Kinetics of MTO-catalyzed oferin epoxidation in ambient temperature ionic liquids: UV/Vis and 2H NMR study

Article abstract of DOI:10.1002/1521-3765(20020703)8:13<3053::AID-CHEM3053>3.0.CO;2-G

The kinetics of oxygen-atom transfer from the peroxo complexes of methyltrioxorhenium (MTO) to alkenes in ionic liquids have been investigated. Noncatalytic conversions of alkenes to epoxide were monitored by UV/Vis at 360 nm, where the monoperoxorhenium

Full text of DOI:10.1002/1521-3765(20020703)8:13<3053::AID-CHEM3053>3.0.CO;2-G

FULL PAPER
Kinetics of MTO-Catalyzed Olefin Epoxidation in Ambient Temperature
2
Ionic Liquids: UV/Vis and H NMR Study**
Gregory S. Owens, Armando Durazo, and Mahdi M. Abu-Omar*[a]
Abstract: The kinetics of oxygen-atom
transfer from the peroxo complexes of
methyltrioxorhenium (MTO) to alkenes
in ionic liquids have been investigated.
Noncatalytic conversions of alkenes to
epoxide were monitored by UV/Vis at
360 nm, where the monoperoxorhenium
(mpRe) and diperoxorhenium (dpRe)
complexes absorb. Water- and peroxide-
free dpRe was prepared in situ by the
reaction of MTO and urea hydrogen
peroxide (UHP) in dry THF. The ob-
served biexponential time profiles in
conjunction with kinetic modeling allow
the assignment of the fast step to the
reaction of olefin with dpRe (k4) and the
slowstep to the analogous reaction with
mpRe (k3). In most of the studied ionic
liquids, k4 ꢀ 5 Â k3. 2H NMR experi-
ments conducted with [D3]dpRe under
non-steady-state conditions confirm the
speciation of the catalytic system in ionic
liquids and assert the validity of the UV/
Vis kinetics. Deuteriated alkenes were
used to study the catalytic epoxidation
and dihydroxylation of alkenes by
2H NMR spectroscopy. The values of k4
for a-methylstyrene in several ionic
liquids exceed what is observed in ace-
tonitrile by an order of magnitude.
While the rate of olefin epoxidation is
unaffected by the nature of the ionic
liquid cation, a discernable kinetic effect
is observed with coordinating anions
such as nitrate.
Keywords: ionic liquids ¥ kinetics ¥
NMR spectroscopy ¥ oxidations ¥
rhenium
polar yet noncoordinating anions (in general), tunable melt-
ing points and solvent miscibilities, and the ability to dissolve
a wide range of inorganic and organic materials are among the
most desirable characterisitics of these room-temperature
liquids.[7]
Introduction
The need for developing environmentally friendly chemical
technologies is more important than ever. At the forefront in
the push for green chemistry are catalytic processes employ-
ing transition metals, benign reagents such as molecular
oxygen and hydrogen peroxide, and nonvolatile, recyclable
reaction media.
Interest in hydrogen peroxide as an oxidant when activated
by transition metal catalysts has expanded rapidly in the past
two decades.[1] One of the most useful organometallic com-
plexes for the activation of hydrogen peroxide is methyltriox-
orhenium (MTO). MTO has been shown in recent years to be
an exceptionally versatile oxygen transfer catalyst when used
with either aqueous hydrogen peroxide or urea hydrogen
UHP was first used by Adam and co-workers as an oxidant
for the MTO catalytic system in chloroform.[3] Recently, the
ionic liquid 1-ethyl-3-methylimidazolium tetrafluoroborate
([emim]BF4) has been shown to be a suitable solvent for the
MTO-catalyzed epoxidation of olefins using UHP.[10] Elec-
tron-rich olefins such as styrenes are particularly easy to
oxidize with this system, giving conversions and yields up to
and exceeding 99%. The use of UHP yields the corresponding
epoxide product, whereas the use of aqueous hydrogen
peroxide (30%) with MTO quantitatively converts the olefin
to the 1,2-diol. These oxidation reactions were found to
proceed on a time scale similar to that reported for the same
reactions in molecular solvents.
To gain further insight into the applicability of ionic liquids
for organic synthesis and industrial oxidations, this group has
undertaken the kinetic study of the MTO/peroxide system in
these relatively unexplored solvents. The interactions of MTO
with aqueous hydrogen peroxide and UHP will be reported
elsewhere.[11] Herein we focus on the kinetics of olefin
oxidation using the MTO catalyst, and examine the effect of
these newmedia on the rates of oxygen-atom transfer from
the catalytically active peroxorhenium complexes.
6]
peroxide (UHP).[2
Ionic liquids have received a great deal of attention as
alternative reaction media. In particular, 1,3-dialkylimidazo-
lium- and 1-alkylpyridinium-based ionic liquids have been the
9]
focus of much investigation.[7 Negligible vapor pressures,
[a] Prof. M. M. Abu-Omar, Dr. G. S. Owens, A. Durazo
Department of Chemistry and Biochemistry
607 Charles E. Young Drive, East University of California
Los Angeles, CA 90095-1569 (USA)
Fax: 310-206-4038
E-mail: mao@chem.ucla.edu
[**] MTO ˆ methyltrioxorhenium.
Both UV/Vis spectrophotometry and NMR spectroscpopy
have been used previously to characterize the MTO/peroxide/
Supporting information for this article is available on the WWW under
http://www.chemeurj.org or from the author.
Chem. Eur. J. 2002, 8, No. 13
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3053 $ 20.00+.50/0
3053

M. M. Abu-Omar et al.
FULL PAPER
olefin system and its kinetics in aqueous and organic media
(Scheme 1).[12 14] Unfortunately, the UV-cutoff for 1,3-dialky-
limidazolium- and 1-alkylpyridinium-based ionic liquids is
approximately 300 nm, which precludes the direct observation
(RTILs) have been documented. Hence, a water-free source
of hydrogen peroxide had to be employed to minimize the
deleterious effects of water as a co-solvent. Second, the UV
cutoff of 300 nm imposed by the ionic media necessitated the
development of a newand reliable method to monitor these
reactions under single turnover conditions. Many solvents
were screened for the preparation of essentially water- and
peroxide-free dpRe. The best results were obtained with
anhydrous THF and UHP, since in this solvent dpRe is stable
for prolonged periods of time and UHP is nearly insoluble.
Hence, filtration of excess UHP affords a concentrated
solution (ca. 40 mm) of dpRe in THF; such solutions are well
behaved and display negligible degradation of dpRe as
observed at 360 nm (lmax for dpRe) even after two days.
For the non-steady-state kinetics with alkenes, the THF
solution of dpRe was diluted into the desired RTIL to about
1 mm such that the final concentration of THF was less than
3% of the ionic solvent. Upon addition of excess alkene (40
150 mm), the reaction progress was followed at 360 nm. The
non-steady-state time profiles are biexponential (a typical
time profile for the epoxidation of a-methylstyrene is shown
in the Supporting Information in Figure S1). The absorbance
at the end of reaction is essentially zero. This indicates that all
of the dpRe and mpRe have reacted with the substrate. The
kinetic traces are fitted using Equation (1), in which both a
and b are constants.
Scheme 1. Reaction of MTO with hydrogen peroxide to form two
peroxorhenium complexes.
of substrate or product spectroscopic features in the UV for
monitoring reaction kinetics. However, it is possible to
monitor the non-steady-state kinetics of these reactions by
observing the spectral changes in the visible region due the
diperoxorhenium complex, dpRe. In this work, the kinetics of
substrate oxidation in water-miscible ionic liquids as well as in
acetonitrile were studied by monitoring the decrease in
absorbance at 360 nm, which is attributable to the reactions
of both dpRe and the monoperoxorhenium complex, mpRe,
with the olefinic substrates. Non-steady-state kinetics was
used to elucidate the rate constants k3 and k4.
ABst À Abs1 ˆ aexp(Àkft) ‡ bexp(Àkst)
(1)
Additionally, while reports of perdeuteriated ionic liquids
have appeared in the literature,[15] these liquids are difficult
and expensive to make. Recently, however, techniques have
been developed in this laboratory to measure kinetics in ionic
liquids using 2H NMR spectroscopy.[16] These techniques
include using deuteriated substrates such as [D8]styrene under
steady-state conditions as well as the peroxo complexes of
[D3]MTO for non-steady-state kinetics. The data collected
from the UV/Vis and NMR studies combined fully describe
the kinetics of olefin oxidation in ionic liquids.
Both kf and ks (the fast and slowpseudo-first-order rate
constants, respectively) exhibit linear dependences on the
alkene concentration. Representative plots of kf and ks versus
[alkene] for the oxidation of trans-b-methylstyrene (TBMS)
in [emim]BF4 (emim ˆ N,N'-ethylmethylimidazolium) are
shown in Figure S2 and S3 in the Supporting Information.
The intercepts of these plots represent kÀ1 and kÀ2
;
however, these measurements are not particularly reliable
since their values are negligible in comparison to the slopes.
The second-order rate constants for the reaction of mpRe (k3)
and for the reaction of dpRe (k4) with the substrate were
obtained from the slopes of the above plots. However, as the
data is presented, the fast and slowsteps cannot be assigned.
The kinetics simulation program KINSIM[18] was used to
determine that the k4 step, the oxidation of alkene with dpRe,
is faster than the k3 step. The details of these simulations are
provided in the Discussion.
Results
Non-steady-state UV/Vis kinetics: The goal of this study is to
determine the rates of oxygen transfer from the catalytically
active peroxorhenium complexes (k3 and k4 in Scheme 1) to
olefinic substrates in ionic liquids, and contrast these values to
those obtained in water and organic solvents. It had become
evident early on that two complications needed to be
addressed here. First, the effects of water on the kinetics[11]
and physical properties[17] of room temperature ionic liquids
The values of k3 and k4 for several olefins have been
determined previously in 1:1 CH3CN:H2O.[13] These constants
are compared with new values for k3 and k4 in acetonitrile and
in [emim]BF4 in Table 1. a-Methylstyrene was chosen as the
Table 1. Values of k3 and k4 [LmolÀ1 sÀ1] in different solvents for several alkenes.
