Deactivation of methylrhenium trioxide - Peroxide catalysts by diverse and competing pathways

Article abstract of DOI:10.1021/ja952305x

The peroxides from methylrhenium trioxide (MTO) and hydrogen peroxide, CH₃ReO₂(η²-O₂), A, and CH₃Re(O)(η²-O₂)₂(H₂O), B, have been fully characterized in both organic and aqueous media by spectroscopic means (NMR and UV-vis). In aqueous solution, the equilibrium constants for their formation are K₁ = 16.1 ± 0.2 L mol^-1 and K₂ = 132 ± 2 L mol^-1 at pH 0, μ = 2.0 M, and 25 °C. In the presence of hydrogen peroxide the catalyst decomposes to methanol and perrhenate ions with a rate that is dependent on [H₂O₂] and [H₃O⁺]. The complex peroxide and pH dependences could be explained by one of two possible pathways: attack of either hydroxide on A or HO₂^- on MTO. The respective second-order rate constants for these reactions which were deduced from comprehensive kinetic treatments are k(A) = (6.2 ± 0.3) x 10⁹ and k(MTO) = (4.1 ± 0.2) x 10⁸ L mol^-1 s^-1 at μ = 0.01 M and 25 °C. The plot of log k(ψ) versus pH for the decomposition reaction is linear with a unit slope in the pH range 1.77-6.50. The diperoxide B decomposes much more slowly to yield O₂ and CH₃ReO₃. This is a minor pathway, however, amounting to <1% of the methanol and perrhenate ions produced from the irreversible deactivation at any given pH. Within the limited precision for this rate constant, it appears to vary linearly with [OH] with k = 3 x 10^-4 s^-1 at pH 3.21, μ = 0.10 M, and 25 °C. Without peroxide, CH₃ReO₃ is stable below pH 7, but decomposes in alkaline aqueous solution to yield CH₄ and ReO₄^-. As a consequence, the decomposition rate rises sharply with [H₂O₂), peaking at the concentration at which [A] is a maximum, and then falling to a much smaller value. Variable-temperature ¹H NMR experiments revealed the presence of a labile coordinated water in B, but supported the anhydride form for A. The peroxides from methylrhenium trioxide (MTO) and hydrogen peroxide, CH₃ReO₂(η²-O₂), A, and CH₃Re(O)(η²-₂)₂(H₂O), B, have been fully characterized in both organic and aqueous media by spectroscopic means (NMR and UV-vis). In aqueous solution, the values of the equilibrium constants for their formation are given. In the presence of hydrogen peroxide the catalyst decomposes to methanol and perrhenate ions with a rate that is dependent on [H₂O₂] and [H₃O⁺]. The complex peroxide and pH dependences could be explained by one of two possible pathways: attack of either hydroxide on A or HO₂^- on MTO. The respective second-order rate constants for these reactions which were deduced from comprehensive kinetic treatments are given. The plot of log k versus pH for the decomposition reaction is linear with the unit slope in the pH range 1.77-6.50. The diperoxide B decomposes much more slowly to yield O₂ and CH₃ReO₃. This is a minor pathway, however, amounting to <1% of the methanol and perrhenate ions produced from the irreversible deactivation at any given pH. Without peroxide CH₃ReO₃ is stable below pH 7, but decomposes in alkaline aqueous solution to yield CH₄ and ReO₄^-. As a consequence, the decomposition rate rises sharply with [H₂O₂], peaking at the concentration at which [A] is a maximum and then falling to a much smaller value. Variable-temperature ¹H NMR experiments revealed the presence of a labile coordinated water in B, but supported the anhydride form for A.

