Sustainable Epoxidation of Electron-Poor Olefins with Hydrogen Peroxide in Ionic Liquids and Recovery of the Products with Supercritical CO2

Article abstract of DOI:10.1002/ejoc.200300442

An efficient procedure is described for the epoxidation of electron-deficient olefins, in particular Vitamin K₃ and analogues, with aqueous basic solutions of hydrogen peroxide in different ambient temperature ionic liquids (ILs) [bmim⁺][X^-] {[bmim ⁺] = 1-butyl-3-methylimidazolium; [X^-] = [BF ₄^-], [CF₃SO₃^-], [PF ₆^-], [N(CF₃SO₂)₂ ^-]}. Various factors affecting epoxide yield (in the range 80-99%), reaction rate and reproducibility have been examined. The almost quantitative extraction of the epoxides from the reaction media has been obtained with supercritical CO₂. The ionic liquid was then recovered and reused in subsequent cycles. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200300442

FULL PAPER
Sustainable Epoxidation of Electron-Poor Olefins with Hydrogen Peroxide in
Ionic Liquids and Recovery of the Products with Supercritical CO2
Olga Bortolini,*[a] Sandro Campestrini,[b] Valeria Conte,*[c] Giancarlo Fantin,[a]
Marco Fogagnolo,[a] and Silvia Maietti[a]
Keywords: Epoxidation / Green chemistry / Ionic liquids / Peroxides / Supercritical fluids
An efficient procedure is described for the epoxidation of
electron-deficient olefins, in particular Vitamin K3 and ana-
logues, with aqueous basic solutions of hydrogen peroxide in
different ambient temperature ionic liquids (ILs) [bmim+][X−]
ducibility have been examined. The almost quantitative ex-
traction of the epoxides from the reaction media has been
obtained with supercritical CO2. The ionic liquid was then
recovered and reused in subsequent cycles.
{[bmim+]
= =
1-butyl-3-methylimidazolium; [X−] [BF4−],
[CF3SO3−], [PF6−], [N(CF3SO2)2−]}. Various factors affecting
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
epoxide yield (in the range 80−99%), reaction rate and repro- Germany, 2003)
Introduction
importance both in the industrial and academic communi-
ties.[15] The two main reasons are: from the one side, H2O2
is nowadays employed in the chemical industry in a safer,
more efficient and innovative mode, and from the other hy-
drogen peroxide is a clean oxidant with limited waste prob-
lems as its reduction product is water.[15]
Room temperature ionic liquids (ILs) are salts, character-
ised by a low melting point and a wide liquidus range, that
possess distinctive physical properties such as non-volatility,
non-flammability and thermal stability. In addition, the
presence of poorly coordinating ions and the ability to dis-
solve a wide range of organic and/or inorganic derivatives
are salient features of these materials.[1Ϫ3] Because of their
peculiar properties ILs have been proposed as solvents in
environmentally benign processes. Selected examples in-
clude hydrogenation,[4] hydroformylation,[5] dimerisation
and oligomerisation of alkenes,[6] Heck CϪC coupling,[7]
FriedelϪCrafts reactions,[8] DielsϪAlder cycloadditions,[9]
selective and asymmetric epoxidations,[10,11] oxidations,[12]
halogenations of double and triple bonds,[13] and biocata-
lytic processes.[14] In most of these reactions ionic liquids
appear to be a good alternative to conventional organic sol-
vents, providing comparable or higher rates and, in some
cases, improved enantioselectivities.
The use of hydrogen peroxide in ionic liquids for the oxi-
dation of organic derivatives is still at an early stage. The
few reported examples include the oxidation of several un-
functionalised alkenes with a ureaϪhydrogen peroxide ad-
duct or hydrogen peroxide in the presence of methyl-
trioxorhenium,[11,16] the epoxidation of simple olefins by the
H2O2/FeIII porphyrin system under biphasic conditions,[17]
the selective oxidation of styrene to acetophenone with
H2O2 mediated by palladium chloride,[12] and the epoxid-
ation of cyclohexene by peroxyoctanoic acid generated in
situ by the immobilized Candida antarctica catalysed reac-
tion of octanoic acid with 60% aqueous H2O2.[18]
Carbon dioxide in the supercritical state is used to re-
place hazardous halogenated and aromatic solvents for ex-
tractions, separations and as a reaction medium in indus-
try.[19] The characteristics which make carbon dioxide an
attractive substance from an environmental point of view
are its chemical inertness, non-flammability and non-tox-
icity. In spite of the fact that carbon dioxide is a ‘‘green-
house gas’’ its uses can be considered environmentally be-
nign if carbon dioxide is recycled so that there is no net
increase in the atmosphere.
