First Epoxidation Reaction of Carbonyl Compounds via Ferrocenyl Sulfur Ylides

Article abstract of DOI:10.1055/s-2003-41445

Epoxidation reactions of carbonyl compounds (aldehydes and a ketone) mediated by sulfanyl ferrocenes have been successfully achieved in a one-pot reaction, under mild conditions. This reaction implies intermediary formation of a sulfonium salt (a ferrocenyl one was observed for the first time by ¹H NMR) and then an ylide. The diastereoselectivity of the formation of stilbene oxide was unusual: the effect of steric hindrance and aromatic nature of the sulfide substituents have been evidenced. A catalytic asymmetric example with a planar chiral ferrocene is described.

Full text of DOI:10.1055/s-2003-41445

SPECIAL TOPIC
2249
First Epoxidation Reaction of Carbonyl Compounds via Ferrocenyl Sulfur
Ylides
a
a
b
a
b
E
S
poxidation
R
t
eaction
é
of Carbon
p
h
s
via Ferroc
a
enyl
S
ulfur
n
Ylides ie Minière, Vincent Reboul, Ramón Gómez Arrayás, Patrick Metzner,* Juan Carlos Carretero
a
Laboratoire de Chimie Moléculaire et Thio-organique (UMR CNRS 6507), ENSICAEN-Université, 6, boulevard du Maréchal Juin,
4050 Caen, France
1
b
Departamento de Química Orgánica, Facultad de Ciencias, Universidad Autónoma de Madrid, 28049, Madrid, Spain
Fax +33(231)452877; E-mail: Metzner@ismra.fr
Received 19 July 2003
1
1
tions. A study of this unreported structure would be de-
sirable. Deprotonation of the sulfonium salt should take
place preferably to the sulfur center (relative to ortho-
metalation on the cyclopentadienyl rings). The expected
Abstract: Epoxidation reactions of carbonyl compounds (alde-
hydes and a ketone) mediated by sulfanyl ferrocenes have been suc-
cessfully achieved in a one-pot reaction, under mild conditions.
This reaction implies intermediary formation of a sulfonium salt (a
1
ferrocenyl one was observed for the first time by H NMR) and then ylide brings a number of questions: structure and stabili-
an ylide. The diastereoselectivity of the formation of stilbene oxide zation through extended delocalization, as well as reactiv-
was unusual: the effect of steric hindrance and aromatic nature of
ity. Here, we report the syntheses of several sulfanyl
the sulfide substituents have been evidenced. A catalytic asymmet-
ferrocenes, sulfonium salts and ylides, and the promising
ric example with a planar chiral ferrocene is described.
results of their first use in epoxidation reactions.
Key words: sulfur , ylides, epoxidations, aldehydes, ferrocene,
stereoselectivity
Ferrocene derivatives have received much attention in the
two last decades. A number of ferrocenylphosphines were
shown to be efficient bidentate ligands for transition-met- Scheme 1
1
–3
al-catalysed asymmetric reactions. However, only a
few examples of reactions using non-phosphino fer-
rocenes as catalysts were described in the literature. Sul-
fur derivatives of metallocenes may find useful
applications in chemistry in connection with a variety of
sulfur functional groups: sulfides, sulfoxides, sulfimines,
thiocarbonyl compounds, etc. Togni reported the synthe-
sis of a variety of ferrocenyl sulfides and some of their us-
First of all, we have worked with achiral sulfides. The in-
troduction of sulfur on the ferrocene molecule may be ef-
ficiently achieved by several methods. The most common
way involves the formation of a ferrocenyllithium (substi-
tuted or not), followed by the introduction of the appropri-
6,12–14
9
ate electrophile, i.e. a disulfide
or a sulfinate. An
alternative involves nucleophilic substitution on a carbon
adjacent to a Cp ring, by reacting a thioacetate with an
4,5
es.
Daï described Tsuji–Trost alkylations with
6
ferrocenyl oxazoline ligands, whereas Bolm used similar
4
3
acetate or an ammonium.
7
ligands for dialkylzinc additions to aldehydes. Bäckvall
Using the first method, Kagan and co-workers described
the formation of methyl and phenyl sulfanyl derivatives
from the reaction of ferrocenyllithium and the corre-
sponding disulfides in moderate to good yields.1 We re-
produced these syntheses with Kagan optimized
conditions15 and extended them to others R groups
and co-workers developed ferrocenyl thiolates as ligands
in the copper-catalyzed substitution of allylic acetates
with Grignard reagents. More recently, 2-amino-substi-
8
2,15
tuted 1-sulfinylferrocenes were used as efficient ligands
for the asymmetric addition of diethylzinc to aromatic al-
1
9
dehydes.
