Comparative kinetic investigations in ionic liquids using the MTO/peroxide system

Article abstract of DOI:10.1016/S1381-1169(02)00236-4

The kinetics and thermodynamics of the reaction of methyltrioxorhenium (MTO) with hydrogen peroxide in ambient temperature ionic liquids have been studied. The rate constant for the formation of the diperoxorhenium complex, CH₃ReO(η²

Full text of DOI:10.1016/S1381-1169(02)00236-4

Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
Comparative kinetic investigations in ionic liquids using the
MTO/peroxide system
Gregory S. Owens, Mahdi M. Abu-Omar∗
Department of Chemistry and Biochemistry, University of California, P.O. Box 95169, Los Angeles, CA 90095-1569, USA
Received 11 February 2002; accepted 7 May 2002
Abstract
Thekineticsandthermodynamicsofthereactionofmethyltrioxorhenium(MTO)withhydrogenperoxideinambienttemper-
ature ionic liquids have been studied. The rate constant for the formation of the diperoxorhenium complex, CH3ReO(␩2-O2)2,
from the monoperoxorhenium complex, CH3ReO2(␩2-O2), and hydrogen peroxide in [bmim]NO3 is 0.053 0.002 M−1 s−1
.
The equilibrium constants for the binding of two peroxide units have been determined in [emim]BF4 to be K1 = 110 28 and
K2 = 160 36. Similar rate constants and equilibrium constants are obtained in other water-miscible, dialkylimidazolium and
alkylpyridinium ionic liquids. The values of the rate constants are highly dependent on the concentration of water in the solvent.
These results indicate that the ionic liquids can behave like organic solvents and aqueous solutions of high salt concentrations.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Ionic liquids; Methyltrioxorhenium; Kinetics; Imidazolium
1. Introduction
including hydrogenation [6,7], hydroformylation
[6,8], Heck reactions [9–11], and oxidations [12,13].
Furthermore, there have been reports of reaction ac-
celeration in ionic liquids as opposed to molecular
solvents. Palladium-catalyzed Suzuki cross-coupling
reactions in 1-butyl-3-methylimidazolium tetrafluo-
roborate ([bmim]BF4) have been described to proceed
at 90–200 times faster than in a mixture of toluene,
water, and ethanol [14]. Rate-enhancement has also
been observed in biphasic Trost–Tsuji couplings when
a biphasic mixture of [bmim]Cl/methylcyclohexane
is used instead of butyronitrile/water [15].
Room temperature ionic liquids have gained a great
deal of attention during the past few years as alterna-
tive reaction media [1–5]. Interest in the dialkylimida-
zolium and alkylpyridinium salts, particularly, is high
because of their negligible vapor pressure and their
stability to both air and moisture. The possibility of
using ionic liquids as truly recyclable reaction media
is exceptionally appealing because of increasing envi-
ronmental concerns with respect to waste production
and containment.
In addition to expanding the applicability of ionic
liquids to various types of organic transformations, a
great deal of physical data and trends has surfaced in
the last 2 years. The specific conductivity of a wide
range of water-immiscible ionic liquids has been stud-
ied in relevance to their potential use in photoelec-
Many transition metal-catalyzed reactions have
been shown to proceed with high yields in both
water-miscible and water-immiscible ionic liquids,
∗
Corresponding author. Fax: +1-310-206-9130.
E-mail address: mao@chem.ucla.edu (M.M. Abu-Omar).
1381-1169/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(02)00236-4

216
G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
trochemical cells [16]. The polarity of a number of
ionic liquids has been investigated recently through
studies that show contributions from both the cation
and the anion [17,18]. The influence of water, organic
solvents, and impurities, especially leftover chloride,
on the viscosity and density of ionic liquids has been
reported recently [19]. Additionally, pulsed-gradient
spin-echo 1H and 19F NMR have been used to measure
the ionic diffusion coefficient of various ionic liquids
[20].
Kinetic studies on reactions in ionic liquids are
conspicuously absent from the growing list of phys-
ical measurements. These studies are an important
part of determining the economic viability of us-
ing ionic liquids as alternative reaction media, es-
pecially concerning industrial applications of ionic
liquids.
2. Experimental section
2.1. Materials
Bromoethane, bromobutane, 1-methylimidazole,
1-butylimidazole, and pyridine were distilled prior to
use. Hydrogen peroxide solutions were standardized
by iodometric and permanganate titrations. UHP was
standardized by permanganate titration as a stock so-
lution in [bmim]BF4. MTO was prepared by the stan-
dard literature method [34]. Stock solutions of MTO
were prepared in ACN and stored at −10 ◦C; con-
centrations were determined spectrophotometrically.
All syntheses were performed under inert atmosphere
conditions, though rigorous water-exclusion methods
are unnecessary.
Our research group is interested in the utility
of ionic liquids for transition metal-catalyzed oxy-
gen atom transfers. We have shown previously that
the ionic liquid [emim]BF4 can be used with urea
hydrogen peroxide (UHP) and the catalyst methyl-
trioxorhenium (MTO) to effect the epoxidation of
olefins [13]. The epoxide yields are comparable to
the yields that are obtained in molecular solvents,
and there seems to be a rate-enhancement compared
to reactions done in dichloromethane and acetonitrile
(ACN).