Solvent
Styrene
trans-b-Methylstyrene
a-Methylstyrene
Cyclohexene
k3 Â 102
k4 Â 102
k3 Â 102
k4 Â 102
k3 Â 102
k4 Â 102
k3 Â 102
k4 Â 102
[emim]BF4[a] 1.3 Æ 0.1
CH3CN[a]
CH3CN/H2O[b]
13 Æ 1
5.8 Æ 0.4
2.1 Æ 0.2
51 Æ 7
26 Æ 1
10 Æ 1
44 Æ 1
3.1 Æ 0.2
3.1 Æ 0.2
13 Æ 1
22 Æ 1
106 Æ 2
0.20 Æ 0.04
1.5 Æ 0.1
11 Æ 1
5.2 Æ 0.2
22 Æ 2
4.5 Æ 0.1
4.5 Æ 0.1
47 Æ 2
[a] Conditions: 0.91 mm dpRe, 38 154 mm substrate, 238C. [b] 1:1 CH3CN/H2O (v/v) ref. [13].
3054
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 0947-6539/02/0813-3054 $ 20.00+.50/0
Chem. Eur. J. 2002, 8, No. 13

Kinetics of MTO-Catalyzed Olefin Epoxidation
3053 3059
Table 3. Values of k4 for the epoxidation and dihydroxylation of several
alkenes as determined by H NMR spectroscopy.
substrate for comparative kinetics in different ionic liquids.
The rate constants k3 and k4 for the oxidation of a-
methylstyrene are shown in Table 2.
2
Entry
Alkene
Solvent[a]
Method[b]
k4 [LmolÀ1 sÀ1
]
1
2
3
4
5
6
7
8
9
[D8]styrene
[D8]styrene
[D8]styrene
styrene
2-fluorostyrene
2,6-difluorostyrene
[D10]cyclohexene
[D10]cyclohexene
[D10]cyclohexene
[emim]BF4
[bupy]BF4
[etpy]BF4
[emim]BF4
[emim]BF4
[emim]BF4
[emim]BF4
[bupy]BF4
[emim]BF4
A
A
B
C
C
C
A
A
B
0.034 Æ 0.004
0.040 Æ 0.008
0.20 Æ 0.02
Table 2. Rate constants for the oxidation of a-methylstyrene in various
ionic liquids.[a]
0.059 Æ 0.006
0.027 Æ 0.002
0.012 Æ 0.001
0.20 Æ 0.01
Entry
Solvent
k3 [LmolÀ1 sÀ1
]
k4 [LmolÀ1 sÀ1
]
1
CH3CN
0.045 Æ 0.001
0.045 Æ 0.001
0.47 Æ 0.02
0.44 Æ 0.01
0.35 Æ 0.02
0.12 Æ 0.02
0.58 Æ 0.01
2[b]
3
1:1 CH3CN/H2O
[emim]BF4
[bmim]BF4
[bmim]NO3
[bupy]BF4
0.23 Æ 0.02
0.10 Æ 0.01
1.75 Æ 0.06
4
5
6
0.074 Æ 0.002
0.016 Æ 0.004
0.13 Æ 0.01
[a] etpy ˆ N-ethylpyridinium. [b] Method
A (steady-state epoxidation
conditions): [UHP] ˆ 1.0 m, [MTO] ˆ 0.008 m, [deuteroalkene] ˆ 0.10 m,
Tˆ 298 K. Method B (steady-state dihydroxylation conditions): [H2O2]
(30% aqueous) ˆ 0.50 m, [MTO] ˆ 8.0  10À4 m, [deuteroalkene] ˆ 0.10 m,
[a] Conditions: 0.91 mm dpRe, 38 154 mm substrate, 238C. bmim ˆ N,N'-
butylmethylimidazolium, bupy ˆ N-butylpyridinium. [b] 1:1 CH3CN/H2O
(v/v). Reference [13].
Tˆ 298 K. Method
C
(non-steady-state epoxidation conditions):
[[D3]dpRe] ˆ 0.048 m, [alkene] ˆ 0.50 m, Tˆ 298 K.
2H NMR kinetics: Deuteriated alkenes can be used as
substrates in proteated ionic liquids to study the MTO-
catalyzed oxidation of alkenes under steady-state conditions.
The integrals of the 2H resonance signals for deuteriated
reactants (alkenes) and products (epoxide or diol) were
monitored over time to obtain kinetic information. UHP is
soluble in RTILs, a fact that makes the MTO/UHP system
homogeneous in ionic liquids and amenable to kinetic
investigations. However, in order to avoid complex kinetics
and minimize catalyst degradation, one is forced to employ
spectroscopy. The relatively long time scale inherent to
NMR kinetic experiments mandated the use of more deac-
tivated styrenes as epoxidation substrates. For this study,
styrene, 2-fluorostyrene, and 2,6-difluorostyrene were used.