Full text of DOI:10.1021/ja952305x

4966
J. Am. Chem. Soc. 1996, 118, 4966-4974
Deactivation of Methylrhenium Trioxide-Peroxide Catalysts by
Diverse and Competing Pathways
Mahdi M. Abu-Omar, Peter J. Hansen,† and James H. Espenson*
Contribution from the Ames Laboratory and Department of Chemistry, Iowa State UniVersity,
Ames, Iowa 50011
ReceiVed July 13, 1995. ReVised Manuscript ReceiVed February 29, 1996X
Abstract: The peroxides from methylrhenium trioxide (MTO) and hydrogen peroxide, CH3ReO2(η2-O2), A, and
CH3Re(O)(η2-O2)2(H2O), B, have been fully characterized in both organic and aqueous media by spectroscopic means
(NMR and UV-vis). In aqueous solution, the equilibrium constants for their formation are K1 ) 16.1 ( 0.2 L
mol-1 and K2 ) 132 ( 2 L mol-1 at pH 0, µ ) 2.0 M, and 25 °C. In the presence of hydrogen peroxide the catalyst
decomposes to methanol and perrhenate ions with a rate that is dependent on [H2O2] and [H3O+]. The complex
peroxide and pH dependences could be explained by one of two possible pathways: attack of either hydroxide on
-
A or HO2 on MTO. The respective second-order rate constants for these reactions which were deduced from
comprehensive kinetic treatments are kA ) (6.2 ( 0.3) × 109 and kMTO ) (4.1 ( 0.2) × 108 L mol-1 s-1 at µ ) 0.01
M and 25 °C. The plot of log kψ versus pH for the decomposition reaction is linear with a unit slope in the pH range
1.77-6.50. The diperoxide B decomposes much more slowly to yield O2 and CH3ReO3. This is a minor pathway,
however, amounting to <1% of the methanol and perrhenate ions produced from the irreversible deactivation at any
given pH. Within the limited precision for this rate constant, it appears to vary linearly with [OH-] with k ) 3 ×
10-4 s-1 at pH 3.21, µ ) 0.10 M, and 25 °C. Without peroxide, CH3ReO3 is stable below pH 7, but decomposes
in alkaline aqueous solution to yield CH4 and ReO4-. As a consequence, the decomposition rate rises sharply with
[H2O2], peaking at the concentration at which [A] is a maximum, and then falling to a much smaller value. Variable-
temperature 1H NMR experiments revealed the presence of a labile coordinated water in B, but supported the anhydride
form for A.
Introduction
few hours, depending on the peroxide concentration and the
pH. Generally speaking, the system is more stable against
decomposition at high [H3O+]. In organic solvents, however,
B is stable at ambient temperature provided hydrogen peroxide
is in excess. We thus undertook a study of how this catalytic
system decomposes in an aqueous medium. Over time, both
methanol and oxygen can be detected, and both methylrhenium
trioxide and hydrogen peroxide decrease in concentration.
Others have claimed8 that the products form together, and have
described the major decomposition pathway by the following
equation:
Methylrhenium trioxide (CH3ReO3, MTO) catalyzes a broad
spectrum of oxygen atom transfer reactions. It commonly assists
in the transfer of an O atom from hydrogen peroxide to a suitable
substrate, S, as represented by the following generalized
chemical equation:
S + H2O2 cat. MTO8 SO + H2O
(1)
A wide range of substrates can participate in this reaction:
alkenes,1,2 phosphines,3 sulfides,4 metal thiolate complexes,5
bromide ions,6 amines,7 and so on. We and others are studying
how MTO functions as a catalyst. We must also define any
limitations to its use. Two catalytically-active peroxorhenium
complexes, with 1:1 and 2:1 peroxide:rhenium ratios, are
involved. They are CH3Re(O)2(η2-O2) (A) and CH3Re(O)(η2-
O2)2(H2O) (B). We address here not the catalytic reactions per
se, but solvolytic and other reactions that the peroxorhenium
complexes undergo on their own, which lead to deactivation of
MTO.
CH3Re(O)(η2-O2)2(H2O)
9
?8 CH3OH + H[ReO4] + 1/2O2
(2)
We shall see, however, that eq 2 is not a correct representation
of the deactivation processes; oxygen is a minor byproduct
compared to methanol and perrhenate ions.