(
Scheme 2, Table 1).
Connecting an ylide moiety to a ferrocene may provide
original structures and reactivity. Indeed, ferrocenyl phos-
phorous ylides are known and used in Wittig reactions.10
Surprisingly, their sulfanylated analogues have not yet
been described in the literature (Scheme 1) and conse-
quently, have never been applied for the epoxidation of
carbonyl compounds. We anticipated sulfonium salt for-
mation adjacent to the ferrocenyl group would be favored Scheme 2
in analogy to the stabilization of ferrocenyl carboca-
Synthesis 2003, No. 14, Print: 02 10 2003. Web: 24 09 2003.
Art Id.1437-210X,E;2003,0,14,2249,2254,ftx,en;C03903SS.pdf.
DOI: 10.1055/s-2003-41445
©
Georg Thieme Verlag Stuttgart · New York

2
250
S. Minière et al.
SPECIAL TOPIC
Monosubtitution of ferrocene has been achieved in good Since the sulfonium salt formation was effective, we in-
yields. In the case of the phenyl compound (entry 2), the vestigated the potential of our derivatives in epoxidation
yield was improved from 40% to 54%. The problem of reactions under the conditions recently reported with a C2
bis-lithiation was encountered, as previously reported.15 symmetrical sulfide.
16,17
They involve a practical one-pot
This was followed by the formation of the undesired dis- procedure under mild conditions: at room temperature us-
ubstituted products 3 (entries 1–5, Table 1). These com- ing a 9:1 mixture of t-BuOH–H O as solvent and a miner-
2
pounds were isolated by slow elution through column al base. In situ formation of the sulfonium salt and the
1
chromatography. Compound 3c (R = t-Bu) decomposed ylide are requisite for a version that is designed to be cat-
during these purification conditions (entry 3).
alytic in sulfide. The carbonyl compounds tested were ar-
omatic and aliphatic aldehydes and a ketone (Scheme 4,
Table 2).
Table 1 Synthesis of Ferrocenyl Sulfides
Entry
R1
Product
2 (%)
67
3 (%)
27
1
2
3
4
Me
Ph
a
b
c
54
40
t-Bu
Cy
Bn
44
–a
d
e
67
22
Scheme 4
5
61
23
a
Decomposed on silica gel.
In all cases the expected oxiranes were formed as judged
1
by H NMR. Attempted isolation by column chromato-
As ferrocenyl sulfonium salts had not been described in graphy of the crude product revealed that they have a po-
the literature, so we desired to monitor their formation by larity very similar to that of ferrocenyl sulfides, requiring
1
H NMR. Methylsulfanylferrocene (2a) and 2 equiv of tedious separations. Then, we chose to oxidize the sulfide
benzyl bromide were mixed together in a 9:1 mixture of into the corresponding sulfoxide with 1 equiv of MCPBA.
CD CN–D O. The conversion of sulfide into sulfonium This reaction was fast and the new derivatives were more
3
2
bromide was measured over 14 h (Scheme 3, Figure 1). polar. This allowed easy purification by silica gel column
After 5 min, a signal appeared at = 3.19, corresponding chromatography.
to the methyl group ( = 2.27 in 2a). The reaction mixture
composition seemed stable after 5 hours and the equilibri-
um ratio sulfide–sulfonium salt was about 7:93.
In most cases, oxiranes were obtained in good to excellent
yields (entries 1–7). For 2a, reaction times were compara-
18
ble to those of a standard thiolane (entries 1 and 7). A
catalytic amount (0.2 equiv) of sulfide 2a could be suc-
cessfully used (entry 5) but needed 7 days for completion.
This time could be significantly reduced (to 1.5 d) by us-
ing 1 equiv of iodide salt (entry 6) to perform halogen ex-
change and in situ formation of the more reactive benzyl
iodide.17 Conversion of an aliphatic aldehyde, CyCHO,
led mainly to cis-oxiranes (entries 7 and 15). This is un-
usual for sulfur ylides but similar trends were noted with
Scheme 3
(
2R,5R)-2,5-dimethylthiolane from the aromatic aldehyde
16
1
series to the aliphatic one. With 2c and 2e (R = t-Bu and
Bn, entries 13 and 17, respectively), no epoxidation reac-
tion took place and degradation of aldehyde and auxiliary
was observed. A ketone was converted into a trisubstitut-
ed oxirane but required prolonged reaction time (entry 8).