Hydrogen peroxide is a clean, cheap and potentially
powerful oxygen-transfer reagent that has recently regained
[a]
`
Dipartimento di Chimica, Universita di Ferrara,
Via Borsari 46, 44100 Ferrara, Italy
Fax: (internat.) ϩ39-0532-240-709
E-mail: brl@unife.it
[b]
[c]
`
Dipartimento di Chimica Organica, Universita di Padova,
Via Marzolo 1, 35131 Padova, Italy
Fax: (internat.) ϩ39-049-827-5239
E-mail: sandro.campestrini@unipd.it
We recently reported an efficient methodology for the
epoxidation of α ,β -unsaturated ketones using basic aque-
ous solutions of hydrogen peroxide, a transformation
known as the WeitzϪScheffer reaction, in ionic liquids.[10]
`
Dipartimento di Scienze e Tecnologie Chimiche, Universita di
Roma Tor Vergata,
Via della ricerca Scientifica, 00133 Roma, Italy
Fax: (internat.) ϩ39-06-7259-4328
E-mail: valeria.conte@uniroma2.it
4804
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300442
Eur. J. Org. Chem. 2003, 4804Ϫ4809

Epoxidation of Electron-Poor Olefins
FULL PAPER
In this paper we examine some of the factors that control strate in various batches of [bmimϩ][PF6Ϫ], including com-
the reactivity by: i) expanding the number of ionic liquids mercially available samples. As an example, the quantitative
employed, ii) optimising the oxidation conditions in terms epoxidation of 2-cyclohexen-1-one in the IL used in the
of the relative substrate-base-H2O2 ratio, and iii) finding the communication[10] is obtained in a few minutes, whereas
most favourable conditions for epoxide extraction using three and five hours are required to complete the oxidation
supercritical CO2 followed by the recovery and reuse of the using the IL obtained from two distinct preparations. Fur-
ionic liquid.[20] This investigation has been extended apply- thermore, no reaction was observed using an as-received
ing the oxidation protocol to other substrates, especially commercial sample of [bmimϩ][PF6Ϫ]. The absence of reac-
models closely related to the Vitamin K class that are tivity could be attributed, in this case, to a high acid content
prothrombogenic and antihemorrhagic derivatives.[21] The in the IL. In fact, quantitative olefin conversion was com-
vitamin K epoxides are also very important because of their pletely restored upon treatment to neutral pH, with a final
higher light stability as compared to the non-oxidised epoxide yield of 98%, although in a few hours. The presence
derivatives.
of water in the medium is an additional problem that has
been encountered with [bmimϩ][PF6Ϫ]. Despite the re-
ported low miscibility of water with this ionic liquid, we
have observed a significant increase of epoxidation rates
and reproducibility after addition of small quantities of
H2O to the solvent. Examples of this effect, using 3,5-di-
methyl-2-cyclohexen-1-one, are given in Table 1.
Results and Discussion
Epoxidation reactions were carried out at room tem-
perature
utilising
1-butyl-3-methylimidazolium·[XϪ]
Ϫ
([bmimϩ][XϪ])
as
solvents.
[bmimϩ][BF4
]
and
Ϫ
[bmimϩ][CF3SO3
]
are hydrophilic ILs, whereas
[bmimϩ][PF6Ϫ] and [bmimϩ][N(CF3SO2)2Ϫ] are only par-
tially miscible in water. These ILs were chosen on the basis
of their stability and availability.