In the early stages of this investigation it became evident that
an array of methods would be needed to characterize the
chemical events taking place. We have employed several
experimental techniques: 1H NMR, stopped-flow, oxygen-
sensing electrodes, UV-vis spectroscopy, kinetics, and gas
chromatography.
Alone in water and in mixed aqueous-organic solvents, dilute
solutions of CH3ReO3 persist for days with minimal (<5%)
decomposition. When CH3ReO3 and H2O2 are present together,
however, decomposition ensues, usually over the course of a
As it turns out, the decomposition reactions in the MTO-
H2O2 system are of interest in their own right, independent of
their implications for catalyst stability. These reactions speak
to the different modes by which peroxo-metal complexes can
react, a subject that has commanded considerable independent
interest.9,10
† On leave from Northwestern College, Orange City, IA.
X Abstract published in AdVance ACS Abstracts, May 1, 1996.
(1) Herrmann, W. A.; Fischer, R. W.; Marz, D. W. Angew. Chem., Int.
Ed. Engl. 1991, 30, 1638.
(2) Al-Ajlouni, A.; Espenson, J. H. J. Am. Chem. Soc. 1995, 117, 9243.
(3) Abu-Omar, M. M.; Espenson, J. H. J. Am. Chem. Soc. 1995, 117,
272.
(4) Vassell, K. A.; Espenson, J. H. Inorg. Chem. 1994, 33, 5491.
(5) Huston, P.; Espenson, J. H.; Bakac, A. Inorg. Chem. 1993, 32, 4517.
(6) Espenson, J. H.; Pestovsky, O.; Huston, P.; Staudt, S. J. Am. Chem.
Soc. 1994, 116, 2869.
(8) Herrmann, W. A.; Fischer, R. W.; Rauch, M. U.; Scherer, W. J. Mol.
Catal. 1994, 86, 243.
(9) Ledon, H.; Bonnet, M.; Lallemand, J.-Y. J. Chem. Soc., Chem.
Commun. 1979, 702.
(7) Zhu, Z.; Espenson, J. H. J. Org. Chem. 1995, 60, 1326.
S0002-7863(95)02305-5 CCC: $12.00 © 1996 American Chemical Society

DeactiVation of Methylrhenium Trioxide-Peroxide Catalysts
J. Am. Chem. Soc., Vol. 118, No. 21, 1996 4967
In addition, A was further characterized spectroscopically.
We sought to resolve the issue of whether A, like B, has a water
molecule coordinated to rhenium. Also, the equilibrium con-
stants for peroxide binding and the spectrum of A were
determined more precisely.
Experimental Section
Materials. HPLC grade tetrahydrofuran (Fisher) was used, and high-
purity water was obtained by passing laboratory-distilled water through
a Millipore-Q water purification system. Hydrogen peroxide solutions,
prepared by diluting 30% H2O2 (Fisher), were standardized by
iodometric titration. The purity of starting reagents was checked by
1H NMR.
Methylrhenium trioxide was prepared by a literature method11 or
purchased from Aldrich. Stock solutions of CH3ReO3 were prepared
in water, protected from light, and stored at -5 °C. The solutions
were used within 3 days. Their concentrations were determined
spectrophotometrically prior to each use.12 Labeled hydrogen peroxide-
18O was purchased from ICON Services, Inc., as a 1.4% solution in
H216O, and 94% H218O from ISOTEC, Inc.