MTO reacts with excess hydrogen peroxide to form
two ␩2-peroxo complexes [21,22]. Both intermedi-
ates are capable of transferring an oxygen atom to
suitable substrates, including olefins [23,24], sulfides
[25–27], phosphines [28], and aromatics [29–31].
The formation of monoperoxorhenium (mpRe) and
diperoxorhenium (dpRe) complexes with accompa-
nying rate constants k1 and k2 has been well estab-
lished in a variety of solvents of differing polarity
[22,28,32,33]. This report focuses on the formation
of the catalytically active peroxorhenium intermedi-
ates, especially dpRe, in ionic liquids and examines
the effect of these unusual media on this well studied
system.
2.2. Solutions
Kinetic and thermodynamic measurements were
obtained on a Shimadzu UV-2501PC at 23 1 ◦C.
Small-volume (0.50 ml) quartz cuvettes with 1.00 cm
optical path length were used. All solutions contained
0.10 M perchloric acid to stabilize the MTO/peroxide
system. The reaction solutions involving aqueous hy-
drogen peroxide were studied by adding perchloric
acid and MTO to the particular solution, obtaining a
baseline, and then adding aqueous hydrogen peroxide
last.
Since UHP is a solid, stock solutions of UHP in the
various ionic liquids were made and then diluted. Per-
chloric acid was then added, a baseline was obtained,
and MTO was added last.
2.3. Kinetics and thermodynamics
The formation of the dpRe complex was monitored
by observing the absorbance change at 360 nm, the
λmax of the dpRe complex (ε = 1100 M−1 cm−1).
Plots of kψ (the pseudo rate constant) versus [H2O2]
were analyzed to obtain values of k2 for the reactions

G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
217
of MTO with hydrogen peroxide and UHP in a variety
3.3. 1-Butyl-3-methylimidazolium bromide
([bmim]Br)
of ionic liquids and salt mixtures. The maximum ab-
sorbance values at 360 nm for variable [H2O2] were
used to calculate the peroxide binding constants K1
and K2. The nonlinear least-squares program Kaleida-
graph was used to analyze the kinetic and thermody-
namic data.
The same procedure was used as for [emim]Br
except that the mixture was heated to 65 ◦C for 2 h
after complete addition of 1-methylimidazole. The
mixture turned cloudy and then separated into two
phases: excess 1-bromobutane on top and [bmim]Br
on bottom. The excess 1-bromobutane was re-
moved under reduced pressure, leaving behind an
extremely viscous, cloudy liquid, [bmim]Br. The
sludgy liquid was washed three times with THF
and dried in vacuo, resulting in a white solid. The
solid was recrystallized from ACN and THF, giv-
3. Synthesis of ionic liquids
3.1. 1-Ethyl-3-methylimidazolium bromide
([emim]Br)
1
[emim]Br was synthesized according to a literature
procedure [16] with the following modifications. With
vigorous stirring, bromoethane was added via addition
funnel over 1.5 h to 1-methylimidazole neat. Ambi-
ent temperature was maintained by the use of a water
bath. The mixture became cloudy after complete addi-
tion of bromoethane, and a white solid formed within
1 h. The excess bromoethane was decanted, and the
white solid, [emim]Br, was washed three times with
tetrahydrofuran (THF) and recrystallized from ACN
and THF in 96% yield. 1H NMR shifts correspond to
those from the literature [16].
ing [bmim]Br in 94% yield. H NMR peaks corre-
spond to those reported previously in the literature
[16].
3.4. 1-Butyl-3-methylimidazolium tetrafluoroborate
([bmim]BF4)
The same synthetic and purification procedure was
used as for [emim]BF4 to yield [bmim]BF4 in 92%
1
overall yield. H NMR peaks correspond to those re-
ported previously in the literature [35].
3.5. 1-Butyl-3-methylimidazolium nitrate
([bmim]NO3)
3.2. 1-Ethyl-3-methylimidazolium tetrafluoroborate
([emim]BF4)
[bmim]NO3 was prepared as described in the liter-
ature by metathesis of [bmim]Br with AgNO3 [19].
The same purification procedure as for [emim]BF4
was used to obtain [bmim]NO3 in 97% overall yield.
1H NMR shifts are in the range for [bmim]+ described
in the literature [16].
[emim]BF4 was synthesized according to the liter-
ature procedure [7] with the following modification.
[emim]Br was dissolved in 9:1 acetone:ACN, and
1.1 equiv. of sodium tetrafluoroborate were added
to the solution. The solution was vigorously stirred
for 60 h at room temperature. Leftover NaBF4 and
the byproduct, sodium bromide, were removed by
filtration, and the solvent was removed under re-
duced pressure to yield [emim]BF4 as a clear, col-
orless liquid. The liquid was redissolved in ACN
and stirred with neutral alumina (Brockman ac-
tivity 1) for 6 h to remove traces of unreacted
1-methylimidazole from the previous step. Leftover
Br− was removed from the product ionic liquid by
precipitation with silver tetrafluoroborate or silver
nitrate. Overall yield, 92%. 1H NMR peaks corre-
spond to those reported previously in the literature
[35].
3.6. 1-Butyl-3-methylimidazolium
trifluoromethanesulfonate ([bmim]OTf)
The literature procedure was used to synthesize
[bmim]OTf as a clear, yellow liquid in 93% yield
[16]. The liquid was redissolved in ACN and stirred
with neutral alumina for 5 h. Removal of the alu-
mina by filtration and the organic solvent by vacuum
yielded the product as a clear, colorless liquid (80%
1
yield). H NMR shifts correspond to those from the
literature [16].

218
G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
3.7. 1-Butyl-3-methylimidazolium
hexafluorophosphate ([bmim]PF6)
3.11. Ethylpyridinium tetrafluoroborate ([etpy]BF4)
The same synthetic procedure as for [bupy]BF4 was
used to obtain [etpy]BF4 as a white solid. The solid
was redissolved in ACN and stirred with neutral alu-
mina to remove organic impurities. The solvent was
then removed to give [etpy]BF4 in 90% overall yield.
1H NMR shifts are the same as for [etpy]Br. 19F NMR
(ACN-d3): −153.0.
[bmim]PF6 was prepared as described in the liter-
ature by metathesis of [bmim]Br with NaBF6 in wa-
ter [19]. After separation from the aqueous phase of
the metathesis mixture, the product ionic liquid was
mixed with THF and stirred with neutral alumina for
2 h. Removal of the alumina and then of the organic
solvent yielded [bmim]PF6 as a clear, colorless liquid
(68%). No residual bromide was detected through as-
1
say with hydrogen peroxide (vide infra). H and 19F
4. Results and discussion
NMR shifts correspond to those from the literature
[7].
4.1. Observations
A word about the color of ionic liquids in general
is relevant here. Since the work described in this pa-
per involves using UV–Vis exclusively to monitor ki-
netics, it was important to be able to produce ionic
liquids that are colorless. The ionic liquids used in
this study can range from colorless to orange (in the
case of [bmim]OTf), and the major difference seems
to be temperature regulation during preparation. The
reaction of 1-methylimidazole with bromoethane is
rather exothermic, and the product [emim]Br will be
colored yellow if the reagents are mixed together too
rapidly. The use of an addition funnel helps reduce this
problem. Furthermore, [emim]Br can be recrystallized
from ACN and THF to form long, white needles. This
recrystallization serves not only to remove the yellow
impurity from the product but also to remove leftover
methylimidazole. Removal of the yellow impurity at
this stage in the ionic liquid synthesis is key because
of the ability to recrystallize the halide precursor. The
1-butyl-3-methylimidazolium salts present a similar
case. While the reaction of 1-methylimidazole with
bromoethane goes to completion without additional
heating, the same reaction with bromobutane requires
heat to go to completion within a few hours. Exces-
sive heating of the methylimidazole/bromobutane re-
action mixture causes the mixture to turn dark yellow,
but recrystallization can again be employed to isolate
[bmim]Br as long, white needles.
3.8. Butylpyridinium bromide ([bupy]Br)
[bupy]Br was synthesized according to a litera-
ture procedure [36] with the following modifications.
Bromobutane was used instead of chlorobutane, and
the reaction was performed neat by adding pyridine
via additional funnel to bromobutane. The mixture
was then heated to 65 ◦C for 6 h, during which time
the mixture turned cloudy and then separated into
two phases. The excess 1-bromobutane was removed
under reduced pressure, leaving behind an off-white
solid, [bupy]Br. The solid was washed three times
with THF and dried in vacuo. The solid was recrys-
tallized from ACN and THF, giving [bupy]Br in 94%
yield. 1H NMR (ACN-d3): δ (ppm) 8.72 (d, 2H), 8.50
(t, 2H), 8.03 (t, 1H), 4.52 (t, 2H), 2.01–1.86 (m, 2H),
1.45–1.27 (m, 2H), 0.98–0.90 ppm (t, 3H).
3.9. Butylpyridinium tetrafluoroborate ([bupy]BF4)
The same synthetic and purification procedure as
for [emim]BF4 and [bmim]BF4 was used to obtain
[bupy]BF4 as a colorless liquid in 93% overall yield.
1H NMR shifts were the same as for [bupy]Br. 19F
NMR (ACN-d3): −152.92.
3.10. Ethylpyridinium bromide ([etpy]Br)
Finally, the synthesis of [bmim]OTf merits special
consideration because it does not proceed through a
solid intermediate, and recrystallization of the product
is not practical. The reaction of methyl triflate with
1-butylimidazole is highly exothermic and should be
The same procedure was used as for [bupy]Br to
obtain [etpy]Br as a white solid in 96% yield. 1H NMR
(ACN-d3): δ (ppm) 8.60 (d, 2H), 8.39 (t, 2H), 7.94 (t,
1H), 4.41 (q, 2H), 1.98 ppm (t, 3H).

G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
219
performed at 0 ◦C with an appropriate solvent such as
trichloroethane or ACN; an addition funnel should also
be employed. The color of the product can range from
pale yellow to dark red, depending on how much care
is taken to control the reaction mixture’s temperature.
Stirring the ionic liquid with neutral alumina at least
once is required to render the ionic liquid colorless.
Solvent purity is of utmost importance for success
in making kinetic measurements on the MTO/peroxide
system in room temperature ionic liquids. The starting
materials 1-methylimidazole, 1-butylimidazole, and
pyridine react with MTO to form yellow complexes.
The influence of pyridine on the MTO/peroxide sys-
tem has been thoroughly examined [33], and it is
assumed that methylimidazole and butylimidazole
present the same advantages (and problems) as those
observed with pyridine, pyrazole, bipyridine, and
related compounds.
is called “hydrophobic” since it is not miscible with
water, the ionic liquid is actually hygroscopic, as
pointed out by a few authors [17,19], and will form
one phase when mixed with very small amounts of
aqueous hydrogen peroxide. Additionally, our system
requires the use of aqueous acid in order to stabi-
lize the ␩2-peroxorhenium complexes, and this small
amount of acid is also miscible with [bmim]PF6.
However, the presence of 0.10 M perchloric acid in
[bmim]PF6 apparently promotes the formation of HF,
as evidenced by etching of the glass container in
which the solution was temporarily stored for kinetics
studies. This observation accords with a simila−r result
recently described involving mixtures of PF6 with
HNO3 [38].
4.2. Kinetics
The kinetics and thermodynamics of this system
were studied by monitoring the absorbance changes at
360 nm, the absorption maximum of dpRe. The spec-
trum of dpRe and MTO in [emim]BF4 is given in
Fig. 1.
The ionic liquids used in this study all have a slight
absorbance at 360 nm in the absence of peroxide and
MTO. This absorbance remains constant at least in
the range of 330–500 nm, so good baselines can be
Leftover bromide is much more problematic than
leftover imidazole. Bromide is oxidized by both per-
oxorhenium complexes to hypobromite, which in
turn forms hypobromous acid [37]. In the presence
of excess peroxide, hypobromous acid catalyzes the
disproportionation of hydrogen peroxide to molecular
oxygen and water. When excess bromide is present,
hypobromous acid reacts to form bromine, which re-
acts with the 1,3-dialkylimidazolium cations, possibly
by brominating at the 2-position. It is important to
remove all traces of halide from the ionic liquids, as
only 1 mol% impurity in the ionic liquid gives an im-
purity concentration of 79 mM. However, treatment of
the ionic liquids with a large excess of Ag(I) can be
problematic as well, since Ag(I) is photosensitive and
will cause the liquids to turn from colorless to brown.
Careful titration of the appropriate amount of Ag(I)
prevents excess silver from coloring the ionic liquids.
Finally, the presence and effects of dissolved salts,
especially NaBF4, in ionic liquids have been docu-
mented [19]. These salts are removed easily by redis-
solving the ionic liquids in dichloromethane, which
causes the precipitation of the solid salt, and subse-
quent filtration.
The ionic liquid [bmim]PF6 presents an interest-
ing case. Since many reports have been published
concerning different reactions employing [bmim]PF6
as the solvent, including catalytic epoxidations
[12], we thought to include it in our study of the
MTO/peroxide chemistry in ionic liquids. Though it
Fig. 1. UV–Vis spectrum of MTO and dpRe in [emim]BF4. Con-
ditions: 0.93 mM MTO, 42 mM H2O2, 0.10 M HClO4, 23 ◦C.

220
G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
obtained. However, the inherent absorbance at 360 nm
changes dramatically with respect to water content,
and the 1,3-dialkylimidazolium-based ionic liquids
are known to be hygroscopic. Additionally, solu-
tions of 30% hydrogen peroxide (70% water) were
employed, and the reactions of MTO with peroxide
themselves liberate two molecules of water per MTO.
Therefore, for kinetic purposes, most of the exper-
iments described here were done with at least 10%
water (v/v) so as to maintain an essentially constant
water concentration.
As pointed out earlier, the MTO/H2O2 system fea-
tures the formation of two catalytically active perox-
orhenium complexes. The rate constants k1 and k2 have
been determined in a number of molecular solvents,
including water [22], ACN [32], methanol [33], and
nitromethane [33]. In aqueous solution, the reactions
of MTO with hydrogen peroxide are so fast that they
have to be studied using stopped-flow techniques. The
kinetic profile of the reaction is biexponential, and ab-
sorbance changes at 305 nm (the λmax of mpRe) and
at 360 nm have been used to determine k1 and k2 (80
and 5.2 M−1 s−1, respectively) [22]. In organic sol-
vents, the reaction has been studied using conventional
UV–Vis and also features biexponential traces.
Because of the viscosity of ionic liquids (generally >
60 cP), we decided against using stopped-flow tech-
niques and resorted instead to conventional UV–Vis
spectroscopy. Furthermore, the kinetics of the reaction
in ionic liquids were studied exclusively at 360 nm
because of the ionic liquids’ absorbance at wave-
lengths below 330 nm. In ionic liquids, the absorbance
change at 360 nm follows a single-exponential curve.
It can be easily seen from the trace in Fig. 2 that the
first part of the reaction, the formation of the mpRe
complex, occurs extremely rapidly; at time = 0.0 s,
the absorbance is already approximately 0.31. This
first step is essentially completed at this stage. Hence,
the reaction that is being monitored is described by
the reversible process in Eq. (1), and the absorption
data at 360 nm obeys Eq. (2), in which Abs0 is the
absorption at the initial dpRe concentration, that is
[dpRe]0, and Abse is the absorption at the equilibrium
concentration [dpRe]e. The rate expression for dpRe
is given by Eq. (3) and the ratio for the equilibrium
concentrations [dpRe]e and [mpRe]e by Eq. (4). With
[mpRe]e + [dpRe]e − [dpRe] in place of [mpRe]
and with the use of Eq. (4), integration of 3 leads to
Fig. 2. Kinetic profiles of the reaction of MTO with aqueous hy-
drogen peroxide in CH3CN and 9:1 [emim]BF4:H2O. Conditions:
[MTO] = 0.85 mM, [HClO4] = 0.100 M, [H2O2] = 42.6 mM,
T = 23 ◦C.
Eq. (5). Equating the exponents in Eqs. (2) and (5)
yields Eq. (6) for the observed pseudo constant kψ .
Plots of kψ versus [H2O2] or [UHP] were then used
to determine k2. A typical plot of kψ versus [H2O2]
is given in Fig. 3. Even though the intercepts from
the plots of kψ versus [H2O2] should, in principle,
provide values for k in different ionic media, the
−2
y-intercepts are significantly smaller than the cor-
responding slopes (Fig. 3), a fact that renders k
negligible within experimental precision. Therefore,
−2
decisive medium effects on the reversal process, k
cannot be discerned.
,
−2
(1)
(2)
Abst = Abse + (Abs0 − Abse) exp(−kψ t)
d[dpRe]
= k2[mpRe][H2O2] − k−2[dpRe]
dt
(3)
(4)
[mpRe]e
[dpRe]e
k
−2
=
k2[H2O2]

G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
221
Fig. 3. Plot of aqueous H2O2 dependence in 9:1 [bmim]NO3:H2O. The slope gives k2, and the intercept gives k−2. Conditions: 0.85 mM
MTO, 0.10 M HClO4, and 4.69–234 mM peroxide.
[dpRe] = [dpRe]e + ([dpRe]0 − [dpRe]e)
× exp{−(k2[H2O2] + k−2)t}
The kinetic treatment used to determine k2 is based
on the assumption that the formation of mpRe is com-
pleted before the time profiles for the formation of
dpRe are obtained, but this treatment does not ac-
count for the fact that mpRe may not have reached
(5)
(6)
kψ = k2[H2O2] + k
−2
Fig. 4. Plot of kψ versus C (defined in text) for the reaction of MTO with aqueous hydrogen peroxide in 90% [bmim]NO3. The slope
gives k2. Conditions: 0.85 mM MTO, 0.10 M HClO4, and 4.69–234 mM peroxide.

222
G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
equilibrium. To verify the validity of this assump-
tion, it is possible to determine k2 using an expres-
sion that incorporates the equilibrium constants K1 and
K2, which are determined independently by thermody-
namic titrations (vide infra). This treatment takes into
consideration the possibility of incomplete reaction of
MTO with H2O2 to form mpRe prior to the formation
of dpRe. Assuming the reaction of MTO with H2O2
to be incomplete gives
liquids in this study range from 66.5 cP ([emim]BF4)
to 154 cP ([bmim]BF4) [19], and rate constants are
expected to decrease with increasing solvent viscos-
ity. However, the rate constants are small enough for
this system that they are not diffusion-limited. The
Stokes–Einstein equation can be used to estimate the
diffusion-controlled rate constants (kdc) in the investi-
gated ionic media [39]. The values of kdc at 298 K lie in
the range of ∼ 1.0 × 108 L mol−1 s−1 for [emim]BF4
and ∼ 4.3 × 107 L mol−1 s−1 for [bmim]BF4. So the
viscosity of the ionic liquids is expected to have minor
effects if any. The anion coordination ability is as fol-
ꢀ
ꢁ
K1[H2O2]2
1 + K1[H2O2] K2
1
kψ = k2
+
= k2C
−
lows: OTf > NO3 > BF4. Higher anion coordina-
Plotting kψ versus C gives a line whose slope
is equal to k2. As shown in Fig. 4 for the case
of 90% [bmim]NO3, the value of k2 obtained is
0.16 0.01 M−1 s−1. When the errors in the equilib-
rium constants K1 and K2 (Table 3) are considered, this
tion ability should lead to a decreased k2 because the
anion will interfere with peroxide binding to the rhe-
nium. Finally, the hydrogen-bonding ability of both the
anion and the cation (from the proton at the 2-position)
may affect the kinetics by interacting with hydrogen
peroxide and with the peroxide group on mpRe. How-
ever, all of these factors are expected to produce very
subtle, minor effects and will not be taken up further.
It is interesting to note that the k2 values are essen-
tially independent of the form of hydrogen peroxide.
UHP has been used previously to effect various ox-
idations with MTO [13,40,41], but its insolubility in
organic solvents has precluded any kinetic studies to
this point. The results in Table 1 are not particularly
surprising considering that UHP is just a stable form
of concentrated hydrogen peroxide.
In order to gain an understanding of how the high
salt concentration of ionic liquids affects the kinetics
of the MTO/hydrogen peroxide chemistry, it was nec-
essary to obtain kinetic data in a number of concen-
trated salt solutions and compare that data to rate con-
stants that have been determined previously in aque-
ous solution and organic solvents. These results are
given in Table 2. The salts used for this study are
[etpy]BF4 (a solid at room temperature), [bmim]NO3
(liquid), and [bmim]BF4. [etpy]BF4 was chosen as a
control because it is expected to have many of the
same properties as all of the ionic liquids used in this
study with respect to ion size, anion coordination abil-
ity, and hydrogen-bond donor (HBD) ability.
value is essentially the same as 0.22 0.01 M−1 s−1
,
the value that was obtained as described above
(Eq. (6)). This result verifies the assumption that the
first step of mpRe formation, that is k1[MTO][H2O2],
is essentially complete prior to the formation of dpRe.
The rate constant k2 was determined in a number of
ionic liquids, ionic liquid/water mixtures, and salt so-
lutions. Values of k2 determined in 9:1 (v/v) solutions
of ionic liquid/water with both aqueous hydrogen per-
oxide and UHP as oxidant are given in Table 1. Note
that the rate constant k2 was not determined for the re-
action of MTO with UHP in [bupy]BF4 because UHP
is only slightly soluble in this ionic liquid.
The rate constant k2 is the same for all five of
the ionic liquids in Table 1. It is likely that this is
so because of several combined factors, including
the viscosity of the liquids, the coordination ability
of the anion, and the hydrogen-bonding ability of
the anion and cation. The dynamic viscosities of the
Table 1
Values of k2 in 9:1 (v/v) solutions of ionic liquid:watera
Ionic liquid
H2O2, k2
(l mol−1 s−1
UHP, k2
(l mol−1 s−1
)
)
[emim]BF4
[bmim]BF4
[bmim]NO3
[bmim]OTf
[bupy]BF4
0.20
0.23
0.22
0.17
0.21
0.01
0.21
0.20
0.23
0.20
–
0.01
0.02
0.01
0.02
0.01
0.01
0.01
0.01
Comparing the rate constants in Table 2 shows
several interesting points. First, the reaction in ionic
liquids is much slower than in water. In the case of
the pure ionic liquid without any added water (en-
try 4), k2 is approximately 1% of that measured in
a Conditions: 0.10 M HClO4, 0.80 mM MTO, 5–250 mM per-
oxide, 10% water by volume, and T = 23 ◦C.

G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
223
Table 2
concentrations. However, relating reaction rates for
the MTO/peroxide system strictly to solution ionic
strength for these experiments is impractical for a
few reasons. First, the effect of ionic strength alone
on these reactions should not be very substantial, as
all reactants and products are neutral. Secondly, these
solutions have such high salt concentrations that all
the ions are unlikely to be solvated by available wa-
ter. Indeed, ion pairing must be considered at these
high salt concentrations, and the extent of ion pairing
versus ion solvation is difficult to estimate in this
regime. For these reasons, relating the kinetic data to
the activity of water is impractical as well.
Values of k2 in various media
Entry
Solvent
k2 (M−1 s−1
)
[H2O] (M)
a
a
1
2
3
4
5
6
7
8
9
5.92 M [etpy]BF4
5.41 M [etpy]BF4
0.20
0.38
0.65
0.053
0.12
0.22
0.36
0.79
1.17
1.90
5.55
8.88
18.7
0.555
2.77
5.55
11.1
22.2
33.3
50.0
a
4.28 M [etpy]BF4
99% [bmim]NO3
95% [bmim]NO3
90% [bmim]NO3
80% [bmim]NO3
60% [bmim]NO3
40% [bmim]BF4
10% [bmim]BF4
10
a The density for this salt was estimated to be 1.25 g/ml.
It has been suggested that the acceleration in aque-
ous solution is attributable to the higher polarity of
water and the possible involvement of water in the
transition state [32]. Furthermore, the involvement of
water in the reaction is clearly illustrated by its co-
ordination to dpRe. The coordination of water onto
mpRe has not been observed directly, but it is difficult
to rule out since little structural information on mpRe
is available.
In light of these phenomena, for kinetic purposes
it is best to consider these concentrated salt solutions
in terms of the concentration of water. For simplicity,
[H2O] for these experiments was calculated without
aqueous solution (k2 = 5.2 M−1 s−1 in 0.1 M HClO4)
[22]. Secondly, k2 in 99% ionic liquid is essentially
the same as k2 in ACN (0.045 M−1 s−1) [32]. This is
not surprising since the polarity of ionic liquids has
been estimated to be very similar to that of ACN and
methanol [17].
It has been observed previously that the reaction
rate of MTO with hydrogen peroxide decreases with
increasing ionic strength [37]. This result seems par-
ticularly relevant to the present study, as ionic liquids
themselves can be thought of as media with high salt
Fig. 5. Plot of k2 versus [H2O] for the reaction of MTO with aqueous hydrogen peroxide in various ionic liquids. Data is presented in
Table 2.

224
G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
Table 3
4.3. Thermodynamics
Peroxide binding constants for the MTO/H2O2 reaction in different
solutionsa
The relationship of Abs360 and [H2O2], derived us-
ing the equilibrium expressions for K1 and K2, is [22]:
Solvent
Form of H2O2
K1
K2
H2Ob
30% aqueous H2O2
30% aqueous H2O2
30% aqueous H2O2
30% aqueous H2O2 110
30% aqueous H2O2
UHP
30% aqueous H2O2
UHP
30% aqueous H2O2
UHP
16.1
209
49
132
660
170
140
130
220
82
50
120
210
CH3CNc
Abs
εAK1[H2O2] + εBK1K2[H2O2]2
1 + K1[H2O2] + K1K2[H2O2]2
=
[emim]BF4
[bupy]BF4
[bmim]BF4
[bmim]BF4
[bmim]NO3
[bmim]NO3
[bmim]OTf
[bmim]OTf
8
29
24
28
80
20
5
[Re]T
17
15
10
7
5
10
12
74
30
28
40
34
25
where [Re]T is the total concentration of Re species
in the solution.
A nonlinear least-squares fit of the data using the
above equation gives values for K1 and K2. The equi-
librium constants K1 and K2 determined from the re-
action of MTO with aqueous hydrogen peroxide and
UHP in 9:1 (v/v) solutions of ionic liquid:water and
in molecular solvents are given in Table 3.
The results of the thermodynamics experiments in-
dicate the same general trends as the results from the
kinetics. Ninety percent ionic liquid solutions tend
to fall between water and ACN with respect to the
equilibration of MTO with hydrogen peroxide. As ex-
pected, K2 > K1 for all cases, indicates cooperativity
in peroxide binding [22]. Also, as expected, there is
no appreciable difference in equilibrium constants for
a particular ionic liquid when UHP is used instead of
aqueous hydrogen peroxide.
30
100
a Ionic liquid samples are 10% water by volume. Conditions:
0.10 M HClO4, 1.00 mM MTO, and 5.0–250 mM peroxide at 23 ◦C.
b Ref. [22].
c Ref. [32].
consideration to the contribution of water from aque-
ous hydrogen peroxide, since that contribution should
be constant in all experiments. The data in Table 2
show that the rate constant k2 increases as [H2O] in-
creases. A plot of k2 versus [H2O], Fig. 5, for all of
the data in Table 2 shows that k2 depends linearly on
[H2O]. This result provides strong evidence that the
particular ionic liquids used in this study do not sub-
stantially affect the formation of dpRe; the most im-
portant consideration is the concentration of water.
The influence of water is not quite as dra-
matic for the equilibrium constants as it is for the
Fig. 6. Equilibration of MTO and H2O2. Data for H2O and ACN are simulated based on known values for K1 and K2. Conditions for
ionic liquid experiment: 1.03 mM MTO, 0.10 M HClO4 in [emim]BF4.

G.S. Owens, M.M. Abu-Omar / Journal of Molecular Catalysis A: Chemical 187 (2002) 215–225
225
rate constant k2. The values for the reaction in
vacuum-dried [emim]BF4 are K1 = 110 28 and
K2 = 160 36. When less water is present in the
ionic liquids, the extent of peroxide binding coop-
erativity seems to be reduced, as is the case when
the reactions are done in ACN. The plot shown in
Fig. 6 contrasts the peroxide binding to MTO in ionic
liquids, water, and ACN.
[9] D.E. Kaufmann, M. Nouroozian, H. Henze, Synlett (1996)
1091.
[10] W.A. Herrmann, V.P.W. Bohm, J. Organomet. Chem. 572
(1999) 141.
[11] A.J. Carmichael, M.J. Earle, J.D. Holbrey, P.B. McCormac,
K.R. Seddon, Org. Lett. 1 (1999) 997.
[12] C.E. Song, E.J. Roh, Chem. Commun. (2000) 837.
[13] G.S. Owens, M.M. Abu-Omar, Chem. Commun. (2000) 1165.
[14] C.J. Mathews, P.J. Smith, T. Welton, Chem. Commun. (2000)
1249.
[15] C. de Bellefon, E. Pollet, P. Grenouillet, J. Mol. Catal. 145
(1999) 121.
5. Conclusions
[16] P. Bonhote, A.-P. Dias, N. Papageorgiou, K.
Kalyanasundaram, M. Gratzel, Inorg. Chem. 35 (1996) 1168.