Table 1. Effect of the presence of water on the epoxidation rates of
3,5-dimethyl-2-cyclohexen-1-one,
with
H2O2/OHϪ
in
[bmimϩ][PF6Ϫ] as IL
A typical reaction, exemplified in Scheme 1, is carried out
at 25 °C by dissolving 1 mmol of the substrate in 1 mL of
the IL. To this mixture, kept under vigorous stirring,
4 mmol of water, 2 mmol of solid NaOH, and 3 mmol of
H2O2 (0.3 mL of commercially available aqueous solution
30% w/w) were subsequently added. The extent of the reac-
tion was monitored by GC analysis (see Exp. Sect. for de-
tails). Chromatographic, and in some cases 1H NMR, com-
parison with genuine epoxides, obtained in the classical re-
action carried out with H2O2/HOϪ in water, was used to
identify the products.[22]
Solvent[a]
Time
(min)
Epoxide
yield (%)
IL batch A
300
90
240
90
99
99
98
97
IL batch A ϩ H2O[b]
IL batch B
IL batch B ϩ H2O[b]
[a] Batch A and B refer to two different IL preparations. NB Several
preparations of common ionic liquids like N,NЈ-dialkylimidazolium
salt derivatives are now available, including solvent-free routes.[24]
We therefore prepared two different batches of [bmimϩ][PF6Ϫ] to
[b]
be used in the oxidation reactions. H2O ca. 4 mmol.
It may be inferred that the addition of water causes a
"saturation" of the IL, which buffers the medium and allows
a more reproducible system as far as solubility and solv-
ation of the reagents is concerned. Taking into account
these findings we have observed a considerable reduction of
the discrepancies described above by following the synthetic
protocol described in the Exp. Sect. mainly related to the
preparation of [bmimϩ][PF6Ϫ], by adding small amounts of
water to the reaction medium and by checking, before use,
all new ILs by a standard reaction. All these precautions
have therefore been used to obtain all the reactivity data
Scheme 1
In our previous communication[10] we reported data re- reported in the tables. For uniformity, we also carried out
lated to some substrates and with two ILs
—
all the reactions in the presence of water.
The relevant results related to the epoxidation of alkenes
[bmimϩ][BF4Ϫ] and [bmimϩ][PF6Ϫ]. Subsequently, a dis-
cussion in the literature was opened concerning the repro- 1Ϫ8 as a function of the ionic liquid, in the optimised oxi-
ducibility of product yields and reaction rates using differ- dation conditions indicated above, are reported in Table 2.
ent batches of ILs.[23] In our case, we observed that the oxi-
Inspection of the data of Table 2 shows that the best re-
dation reactions performed in the same batch of ionic liquid sults in terms of product yield and reaction rate with all
give consistent results in terms of both reaction rate and substrates were obtained with [bmimϩ][CF3SO3Ϫ]. A sig-
yield. Similar yields with different — in some cases very nificant example refers to the methyl-substituted 1,4-naph-
different — reaction times are observed for the same sub- thoquinone 8 (Vitamin K3), which is almost quantitatively
Eur. J. Org. Chem. 2003, 4804Ϫ4809
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4805

O. Bortolini, S. Campestrini, V. Conte, G. Fantin, M. Fogagnolo, S. Maietti
FULL PAPER
drogen-bond interactions of the substrates with the solvents
are probably playing a role in determining the observed dis-
crepancies. This point is currently under investigation with
the aid of computational calculations.
Table 2. Base-catalysed oxidation of selected electrophilic alkenes
with H2O2 in ionic liquids as a function of the different [XϪ] coun-
terions
A study dedicated to the optimisation of the oxidation
protocol aimed at obtaining a more appealing procedure
was then performed. To this aim 2-cyclohexen-1-one (1) was
Ϫ
used as model substrate and [bmimϩ][CF3SO3
] or
[bmimϩ][PF6Ϫ] as solvent. The relevant results are shown
in Table 3.
Table 3. Base-catalysed oxidation of 2-cyclohexen-1-one (1) with
H2O2 in ionic liquids
Entry
H2O2
(equiv.)
OHϪ
(equiv.)
Time
(min)
Epoxide
yield (%)
Ϫ
[bmimϩ][CF3SO3
]
1
2
3
4
3
3
3
1.5
2
0.5
0
5
5
1200
5
99
99
traces
85
0.5
Ϫ
[bmimϩ][PF6
]
5
6
7
8
3
3
3
1.5
2
0.5
0
5
5
1200
5
98
95
80
90
0.5
As the first step of this optimisation we reduced the
amount of base from 2 to 0.5 equivalents with no signifi-
cant reduction of the substrate conversion, which is com-
plete in minutes at room temperature (entries 2 and 6). Un-
der these conditions, halving the amount of hydrogen per-
oxide gives rise to a decrease of the epoxide yield from al-
most quantitative to 85% and 90%, respectively. It is
worthwhile to note that the uncatalysed oxidation proceeds
at a much slower rate in both ionic liquids (entries 3 and
7). With all substrates, however, a reduction of the base
content to 0.5 equivalents causes insignificant differences in
product yields and reaction times provided that three equiv-
alents of H2O2 are used.