Instrumentation. 1H NMR spectra were obtained with either a
Nicolet 300 MHz or a Varian VXR-300 spectrometer. NMR spectra
recorded in THF-d8 were referenced to Me4Si. tert-Butyl alcohol (δ
1.27) was used as an internal reference for spectra obtained in D2O.13
For spectrophotometric measurements, conventional (Shimadzu UV-
2101PC and UV-3101PC) and stopped-flow (Sequential DX-17MV,
Applied Photophysics Ltd.) instruments were used with quartz cells (1
or 2 cm optical path length). The pH was measured using a Corning
pH meter, model 320. Molecular oxygen production was measured
using a YSI Model 5300 biological oxygen monitor with a Model 5331
oxygen probe. All measurements were made at 25.0 ( 0.2 °C, and
the instrument was calibrated using air-saturated water assuming [O2]
) 0.27 mM.14 Methane production was measured with an HP 5790A
Series gas chromatograph using an Alltech VZ-10 packed column and
an HP 3390A integrator. Electrospray mass spectra of perrhenate were
obtained using a Finnigan TSQ 700 operated in the single quad mode
for negative ions.
Figure 1. Kinetic traces obtained by NMR spectroscopy for the
formation of A in THF-d8. Conditions: [MTO] ) [H2O2] ) 50 mM at
25.0 °C. The fit for A is based on the model MTO + H2O2 f A f
-
MeOH + H+ + ReO4
.
since the formation of A and B yields water, free or coordinated.
Hence, complete exclusion of water was not attempted; the
amount of water present, however, was limited to that introduced
by the 30% H2O2. Figure 1 shows the time profile for the
formation of A (2.4 ppm) from MTO (2.1 ppm) in THF-d8 from
an equimolar mixture of MTO and H2O2 (50 mM). The curve
for A constitutes a rise-and-fall pattern according to which A
first forms from MTO and H2O2, and then the [A] drops as
MeOH (3.3 ppm) is produced from catalyst degradation. In
THF, the pseudo-first-order rate constant for the formation of
A is (3.7 ( 0.2) × 10-3 s-1, and that for its decomposition is
(3.1 ( 0.3) × 10-4 s-1
.
At high [H2O2], where A forms very rapidly, the reaction of
A with one additional mole of peroxide to form B could also
1
be followed by H NMR. The first-order rate constant for the
formation of B from A in THF-d8 with 40 mM MTO and 0.30
M H2O2 is (2.05 ( 0.04) × 10-3 s-1
.
Is Water Coordinated to A and B? This issue was
examined by variable-temperature 1H NMR. Solutions of H2O2
in THF-d8 were dried over MgSO4. Although fully-dried
peroxide was not obtained, the water concentration was quite
significantly reduced. In one set of conditions, high [H2O2] )
1.0 M and [MTO] ) 0.08 M were used, so that only B would
be present. The temperature was varied from +20 °C to -55
°C. At -40 °C two water peaks were evident, one for free
water at 5.3 ppm and the other at 6.2 ppm for coordinated water.
As shown in Figure 2, the two are well separated. The low
temperature was needed to separate the coordinated water peak
from that of free water, since this water is reasonably labile.
This finding contradicts another report,15 wherein water is said
to be tightly bound to B and not exchanging, with a chemical
shift of 9.4 ppm. We found no chemical shift that high in the
variable-temperature NMR experiments. In general, we find
that the peaks broadened and coalesced as the temperature was
raised. The rate of water exchange proved to be dependent on
the concentration of water, consistent with the following
exchange process:
Results
Identification of A by NMR. The chemical shift of A in
THF-d8 was recently reported to be 3.2 ppm.15 When we mixed
equimolar amounts of MTO and H2O2 (50 mM each) in THF-
d8 and recorded the 1H NMR spectrum, three CH3 signals were
apparent: δ(CH3) ) 2.1 (MTO), 2.4 (A), and 3.3 (CH3OH, from
partial decomposition) ppm. The peak of A at 2.4 ppm
increased with time, as did the methanol signal at 3.3 ppm,
although at different rates. When more hydrogen peroxide was
added to this solution (10× the initial MTO concentration), the
peaks of MTO and A diminished and a new peak emerged at
2.7 ppm, corresponding to the di(η2-peroxo) complex B. The
signal for methanol did not continue to increase once A and
MTO were converted to B by the addition of a sufficiently high
concentration of hydrogen peroxide. We infer that the first-
reported15 chemical shift of A was in fact that of the methanol
formed by partial decomposition.
The rates of formation of A and B in THF depend on the
water concentration; the source of water in this experiment was
the 30% H2O2. Drying H2O2 solutions in THF proved difficult,
CH3Re(O)(η2-O2)2(H2O) + H2O* a
CH3Re(O)(η2-O2)2(H2O*) + H2O (3)
(10) Brown, S. N.; Mayer, J. M. Inorg. Chem. 1992, 31, 4091.
(11) Herrmann, W. A.; Ku¨hn, F. E.; Fischer, R. W.; Thiel, W. R.; Ramao,
C. C. Inorg. Chem. 1992, 31, 4431.
(12) Kunkely, H.; Turk, T.; Teixeira, C.; Bellefon, C. d. M. d.; Herrmann,
W. A.; Volger., A. Organometallics 1991, 10, 2090.
(13) Gordon, A. J.; Ford, R. A. The Chemist’s Companion; Wiley: New
York, 1972.
(14) Fogg, P. C. T.; Gerrard, W. Solubility of Gases in Liquids; Wiley:
New York, 1991; p 293.
A second set of experiments employed a low [H2O2] ) 80
mM and [MTO] ) 80 mM to give a solution containing
predominantly A and MTO and little B. The temperature was
again varied from +20 °C to -55 °C. The separation of a
coordinated water peak for A was not observed. We conclude,
therefore, either that A lacks a coordinated water or that its
(15) Herrmann, W. A.; Fischer, R. W.; Scherer, W.; Rauch, M. H. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1157.

4968 J. Am. Chem. Soc., Vol. 118, No. 21, 1996
Abu-Omar et al.
Scheme 1
are comparable concentrations of all three rhenium species, for
example, 60 mM MTO, 70 mM H2O2, and 1.0 M DClO4 in
1
D2O. The H NMR spectrum was run as soon as practicable
(ca. 4 min after mixing). Along with a substantial amount of
CH3OD (δ 3.4 ppm), the spectrum showed MTO (δ 2.5), A (δ
2.6), and B (δ 3.0) in ratios that agree with those predicted
from the equilibrium constants.
When this experiment was repeated with a large excess of
H2O2, B was the only methylrhenium compound detected, and
little methanol was seen.
Catalyst Decomposition. Kinetics. This system is suf-
ficiently intricate in its pattern of kinetic behavior that it seems
worthwhile to display all the plausible reactions before present-
ing the data from which they were constructed. The reactions
are given in Scheme 1.
The two reactions in Scheme 1 that account for the production
of methanol and perrhenate ions are kinetically indistinguishable,
since they both give rise to an identical transition state, namely,
[CH3ReO5H-]q. The rate equations for this scheme will be
written so as to presume the two peroxide binding steps are
sufficiently rapid that they can be treated as prior equilibria, a
point that we verified independently. It should be noted from
this scheme that the irreversible decomposition of the catalyst
proceeds via the reaction of either A with hydroxide or MTO
with HO2-, whereas B regenerates MTO upon expulsion of
dioxygen. The kinetic equations for Scheme 1 will be derived
here, since we are not aware of a prior derivation. Let Ct
represent at any time t the concentration of all the forms of
rhenium other than perrhenate ions:
1
Figure 2. Variable-temperature H NMR spectra of B (generated in
situ) in THF-d8. Conditions: 80 mM MTO and 1.0 M H2O2. The
asterisk indicates coordinated H2O.
coordinated water is much more labile than that of B. Because
of other results from the pH-dependent kinetics, as presented
in a later section, we are inclined to the first interpretation.
