Organic Letters
Letter
were used for the substrate scope. The use of other additives
and ligands was not beneficial (see SI, Table S1). Similarly, the
use of the other bidentate-directing groups under standard
reaction conditions also proved unsuccessful (see SI, Table
S2). To illustrate the practicality of this catalytic process, a
gram-scale reaction was conducted under the standard
conditions using 1.0 g of 1a (3.8 mmol) with 2a to isolate
the silylated product 3a in an impressive 79% yield (see SI for
more details).
formed in 11% yield, showing the limitation of this method to
ortho-substituted benzamides. The reason behind this could be
the better chelating property of the substrate for facile C−H
activation when an ortho-substituent is present. Very
interestingly, aliphatic benzamide could also be successfully
silylated at the β-position under the optimized conditions,
although with a paltry yield (3ai, 13%). Many other substrates
were attempted under the present protocol, including the
naphthyl substrate, only meta-substituted benzamides, olefinic,
as well as other heterocyclic variants, but all of them either
gave low yields or did not furnish any silylation product (see
With the optimized conditions in hand, we then probed the
substrate scope of benzamides (1) in order to check the
generality of the silylation protocol (Scheme 2). In general, the
Adopting the standard reaction conditions, we next pursued
the scope of other organosilicon sources. Using bulky silanes
like (Me2PhSi)2 or (MePh2Si)2 in place of (Me3Si)2, moderate
to good conversion was seen, affording the corresponding
silylation products 3ag and 3ah in 67% and 47% yields,
respectively. No desired product was isolated with (Ph3Si)2,
which may be due to additional steric hindrance (see SI, Table
Scheme 2. Scope of Arylamides (1) with Disilanes (2)
In order to shed light on the reaction mechanism and the
catalytic cycle, several control experiments were executed
(Scheme 3 and SI). A deuterium−proton exchange experiment
Scheme 3. Mechanistic Experiments
reactivity of amides para-substituted with electron-donating
and electron-withdrawing groups was slightly superior to that
of the meta-substituted ones. Functional groups like Me, OMe,
t
F, Cl, Br, CF3, and Bu were all well tolerated. Beginning with
was conducted under standard conditions using 1a in the
absence of HMDS 2a and quenching with excess CD3OD
(Scheme 3a), revealing no deuterium exchange in the
recovered starting material, which suggested that the C−H
bond cleavage step was irreversible. This conclusion was
reassured when the same experiment was performed with the
deuterated version 1a-d1 (92% D), even when external H
sources like excess H2O or CH3OH were used since no H-
incorporation was observed at the ortho C−H bond of the
recovered starting material (Scheme 3a) (SI for more details).
In addition, a pronounced substituent effect was observed with
electron-withdrawing group bearing substrates undergoing
reaction at a faster rate than the electron-donating ones at
both meta and para positions under the optimized reaction
conditions (Scheme 3b and 3c, respectively, and SI).
the meta-substituted amides of o-toluic acid (1b−1f) for this
transformation, we found that electron-donating groups such
as Me and OMe gave the desired products 3b and 3c in high
yields (73% and 78%, respectively). Even an electron-
withdrawing substituent like fluoro produced the silylated
product 3d in a very good yield (75%), though chloro and
bromo substituents furnished the products 3e and 3f only in
moderate yields (54% and 37%, respectively). The location of
variously substituted aryl groups (1g−1n) at this position
provided moderate to satisfactory yields of products ranging
from 36 to 50%. Among amides carrying substituents at the
para position (1o−1aa), those with different aryl groups also
showed good conversion and provided the products (3o−3aa)
to the extent of 43−82%.
Changing the ortho-methyl group of benzamide with phenyl,
ethyl, or isopropyl (1ab−1ad) continued to deliver adequate
yields. Notably, the current protocol also allowed the use of the
challenging heterocycle thiophene (substrate 1ae) albeit with a
poor yield (3ae, 15%). However, the unsubstituted benzamide
1af fared worse, with only monosilylated product 3af being
On the basis of the preceding experimental results and
previous literature reports,12−14 a reaction pathway is proposed
as in Scheme 4. In accordance with the known coordination
properties of Cu(II), benzamide 1a first generates the
intermediate A in the presence of CuX2 (X = acetate/
carbonate) with the aid of complexation by an N,N-bidentate
4523
Org. Lett. 2021, 23, 4521−4526