No epoxidation reaction occurred with the phenylsulfanyl
ferrocene (2b) (entry 10). The reaction was conducted in
the solvent mixture MeCN–H O (9:1) to improve the sol-
2
ubility of 2b. However, after 14 days, only tribenzylamine
Figure 1 Formation of ferrocenyl sulfonium salt
was formed from acetonitrile and benzyl bromide in the
16
presence of NaOH. The use of the benzylsulfanyl deriv-
Deprotonation of the sulfonium salt was then monitored
by adding 2 equiv of KOD. After 10 min, the H NMR
ative 2e (entry 16) showed moderate conversion, even af-
1
ter 14 days. Consequently, sulfides 2a, 2c and 2d
1
spectrum showed that the two benzylic hydrogens were
exchanged. So, the deprotonation was fast and reversible,
(
R = Me, t-Bu and Cy) are efficient for epoxidation reac-
tion.
1
6
as expected from a previous study.
Synthesis 2003, No. 14, 2249–2254 © Thieme Stuttgart · New York

SPECIAL TOPIC
Epoxidation Reaction of Carbonyl Compounds via Ferrocenyl Sulfur Ylides
2251
Table 2 Oxirane Synthesis
Diastereoselectivity is a synthetic key issue. It is not yet
fully controlled for the epoxidation reaction, and deserves
further studies. In our case, when the size of the R group
increases, we observed a higher trans diastereoselectivity.
En- R1
try
Sulfide (solvent or ad-R R3 Ox- TimeYield dr
ditive)
a
2
irane (d) (%) (trans–
b
cis)
For benzaldehyde, the diastereomeric ratio was about
c
1
1
2
3
4
5
6
Me 2a
Ph
H
H
H
H
H
H
4
4
4
4
4
4
3.5 100 62:38
6
0:40 with 2a (R = Me, entry 1) but rose to 80:20 with 2c
1
and 2d (R = t-Bu and Cy, entries 11 and 14, respective-
ly). Influence of the solvents was also studied, with the ex-
ample of methylsulfanyl ferrocene (2a). Three different
Me 2a (CH Cl –H O, 9:1) Ph
7
1
92 81:19
78 77:23
2
2
2
Me 2a (MeCN–H O, 9:1) Ph
2
solvents of increasing polarity as mixture with H O were
2
Me 2a (DMSO–H O, 9:1) Ph
0.7 82 75:25
74 62:38
1.5 98 62:38
2
used (CH Cl < MeCN < DMSO, entries 2–4). Results
2
2
Me 2a (0.2 equiv)
Ph
7
showed that the reactions were faster in more polar sol-
vents (entries 2 and 4), probably due to a better stabiliza-
tion of the charged intermediates (sulfonium salts, ylides
and betaines). Diastereomeric ratios were higher in MeCN
and DMSO (entries 3–4) than in t-BuOH (entry 1), in ac-
cordance to previous observations.19 Somewhat surpris-
ingly, the best dr was obtained in a biphasic CH Cl –H O
Me 2a (0.2 eq) + Bu NI Ph
4
(
1 equiv)
7
8
9
0
Me 2a
Me 2a
Me 3a
Cy
H
5
1
14
1
86 33:67
56d 50:50
90d 60:40
Ph Me 6
Ph
2
2
2
H
4
system (entry 2).
How can we rationalize the effect of steric hindrance and
nature of the solvent on diastereoselectivity? What do we
know of kinetic versus thermodynamic control?