Catalyst Deactivation. General Observations. Compound
B is stable in organic solvents including tetrahydrofuran,
acetonitrile, and acetone, especially with excess peroxide. In
aqueous solutions, however, the peroxides are prone to decom-
pose, although the addition of acid greatly improves their
stability. NMR experiments in THF-d8, similar to the ones
described previously, revealed that, at low peroxide (generally
e[Retot], such that [A] . [B]), methanol is the decomposition
product. Perrhenate ions were detected from the distinctive UV
band at 225 nm after decomposing the interfering hydrogen
peroxide by adding NaOH and heating to 100 °C or by adding
MnO2. Perrhenate ions were also detected by their distinctive
UV absorption without the need to decompose hydrogen
peroxide when [H2O2] was low enough, that is e5 mM. Hence,
methanol and perrhenate ions are formed concurrently by either
one of the following net reactions:
-
[Re]T ) Ct + [ReO4 ]t
(6)
By differentiating this equation, and assuming perrhenate ions
originate from A only, we have
-
d[ReO4 ]
dC
dt
) -
) kA[A][OH-] ) kAfA[OH-]Ct (7)
dt
where fA ) [A]t/Ct is given by
K1[H2O2]
1 + K1[H2O2] + K1K2[H2O2]2
fA )
(8)
Integration of this equation, noting that C0 ) [Re]T, affords
Ct ) [Re]T exp(-kAfA[OH-]t)
(9)
-
CH3Re(O)2(η2-O2) + H2O f CH3OH + H+ + ReO4 (4)
On the other hand, assuming perrhenate ions instead originate
from the reaction of MTO and HO2-, we obtain
-
CH3ReO3 + HO2- f CH3OH + ReO4
(5)
-
d[ReO4 ]
Even under acidic aqueous conditions the same reaction
occurred, albeit more slowly. On the basis of the equilibrium
constants for peroxide binding (K1 and K2 were determined here
as subsequently described), we settled on conditions where there
)
dt
dC
dt
-
-
-
) kMTO[MTO][HO2 ] ) kMTOfMTO[HO2 ]Ct (10)

DeactiVation of Methylrhenium Trioxide-Peroxide Catalysts
J. Am. Chem. Soc., Vol. 118, No. 21, 1996 4969
K1K2[H2O2]2
fB )
(14)
1 + K1[H2O2] + K1K2[H2O2]2
Since kA[OH-] ) kψ/fA, regardless of whether kψ is measured
spectrophotometrically or from O2 buildup, the value of kB′ can
be obtained from the yield of oxygen at the end of the reaction.
Integration of eq 13 and rearrangement afford the expression
fAkA[OH-] [O2]∞
kB′ )
(15)
fB
[Re]T
Similar expressions, of course, could be written using the MTO
-
and HO2 reaction as the decomposition pathway by relating
kA to kMTO as shown in the following relation:
Figure 3. Pseudo-first-order rate constant for the decomposition of
the catalyst as a function of [H2O2]. The fit to eq 9 (or to the integral
kMTOKHa O
2
2
of eq 10) yields kA[OH-] (or kMTOKH O /K1Kw)[OH-]) ) 0.0062 (
2
2
a
kA )
(16)
0.0003 s-1. Conditions: 1.0 mM MTO, 0.010-0.50 M H2O2, and µ )
[HClO4] ) 0.010 M at 25.0 °C.
K1Kw
The overall relationship between k′B and [O2] remains un-
changed. The only difference lies in expressions of rate and
equilibrium constants for the formation of methanol and
perrhenate.
where fMTO ) [MTO]t/Ct is given by
1
fMTO
)
(11)
1 + K1[H2O2] + K1K2[H2O2]2
Oxygen buildup, detected with an oxygen electrode, was used
to monitor the decomposition processes at high [H2O2], 0.2-
0.8 M. These experiments had [Re]T ) 1.00 mM at pH 3.21
(H3PO4/KH2PO4 buffer) and 25 °C. The final O2 concentration
varied linearly with [H2O2], and the plot of kψ shows a
hyperbolic dependence on [H2O2]. These relationships are
depicted in Figure 4. Since the expression for fA at high [H2O2]
reduces to fA ) fA/fB ) 1/K2[H2O2], these forms are expected
from, and are consistent with, Scheme 1. The average value
and [HO2-] expressed relative to [H2O2] is
KHa O [OH ][H2O2]
-
2
2
-
[HO2 ] )
(12)
Kw
Therefore, eqs 7 and 10 differ only in the identities of the
equilibrium and rate constants while maintaining the same
dependences on hydrogen peroxide, hydroxide, and rhenium
concentrations. In summary, both methanol-producing reactions
in Scheme 1 give rise to indistinguishable rate laws as shown
by eqs 7 and 10. This treatment accounts for the main pathway
following first-order kinetics at constant pH and at constant fA
or fMTO, which also implies at constant [H2O2], by virtue of it
being taken in excess.