[a]
[substrate]:[H2O2]:[OHϪ] ϭ 1:3:2, 4 mmols of water were added
to all the reaction mixtures. [b] These reactions were not performed
because of the observed low stability of [bmimϩ][BF4Ϫ] at longer
reaction times.
As reported by many authors, the classical epoxidation
converted into the corresponding epoxide in minutes de- of α ,β -unsaturated carbonyl compounds using hydrogen
pending on the solvent. The same substrate is oxidized over peroxide in a basic medium suffer some drawbacks, so that
much longer reaction times[25,26] in various experimental several new methods have been proposed in recent
conditions (from two hours up to four days). Our oxidation years.[29Ϫ31] In almost all cases, however, consistent differ-
procedure also appears to be efficient with other ILs, like ences are found between the reaction yield, as obtained by
[bmimϩ][PF6Ϫ], the reaction rates being higher than those GC for example, and the isolated product yield. These dif-
observed
in
the
conventional
alkaline-aqueous ferences are usually attributed to the difficulty of recovering
conditions.[22Ϫ27] As a further comment we would like to the epoxide.
point out that with the classical alkaline-aqueous con-
As the final part of this study we checked the viability of
ditions side reactions that strongly influence the isolated the extraction of the epoxides formed with scCO2, with the
yields of epoxides are often observed.[28] Such a problem aim of obtaining an even "greener" procedure for the epox-
has never been observed in our experimental conditions.
idation of our electrophilic alkenes. To this purpose we car-
Comparison of the results obtained with structurally ried out some epoxidation reactions of 2-cyclohexen-1-one
similar substrates appears to suggest that a significant inter- (1), using the optimised reaction conditions as far as the
action between the substrate and the solvent is occurring. molar ratio of reagents is concerned, save for having a dif-
Such an effect is particularly evident with 6 (81% epoxide ferent total volume of IL ([bmimϩ][PF6Ϫ]) solvent. In par-
yield in 300 min) and 7 (99% epoxide yield in 5 min). Differ- ticular, three different reactions were carried out reducing
ences in the preferred conformation of the substrate or hy- the total amount of IL used from 0.50 mL, down to
4806
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 4804Ϫ4809

Epoxidation of Electron-Poor Olefins
FULL PAPER
0.25 mL and 0.10 mL. The reaction mixtures were left to solubility of ca. 1.1 mg/mL. This outcome is probably re-
react, in a suitable glass vessel, for about one hour in order lated to a change of scCO2 solvent power induced by the
to assure a quantitative substrate conversion. Subsequently, standard (n-decane), which acts as a solvent modifier. Inci-
after the addition of the appropriate amount of a chromato- dentally, these results suggest a possible method to improve
graphic standard completely soluble in the scCO2 phase (n- epoxide solubility in industrial applications exploiting batch
decane), the vessel was inserted into a dedicated steel reac- reactors. Alternatively, efficient epoxide extractions may be
tor with an internal volume of 8.5 mL and pressurised with obtained even in the case of very low solubilities by taking
CO2 until the final conditions of 40 °C and 200 atm. Under advantage of continuous flow reactors, in which a stream
these experimental conditions, CO2 is in the supercritical of scCO2 is employed and recycled after an expansion step,
state and shows a density of 0.84 g/mL — very close to that which allows the separation of the epoxide. With these re-
of a saturated hydrocarbon; it therefore exhibits a similar sults at hand it is possible to underline that the coupling of
solvent power. After a contact time of about 19 hours, the the use of IL liquids as solvents, hydrogen peroxide as oxi-
amount of epoxide extracted into the scCO2 was deter- dant and extraction of products with scCO2 is a viable and
mined through the withdrawal and GC analysis of sustainable procedure for the epoxidation of electron-poor
supercritical-phase samples (see Exp. Sect. for details).
The data collected in Table 4 confirm that all the reac-
olefins.