1
Ph
2b (MeCN–H O, 9:1) Cy, H 4,5 14
–
–
2
Ph
11
12
13
14
15
16
17
t-Bu 2c
Ph
Ph
Cy
Ph
Cy
Ph
Cy
H
H
H
H
H
H
H
4
4
5
4
5
4
5
14
14
7
50d 81:19
68d 81:19
The reversibility of the formation of the syn and anti be-
t-Bu 2c + NaI (1 equiv)
t-Bu 2c
taines has been investigated by equilibration and cross-
over reactions.1
9,20
We wished to carry out such experi-
–
–
ments with our ferrocenyl epoxidation system. We have
thus synthesized independently the corresponding diaste-
reomerically pure sulfonium salts by reaction of a fer-
rocene thiolate separately with trans- and cis-stilbene
Cy 2d
14
14
14
14
95d 80:20
50d 44:56
36d 72:28
Cy 2d
21
Bn 2e
oxides, and S-alkylation of the resulting hydroxysulfides
with methyl triflate.22 Treatment with sodium hydroxide
provided stilbene oxides. To discount a possible base-cat-
alyzed epimerization, a second set of experiments was
performed by adding 3 equiv of a more reactive aldehyde,
p-nitrobenzaldehyde (Scheme 5).
Bn 2e
–
–
a
b
c
d
t-BuOH–H O (9:1) unless otherwise noted.
Determined on the crude H NMR spectra.
2
1
9
0% conversion after 1 day and dr 60:40.
Conversion (%).
With both solvents (t-BuOH and CH Cl ) the anti betaine
2
2
7
8
selectively gave the trans epoxide, whereas the syn one
gave cis (major) epoxide accompanied by the trans iso-
We have also compared the epoxidation reaction with
monosulfanyl 2a and disulfanyl 3a derivatives (entries 1
and 9). No change was observed from 2a to 3a. Thus, di-
substitution was not kinetically significant.
mer [15% in CH Cl –H O (9:1), 11% in t-BuOH–H O
2
2
2
2
(9:1)]. In CH Cl –H O, cross-over product with p-ni-
2 2 2
Scheme 5
Synthesis 2003, No. 14, 2249–2254 © Thieme Stuttgart · New York

2
252
S. Minière et al.
SPECIAL TOPIC
trobenzaldehyde was observed (trans isomer, 25%), with Unhindered sulfides, thiolane and dimethyl sulfide, led to
a 1:3 ratio over the cis-stilbene oxide. In t-BuOH, 4% of stilbene oxide with 85:15 and 83:17 (trans–cis) ratios (en-
cross-over product (trans only) was observed.
tries 1 and 2). With a more hindered compound, t-butyl
methyl sulfide, the ratio rises to 94:6 (entry 4), confirming
again that steric hindrance enhances the trans selectivi-
These experiments showed that formation of the anti be-
taine is irreversible whereas formation of the syn one was
reversible to some extent, in both dichloromethane and
protic solvent, t-BuOH. Our observations are somewhat
different of those of Aggarwal and his group with dimeth-
2
4
ty. These data are far from the ratios observed with fer-
rocenyl sulfides. For example, compound 2a (FcSMe)
exhibits a ratio of 62:38 (trans–cis). An effect, other than
steric, is thus operative here. Anticipating an electronic ef-
fect, we examined a sulfide with an aromatic group.2
Interestingly, thioanisole gave a 64:36 ratio (entry 3), very
similar to that of ferrocenyl methyl sulfide 2a. The aroma-
ticity of this structure, and consequently its ability to sta-
bilize charged intermediates by delocalization seems to be
a key point. This stabilization would reduce the difference
in the syn and anti betaine transition state energies and in-
duce less stereochemical discrimination than in the case of
non-aromatic auxiliaries. In summary, for ferrocene sul-
fides, we have almost no kinetic stereoselectivity, and this
is only slightly counterbalanced (in favor of the trans iso-
mer) by a minor equilibration of the syn betaine. This
stands in contrast with Aggarwal’s observations of a
1
9
yl sulfide. They noticed that equilibration was more sig-
4,25
nificant in CH Cl –H O, with 42:58 (trans/cis) ratio from
2
2
2
the syn betaine. In DMSO, equilibration was complete.
We propose a tentative mechanism to explain the ob-
served diastereomeric preferences, using a cisoid ap-
proach (Scheme 6). Indeed, recent calculations have led to
the proposal that,2 in the case of dimethyl sulfide and
benzaldehyde, the first intermediates were formed
through a ‘cisoid’ approach, imposed by coulombic inter-
actions between the sulfonium and the oxy group. Subse-
quent torsional rotation around the C–C bond gave the
rotamer betaine where the charged groups were anti. Fi-
nally, trans elimination led to the oxiranes.
3
The higher trans selectivity noticed with ferrocenyl sulfur largely predominant trans selectivity and significant ther-
ylides of increasing steric hindrance may be due to steric modynamic contribution.