Over the course of a reaction, the intensity of the [D3]dpRe
resonance signal (d ˆ 2.9 ppm) decreases and the intensity of
the [D3]MTO resonance signal (d ˆ 2.4 ppm) increases. The
2H resonance signals due to products of peroxorhenium
complex decompositions, namely CD3OH and CD3OOH (d ˆ
3.4 and 3.3 ppm, respectively), increase during the course of
reaction.[19, 20] An NMR stack plot for the epoxidation of
2-fluorostyrene is shown in Figure 2.
high hydrogen peroxide concentrations (0.5 1.0 m).[19]
A
representative NMR stack plot for the epoxidation of
[D10]cyclohexene in [emim]BF4 is shown in Figure 1. Under
these conditions, the kinetic traces followa single-exponential
decay and are fit with Equation (2).
The integration for the [D3]dpRe resonance signal is plotted
as a function of time. The kinetic traces obtained are
exponential, and the integration for the [D3]dpRe resonance
It À I1 ˆ aexp(Àkyt)
(2)
The kinetics are first-order in
[alkene] and showno depend-
ence on [H2O2]. Hence, ky ˆ k3
[mpRe] ‡ k4 [dpRe] simplifies
to ky ꢀ k4 [dpRe], since at
these high [H2O2], the major
rhenium species is dpRe and (as
shown above by UV/Vis kinet-
ics) k3 < k4 in RTILs. Table 3
displays k4 values obtained by
2H NMR spectroscopy for the
epoxidation and dihydroxyla-
tion of [D8]styrene and [D10]cy-
clohexene.
Figure 1. 2H NMR stack plot for the steady-state epoxidation of [D10]cyclohexene in [emim]BF4. Conditions :
[Re]T ˆ 5.0 mm, [[D10]cyclohexene] ˆ 100 mm, [UHP] ˆ 1.0 m. Spectra are shown at 60 s intervals.
In an alternative method that
more closely resembles the con-
ditions used in the UV/Vis
kinetic
studies,
CD3ReO3
([D3]MTO) was alowed to react
with two equivalents of UHP in
RTILs to generate [D3]dpRe.
The latter was used for single
turnover reactions with alkenes,
and monitored by 2H NMR
Figure 2. 2H NMR stack plot for the single-turnover epoxidation of 2-fluorostyrene in [emim]BF4. Conditions :
[Re]T ˆ 36 mm, [2-fluorostyrene] ˆ 364 mm, [UHP] ˆ 72 mm. Spectra are taken at 165 s intervals.
Chem. Eur. J. 2002, 8, No. 13
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3055 $ 20.00+.50/0
3055

M. M. Abu-Omar et al.
FULL PAPER
k4‰alkeneŠ
dpRe
mpRe
dpRe
mpRe ‡ epoxide
MTO ‡ epoxide
mpRe ‡ H2O2
(3)
(4)
(5)
signal is nearly zero at the end of the reactions, which
demonstrates that the reaction between [D3]dpRe and
substrate proceeds to completion. The kinetic traces are
treated with a standard exponential decay equation (vide
supra) to obtain k4 from the pseudo-first-order rate constant,
ky ˆ k4 [alkene]. Table 3 lists the values of k4 obtained for the
epoxidation of three different styrenes alongside the data
acquired under steady-state conditions with deuteriated
substrates. The k4 values reflect a multiplicative decrease in
rate depending on the number of fluorine atoms present in the
ortho positions on the aromatic ring.
À!
k3‰alkeneŠ
À!
k
À2
À!
value for kÀ2 ꢀ 2.0 Â 10À3 sÀ1 in [emim]BF4.[11] For simplicity,
the disproportionation of mpRe to H2O2 and MTO and the
decomposition pathways for dpRe and mpRe were omitted
from the mechanism. Figure 3 shows the reaction of dpRe and
the formation and subsequent reaction of mpRe.
Discussion
UV/Vis kinetic experiments and simulations: MTO reacts
with hydrogen peroxide to form two peroxorhenium com-
plexes, as shown in Scheme 1. In ionic liquids, the peroxide
binding constants are approximately K1 ˆ 40 and K2 ˆ 120,
and the rate constant k2 is approximately 0.20 mÀ1sÀ1.[11] These
constants necessitate a large excess of hydrogen peroxide to
convert all of the rhenium to dpRe. Hence, a typical
procedure involves making dpRe in situ and then reacting it
with an olefinic substrate. It is important to note that the
concentrations needed for this type of experiment are
generally 0.75 1.00 mm MTO, ꢁ 50 mm H2O2, and 50
300 mm substrate; the latter would be added after the
complete formation of dpRe. However, because of the excess
of peroxide used to form dpRe, H2O2 remains in the solution
after the substrate is added. Thus, once steady state is
achieved, there is an appreciable concentration of dpRe as
well as mpRe in the solution due to the reactions with H2O2
(k1 and k2), and hence the rate constants k3 and k4 are not
easily resolved.
