Communication
trigued by Bolm and co-workers’ use of the inexpensive triple
salt, oxone, with the widely used nitroxide radical, TEMPO (3),
to oxidize alcohols catalytically in the presence of a halide
that the bromide additive must be controlled in some manner
to prevent this detrimental side reaction.
Although control experiments revealed competitive process-
es, the addition of ACT resulted in a faster, more selective oxi-
dation to access nitriles. The use of a slight molar excess of
ACT relative to the tetrabutylammonium bromide effectively
sequesters the bromide in the generated oxoammonium salt,
reducing the likelihood of bromination of the amine substrate.
These modifications resulted in the formation of benzonitrile
(4a) in 87% isolated yield, accompanied by ꢀ1% of benzalde-
hyde (4b; Table 1, entry 4) likely due to the presence of residu-
al water. However, when strictly anhydrous conditions were
employed the reaction failed to proceed. Apparently, a catalytic
amount of water is necessary to solubilize a small portion of
the oxone. The use of commercially available reagent-grade
CH Cl without any additional drying was sufficient for this pur-
[
11]
source. While Bolm’s chemistry offers a cheap and efficient
route to generate an oxoammonium salt, oxone is known to
[
12]
react with a variety of substrates, including amines. Addi-
tionally, the oxone/TEMPO system is multifaceted as it relies on
the in situ generation of hypobromous acid, which is responsi-
ble for the turnover of the TEMPO catalyst as reported by
[13]
Rychnovsky and co-workers.
Exploratory experiments with
benzylamine (4), chosen as a model substrate, were performed
as highlighted in Table 1.
Table 1. Optimization and control studies.
2
2
pose, and any aldehyde byproduct (typically less than 3%) is
[
15]
easily removed upon work-up. Decreasing the catalyst load-
ing to 5 mol % ACT, 4.5 mol % tetrabutylammonium bromide,
and 6 molar equivalents of pyridine relative to the starting
amine resulted in a slight increase in the isolated yield of 4a
(entry 5). Owing to the extremely hygroscopic nature of tetra-
butylammonium bromide, it was necessary to constantly re-dry
the bromide to prevent substantial amounts of water from
being introduced into the reaction mixture resulting in the for-
mation of unwanted aldehyde. As a practical matter, complete
removal of the tetrabutylammonium bromide upon comple-
tion of the reaction was problematic. When anhydrous potassi-
um bromide was used, the system failed to generate hypobro-
mous acid, and the intentional addition of a small amount of
water to help solubilize the bromide resulted in the loss of se-
lectivity (entries 6–7). Pyridinium bromide was found to be op-
timal for our purposes (entry 8). In addition to complementing
our choice of base, the salt is easily prepared in anhydrous
Entry Halide
source
Pyridine ACT
[equiv] (1)
Result
(
mol%)
[mol%]
[
a]
a]
1
2
None
None
None
8
None no change in 4
[
None equal amounts of 4 and 4c ob-
served in addition to pyridine N-
oxide
[
[
[
a]
b]
b]
+
À
À
À
3
4
5
6
7
8
n-Bu
9)
n-Bu
5)
n-Bu
4.5)
4
4
4
N Br
8
8
6
8
8
6
None 4 remained, pyridine N-oxide, and
ꢀ8% of 4d detected
(
+
N Br
10
87% isolated yield of 4a with 1%
4b
95% isolated yield of 4a
(
+
N Br
5
(
KBr (9)
10
10
5
failed to generate hypobromous
acid
4a along with substantial amounts
of 4b and 4d
[a]
KBr (9) +
[
16]
form.
5
0 mL H
pyridine·HBr
4.5)
2
O
[
b]
92% isolated yield 4a
Oxidation of primary amines to nitriles under our catalytic
(
conditions is achieved as follows: Addition of a 0.5m solution
À1
of amine substrate in CH Cl at a rate of 10 mLh by syringe
[
a] The crude reaction mixture was analyzed by GC/MS both before and
2
2
[
17]
after extraction with 1m NaOH to ensure all basic substrates were detect-
ed. [b] Isolated yields on a 10 mmol scale.
pump to a stirred slurry of 4.4 molar equivalents of oxone,
mol % of ACT, 4.5 mol % of pyridinium bromide, and 6 molar
equivalents of pyridine in 100 mL of CH Cl . Although most re-
5
2
2
actions were completed within a few hours after the conclu-
sion of amine addition to the reaction mixture, product isola-
tion was substantially easier when reaction mixtures were
When 4, as a 0.5m solution in CH Cl , was added at a rate of
0 mLh to a stirred slurry of oxone in CH Cl and allowed to
2 2
2
2
À1
1
[
18]
stir for 12 h at room temperature, no change was observed
stirred at room temperature for 12 h. Passage of the crude
reaction mixture through a small bed of silica gel, and removal
of the solvent, afforded pure nitrile products (Table 2).
[14]
(
Table 1, entry 1). Not surprisingly, when the same reaction
was conducted under basic conditions in the presence of
equivalents of pyridine, substantial amounts of pyridine N-
8
The oxidation works well for a variety of electron-deficient
and -rich benzylamines affording isolated yields in the range of
86–97% (Table 2, entries 1–9). Oxidation of cinnamylamine
under the general reaction conditions resulted in a moderate
isolated yield of 72% (entry 10). However, upon increasing the
amount of ACT and pyridinium bromide to 10 and 9 mol %, re-
spectively, the reaction proceeded efficiently resulting in 89%
isolated yield of cinnamonitrile (entry 10). Similarly, when the
alkene moiety is removed from conjugation and is placed in
a terminal location, the yield also suffers (entry 16), perhaps
oxide resulted; additionally, approximately half of the starting
material, 4, was consumed and the formation of 4c was ob-
served (entry 2). The introduction of tetrabutylammonium bro-
mide into the reaction mixture (entry 3) resulted in the forma-
tion of a small amount of 4d, likely formed by a bromination
of the amine and subsequent elimination under the basic con-
ditions to generate an aldimine, which rapidly undergoes reac-
tion with another equivalent of 4 with loss of ammonia to gen-
[
7]
erate 4d. From this observation, it quickly became apparent
Chem. Eur. J. 2016, 22, 5156 – 5159
5157
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim