T. Sanji et al. / Tetrahedron Letters 57 (2016) 1921–1924
1923
or even unreactive under these conditions (entries 6 and 7). The
reaction of 1-phenyl-2-(trimethylsilyl)acetylene 1b or 1-(4-trifluo-
romethylphenyl)-2-(trimethylsilyl)acetylene 1c and 2a with
CsF/18-crown-6 (1 equiv) gave 3b and 3c, respectively, in moder-
ate yields under the conditions (entries 8, 9).
fluorosilicate and fluoride ions occurred to give nucleophilic
ethoxide ions. The resulting ethoxide ion attacked the trifluo-
rophenyl group of 2b to afford 5.
For the reaction of 1a and 2b, the best reaction condition was
CsF/18-crown-6 as the base in DMSO (entries 13 or 15). This was
because no byproducts were found and the isolation of the product
was easy.
Next, the reaction of a 2,4,6-trifluorophenyl derivative was
examined. The addition of 1-(2,4,6-trifluorophenyl)-2-pheny-
lacetylene 2c to 1a with CsF/18-crown-6 (1 equiv) at room
temperature in DMSO gave a mixture of 1,3-difluoro-5-((4-meth-
oxyphenyl)ethynyl)-2-(phenylethynyl)benzene 3g-p and 1,5-diflu-
oro-2-((4-methoxyphenyl)ethynyl)-3-(phenylethynyl)benzene 3g-
o in 47% yield (entry 24). The regioisomers were easily separated
from each other using preparative HPLC, the isomers 3g-p and -o
being obtained in 3% and 26% yields, respectively. The structure
of the major isomer was identified to be 3g-o. The product ratio
of 3g-o/3g-p was estimated to be 2.4 from the integration ratio
of the 19F NMR spectra of the isolated mixture. These indicate that
the reaction is not regioselective.
Finally, the reaction of 1a and 1-(3,4-difluorophenyl)-2-pheny-
lacetylene 2d with CsF/18-crown-6 (1 equiv) at room temperature
in DMSO gave 3h in 8.6% yield (entry 25). After the reaction, 2d
was recovered in 54% and no regioisomers were found. As
expected, the reactivity of 2d was quite low.
In summary, we have demonstrated the Lewis base-promoted
alkynylation of fluorobenzenes with trimethylsilylacetylenes. The
reactivity of the fluorobenzenes as a function of the number of flu-
oride groups was also noted. The reaction of pentafluorophenyl
and 3,4,5-trifluorophenyl derivatives and trimethylsilylacetylene
with CsF/18-crown-6 gave the alkynylated products in moderate
to good yields with high regioselectivity under the mild conditions.
However, the 3,4-difluorophenyl derivative showed low reactivity.
We believe these results provide a rather detailed sketch of the
landscape in the Lewis base-promoted alkynylation of electron-
deficient benzenes with trimethylsilylacetylenes in future
applications.
With adequate reaction conditions in hand, we then examined
the reaction of 1-(3,4,5-trifluorophenyl)-2-phenylacetylene 2b. In
the reaction of 1a and 2b with CsF/18-crown-6 (1 equiv) in DMSO
at room temperature (entries 10–12), the conversion of 2b reached
only 35% after 2 h when monitoring the reaction using 19F NMR. A
longer reaction time (4 and 6 h) did not show an improvement and
the resulting isolated yield of 3d was 37%. The reaction at 80 °C
gave 3d in 44% yield (entry 13). When 10 mol % CsF/18-crown-6
was used, 3d was obtained in 33% yield (entry 14). Under these
conditions, the yield was almost the same as that found after
18 h. Alternatively, the addition of 2b to a mixture of 1a and
CsF/18-crown-6 in DMSO impacted on the reaction outcome (entry
15). The reaction proceeded at room temperature to give 3d in 53%
yield. However, the reaction at 80 °C resulted in 10% yield (entry
16). It is noted that no other products were found and unreacted
2b was recovered quantitatively after the reaction under the con-
ditions. Potassium t-butoxide also worked as the base to afford
3d, but the reaction features were different. In the reaction of 1a
and 2b with 1 equiv of potassium t-butoxide, the conversion of
2b was estimated to be 80% (entry 17). After the reaction, 3d
was obtained in 40% yield and 1-(4-t-butoxy-3,5-difluorophenyl)-
2-phenylacetylene 4 was also found in 33% yield. This indicates
that the reaction of the t-butoxide ion and the 3,4,5-trifluo-
rophenyl group of 2b occurred. Thus, the t-butoxide ion worked
as the base to the trimethylsilyl group of 1a and also as the nucle-
ophile to 2b in the reaction. When the reaction with potassium
t-butoxide was run in the presence of cryptand[2.2.2], 3d was
obtained in 52% yield and 4 was also found in 26% yield (entry
18). The addition of 2b to a mixture of 1a and potassium t-butox-
ide/cryptand[2.2.2] at room temperature or 80 °C was not effective
(entry 19). On the other hand, TBAF did not work well (entry 20).
As examples of the substrate scope of the reaction, the reaction
of 1-phenyl-2-(trimethylsilyl)acetylene 1b or 1-(4-trifluo-
romethylphenyl)-2-(trimethylsilyl)acetylene 1c and 2b with
CsF/18-crown-6 (1 equiv) in DMSO at 80 °C was tolerated and
the yields of 3e and 3f were 37% and 32%, respectively (entries
21, 22).
Acknowledgment
This work was partially supported by JSPS KAKENHI Grant
Number 15H03813.
In the reaction, the activation of
a carbon–silicon bond
catalyzed by fluoride ions to form nucleophilic pentacoordinated
fluorosilicates is involved. Considering this, 1-triethoxysilyl-2-
phenylacetylene 1d was used because alkoxy groups on the silicon
are usually accessible to a stable pentacoordinated fluorosilicate in
the presence of a Lewis base (entry 23).11 However, 1-(4-ethoxy-
3,5-difluorophenyl)-2-phenylacetylene 5 was obtained in 37%
yield as the major product. A plausible mechanism is as follows
(Scheme 2). Because the energy of the Si–F bond is much larger
than that of the Si–O bond,5 exchange of the fluoride and ethoxide
ions on the difluorosilicate resulting from the reaction of the
Supplementary data
Supplementary data associated with this article can be found, in
References and notes
F
CsF/18-crown-6 (1)
EtO
1d
2b
+
DMSO, rt, 2 h
F
5
(37%)
F−
2b
(EtO)2SiF2 EtO−
Si(OEt)3F
(EtO)3SiF2
+
F−
Scheme 2. Reaction of 1d with 2b in the presence of CsF/18-crown-6 (1 equiv).
Sanji, Takanobu