[17] S.N.V.K. Aki, J.F. Brennecke, A. Samanta, Chem. Commun.
(2001) 413.
This work represents the first detailed study of
the kinetics of a transition metal reaction in ionic
liquids. With respect to formation constants and
equilibrium constants of the MTO/H2O2 system, the
water-miscible dialkylimidazolium and alkylpyri-
dinium ionic liquids behave like ACN at low water
concentrations. As [H2O] increases, the liquids be-
have more like aqueous solutions with high salt con-
centrations. More studies of this nature are needed to
determine the viability of ionic liquids as efficient re-
action media. Current efforts in this lab are focused on
determining the rate constants of catalytic reactions
using the MTO/H2O2 system in ionic liquids.
[18] M.J. Muldoon, C.M. Gordon, I.R. Dunkin, J. Chem. Soc.,
Perkin Trans. 2 (2001) 43.
[19] K.R. Seddon, A. Stark, M.-J. Torres, Pure Appl. Chem. 72
(2000) 2275.
[20] A. Noda, K. Hamayizu, M. Watanabe, J. Phys. Chem. B 105
(2001) 4603.
[21] W.A. Herrmann, R.W. Fischer, W. Scherer, M.U. Rauch,
Angew. Chem. Int. Ed. Engl. 32 (1993) 1157.
[22] S. Yamazaki, J.H. Espenson, P. Huston, Inorg. Chem. 32
(1993) 4683.
[23] W.A. Herrmann, R.W. Fischer, D.W. Marz, Angew. Chem.
Int. Ed. Engl. 30 (1991) 1638.
[24] A.M. Al-Ajlouni, J.H. Espenson, J. Am. Chem. Soc. 117
(1995) 9243.
[25] K.A. Vassell, J.H. Espenson, Inorg. Chem. 33 (1994) 5491.
[26] W. Adam, C.M. Mitchell, C.R. Saha-Moller, Tetrahedron 50
(1994) 13121.
[27] S. Yamazaki, Bull. Chem. Soc. Jpn. 69 (1996) 2955.
[28] M.M. Abu-Omar, J.H. Espenson, J. Am. Chem. Soc. 117
(1995) 272.
Acknowledgements
This research was supported by the University of
California Toxic Substances Research and Teach-
ing Program (UCTSR & TP), the Beckman Foun-
dation through a BYI to M.M.A.-O., and the NSF
(CHE-9874857-CAREER).
[29] W. Adam, W.A. Herrmann, J. Lin, C.R. Saha-Moller, J. Org.
Chem. 59 (1994) 8281.
[30] W. Adam, W.A. Herrmann, C.R. Saha-Moller, M. Shimizu,
J. Mol. Catal. 97 (1995) 15.
[31] J. Jacob, J.H. Espenson, Inorg. Chim. Acta 270 (1998) 55.
[32] W.-D. Wang, J.H. Espenson, Inorg. Chem. 36 (1997) 5069.
[33] W.-D. Wang, J.H. Espenson, J. Am. Chem. Soc. 120 (1998)
11335.
References
[34] W.A. Herrmann, F.E. Kuhn, R.W. Fischer, W.R. Thiel, C.C.
Romao, Inorg. Chem. 31 (1992) 4431.
[35] J.D. Holbrey, K.R. Seddon, J. Chem. Soc., Dalton Trans.
(1999) 2133.
[1] T. Welton, Chem. Rev. 99 (1999) 2071.
[2] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 39 (2000)
3772.
[36] J. Robinson, R.A. Osteryoung, J. Am. Chem. Soc. 101 (1979)
[3] M.J. Earle, K.R. Seddon, Pure Appl. Chem. 72 (2000) 1391.
[4] M.M. Abu-Omar, Green Chemistry, 2002 McGraw-Hill
Yearbook of Science and Technology, McGraw-Hill, New
York, 2001, p. 148.
[5] G.S. Owens, M.M. Abu-Omar, Adv. Chem. Ser. 818 (2002),
in press.
[6] Y. Chauvin, L. Mussman, H. Olivier, Angew. Chem. Int. Ed.
Engl. 34 (1995) 2698.
[7] P.A.Z. Suarez, J.E.L. Dullius, S. Einloft, R.F. de Souza, J.
Dupont, Polyhedron 15 (1996) 1217.
323.
[37] J.H. Espenson, O. Pestovsky, P. Huston, S. Staudt, J. Am.
Chem. Soc. 116 (1994) 2869.
[38] A.E. Visser, R.P. Swatloski, W.M. Reichert, S.T. Griffin, R.D.
Rogers, Ind. Eng. Chem. Res. 39 (2000) 3596.
[39] J.H. Espenson, Chemical Kinetics and Reaction Mechanisms,
second ed., McGraw-Hill, New York, 1995, pp. 199–203.
[40] W. Adam, C.M. Mitchell, Angew. Chem. Int. Ed. Engl. 35
(1996) 533.
[41] T.R. Boehlow, C.D. Spilling, Tetrahedron Lett. 37 (1996)
2717.
[8] J.F. Knifton, J. Mol. Catal. 43 (1987) 65.