieve good
(
3q, 3r) and alcohols (3s) are also suitable substrates.
Substrates bearing multiple aryl groups react selectively on
the more electron-rich arene, without any diamination being
observed (3f, 3t). Collectively, these results suggest that the
1
0b
reactivity, we next turned our attention to the synthesis of a
new reagent bearing a tosylate (OTs) group (2b). This proved
to be a crucial change, and the product 3a was obtained in
moderate yields (entry 2).
17
reaction could be employed in late-stage functionalization.
To support this hypothesis, pharmaceuticals or derivatives
thereof were used as substrates for the reaction. Nimesulide
(
3t) and ibuprofen methyl ester (3u) both underwent the N-
Methylamination of Benzene
a
methylamination reaction to afford the corresponding aniline.
The preparative value of the methodology could be further
demonstrated in a scale-up experiment where mesitylene was
aminated on a 5 mmol scale to afford 3c in an even slightly
higher yield than on small scale.
Control reactions revealed that the two very electron-rich
anilines 3e and 3f can be obtained from the corresponding
arenes without the use of FeSO ·7H O in a similar yield
b
entry
iron salt
solvent (ratio)
HFIP
HFIP
TFE
R
yield (%)
4
2
1
2
3
4
5
6
7
8
9
FeSO ·7H O
Piv
Ts
Ts
Ts
Ts
Ts
Ts
Ns
Ns
Ns
<5
39
<5
<5
37
38
25
67
9
4
2
compared to the optimized reaction conditions. Due to the
high electrophilicity of the reagent 2c, the iron-free reaction
presumably proceeds in these two cases through a direct
electrophilic aromatic substitution pathway. However, for less
Information for details). In light of these results, we next
performed further control reactions using common radical
Information for details). When 5% benzoyl peroxide were used
instead of the iron salt, N-methylaniline (3a) was obtained in a
decreased yet significant 45% yield, while the use of AIBN and
Cu(I) salts did not deliver the product. Furthermore, when the
iron-free reaction is heated to 60 °C instead of 40 °C, 3a is
obtained in 66% yield. The reaction, in this case, is likely
initiated by thermal homolysis of the N−O bond. However, we
FeSO ·7H O
4
2
FeSO ·7H O
4
2
FeSO ·7H O
MeCN/H O (2:1)
4
2
2
Fe(acac)2
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
FeCl2
FePc
FeSO ·7H O
4
2
c
1
0
FeSO ·7H O
21
4
2
a
Reaction conditions: benzene (0.20 mmol), aminating reagent (2)
0.50 mmol), iron salt (0.01 mmol), solvent (0.2 mL), 40 °C, 16 h,
under air. Yields in percent obtained by GC-FID using n-dodecane
(
b
c
as internal standard. Boc-protected aminating reagent precursor,
triflic acid (0.50 mmol), 1 mL solvent. Ns = nosyl. Pc =
phthalocyanin. Piv = pivaloyl. TFE = 2,2,2-trifluoroethanol. Ts =
tosyl.
1
423
Org. Lett. 2021, 23, 1422−1426