Organic Letters
Letter
Procter’s group introduced an interrupted Pummerer approach
allowing C3 arylation and allylation processes on benzothio-
phene S-oxides (Scheme 1a).7 Here we report a mechanisti-
cally distinct benzothiophene functionalization method that
uses simple alkynes to access complementary functionality at
C3 through a gold-catalyzed oxyarylation process (Scheme
1b).
Gold- and subsequently Brønsted acid-catalyzed C−C bond-
forming rearrangements between sulfoxides and alkynes have
shown great promise for atom-efficient aryl C−H functional-
ization and other transformations.8 For the desired trans-
formation, the vinylgold carbenoid intermediate B should
evolve through a [3,3]-sigmatropic rearrangement and onto C3
rather than the C7 benzenoid position (Scheme 1b).
Rearrangement must occur before the electrophilic organogold
species reacts with a second equivalent of sulfoxide.8c Our
interest in gold-catalyzed reactions with sulfoxides9 led us to
investigate the potential of oxyarylation for C3 benzothiophene
elaboration.
Our study started with 2-methylbenzothiophene S-oxide (1)
and 1-hexyne (2a), as unsubstituted benzothiophene S-oxide is
unstable and decomposes out of solution (vide infra).10 All
reactions were performed under non-inert conditions with
undried, commercial solvents. A range of Au(I) catalysts was
tested (Table 1, entries 1−6). In all cases the product 3a from
reaction at C3 was preferred, but some of the C7-function-
alized product 4a was also seen. Low reactivity occurred with
SPhosAuNTf2 and was not improved with higher temperatures
(Table 1, entries 1 and 2). Cleaner reactions and greater
conversion to the desired product were seen with the
phosphite-containing catalyst [DTBPAu(PhCN)]SbF6 over
those derived from more electron-rich ligands, with small
amounts of benzothiophene 5 also observed in the latter case
(Table 1, entries 1−5). Use of a more coordinating tosylate
counterion saw reduced activity (Table 1, entry 6).
Improved yields were obtained in CH2Cl2 and fluoroben-
zene (Table 1, entries 7 and 8). A more substantial solvent
effect was seen upon switching from hexyne to phenylacetylene
(2b), with fluorobenzene giving a superior outcome with a
90% combined yield of oxyarylation products and much higher
regioselectivity (Table 1, entry 10 vs 11). A modest reduction
in yield was seen at a catalyst loading of 2.5 mol % (Table 1,
entry 12). Reducing the equivalents of alkyne lowered the
overall yield of 3 and 4 with both 1-hexyne and phenyl-
acetylene (Table 1, entries 7 vs 9 and 11 vs 13). Reversing the
stoichiometry of sulfoxide 1 and phenylacetylene 2b further
reduced the combined yield of 3b and 4b (Table 1, entry 14).
The best conditions were then applied across a range of
alkynes and benzothiophene S-oxides (Scheme 2), with the
reactions typically performed on a 0.2−0.5 mmol scale. On the
1.0 mmol scale, isomerically pure 3b was isolated in 68% yield.
Both electron-rich and electron-deficient aryl alkynes under-
went oxyarylation successfully. Appreciable amounts of the free
benzothiophene were observed with more electron-rich aryl
alkynes (Scheme 2, compounds 6−9 and 24). A low yield and
poor C3:C7 regioselectivity were seen with an o-methoxy
group on the aromatic ring. Aliphatic alkynes provided
excellent yields with lower regioselectivities compared with
aromatic alkynes (Scheme 2, compounds 15−18 and 23).
Good functional group tolerance was shown across these
reactions with aryl halides, primary alkyl bromide, a tertiary
amine, phthalimide, and carboxylic esters all readily incorpo-
rated.
Table 1. Reaction Optimization Study between 2-
Methylbenzothiophene S-Oxide and Hex-1-yne or
Phenylacetylene
The formation of 12−15 and 25 (Scheme 2) required
higher temperatures for oxyarylation to occur. During the
formation of 15 with 2.0 equiv of alkyne, the alkyne hydration
product was also formed and was inseparable from the desired
product. Using 1.0 equiv of alkyne saw no formation of the
hydration side product, and 15 was isolated in good yield. No
oxyarylation product was formed when diphenylacetylene and
1 were submitted to the reaction conditions. Only the benzil
product from double oxidation of the alkyne was isolated (98%
based on sulfoxide 1).
Benzothiophene S-oxides with different substitution patterns
also performed well. The reaction proceeds well in the absence
of a C2 substituent, and halogen substituents are also tolerated
(Scheme 2, 19−25). In several cases very high regioselectiv-
ities were observed (>20:1), so simple trituration with
methanol was sufficient to obtain the single C3 regioisomer
(Scheme 2, 21, 24, and 25). Good outcomes were also
observed with an ester group at C2 (Scheme 2, 26 and 27).
The presence of a substituent at C3 did not lead to efficient
formation of the C7 oxylarylation product, as a complex
mixture was formed upon reaction of 3-methylbenzothiophene
S-oxide with phenylacetylene 2b.
a
b
entry 2 (equiv)
catalyst
solvent
yields (1:3:4:5) (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
2a (2)
2a (2)
2a (2)
2a (2)
2a (2)
2a (2)
2a (2)
2a (2)
2a (1)
2b (2)
2b (2)
2b (2)
2b (1)
2b (1)
A
PhMe
PhMe
PhMe
PhMe
PhMe
PhMe
CH2Cl2
C6H5F
CH2Cl2
CH2Cl2
C6H5F
C6H5F
C6H5F
C6H5F
70:16:2:0
74:16:2:0
32:39:10:−
A (23 °C)
c
B
C
D
c
46:33:4:−
d
− :55:23:0
e
E/AgOTs
88:6:4:0
0:64:23:−
c
D
D
D
D
D
7:63:24:0
f
0:50:17:<6
d
c
- :51:9:−
f
0:80:10:<7
g
g
f
D (2.5 mol %)
0:67:12:<10
0:62:9:8
− :50:9:−
D
D
d
c
a
A, SPhosAuNTf2; B, IPrAuNTf2; C, [JohnPhosAu(MeCN)]SbF6; D,
[DTBPAu(PhCN)]SbF6; E, DTBPAuCl; DTBP
= (2,4-
b
(tBu)2C6H3O)3P. Yields and ratios were determined by H NMR
1
As unsubstituted benzothiophene S-oxide is unstable when
neat, a protocol was developed that telescoped together the S-
oxidation of benzothiophene with the gold-catalyzed oxy-
arylation. C3-substituted benzothiophene 28 was obtained in
details).
spectroscopy using a known concentration of 1,2,4,5-tetramethylben-
zene. 5 was present, but the yield could not be determined because of
overlapping resonances. 1 was present, but the yield could not be
determined because of overlapping resonances. E (0.05 equiv) and
AgOTs (0.10 equiv). Represents the maximum possible yield because
of overlapping resonances. 2 equiv of 1.
c
d
e
f
g
643
Org. Lett. 2021, 23, 642−646