of kB at pH 3.21 and µ ) 0.10 M is (2.9 ( 0.6) × 10-4 s-1
.
Increasing [H2O2] results in an increased ratio of B to A and
MTO, thus enhancing the relative importance of the O2-
producing reaction of B. At the same time, however, the
destruction of the catalyst is slowed since A decomposes
irreversibly. Nonetheless, decomposition of B is minor (<1%)
compared to the irreversible decomposition leading to methanol
and perrhenate ions. This means that the kinetic data for this
very minor pathway are of considerably lower quality; this point
must be kept in mind in considering the data obtained.
The results of these experiments are given in Table 1, which
includes values of the rate constants that were calculated from
values of K1 and K2 determined at pH 0 (see later). The
peroxide-binding equilibria are not noticeably pH-dependent,
and this treatment should be fully reliable.
The quantitative kinetic study of the decomposition pathways
was carried out spectrophotometrically by monitoring the
absorbance at 360 nm, where A and especially B (ꢀmax ) 1.1
× 103 L mol-1 cm-1) absorb. In an extensive series of
experiments [Re]T was maintained at 1.00 mM and the pH at
2.0 with perchloric acid. The concentration of hydrogen
peroxide was varied in the range 0.007-0.50 M. The data from
each experiment followed first-order kinetics, and the value of
kψ was evaluated from nonlinear least-squares fitting of the
absorbance-time curves to a single exponential function.
A plot of kψ versus [H2O2] was made, as shown in Figure 3.
The rate constant rises from very near zero without peroxide,
and after peaking it falls, following the proportion of A.
According to eqs 7, 8, and 10-12. The fit to eq 9 yielded
kA[OH-] ) (6.2 ( 0.3) × 10-3 s-1 or kMTO[OH-] ) (4.1 (
0.2) × 10-4 s-1 at pH 2.00 and µ ) 0.01 M.
The possibility was considered that methane formation might
accompany the decomposition of B:
CH3Re(O)(η2-O2)2(H2O) f CH4 + H+ + ReO4- + O2 (?)
(17)
Although this would yield stable products, we reject this
alternative on the basis of the results of GC analyses for
methane. Only a trace of CH4 could be detected from a reaction
between MTO (1.0 mM) and H2O2 (0.8 M) at pH 3.21. It
amounted to only 2.2((0.3)% of the amount of O2 formed under
these conditions, after calibration of the GC with pure methane
from the decomposition of MTO in NaOH solutions (see later).
Catalyst Decomposition. pH Profile. The kinetic depen-
dence of the rate of decomposition on [H3O+] was investigated
with [Re]T ) 0.10 mM and at [H2O2] ) 2.00 mM to maximize
the accuracy. At this concentration fA reduces to fA ) K1[H2O2],
As shown in Scheme 1, B decomposes to yield O2. Again
from this scheme the rate of buildup of oxygen, for the moment
using k′B to symbolize the rate constant at a particular pH, is
given by
d[O2]/dt ) k′B[B] ) fBk′BCt ) fBk′B[Re]Te-f k [OH-]t (13)
A
A
2
2
where
and kψ to kA[OH-]K1[H2O2] or to (kMTOKH O /Kw)[OH-]-
a

4970 J. Am. Chem. Soc., Vol. 118, No. 21, 1996
Abu-Omar et al.