1
interactions between R and Ph in the syn betaine. Conse-
Having shown that ferrocenyl sulfides efficiently mediate
quently, the torsional rotation in this betaine was disfa-
the epoxidation reaction, we wished to demonstrate that
vored, leaving starting material (ylide and aldehyde). As a
asymmetric induction was feasible, using an enantiopure
sulfide bearing only planar chirality. We have tested the
result, more trans oxirane was formed.
Now, when comparing to standard sulfides, the stereose- behavior of ferrocene 9, bearing a sulfanyl group with a
lectivities are puzzling. This led us to some complementa- sterically hindered tert-butyl, and an ortho-substituent, an
ry experiments with model sulfides, in our standard amino group, whose nucleophilicity was reduced by an
epoxidation conditions (Table 3).
electro-withdrawing substituent. The sulfoxide, corre-
sponding to 9, was an efficient catalyst for the asymmetric
addition of diethylzinc to aldehydes.9 Under our stan-
dard conditions, a reaction has been performed with 0.2
equiv of sulfide 9 to lead to stilbene oxide in reasonable
yield (Scheme 7). Again, an unexpected diastereomeric
ratio (trans–cis, 56:44) was observed. The trans isomer
was produced in a promising 67% enantiomeric excess, in
favor of the S,S enantiomer.
,26
Table 3 Epoxidation Reaction of Benzaldehyde with Standard Sul-
fides and Benzylbromide
Entry
Sulfide
Time (d)
Stilbene Oxide dr
Yield (%)
(trans–cis)
1
2
3
4
thiolane
MeSMe
PhSMe
1
95
86
83
98
85:15
83:17
64:36
94:6
0.3
4
Our strategy is thus feasible. This is the first example of
an asymmetric conversion of aldehyde into oxirane medi-
ated by sulfur ylides with planar chirality. Current studies
are underway to explore the scope of this new type of
MeSt-Bu
2.15
Scheme 6
Synthesis 2003, No. 14, 2249–2254 © Thieme Stuttgart · New York

SPECIAL TOPIC
Epoxidation Reaction of Carbonyl Compounds via Ferrocenyl Sulfur Ylides
2253
Phenyl Sulfanyl Ferrocene (2b)27
Prepared from 1 (1.030 g; 5.54 mmol) in 54% yield as detailed for
2
0
a. Flash chromatography (petroleum ether–EtOAc, 100:0 to
:100) gave 2b [yield: 887 mg (3.01 mmol); orange crystalline
2
8
solid] and 3b [yield: 900 mg (2.24 mmol, 40%); dark orange
crystalline solid]. Compound 3 (31 mg, 0.16 mmol, 3%) was recov-
ered.
t-Butyl Sulfanyl Ferrocene (2c)29
Prepared from 1 (501 mg, 2.69 mmol) in 44% yield as detailed for
Scheme 7
2
a. Flash chromatography (petroleum ether–EtOAc, 100:0 to
50:2) gave 2c [yield: 326 mg (1.19 mmol); orange powder]. Com-
2
pound 3c was not isolated. Compound 1 (38 mg, 0.20 mmol, 8%)
was recovered.
chiral catalysts and its design towards a higher stereo-
selectivity.
Cyclohexyl Sulfanyl Ferrocene (2d)
Prepared from 1 (1.000 g, 5.37 mmol) in 67% yield as detailed for
In summary, we have achieved the first epoxidation reac-
tion of carbonyl compounds via ferrocenyl sulfur ylides.
It involved formation of ferrocenyl sulfonium salts, ob-
2
a. Flash chromatography (petroleum ether–EtOAc, 100:0 to 99:1)
gave 2d [yield: 1.075 g (3.58 mmol); orange powder] and 3d [yield:
97 mg (1.20 mmol, 22%); pale orange powder). Compound 1 (23
1
served for the first time and monitored by H NMR.
4
Deprotonation and reaction with aldehydes led to oxiranes mg, 0.12 mmol, 2%) was recovered.
in good yields. The stereochemical course kinetically fa-
Cyclohexyl Sulfanyl Ferrocene (2d)
Mp 80 °C.
vors more cis isomer than usual with other sulfides, prob-
ably related to an earlier transition state. Steric hindrance
of the sulfide provided enhanced percentage of the trans
isomer, with some reversibility of the cis precursor forma-
IR (KBr): 2924, 2846, 1646, 1444, 1406, 1330, 1258, 1204, 1158,
–
1
1
102, 996, 886, 830, 806 cm .
tion. An enantioselective version has been successfully 1H NMR (250.13 MHz, CDCl
): = 4.28 (m, 2 H, CpH), 4.20 (m,
H, CpH), 2.55 (m, 1 H, Cy), 1.86 (m, 2 H, Cy), 1.70 (m, 2 H, Cy),
.22 (m, 4 H, Cy), 0.86 (m, 2 H, Cy).
3
7
1
achieved for the first time with planar chirality, and others
structures are under investigation.
13C NMR (62.9 MHz, CDCl3): = 75.1 (Cp), 69.4 (Cp), 69.3 (Cp),
THF was freshly distilled from sodium-benzophenone before use.
Hexane was distilled from P O . All non-aqueous reactions were
4
8.3 (Cy), 33.5 (Cy), 26.1 (Cy), 25.7 (Cy).
2
5
+
+
carried out in oven-dried septum capped flasks, and under
MS (EI): m/z (%) = 300 (100, M ·), 218 (81, [M – Cy] ).
atmospheric pressure of N . Commercial reagents were used
2
Anal. Calcd for C H FeS: C, 64.01; H, 6.71; S, 10.68. Found: C,
1
6
20
directly as received. All liquid reagents were transferred via oven–
dried syringes. Lithium bases were purchased from Aldrich and
concentractions were checked before experiments by titration with
diphenylacetic acid. All reactions were monitored by TLC carried
out on analytical silica gel TLC plates purchased from Merck silica
gel and visualised with UV light and phosphomolybdic acid (1g in
6
3.71; H, 6.57; S, 10.91.
1
,1 -Bis(Cyclohexyl Sulfanyl)Ferrocene (3d)
Mp 89 °C.
IR (KBr): 2924, 2850, 1652, 1446, 1384, 1340, 1258, 1206, 1166,
–
1
1
022, 996, 884, 830, 818, 668 cm .
1H NMR (250.13 MHz, CDCl3): = 4.30–4.22 (m, 8 H, CpH), 2.55
m, 2 H, Cy), 1.85 (m, 4 H, Cy), 1.69 (m, 4 H, Cy), 1.55 (m, 2 H),
1
00 mL i–PrOH). Preparative flash liquid chromatography was per-
formed with Merck 60 silica gel (40–63 microns) in the eluting
solvents indicating below. Petroleum ether refers to the fraction
(
1
with bp 35–60 °C. H NMR spectra were recorded on Bruker DPX
1.21 (m, 10 H, Cy).
13C NMR (62.9 MHz, CDCl3): = 78.2 (Cp), 76.2 (Cp), 71.0 (Cp),
2
50 spectrometer. 13C NMR spectra were determined with the same
1
spectrometer operating with broad band H decoupling. IR absorp-
tion spectra were run on Perkin–Elmer 684 and 16 PC FT–IR. Mass
spectra were obtained with a Varian 3800 (GC) and Saturn 200
4
8.3 (Cy), 33.5 (Cy), 26.1 (Cy), 25.7 (Cy).
+
+
MS (EI): m/z (%) = 414 (100, M ·), 332 (28, [M – Cy] ), 250 (27,
[M – Cy – Cy] ).
+
(MS) spectrometers. Elemental analyses were performed on
Thermoquest apparatus. Melting points were measured on an
Electrothermal 9200 apparatus and are uncorrected.
Benzyl Sulfanyl Ferrocene (2e)
Prepared from 1 (1.004 g, 5.40 mmol) in 61% yield as detailed for
2a. Flash chromatography (petroleum ether–EtOAc, 100:0 to 98:2)
gave 2e [yield: 1.020 g (3.31 mmol); orange crystalline solid] and
Methyl Sulfanyl Ferrocene (2a);27 General Procedure
A mixture of ferrocene 1 (1.000 g, 5.37 mmol) in an anhyd mixture
of THF–hexane (1:1, 2.50 mL) was stirred at r.t. for 30 min and then
cooled to 0 °C. t-BuLi (10.00 mL, 10.70 mmol, 1.07 M in pentane)
was then added at a rate of approximately 1 mmol/min. After 30
min, methyl disulfide (1.00 mL, 11.25 mmol) was added, the mix-
ture was allowed to warm to r.t. and stirred until the reaction was
2
8
3
e [yield: 534 mg (1.24 mmol, 23%); pale orange powder]. Com-
pound 1 (54 mg, 0.29 mmol, 5%) was recovered.
Benzyl Sulfanyl Ferrocene (2e)
Mp 51 °C.
complete as judged by TLC (petroleum ether–Et O, 95:5).
Hydrolysis was performed with aq NaOH (2 M; 10 mL), the aq lay-
IR (KBr): 2922, 1598, 1490, 1450, 1408, 1232, 1198, 1162, 1104,
1066, 1000, 888, 826, 804, 766, 696 cm .
1H NMR (250.13 MHz, CDCl3): = 7.25–7.11 (m, 5 H, Ph), 4.16
2
–
1
er was extracted with Et O (3 × 10 mL). The organic layer was
2
washed with brine (25 mL) and H O (25 mL), dried (MgSO ), fil-
2
4
(
m, 9 H, Cp-H), 3.75 (s, 2 H, CH2).
13C NMR (62.9 MHz, CDCl
): = 138.7 (Ph), 128.8 (Ph), 128.2
(Ph), 126.7 (Ph), 79.1 (Cp), 73.9 (Cp), 69.3 (Cp), 69.1 (Cp), 42.7
(CH ).
tered then concentrated. Column chromatography (petroleum
ether–Et O, 100:0 to 99:1) gave sulfide 2a [yield: 839 mg (3.61
3
2
2
8
mmol, 67%); orange oil] and disulfide 3a [yield: 402 mg (1.44
mmol, 27%); dark orange oil].
2
Synthesis 2003, No. 14, 2249–2254 © Thieme Stuttgart · New York

2
254
S. Minière et al.
SPECIAL TOPIC
+
+
MS (EI): m/z (%) = 308 (100, M ·), 217 (67, [M – Bn] ), 91 (12,
(3) Richards, C. J.; Locke, A. J. Tetrahedron: Asymmetry 1998,
9, 2377.
+
[
PhCH ] ).
2
(
(
4) Togni, A.; Haüsel, R. Synlett 1990, 633.
5) Barbaro, P.; Bianchini, C.; Togni, A. Organometallics 1997,
Anal. Calcd for C H FeS: C, 66.25; H, 5.23; S, 10.40. Found: C,
17
16
6
6.12; H, 5.21; S, 10.73.
16, 3004.
(
6) You, S.-L.; Zhou, Y.-G.; Hou, X.-L.; Dai, L. X. Chem.
Commun. 1998, 2765.
7) Muniz, K.; Bolm, C. Chem.–Eur. J. 2000, 6, 2309.
8) Karlström, A. S. E.; Huerta, F. F.; Meuzelaar, G. J.;
Bäckvall, J.-E. Synlett 2001, 923.
Epoxidation Reaction; Typical Procedure
To a solution of 2 (0.50 mmol) in t-BuOH–H O (9:1, 830 L) was
added benzyl bromide (120 L, 1.00 mmol), benzaldehyde (50 L,
2
(
(
0
.50 mmol), NaOH (45 mg, 1.12 mmol) (and iodide salt). The reac-
tion mixture was stirred at r.t. and monitored by TLC (petroleum
ether–Et O, 98:2). H O (2 mL) was then added and the aq phase was
(
9) Priego, J.; Mancheno, O. G.; Cabrera, S.; Carretero, J. C. J.
Org. Chem. 2002, 67, 1346.
2
2
extracted with Et O (3 × 3 mL). The combined organic phases were
2
(
10) McEwen, W. E.; Sullivan, C. E.; Day, R. O.
dried (MgSO ) and concentrated to dryness.
4
Organometallics 1983, 2, 420.
To a cooled solution of the crude oxiranes in CH Cl (2 mL) was
slowly added MCPBA (135 mg, 0.55 mmol) in CH Cl (3 mL). The
mixture was stirred at 0 °C for 45 min and then allowed to warm to
r.t. The reaction was monitored by TLC (petroleum ether–EtOAc,
2
2
(11) Watts, W. E. In Comprehensive Organometallic Chemistry,
Vol. 8; Wilkinson, G.; Stone, F. G. A.; Abel, E. W., Eds.;
Pergamon: Oxford, 1982, 1013–1071.
12) Rebière, F.; Samuel, O.; Kagan, H. B. Tetrahedron Lett.
1990, 31, 3121.