As a further improvement of our procedure, the possibil-
tions are characterised by a very high degree of substrate ity of performing subsequent cycles of reaction/scCO2 ex-
conversion independent of the counteranion present in the traction of products by reusing the IL solvent was explored.
solvent. On the other hand, as the total volume of the reac- To this end, two series of reactions, whose results are col-
tor is constant, the efficiency of the extraction procedure lated in Table 5, were carried out by using the experimental
increases as the volume of the reaction mixture decreases. conditions of run 3 in Table 4.
The data also indicate that the epoxide solubility is scarcely
Table 5. Recycling of the solvent in the epoxidation of 2-cyclo-
hexen-1-one (1) in [bmimϩ][PF6Ϫ] with H2O2 in the presence of
NaOH
influenced by the nature of the IL counterion. Moreover, in
the case of [N(CF3SO2)2Ϫ] the possibility of a reversible
reaction with CO2 that changes the counterion into a carba-
mate-like one should be taken into account.
Cycles Steel reactor
Glass reactor
Conv. (%) Extrac. yield (%) Conv. (%) Extrac. yield (%)
Table 4. Conversion of substrate and efficiency of the extraction
with scCO2 of cyclohexanone-2-epoxide formed by the reaction of
2-cyclohexen-1-one (1) with H2O2 in the presence of NaOH in vari-
ous [bmimϩ][XϪ][a]
1
2
82
78
35
79
90
Ͼ99
97
91
46
99
89
91
78
3
4[a]
Ͼ98
Run React. volume Conversion [XϪ]
Epoxide
extraction [a] IL washed with water before being recycled in order to eliminate
(mL)
(%)
(%)
Na2CO3 formed after extraction of the basic reaction mixture
with scCO2.
Ϫ
Ϫ
Ϫ
1
2
3
4
5
0.50
0.25
0.10
0.10
0.10
97
[PF6
[PF6
[PF6
]
]
]
66
71
96
94
92
97
Ͼ99
Ͼ99
Ͼ99
In the first series the subsequent reactions were carried
out directly in the steel reactor illustrated in the Exp. Sect.
In the second series, the reactions were carried out in a glass
vessel which, after completion of oxidation, was introduced
into the steel vessel for scCO2 extraction; details are re-
ported in the Exp. Sect.
Ϫ
[CF3SO3
[N(CF3SO2)2
]
Ϫ
]
[a]
[1]:[H2O2]:[NaOH] ϭ 1:1.5:0.5.
Once we had verified the complete solubility of n-decane
As far as the utilisation of the steel reactor as reaction
in the supercritical phase under all the experimental con- vessel is concerned, it is worthwhile to note that the sub-
ditions of Table 4, the above reported data allowed us to strate conversions are lower than those observed for reac-
estimate the solubility of cyclohexanone-2-epoxide in scCO2 tions carried out in the glass vessel. This is probably related
(40 °C, 200 atm). In fact, run 1 of Table 4 indicates that to a surface metal catalysed radical decomposition of the
66% of the initial amount of 2-cyclohexen-1-one (50 mg) oxidant as observed previously.[32] This observation thus il-
was extracted as cyclohexanone-2-epoxide in 8.5 mL of a lustrates one of the problems that have to be taken into
supercritical phase volume, leading to a solubility of ca. consideration if this oxidation procedure is to be used. The
3.9 mg/mL. Runs 2 and 3 in Table 4 show that upon further decrease of substrate conversion observed after re-
decreasing the total amount of reaction mixture, and conse- cycling the IL is probably a consequence of NaOH neutral-
quently decreasing the amount of reactants and standard, isation by CO2 during the extraction step, with the forma-
the extraction efficiency increases, passing from 66 to 71 tion of Na2CO3. Another unsatisfactory feature of the reac-
up to 96% but, at the same time, epoxide solubility in the tions carried out in the steel vessel is the non-quantitative
supercritical phase decreases. For instance, run 3 of Table 4 recovery of organic materials by scCO2 extraction. A pos-
shows that 96% of the starting 2-cyclohexen-1-one (10 mg) sible rationale for this is the formation of highly polar prod-
was extracted as the corresponding epoxide, leading to a ucts, in particular diols, which are not extracted into scCO2
Eur. J. Org. Chem. 2003, 4804Ϫ4809
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4807

O. Bortolini, S. Campestrini, V. Conte, G. Fantin, M. Fogagnolo, S. Maietti
FULL PAPER
to the reaction mixture. The identity of the epoxides was deter-
mined on products isolated from reactions carried out on a larger
under the experimental conditions selected. Conversely, in
the reactions carried out in the glass reactor, the decompo-
