a
a
Table 1 Reaction of trimethylsulfonium methyl sulfate and dimethyl sulfide with mineral acids and Amberlyst resin
Ϫ
1
Acid–MeSO4 (mol.
H NMR integration
ϩ
Ϫ
4
Acid
equiv.–mol. equiv.)
Time/h
Temp./ЊC
ratio (CH ) S :CH SO
% Reaction
3
3
3
c
HCl (36%)
1
1.5
6
6
6
6
40
40
35–40
35–40
29:1
5.7:1
~80%
25%
100%
~20%
H PO (85%)
3
4
Ϫ
4
PhSO H
4
No peak for CH SO
3
3
b
Amberlyst
4:1
a
b
Trimethylsulfonium methyl sulfate (1.0 equiv.) and dimethyl sulfide (2.0 equiv.). For a polymer supported acid, the mole equivalent cannot be
c
calculated. Mixture used in subsequent epoxidation reaction.
1
and H NMR analysis indicated a 30% conversion to trimethyl-
spectrometer; chemical shifts are quoted in ppm downfield from
tetramethylsilane and 3-(trimethylsilyl)propanesulfonic acid,
sodium salt (DSS), and coupling constants are in Hertz.
Amberlyst 15 was purchased from Aldrich.
sulfonium hydrogen sulfate and methyl hydrogen sulfate after
8 h at room temperature. The reaction of Me S MeSO –
Me S–concentrated H SO was then repeated (40 ЊC, 6 h) with
ϩ
Ϫ
4
3 4
2
2
4
a molar ratio of 1:1.5:2, respectively. After 6 h at 40 ЊC
7
and standing at room temperature, the product consisted of
1
-(2,4-Dichlorophenyl)pentan-1-one (5) and 2-butyl-2-(2,4-
1
8
two layers with the lower layer at pH zero; H NMR analysis of
dichlorophenyl)oxirane (6)
this layer indicated an almost quantitative conversion to tri-
methylsulfonium hydrogen sulfate. The latter was then used
Compounds 5 and 6 were prepared by known methods. NMR
spectral data are as follows: Compound 5 δH (CDCl –TMS)
6
3
for epoxidation using our procedure described earlier: 1-(2,4-
0
3
.75–1.9 (m, 7H, (CH ) CH ), 2.87 (t, 2H, CH ), 7.15–7.5 (m,
2 2 3 2
dichlorophenyl)pentan-1-one (5) [1 mol equiv.], potassium
H, aromatic H).
Compound 6 δ (CDCl –TMS) 0.5–2.4 (m, 9H, (CH ) CH ),
H
3
2
3
3
2
.6–3.0 (m, 2H, CH O), 7.0–7.5 (m, 3H, aromatic H).
2
Reaction of trimethylsulfonium methyl sulfate, dimethyl sulfide
and sulfuric acid
Dimethyl sulfide (3.60 g, 0.058 mol) was added to a solution of
trimethylsulfonium methyl sulfate (5.97 g, 0.029 mol) in water
(0.5 ml). 98% Sulfuric acid (5.8 g, 0.058 mol) was added drop-
wise to the stirred mixture which was then heated to 40 ЊC and
stirred for 6 h. The product was held unagitated for 16 h after
which it consisted of two layers with a large lower aqueous layer
hydroxide (5 equiv.) and trimethylsulfonium cation (1.05 equiv.)
were allowed to react in tert-butyl alcohol at room temperature
to afford the epoxide 6 in 73% yield. The use of hydrochloric,
phosphoric, benzene sulfonic acid and a strong cation exchange
resin (Amberlyst 15) was then evaluated in respect of the
pre-epoxidation step as described above (see Table 1).
1
of pH = 0. A sample of this layer was removed for H NMR
Ϫ
The use of hydrochloric acid showed a significant rise in the
analysis. δ (DMSO d –DSS) 2.1 (s, 1H, (CH ) S), 3.0 (s, 13H,
H
6
3 2
ϩ
Ϫ
4
ϩ
Ϫ
(
CH ) S :CH SO , ratio but phosphoric acid showed only a
(CH ) S ), 3.5 (s, 0.1H, CH SO ).