2
2
(
9
5:5). The mixture was washed with sat. aq NaHCO (2 × 4 mL) and
3
brine (4 mL). Column chromatography (petroleum ether–EtOAc,
(
13) Tsukuzaki, M.; Tinkl, M.; Roglans, A.; Chapell, B. J.;
Taylor, N. J.; Snieckus, V. J. Am. Chem. Soc. 1996, 118,
2
4
9
9:1 to 95:5) afforded stilbene oxiranes 4.
6
85.
Yield: 99 mg (0.50 mmol, 99%); white oily solid.
(
14) Metallinos, C.; Snieckus, V. Org. Lett. 2002, 4, 1935.
(15) Guillaneux, D.; Kagan, H. B. J. Org. Chem. 1995, 60, 2502.
16) Julienne, K.; Metzner, P.; Henryon, V. J. Chem. Soc., Perkin
Trans. 1 1999, 731.
17) Zanardi, J.; Leriverend, C.; Aubert, D.; Julienne, K.;
Metzner, P. J. Org. Chem. 2001, 66, 5620.
18) Julienne, K.; Metzner, P.; Henryon, V.; Greiner, A. J. Org.
Chem. 1998, 63, 4532.
19) Aggarwal, V. K.; Calamai, S.; Ford, G. J. J. Chem. Soc.,
Perkin Trans. 1 1997, 593.
-Cyclohexyl-3-phenyloxirane 524
2
(
(
(
(
(
Prepared from cyclohexanecarboxaldehyde (61 L, 0.50 mmol) in
6% yield as detailed for 4. Flash chromatography (petroleum
8
ether–EtOAc, 95:5) gave 5.
Yield: 87 mg (0.43 mmol); colourless oil.
2
-Methyl-2,3-diphenyloxirane 624
Prepared from acetophenone (60 L, 0.50 mmol) in 56% yield as
detailed for 4. Flash chromatography (petroleum ether–EtOAc,
20) Townsend, J. M.; Sharpless, K. B. Tetrahedron Lett. 1972,
9
5:5) gave 6.
3313.
Yield: 58 mg (0.28 mmol); pale yellow oil.
(21) Bernardi, L.; Bonini, B. F.; Comes-Franchini, M.; Fochi, M.;
Mazzanti, G.; Ricci, A.; Varchi, G. Eur. J. Org. Chem. 2002,
2
776.
Acknowledgments
(
(
(
(
(
(
(
(
22) Solladié-Cavallo, A.; Diep-Vohuule, A.; Sunjic, V.;
Vinkovic, V. Tetrahedron: Asymmetry 1996, 7, 1783.
23) Aggarwal, V. K.; Harvey, J. N.; Richardson, J. J. Am. Chem.
Soc. 2002, 124, 5747.
24) Aggarwal, V. K.; Abdel-Rahman, H.; Fan, L.; Jones, R. V.
H.; Standen, M. C. H. Chem.–Eur. J. 1996, 2, 1024.
25) Johnson, A. W.; Hruby, V. J.; Williams, J. L. J. Am. Chem.
Soc. 1964, 86, 918.
26) Priego, J.; Mancheno, O. G.; Cabrera, S.; Carretero, J. C.
Chem. Commun. 2001, 2026.
27) Diter, P.; Tausien, S.; Samuel, O.; Kagan, H. B. J. Org.
Chem. 1994, 59, 370.
We gratefully acknowledge the ‘PunchOrga’ Network (Pôle Uni-
versitaire Normand de Chimie Organique), the ‘Ministère de la Re-
cherche et des Nouvelles Technologies,’ CNRS (Centre National de
la Recherche Scientifique), the ‘Région Basse-Normandie’ and the
European Union (FEDER funding) for financial support. We also
thank warmly Dr Jean-François Brière (Caen), Prof. Serge Piettre,
and Dr Isabelle Châtaignier (Rouen) for fruitful discussions and col-
laboration.
References
28) McCulloch, B.; Ward, D. L.; Woollins, J. D.; Brubaker, C.
H. Organometallics 1985, 4, 1425.
29) Rebière, F.; Riant, O.; Ricard, L.; Kagan, H. B. Angew.
Chem., Int. Ed. Engl. 1993, 32, 568.
(
(
1) Togni, A.; Hayashi, T. Ferrocenes; VCH: Weinheim, 1995.
2) Togni, A. Chimia 1996, 50, 86.
Synthesis 2003, No. 14, 2249–2254 © Thieme Stuttgart · New York