.
Angewandte
Communications
acetonitrile (Table 1, entry 1). Other palladium catalysts, such
as Pd(OTFA)2 and [Pd(dba)2], also worked for this reaction
(entries 2 and 3). However, product 2 was not detected when
the salts of other metals, including Cu, Fe, Co, Au, Zn, Sc, and
Al, were employed as the catalysts (entry 4). Control experi-
ments also showed that no cyanation product is formed in the
absence of a metal catalyst (entry 5). In addition, when the
reaction was carried out in the absence of either TBN or
NHPI, product 2 was also not observed (entries 6 and 7). We
found that the yield of product 2 diminished slightly as the
loading of TBN increased, or when a smaller amount of NHPI
was used (entries 8 and 9). We then carried out further
optimization of the reaction conditions. To our delight, we
found that the loading of Pd(OAc)2 could be reduced to
5 mol%, when adjusments were made to the reaction
temperature and concentration (entries 11–13).
With the optimized reaction conditions, we proceeded to
explore the substrate scope of this transformation. As the
reaction conditions are mild, we expected that various
functional groups would be tolerated. To our delight, we
could indeed show that a wide range of functional groups are
tolerated under the reaction conditions (Scheme 2). Reac-
tions with toluene derivatives bearing electron-donating
groups such as MeO and Me proceeded efficiently in good
to excellent yields (2, 4–7, 22–25; Scheme 2). Substrates
substituted with weakly electron-withdrawing groups, such as
F, Cl, Br, and I, also worked well and afforded the desired
products (8–15), which could be used in further coupling
reactions. When there are two methyl substituents, only one
methyl group is converted into a cyano group. However, with
a high loading of NHPI at 808C, the second methyl group can
also be converted into a cyano group (32).
The reaction also works with toluene derivatives bearing
strongly election-withdrawing substituents, such as NO2,
CO2Me, Ac, and CN (29–32). However, in these cases, the
reaction needs to be carried out with a high loading of NHPI
(100 mol%) and at slightly elevated temperatures (808C).
Notably, this cyanation method also works for toluene
derivatives bearing more than one substitutent or an oxida-
tion-sensitive group, such as Bpin (27 and 28; Bpin = pina-
colborane).
Next, we extended this transformation to polycyclic
aromatic hydrocarbons and heteroaromatic compounds
(Scheme 3). Polycyclic aromatics are highly reactive sub-
strates and gave the expected cyanation products in excellent
yields (33–35; Scheme 3). For indole derivatives, the reaction
also worked well (36–38). For heteroaromatic compounds
containing electron-deficient pyridine rings, the reactions
gave diminished yields (39 and 40).
Scheme 2. Direct conversion of methyl arenes into aromatic nitriles. If
not otherwise noted, the reaction conditions are as follows: methyl
arene (0.5 mmol), TBN (1.5 mmol), [Pd(OAc)2 (0.025 mmol), and
NHPI (0.15 mmol) in MeCN (0.5 mL) at 708C under N2 for 24 h.
Yields of isolated products are given. [a] TBN (2.0 equiv) was used.
[b] The isolated yield for a reaction performed on a 1.0 gram scale is
shown in parentheses. [c] NHPI (50 mol%) was used. [d] The yield is
based on GC analysis. [e] These reactions were carried out with an
increased loading of NHPI (100 mol%) in MeCN (0.5 mL) at 808C for
24 h. Ac=acetyl, Boc=tert-butoxycarbonyl, pin=pinacolato.
To gain insight into the mechanism, several control
experiments were carried out. First, when the reaction was
conducted in the presence of stoichiometric amounts of
2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO), the transfor-
mation of 1 was almost completely inhibited and only trace
amounts of product could be detected by GC analysis
(Scheme 4a). This result indicates that the transformation
may proceed through a radical intermediate.
yield (Scheme 4b). Hence, we hypothesized that an aldehyde
might be an intermediate of the cyanation reaction. However,
when benzaldehyde (43) was submitted to the reaction
conditions, it was recovered unchanged (Scheme 4c).
Subsequently, we hypothesized that an aldoxime might be
the key intermediate of this transformation. To verify this
hypothesis, several experiments were carried out. At first, we
found that aldoxime 44 could be effectively transformed into
the nitrile 26 in 90% yield under the optimized reaction
conditions, with minor formation of benzaldehyde (43,
Second, when diphenyl methane 41 was employed as the
substrate, the reaction produced benzophenone 42 in 80%
2
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
These are not the final page numbers!