sition of the peroxide does not take place and therefore in
the first cycle we observe the expected high conversion of
the substrate and high yield (almost quantitative) of epox-
1
scale by comparison of the H NMR spectra with those obtained
for genuine epoxides prepared using the classical H2O2/HOϪ
reaction.[22Ϫ28]
ide extraction. However, in the second cycle, a decrease in General Procedure for Extraction of Epoxides with scCO2: An
amount of n-decane (internal standard) depending on the volume
of reaction mixture employed was added to the mixtures obtained
after reaction either in a glass vessel or directly in the steel vessel
under the experimental conditions of Table 3 and 4, and the prod-
uct was extracted through an overnight exposure to scCO2 by using
the apparatus shown in Figure 1. The system core is a modified
SFC 3000 series CarloϪErba Instrument originally designed for
capillary supercritical fluid chromatography. The capillary column
and splitting valve were replaced by a 316 stainless steel 8.5-mL
high-pressure vessel (10000 psig) connected to a restrictor (1/16"
OD, 0.005" ID) through a six-way HPLC valve (Rheodyne 7125)
for withdrawing reaction samples. The apparatus consists of a SFC
300 CarloϪErba Instruments pump, a chromatographic SFC 3000
CarloϪErba Instruments oven containing the reaction vessel and
an ICU 600 from CarloϪErba Instruments for the temperature
control of the oven and restrictor. The SFC 300 is a 150 mL ca-
pacity pulse-free syringe pump developed for pressures ranging up
to 50 MPa and flow rates from 1 to 7000 µL/min. The pump cylin-
der is thermostatted at 7 °C by an external liquid circulation ther-
mostat thus permitting the syringe to be easily refilled with liquid
carbon dioxide. A sample of the supercritical phase was withdrawn
through the stainless steel restrictor maintained at 80 °C, by means
of a six-way HPLC valve and trapped in dichloromethane. The
extraction yield and conversions were then determined by GLC
analysis with the internal standard method on the basis of pre-
viously measured response factors.
the conversion due to Na2CO3 precipitation during the ex-
traction step is detectable (91% vs. 97%). An even more
marked decrease in conversion (46%) was observed during
the third cycle.
At this stage the IL was washed with water to eliminate
Na2CO3, and dried before being reused in a fourth cycle
where hydrogen peroxide and base were reloaded. The con-
version and extraction yield thus obtained were again very
good.
Conclusion
The oxidation procedure outlined in this paper is charac-
terised by a number of advantages that make it a very inter-
esting environmentally friendly process for the base-cata-
lysed oxidation of electrophilic alkenes. In particular, we
have reported a successful combination of new and less-
threatening solvents {i.e. [bmimϩ][XϪ] with [XϪ] ϭ [PF6Ϫ],
[CF3SO3Ϫ], [N(CF3SO2)2Ϫ]}, H2O2 — one of the safer, least
wasteful and cheaper oxidants — and an efficient, clean and
safe extraction of the products with scCO2. Worthy of note
is the fact that complete consumption of the substrates is
usually observed, in short reaction times, in the presence of
small excesses of the oxidant and a submolar quantity of
the base at room temperature. The efficiency of the extrac-
tion with scCO2 is almost quantitative as well. We have also
showed that recovery and reuse of the ionic liquid is achiev-
able.
The use of ILs in the base-catalysed epoxidation of elec-
tron-poor olefins appears also to be of interest for the en-
antioselective versions of such reaction.[25,26,33]
Experimental Section
Figure 1. Home made apparatus for extraction with scCO2; 1: CO2
reservoir; 2: cryogenic pump (CarloϪErba SFC 300); 3: thermo-
static oven (CarloϪErba SFC 3000); 4: reactor; 5: six-way valve; 6:
alternating field magnetic stirrer; 7: magnetic stirrer control unit;
8: restrictor and oven temperature control unit (CarloϪErba ICU
600); 9: restrictor
Instruments: GC analyses were performed on a FISONS GC 9000
series, using a DETTBS column (Mega). 1H NMR spectra were
recorded on a Varian Gemini 300 MHz. Thermostatted liquid cir-
culation was obtained using a Haake GH refrigerator controlled
by a Haake D8 head. Extraction of epoxides with scCO2 was per-
formed on a modified SFC 3000 series CarloϪErba instruments
originally designed for capillary supercritical fluid chromatography.