Making dpRe in THF (in which UHP is nearly insoluble)
alleviates the problem described above. When dpRe is
isolated by removing the THF solution from the solid UHP,
the only Re compound in solution is dpRe; essentially zero
hydrogen peroxide is present in the solution. Hence, when
dpRe reacts with the substrate to form mpRe, there is
negligible amount of peroxide to re-form dpRe. Therefore,
the change in absorbance is attributable only to the reactions
of both peroxorhenium complexes with the substrate. It
should be noted that dpRe re-establishes the equilibrium with
mpRe (and mpRe with MTO) once the THF solution is
removed from the solid UHP and diluted into the ionic liquid.
However, the re-establishment of these equilibria is much
slower than the reactions of the peroxorhenium complexes
with the substrate, so the effects on the absorbance at 360 nm
(A360) are essentially negligible with respect to reactions with
substrate. Therefore, the major reactions taking place using
this method are the epoxidation of olefin by dpRe (k4) and by
mpRe (k3).
Figure 3. Simulation of the reaction of dpRe and mpRe with a-methyl-
styrene using KINSIM. Conditions: [dpRe]0 ˆ 0.98 mm, [mpRe]0 ˆ
0.00 mm, [AMS] ˆ 96 mm, k4 ˆ 0.44 mÀ1sÀ1
,
k3 ˆ 0.10 mÀ1sÀ1
,
and kÀ2
ˆ
2.0 Â 10À3 sÀ1
.
Since both mpRe and dpRe absorb at 360 nm,[11, 19] their
simulated contributions were added together to generate a
simulated kinetic trace that should, in theory, be similar to the
time profile that is observed experimentally at comparable
conditions. Since it was still unknown up to this point which
rate constant was faster, simulations were performed using
both k4 > k3 and k3 > k4. The contributions of both mpRe and
dpRe were added together as described above, and the results
were plotted alongside the actual experimental data for
comparison. Figure 4 compares the experimental data to the
simulated kinetic traces for k4 > k3 and k3 > k4.
To gain insight into which is the faster step, the modeling
program KINSIM was used to create simulated kinetic
traces.[18] The mechanism shown in Equations (3) (5) was
used in the simulations, and the rate constants for the
oxidation of a-methylstyrene (AMS) with both of the
peroxorhenium complexes were used alongside the known
*
Figure 4. Comparison of experimental data ( ) and simulated kinetic
traces (k4 > k3, ––; and k3 > k4,
) for the oxidation of a-methylstyrene
(AMS). Simulation conditions: [dpRe]0 ˆ 0.98 mm, [mpRe]0 ˆ 0.00 mm,
[AMS] ˆ 96 mm, k4, k3 ˆ 0.44 and 0.10 mÀ1sÀ1, and kÀ2 ˆ 2.0  10À3 sÀ1
.
Experimental conditions: [dpRe]0 ˆ 0.98 mm, [AMS] ˆ 96 mm in
[emim]BF4 at 238C.
3056
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3056 $ 20.00+.50/0
Chem. Eur. J. 2002, 8, No. 13

Kinetics of MTO-Catalyzed Olefin Epoxidation
3053 3059
À
Figure 4 shows good agreement between the experimental
data and the simulated kinetic trace when k4 > k3. The
simulated and experimental profiles begin to deviate slightly
after 150 s. This deviation is probably attributable to the
decomposition of mpRe to methanol and methyl hydroper-
oxide as observed by 2H NMR spectroscopy. These reactions,
due to their minimal contribution and for the sake of
simplicity, do not appear in the mechanism used for the
KINSIM simulations.
The point to be made here is that Figure 4 shows rather
poor agreement between the experimental data and the
simulated kinetic trace when k3 > k4. The two profiles deviate
substantially from the outset and never agree at any point in
the reaction/simulation. Therefore, in light of the good
agreement between experiment and simulation when k4 >
k3, we concluded that k4 is greater than k3 for the oxidation
of a-methylstyrene in [emim]BF4. The kinetic data for a-
methylstyrene in other RTILs and for other alkenes in
[emim]BF4 are again in excellent agreement with simulations
that employ k4 > k3 values.
As shown in Table 1, the k4 values in [emim]BF4 for the
three styrene substrates are quite similar to those values in 1:1
CH3CN/H2O. The notable exception is the oxidation of
cyclohexene for which k4 in [emim]BF4 is lower by an order
of magnitude than k4 in 1:1 CH3CN/H2O and lower even than
k4 in CH3CN. This result is anomalous but is explained by
considering the limited solubility of cyclohexene in
[emim]BF4. This limited solubility lowers the effective con-
centration of cyclohexene in the ionic liquid, which in turn
causes the k4 value to be artificially low. The low solubility of
several other substrates, including 1-methylcyclohexene,
1-phenylcyclohexene, and 1-decene, precluded their use in
kinetic studies.
