Journal of the American Chemical Society
Communication
sulfoxide4b to examine this hypothesis (Scheme 1d). This
design was based on the following considerations: (1) the 3-
silyl group on benzyne is known to serve as an inductively
electron-donating group (EDG)7 that could direct the
incoming sulfoxide oxygen to attack its ortho-position; (2) it
might then act as a “blocking group” after the allyl group
migrates to the C3-position and trigger a subsequent
rearrangement, presumably a Cope rearrangement, on
intermediate v.
a
Table 1. Reaction Optimization of 1,3-Silyl Migration
entry “F” (equiv) additive (equiv) solvent temp (°C) yield (%)
1
2
3
4
5
6
7
8
CsF (4.0)
CsF (4.0)
CsF (4.0)
CsF (2.0)
CsF (4.0)
CsF (4.0)
KF (4.0)
TBAF (4.0)
CsF (4.0)
CsF (4.0)
CsF (4.0)
CsF (4.0)
no
no
no
no
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
toluene
dioxane
THF
rt
49
83
80
50
78
29
41
10
nr
50
80
50
50
50
50
50
100
80
50
50
As shown in Scheme 2, a series of 3-silylbenzyne precursors
1a−1f were prepared, which could generate the corresponding
Cs2CO3 (2.0)
18-c-6 (2.0)
18-c-6 (2.0)
no
no
no
no
no
Scheme 2. Reactions of 3-Silylbenzynes with Allyl Sulfoxide
9
10
11
12
trace
64
68
DME
a
Conditions: 1e (0.2 mmol), 2a (0.4 mmol), “F”, and BrCH2CO2Et
(0.4 mmol) in solvent (2 mL) overnight.
yield (entries 5 and 6). Other fluoride salts could not enhance
the yield as well (entries 7 and 8). At last, different solvents
were tested and all of them gave lower yields than that with
MeCN (entries 9−12). Finally, the optimal conditions for this
transformation were determined to be CsF in MeCN at 50 °C,
furnishing 3a in 83% yield with complete suppression of the
desilylation pathway (entry 2).
arynes upon activation with CsF in acetonitrile. These
precursors were then subjected to the reaction with p-tolyl
allyl sulfoxide (2a) and ethyl bromoacetate at room temper-
ature. When 1a with a 3-trimethylsilyl (3-TMS) group was
used, the reaction only afforded a desilylation product 4 in 54%
yield. By altering aryne precursors to 1b with a 3-triethylsilyl
(3-TES) group, 1c with a 3-tri(n-butyl)silyl (3-n-Bu3Si) group,
or 1d with a 3-triphenylsilyl (3-TPS) group, compound 4
remained the only obtained product. Intriguingly, when aryne
precursors containing a 3-tert-butyldimethylsilyl (3-TBS)
group (1e) and 3-tris(isopropyl)silyl (3-TIPS) group (1f)
were examined, the main reaction path changed dramatically,
affording 3a and 3b in 49% and 36% yields, respectively, along
with certain amount of the desilylation product 4. Notably, no
product via Cope rearrangement was detected. Alternatively,
the formation of 3a and 3b should proceed through a net 1,3-
silyl migration on the benzene ring. In the presence of a
fluoride ion, those silyl groups with less sterically hindered
substituents are vulnerable, whereas both TBS and TIPS
groups could sustain. Although the thermal 1,3-silyl shift of
allylsilanes has been studied by Kwart8 and Kira9 and has
attracted some theoretical attention,10 harsh reaction con-
ditions (>350 °C) for this transformation prohibited its
synthetic application. Besides the much milder reaction
temperature, our serendipitous discovery is of great interest,
because, to the best of our knowledge, there was no precedent
of thermal 1,3-silyl migration on aromatic systems. Moreover,
this type of transformation provides a convenient means to
deliver an aromatic silyl group to its meta-position specifically
under certain circumstances.
We then explored the substrate scope for this trans-
formation. As shown in Scheme 3, this reaction can be also
scaled up to gram-scale, giving rise to 3a in 79% yield. When 3-
TIPS-substituted benzyne precursor 1f was examined under
the optimal conditions, 3b was obtained in 60% yield. By
altering ethyl bromoacetate to allyl bromide, cinnamyl
bromide, and di-tert-butyl dicarbonate (Boc2O), the corre-
sponding products 3c−3e were obtained as well. The
employment of different aryl groups on sulfoxide afforded
the desired products 3f−3j in moderate to good yields.
Sulfoxides with substituted allyl groups were tested, and
products 3k−3p were readily achieved. Among them, the X-ray
crystallographic structure of 3k unequivocally confirmed this
1,3-silyl migration event. Notably, desilylation reactions were
suppressed, despite the fact that silyl groups are naturally
vulnerable to the fluoride ion. We reasoned that the presence
of either the tert-butyl group or triple isopropyl groups on
silicon would kinetically disfavor the desilylation reaction.
Consequently, the 1,3-silyl shift turned out to be the dominant
path under the reaction conditions.
Distinctively, this cascade process still obeys the regiose-
lective regulation when additional substituents are properly
positioned on the benzyne ring, furnishing arenes with up to
five substituents. For instance, when Kobayashi precursor 1g
was employed, pentasubstituted benzene 3q was harvested in
56% yield (Scheme 4). Similarly, pentasubstituted benzenes 3r
and 3s could be obtained in good yields as well from the
corresponding aryne precursors 1h and 1i. The examples in
Schemes 3 and 4 demonstrate that aromatic 1,3-silyl migration
is a general scenario in this transformation, leading to an
unprecedented tandem maneuver toward polysubstituted
benzenes with substituents of different types.
With the above observation in hand, we decided to optimize
the reaction conditions. As shown in Table 1, in the presence
of ethyl bromoacetate, p-tolyl allyl sulfoxide (2a) could react
with benzyne precursor 1e to afford 3a. By screening the
reaction temperature (entries 1−3, Table 1), it was found that
50 °C was the best one. Altering the stoichiometry of CsF to
2.0 equiv gave only 50% of 3a (entry 4). Additives, such as
Cs2CO3 and 18-crown-6, were used but did not improve the
2179
J. Am. Chem. Soc. 2021, 143, 2178−2184