3
3
3
3 3 3 4
slight increase under the same conditions. Benzenesulfonic acid,
using a four-fold molar excess, resulted in complete conversion,
with no methyl sulfate anion remaining. The use of Amberlyst
Epoxidation of 1-(2,4-dichlorophenyl)pentan-1-one (5) using the
reaction mixture from above
1
5 resulted in a slight increase in the ratio to 4:1, but the resin
To the above reaction mixture containing trimethylsulfonium
cation (assumed 0.058 mol) was added tert-butyl alcohol (4.10
g, 0.055 mol) and 1-(2,4-dichlorophenyl)pentan-1-one (13.38 g,
absorbed the liquid and the reaction mixture became very
viscous. Further work would be required to optimise the
quantity of Amberlyst 15 required.
0
.055 mol). This mixture was stirred and potassium hydroxide
The reaction mixture produced from the use of hydrochloric
acid was also used for epoxidation as described above. Two
equivalents of 1-(2,4-dichlorophenyl)pentan-1-one (5) were
used per equivalent of trimethylsulfonium methyl sulfate and
the product epoxide 6 was isolated in 54% yield. This yield
could be improved by allowing the reaction of trimethyl-
sulfonium methyl sulfate and dimethyl sulfide to go to com-
pletion, either by allowing a longer reaction time, by increasing
the temperature or by adding a larger excess of hydrochloric
acid. It may be noted that the mixture produced from this
reaction will contain the trimethylsulfonium cation with two
(
15.4 g, 0.275 mol) was added in portions. The colour of
the mixture changed quickly to yellow and an exotherm was
observed; cooling with an ice bath was applied. Half of the
potassium hydroxide was added and the mixture was stirred for
1
h, then the remainder was added over 20 min. The mixture
was very viscous but became less so during the 3 h stirring
period. The mixture was held unagitated over 2 days. Samples
were removed periodically for gas chromatographic analysis
(
Shimadzu GC-9A instrument with a Shimadzu C-R3A
Chromatopac Integrator). The column was of fused silica
(2 m × 3 mm) with 5% OVI7 stationary phase; the oven tem-
perature was 180 ЊC. Samples (~0.5 ml) were added to a mixture
of dichloromethane (~1 ml) and water (~1 ml) and shaken. The
lower dichloromethane layer (1 µl) was injected, showing reten-
tion times for 1-(2,4-dichlorophenyl)pentan-1-one (5) and the
epoxide 6 = 8.4 and 7.7 min, respectively.
Ϫ
Ϫ
4
different counter ions (Cl and HSO ).
Conclusions
A new, efficient epoxidation method has been devised based on
dimethyl sulfate (1 equiv.) and dimethyl sulfide (2 equiv.) to
produce trimethylsulfonium cation (2 equiv.), utilizing both
methyl groups present in dimethyl sulfate. Sulfuric acid is used
to generate methyl hydrogen sulfate as an intermediate, but
other acids, e.g. hydrochloric acid, can also fulfil the same role.
The progress of reaction was monitored by GC % areas for
ketone 5 and epoxide 6 at the following times: end of KOH
addition (1 h, 20 min; ratio of 5:6 = 86:13), 3 h (5:6 = 25:68)
and 48 h (5:6 = 9:89); the yield of 6 after 48 h was 73%. 3
To the bulk reaction mixture was added water (50 cm ) and
the solvents were then evaporated by distillation, up to a
Experimental
3
column head temperature of 55 ЊC. Water (ca. 250 cm ) was
1
H NMR spectra were recorded on a Bruker WP-SY (80 MHz)
added and the lower organic layer was separated and washed
3
334
J. Chem. Soc., Perkin Trans. 1, 1999, 3333–3335