Preparation of [bmim؉][PF6Ϫ]: 1-Butyl-3-methylimidazolium chlo-
ride was prepared by reaction of 1-methylimidazole (ca. 10 g) with
chlorobutane (60 mL) under reflux for 24 hours. The disappearance
General Procedure for the Epoxidation of Alkenes: In a typical reac-
tion 1 mmol of the substrate was dissolved in 1 mL of 1-butyl-3-
methylimidazolium hexafluorophosphate [bmimϩ][PF6Ϫ] at 25 °C of 1-methylimidazole was followed by TLC using CHCl3/CH3OH/
in a 5 mL volumetric flask. Water (4 mmol), 2 mmol of solid
NaOH, and 0.3 mL of commercially available aqueous solution of
H2O2 (30% w/w) 3 mmol were then added with vigorous stirring.
At increasing reaction times, samples of the mixture were taken
NH4OH (90:15:1) as eluent. The excess of chlorobutane was re-
moved by evaporation under vacuum. To prepare [bmimϩ][PF6Ϫ],
1-butyl-3-methylimidazolium chloride (5 g) was dissolved in 70 mL
of water, the solution was cooled in an ice-cold bath, and hexa-
out, poured into water (1 mL), extracted with ethyl acetate, dried fluorophosphoric acid (1.3 equivalents) was slowly added over a
over MgSO4 and analysed by GC. No changes were observed when
NaOH was dissolved in aqueous hydrogen peroxide before addition
period of 2 hours. After stirring for 12 hours and warming to room
temperature, the upper acidic aqueous layer was decanted and the
4808
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 4804Ϫ4809

Epoxidation of Electron-Poor Olefins
FULL PAPER
[16]
[17]
[18]
[19]
G. S. Owens, A. Durazo, M. M. Abu-Omar, Chem. Eur. J. 2002,
8, 3053Ϫ3059.
K. A. Srinivas, A. Kumar, S. M. S. Chauhan, Chem. Commun.
2002, 2456Ϫ2457.
R. Madeira Lau, F. van Rantwijk, K. R. Seddon, R. A. Shel-
don, Org. Lett. 2000, 2, 4189Ϫ4191.
Supercritical Fluids, Extraction and Pollution Prevention (Eds.:
M. A. Abrahm, A. K. Sunol), ACS Symposium series, vol. 670,
American Chemical Society, Washington, DC, 1997.
S. V. Dzyuba, R. A. Bartsch, Angew. Chem. 2003, 115,
158Ϫ160, Angew. Chem. Int. Ed. 2003, 42, 148Ϫ150, and refs.
cited therein.
lower ionic-liquid portion was washed with water to neutral pH.
The ionic liquid was then heated under vacuum at 70 °C to remove
any excess water.[34] All new batches of [bmimϩ][PF6Ϫ] were
checked in a standard reaction before use. The synthesis of
Ϫ [36]
[bmimϩ][XϪ] {X ϭ [BF4Ϫ],[35] [CF3SO3
]
Ϫ [36] and [N(CF3SO2)2
]
}
was performed according to literature methods.
Acknowledgments
[20]
[21]
We acknowledge financial support from the Italian Ministry of
University and Scientific and Technological Research to the pro-
ject: ‘‘Activation of Peroxidic Species in Innovative Selective Oxi-
dation Processes’’. We wish to thank Prof. K. R. Seddon for helpful
discussions and for a generous gift of ILs.
P. Dowd, R. Hershline, S. W. Ham, S. Naganathan, Nat. Prod.
Rep. 1994, 11, 251Ϫ264.
Org. Synth. Coll. Vol. 1963, 4, 552Ϫ553.
[22]
[23]
[24]
V. Farmer, T. Welton, Green Chem. 2002, 4, 97Ϫ102.
J. H. Davis Jr., K. J. Forrester, T. Merrigan, Tetrahedron Lett.
1998, 39, 8955Ϫ8958.
S. Colonna, A. Manfredi, R. Annunziata, N. Gaggero, J. Org.
Chem. 1990, 55, 5862Ϫ5866.