The values of k3 in [emim]BF4 for the oxidation of the
styrenes are only slightly higher than the corresponding k3
values in acetonitrile. There is, however, a significant differ-
ence between the k3 values obtained in [emim]BF4 versus
those in 1:1 CH3CN/H2O. This is in contrast to k4 for which the
values in the ionic liquid and in the CH3CN/H2O mixture are
virtually the same.
considering that NO3 has much higher coordination ability
À
than BF4 and thus interferes with the oxidation reactions.
In contrast to what is observed here for RTILs and CH3CN,
k3 in general is larger than k4 for reactions done in 1:1 CH3CN/
H2O.[5] The relative reactivities of mpRe and dpRe are
apparently dependent on the water content of the solvent.
This is likely so because of a couple of different factors: 1)
Additional water in the solution makes the solvent more
polar, which stabilizes the more polar dpRe over mpRe
toward reaction with the substrate. 2) Water participates in
hydrogen bonding more with dpRe than with mpRe because
of the two peroxo groups, rendering dpRe less reactive.
2H NMR studies under steady-state conditions: Deuteriated
alkenes were used to monitor the kinetics of alkene epox-
idation (with UHP) and dihydroxylation (with aqueous H2O2)
under steady-state conditions. The advantages of using
2H NMR spectroscopy include the ability to observe prod-
uct(s) formation in real time, characterize the kinetics of
reaction under catalytic conditions, and use easily prepared
proteated ionic liquids. Nevertheless, this method differs from
the non-steady-state kinetics in that the concentration of
water in the ionic liquid is significant. Even for the MTO/
UHP system, the concentration of water, being a by-product,
increases as the reaction proceeds. The profound effects of
water on the physical properties of RTILs have been
documented previously.[17] Since water increases the polarity
of the ionic solvent, the rate of oxygen transfer from dpRe to
the substrate is expected to increase as [H2O] increases. This
effect is reflected in the k4 values for the dihydroxylation of
[D8]styrene and [D10]cyclohexene (entries 3 and 9, Table 3) in
comparison to the values for epoxidation of the same
substrates (entries 1, 2, 7, and 8, Table 3).
2
The k4 values determined by H NMR spectroscopy under
steady-state conditions for [D8]styrene and [D10]cyclohexene
(Table 3) are not in full agreement with those measured by
UV/Vis kinetics (Table 1). In the case of styrene, the k4 value
from the 2H NMR experiments is one third the value
determined by UV/Vis. On the other hand, in the case of
2
cyclohexene, the k4 value from the H NMR experiments is
The results in Table 1 showclearly that k4 is significantly
greater than k3 when the reactions are done in [emim]BF4.
This trend holds for the same reactions done in other water-
miscible ionic liquids, shown in Table 2. The reactivity of the
styrenes with the peroxorhenium complexes shows the
expected trend for electrophilic oxygen transfer from rheni-
um: styrene < TBMS < a-methylstyrene (AMS). The addi-
tional methyl groups of TBMS and AMS increase the
twice the value determined by UV/Vis. These discrepancies
are attributed to the differences in solvent composition
between the NMR and UV/Vis experiments. While the UV/
Vis experiments are essentially anhydrous and contain < 3%
THF, the water concentration in the NMR experiments
increases as more product (epoxide) forms. Even though
water is expected to enhance the rate of alkene oxidation by
dpRe, it also accelerates the degradation of the peroxorhe-
nium complexes. Furthermore, the need for higher [MTO] in
the steady-state experiments results in an ionic liquid that
contains 10 20% CH3CN; an MTO stock solution in
acetonitrile is normally diluted into the ionic liquid in order
to maintain the homogeneity of the catalyst at these concen-
trations. The presence of an acetonitrile co-solvent in the
2H NMR experiments also explains the improved rates for
cyclohexene epoxidation, since the solubility of cyclohexene
is improved.
ˆ
styrenes' reactivities relative to PhCH CH2, and the sterics
of TBMS make this substrate slightly less reactive than AMS.
Table 2 shows the k3 and k4 values that have been
determined for the oxidation of a-methylstyrene in several
ionic liquids and in molecular solvents. Entries 3, 4, and 6 in
Table 2 showessentially the same rate constants for the ionic
liquids [emim]BF4, [bmim]BF4, and [bupy]BF4. This indicates
that the cations of these liquids do not affect the reaction of
the rhenium complexes with the olefin. However, both k3 and
k4 are substantially lower when the reaction is done in
[bmim]NO3 (Table 2, entry 5). This result is rationalized by
Despite all of the mentioned differences between the UV-
vis (non-steady-state) and 2H NMR (steady-state) kinetic
Chem. Eur. J. 2002, 8, No. 13
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3057 $ 20.00+.50/0
3057

M. M. Abu-Omar et al.
FULL PAPER
experiments, the general agreement within a factor of 2
between the two sets of rate constants is remarkable.
