Organic Letters
Letter
Scheme 1. Sigmatropic Rearrangement of Sulfonium
Intermediates for the Synthesis of Carbonyl-Containing
Compounds
Table 1. Optimization of the Reaction Conditions
a
a
entry
[Au]
AuCl
IPrAuCl
additive Conv. (%) (3a/3′a) yield of 3a (%)
1
2
3
4
5
6
−
−
−
−
−
−
−
H2O
H2O
H2O
90 (1/1.1)
90 (1/1)
>98 (1/1)
90 (1.1/1)
83 (1.8/1)
>98 (1/1)
<5
>98 (7/1)
50 (1/1)
>98 (10/1)
37
43
50
47
53
48
nd
72
25
(PhO)3PAuCl
b
L1AuCl
L2AuCl
c
Ph3PAuCl
Ph3PAuCl
Ph3PAuCl
−
d
7
e
e
e
8
9
f
g
10
Ph3PAuNTf2
76 (74 )
a
Conversions, 3a/3′a ratios, and yields of 3a were determined by 1H
NMR analysis of the crude reaction mixtures with an internal
standard. L1 = di-1-adamantyl-2-morpholinophenylphosphine. L2 =
tris(4-methoxyphenyl)phosphine. No silver salt was added. 3.0
b
c
d
e
f
equiv of H2O was added. Only the silver salt AgNTf2 was used.
g
Isolated yield.
necessary to form the cationic gold catalyst (entry 7), we added
some water to the reaction mixture to facilitate the hydrolysis of
the thionium ion and to favor the formation of 3a (entry 8). The
expected effect was verified, as the yield of 3a increased from 48
to 72%. Silver salt alone did not efficiently catalyze the reaction
(entry 9). It is noteworthy that the use of HNTf2 as the catalyst
gave the same disappointing results (38% yield). Among all of
the different solvents and counteranion tested, AgNTf2 proved
to be the best silver salt and DCE the best solvent (see Table
S1). Finally, the optimal result was obtained with the use of
preformed Ph3PAuNTf2 complex (10 mol %) in the presence of
water at room temperature for 14 h (entry 10). The desired
compound 3a was isolated in pure form in 74% yield.
Under the optimized reaction conditions, we next examined
the substrate scope, and the results are summarized in Scheme 2.
We started our study with the use of various substrates 2a−e (R1
= alkyl, R2 = H) (Scheme 2a). The products 3a−3e were isolated
in 44−74% yield. Product 3f, with a cyclohexyl group, was
isolated in 64% yield. Interestingly, the methodology could be
efficiently extended to the 2-arylvinyl sulfoxide substrates 2g−i
(R1 = aryl, R2 = H). 4-(4-Methoxyphenyl)-4-oxo-2-phenyl-
butanal (3g) was synthesized in 77% yield on a 1 mmol scale.
Regardless of the electron-donating p-methoxy and electro-
negative p-fluoro substituents, the corresponding products 3h
and 3i were obtained in 67−70% yield. A heteroaryl substituent
such as a thiophene group was also well-tolerated (compound 3j,
65% yield). The second step was then to expand the substrate
scope to various aryl alkynes 1b−h (Scheme 2b). Seven different
aryl-substituted alkynes reacted with vinyl sulfoxide 2a. Indeed,
1,2-dimethoxybenzene, 6-methoxynaphthalene, and ferrocene
alkynes reacted smoothly to give compounds 3k−m in up to
73% yield. Substrates possessing hydroxyl groups exhibited
moderate yields in this transformation because of the instability
of both the alkyne substrates and the compounds 3n and 3p
under air atmosphere. The use of a TBS protecting group
oxygen atom of the sulfoxide partner 27h (Scheme 1d).
Subsequently, the sulfonium intermediate (IV) should perform
a [3,3]-sigmatropic rearrangement to deliver a thionium ion
(V). After hydrolysis, the corresponding 1,4-keto aldehydes 3
should be isolated.
In this paper, we report our efforts in this direction and the
extension of this methodology to the use of 1-cycloalkenyl
sulfoxides 5 as starting materials, whose potential has been
underexploited to date. The corresponding highly function-
alized cycloalkyl ketones 6 have also been used as key building
blocks for the rapid synthesis of five active pharmaceutical
ingredients (APIs).
Our study was initiated by reacting 1-ethynyl-4-methox-
ybenzene (1a) and sulfoxide 2a in the presence of a 10 mol %
loading of different gold(I) precatalysts in dichloroethane at
room temperature (Table 1). After activation with silver
bis(trifluoromethanesulfonyl)imide, the corresponding cationic
gold catalysts without ligand (entry 1) or coordinated with
structurally diverse ligands such as an N-heterocyclic carbene
(entry 2), triphenyl phosphite (entry 3), or phosphine ligands
(entries 4−6) furnished the desired 2-(2-aryl-2-oxoethyl)-
pentanal (3a) in moderate yields. The formation of the
bis(arylsulfane) byproduct 3′a was also observed. Slight
differences in both conversion and product selectivity (3a/
3′a) were ascertained depending on the catalyst used. We
decided to continue our study with commercially available
Ph3PAuCl as the catalyst (entry 6). After verifying that it was
B
Org. Lett. XXXX, XXX, XXX−XXX