S. Arai, H. Tsuge, M. Oku, M. Miura, T. Shioiri, Tetrahedron
2002, 58, 1623Ϫ1630.
For a recent paper, see: K. Yamaguchi, K. Mori, T. Mizugaki,
K. Ebitani, K. Kaneda, J. Org. Chem. 2000, 65, 6897Ϫ6903
and references cited therein.
See the Exp. Sect. of: J. Eling, H. Lempers, R. A. Sheldon, Eur.
J. Org. Chem. 2000, 1905Ϫ1911.
J. M. Fraile, J. I. Garcia, C. Cativiela, J. A. Mayoral, F. Figu-
eras, Tetrahedron Lett. 1996, 37, 5995Ϫ5996.
T. S Straub, Tetrahedron Lett. 1995, 36, 663Ϫ664.
A. Sharifi, M. Bolourtchian, F. Mohsenzadeh, J. Chem. Res.
(S) 1998, 668Ϫ669.
[1]
T. Welton, Chem. Rev. 1999, 99, 2071Ϫ2083.
[2]
P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112,
[25]
[26]
[27]
3926Ϫ3945; Angew. Chem. Int. Ed. 2000, 39, 3772Ϫ3789.
[3]
R. A. Sheldon, Chem. Commun. 2001, 2399Ϫ2407.
[4]
Y. Chauvin, L. Mussmann, H. Oliver, Angew. Chem. 1995, 107,
2941Ϫ2943; Angew. Chem. Int. Ed. Engl. 1995, 34, 2698Ϫ2700.
[5]
E. G. Kuntz, CHEMTECH 1987, 570Ϫ576.
[6]
J. E. L. Dullius, P. A. Z. Suarez, S. Einloft, R. F. de Souza, J.
Dupont, J. Fisher, A. De Cian, Organometallics 1998, 17,
[28]
[29]
815Ϫ819.
[7]
A. J. Carmichael, M. J. Earle, J. D. Holbrey, P. B. McCormac,
K. R. Seddon, Org. Lett. 1999, 1, 997Ϫ1000.
J. A. Boon, J. A. Levisky, J. L. Pflug, J. S. Wilkes, J. Org. Chem.
[8]
[30]
[31]
1986, 51, 480Ϫ483.
J. Howarth, K. Hanlon, D. Fayne, P. McCormac, Tetrahedron
[9]
Lett. 1997, 38, 3097Ϫ3100.
O. Bortolini, V. Conte, C. Chiappe, G. Fantin, M. Fogagnolo,
[32]
[33]
[10]
S. Campestrini, U. Tonellato, Adv. Synth. Catal. 2001, 343,
819Ϫ825.
S. Maietti, Green Chem. 2002, 4, 94Ϫ96.
[11]
M. W. Cappi, W. P. Chen, R. W. Flood, Y. W. Liao, S. M.
Roberts, J. Skidmore, J. A. Smith, N. M. Williamson, Chem.
Commun. 1998, 1159Ϫ1160.
J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser,
R. D. Rogers, Chem. Commun. 1998, 1765Ϫ1766.
P. A. Z. Suarez, J. E. L. Dullius, S. Einloft, R. F. De Souza, J.
Dupont, Polyhedron 1996, 15, 1217Ϫ1219.
P. Bonhote, A.-P. Dias, N. Papageorgiou, K. Kalyanasunda-
ram, M. Gratzel, Inorg. Chem. 1996, 35, 1168Ϫ1178.
Received July 16, 2003
S. Owens, M. M. Abu-Omar, Chem. Commun. 2000,
1165Ϫ1166.
[12]
V. V. Namboodiri, R. S. Varma, E. Sahle-Demessie, U. R. Pil-
[34]
[35]
[36]
lai, Green Chem. 2002, 4, 170Ϫ173.
O. Bortolini, M. Bottai, C. Chiappe, V. Conte, D. Pieraccini,
[13]
Green Chem. 2002, 4, 621Ϫ627.
R. A. Sheldon, R. Madeira Lau, M. J. Sorgedrager, F. van
[14]
Rantwijk, K. R. Seddon, Green Chem. 2002, 4, 147Ϫ151.
C. W. Jones, Application of Hydrogen Peroxide Derivatives, RSC
[15]
London, 1999.
Eur. J. Org. Chem. 2003, 4804Ϫ4809
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4809