Non-steady-state kinetics by H NMR spectroscopy: The
Experimental Section
Materials: MTO was prepared by the standard literature method.[21]
(nBu)3Sn(CD3)[22] was used in the synthesis of [D3]MTO. Stock solutions
of MTO were prepared in tetrahydrofuran and stored at À108C;
concentrations were determined spectrophotometrically. The ionic liquids
[emim]BF4, [bmim]BF4, [bmim]NO3, [etpy]BF4, and [bupy]BF4 were
prepared as described previously.[11] Hydrogen peroxide solutions were
standardized by iodometric and permanganate titrations. Urea hydrogen
peroxide was standardized by permanganate titration as a stock solution in
[bmim]BF4. The olefinic substrates were used as received (Aldrich).
2
addition of a stock solution of [D3]MTO in anhydrous THF to
UHP in
a
RTIL yields deuteriated diperoxorhenium
2
([D3]dpRe). The H spectrum shows the exclusive presence
of [D3]dpRe (d ˆ 2.9 ppm). However, over time the deactiva-
tion products CD3OH (d ˆ 3.4 ppm) and CD3OOH (d ˆ
3.3 ppm) become evident.
In the absence of excess peroxide, the monoperoxorhenium
complex ([D3]mpRe) decomposes quickly, and hence no
NMR signal for this species is observed. Due to the decom-
position of [D3]mpRe the observed stoichiometry of epox-
idation is always above 1:1 and below 2:1–namely, 1 mole of
[D3]dpRe will convert between 1 and 2 moles of alkene into
epoxide. The stoichiometry depends on both the concentra-
tion and the nature of the alkene undergoing epoxidation.
UV/vis kinetics: Kinetic measurements were obtained on a Shimadzu UV-
2501PC at 23 Æ 18C. Small-volume (0.5 mL) quartz cuvettes with 1.0 cm
optical path length were used. The formation and reaction of the
diperoxorhenium complex was monitored by observing the absorbance
change at 360 nm, the lmax of the dpRe complex (e ˆ 1100 mÀ1 cmÀ1). The
monoperoxorhenium complex mpRe also features an absorbance at
360 nm (e ˆ 500–700 mÀ1 cmÀ1).[12, 19] The kinetics of the reactions of dpRe
and mpRe with substrates were measured by monitoring the decline in
absorbance at 360 nm using the following method.
2
MTO (40.0 mm in dry THF) was mixed with an excess of urea hydrogen
peroxide (UHP), which is insoluble in tetrahydrofuran. After the reaction
was complete, the solution containing the dpRe was removed from the
remaining solid UHP. An aliquot of the 40.0 mm dpRe solution was added
to the ionic liquid of interest to make about 1.0 mm solution. The initial
absorbance was recorded, the substrate was added, and the absorbance
change at 360 nm was monitored with respect to time. Kaleidagraph 3.0 was
used to analyze the kinetic data. The kinetics simulation program KINSIM
was used to model the reactions.[18]
2H NMR kinetics: Kinetic measurements were obtained on Br¸ker Avance
and ARX 500 MHz (1H) spectrometers at ambient temperature. A sealed
capillary containing a 0.346 m solution of [Cr(acac)3] in CD3CN/toluene
(5% v/v CD3CN) was used as a standard for chemical shifts (d ˆ 9.80 ppm)
and integrations.
Based on H NMR kinetic data, the value of k4 for the
epoxidation of styrene in [emim]BF4/THF (27% v/v THF) is
0.059 mÀ1sÀ1 (entry 4, Table 3). This value is half the value of k4
obtained from UV/Vis kinetics for this epoxidation in
[emim]BF4 (0.13 mÀ1sÀ1). The modest decrease in rate be-
tween the NMR study and the UV/Vis study in [emim]BF4 is
attributable to the relatively large amount of THF used in the
NMR study, which is imposed by the need of concentrated
[D3]dpRe (ca. 40 mm) for reliable detection over short (ca.
3 min) collection times.
The following is a typical procedure for collecting steady-state kinetics with
a deuteriated alkene ([D8]styrene): An NMR tube was charged with
[emim]BF4 (0.5 mL), UHP (47 mg), and about 100 mL of a 0.040 m MTO
stock solution in CH3CN. After 10 min, a sealed capillary tube containing
the external standard was added, along with [D8]styrene (6 mL). The
contents of the tube were mixed well, and 2H NMR spectra were
subsequently collected about every 5 min.
Conclusion
The reaction of MTO and UHP in dry THF yields essentially
water- and peroxide-free methyldiperoxorhenium (dpRe),
which can be diluted into room temperature ionic liquids
(RTILs) after the removal of the remaining solid UHP. The
homogeneous solution of dpRe in RTILs was reacted with
various olefins under pseudo-first-order conditions. The
kinetic profiles are biexponential, and the observed rate
constants display first-order dependences on [olefin]. Based
on kinetic simulations and their agreement with experimental
data, the fast step was assigned to the reaction of olefin with
dpRe (k4) and the slowstep to the reaction of olefin with the
monoperoxorhenium complex (mpRe) (k3). While the mpRe,
in general, is more reactive towards alkenes than the dpRe in
1:1 CH3CN/H2O, the opposite is true for the reactions
conducted in several RTILs. For example, in all of the
investigated RTILs, k4 ˆ 4.5  k3. The rate constants k3 and k4
are unaffected by the cation of the ionic liquid, but are
sensitive to the anion. The kinetics of the reaction of dpRe
with deuteriated alkenes have been characterized by 2H NMR
spectroscopy under steady-state conditions. Also, [D3]dpRe,
prepared in situ from [D3]MTO and UHP, has been employed
to characterize the system and determine k4 by 2H NMR
spectroscopy under single turnover experiments. The NMR
rate constants are in general agreement (within a factor of
about 2) with those obtained by UV/Vis spectroscopy. The
difference in solvent composition between the NMR study
and the UV/Vis study accounts for the modest variations in
rate constants.
For non-steady-state kinetics, 150 mL of 0.13 m [D3]MTO solution in dry
THF was added to 0.10 m UHP in 400 mL of [emim]BF4. The reaction
solution was mixed well, and monitored by 2H NMR spectroscopy until all
the starting [D3]MTO had converted to [D3]dpRe. At this time, a tenfold
molar excess of an olefinic substrate was added to ensure pseudo-first order
conditions.
Under these conditions, the 2H resonance signal of [D3]dpRe appears at
d ˆ 2.9 ppm. The extent of reaction between [D3]dpRe and substrates is
indicated by the decrease in the integral for this resonance signal over time.
Kaleidagraph 3.0 was used in the analysis of the kinetic data.
Acknowledgements
Support for this research from the University of California Toxic
Substances Research and Teaching Program (UCTSR&TP), the Arnold
and Mabel Beckman Foundation (BYI to M.M.A.), and the NSF (CHE-
9874857) is gratefully acknowledged.
[1] G. Strukul in Catalytic Oxidations with Hydrogen Peroxide as Oxidant
(Ed.: G. Strukul), Kluwer Academic Publishers, Dordrecht 1992,
pp. 1 11.
[2] W. A. Herrmann, R. W. Fischer, D. W. Marz, Angew. Chem. Int. Ed.
Engl. 1991, 30, 1638 1641; Angew. Chem. 1991, 103, 1706 1708.
[3] W. Adam, C. M. Mitchell, Angew. Chem. 1996, 108, 578–581; Angew.
Chem. Int. Ed. Engl. 1996, 35, 533 535.
3058
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3058 $ 20.00+.50/0
Chem. Eur. J. 2002, 8, No. 13

Kinetics of MTO-Catalyzed Olefin Epoxidation
3053 3059
[4] C. C. Romao, F. E. Kuhn, W. A. Herrmann, Chem. Rev. 1997, 97,
3197 3246.
[5] J. H. Espenson, Chem. Commun. 1999, 479 488.
[6] G. S. Owens, J. Arias, M. M. Abu-Omar, Catal. Today 2000, 55, 317
363.
[7] T. Welton, Chem. Rev. 1999, 99, 2071 2083.
[8] P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112, 3926 3945;
Angew. Chem. Int. Ed. 2000, 39, 3772 3789.
[9] R. Sheldon, Chem. Commun. 2001, 2399 2407.
[10] G. S. Owens, M. M. Abu-Omar, Chem. Commun. 2000, 1165 1166.
[11] G. S. Owens, M. M. Abu-Omar, J. Mol. Catal. A; Chemical, 2002, in
press.
[12] W. A. Herrmann, R. W. Fischer, W. Scherer, M. U. Rauch, Angew.
Chem. 1993, 105, 1209 1212; Angew. Chem. Int. Ed. Engl. 1993, 32,
1157 1160.
[14] A. Al-Ajlouni, J. H. Espenson, J. Org. Chem. 1996, 61, 3969 3976.
[15] C. Hardacre, J. D. Holbrey, J. S. E. McMath, Chem. Commun. 2001,
367 368.
[16] A. Durazo, M. M. Abu-Omar, Chem. Commun. 2002, 66 67.
[17] K. R. Seddon, A. Stark, M.-J. Torres, Pure Appl. Chem. 2000, 72,
2275 2287.
[18] B. A. Barshop, C. F. Wrenn, C. Frieden, Anal. Biochem. 1983, 130,
134 145.
[19] M. M. Abu-Omar, P. J. Hansen, J. H. Espenson, J. Am. Chem. Soc.
1996, 118, 4966 4974.
[20] W.-D. Wang, J. H. Espenson, Inorg. Chem. 1997, 36, 5069 5075.
[21] W. A. Herrmann, F. E. Kuhn, R. W. Fischer, W. R. Thiel, C. C. Romao,
Inorg. Chem. 1992, 31, 4431 4432.
[22] B. Jousseaume, M. Lahcini, M.-C. Rascle, Organometallics 1995, 14,
685 689.
[13] A. Al-Ajlouni, J. H. Espenson, J. Am. Chem. Soc. 1995, 117, 9243
9250.
Received: March 13, 2002 [F3943]
Chem. Eur. J. 2002, 8, No. 13
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0813-3059 $ 20.00+.50/0
3059