Figure 6. [O2]∞ as a function of (K2[Re]T[H2O2])/kA (see eq 18).
Conditions: 1.0 mM MTO, 0.15-1.00 M H2O2, pH 2.6-3.5 (H3PO4/
KH2PO4 buffer), 25.0 ( 0.2 °C, µ ) 0.10 M (+), 0.20 M (O).
The pH profile for decomposition of B is harder to define
with certainty, since so little decomposition leads to oxygen.
The conclusion we have reached is that B probably reacts with
a first-order dependence on [OH-], just like A and MTO do.
Rearrangement of eq 15, after substitution of the expressions
for fA and fB, affords the following relation between [O2]∞ and
[H2O2]:
Figure 4. Pseudo-first-order rate constant from O2 measurements as
a function of [H2O2]. Conditions: 1.0 mM MTO, 0.2-0.8 M H2O2,
pH 3.21, µ ) 0.10 M at 25.0 °C. Inset: Linear relationship between
the amount of O2 produced and the [H2O2] as required by eq 18.
Table 1. Kinetic Results and Yields from the Measurement of O2
Productiona
[H2O2]/
M
kψ/
[O2]∞/
mM
k′B/
(10-3 s-1
)
k′A/s-1
(10-4 s-1
)
fA
fB
k′B K2[Re]T[H2O2]
[O2]∞ )
(18)
0.20
0.30
0.40
0.60
0.80
5.7
3.4
2.7
1.8
1.5
0.035
0.073
0.12
0.17
0.23
0.16
0.14
0.14
0.15
0.16
2.1
2.5
3.2
3.1
3.6
0.036
0.025
0.019
0.013
0.95
0.97
0.98
0.99
[OH-]
kA
The plot suggested by this equation, [O2]∞ versus K2[Re]T-
[H2O2]/kA, where each individual kA value and not the average
was taken, is shown in Figure 6 for data taken at two ionic
strengths, 0.1 and 0.2 M. Each of the plots approximates a
straight line, but the precision is evidently fairly low. Again,
this reflects the fact that oxygen evolution is a quite minor
pathway. At any rate, within the precision of the data, the values
determined over the accessible pH interval (2.6-3.5) fall on a
straight line, indicating that the slope is roughly pH-independent.
For that to be the case, the denominator term in [OH-] must be
0.0094 0.99
av: 0.15 2.9 × 10-4
a Experimental conditions: 1.0 mM MTO, pH 3.21 (H3PO4/KH2PO4
buffer), and µ ) 0.10 M at 25 °C. The fA and fB values were calculated
with K1 ) 16.1 L mol-1 and K2 ) 132 L mol-1
.
canceled by a corresponding term in the numerator, or k′B
)
kB[OH-]. The values of kB are (3.0 ( 0.2) × 107 and (2.0 (
0.1) × 107 L mol-1 s-1, at µ ) 0.10 and 0.20 M, respectively.
Labeling Studies. Since all the kinetic data presented thus
far for the irreversible decomposition reaction could be rational-
-
ized by either the reaction of HO2 upon MTO or that of
hydroxide upon A, we carried out labeling experiments utilizing
18O enriched hydrogen peroxide and water in an effort to
distinguish between the two pathways. The reaction of A with
hydroxide ions would result in the incorporation of two peroxide
oxygens into the final perrhenate ions (eq 20); on the other hand,
-
if MTO reacts with HO2 and a peroxide oxygen is used to
Figure 5. kψ for the irreversible decomposition reaction as a function
of pH. Conditions: 0.10 mM MTO, 2.0 mM H2O2, and µ ) 0.05 M
at 25.0 °C. Inset: kψ versus [H+] (log scale). The fit shown is that to
kψ ) k′/[H+].
form MeOH, then only one peroxide oxygen would remain on
the final perrhenate (eq 19). It is worth noting here that the
peroxide oxygens of A and B in this catalytic system exchange
with neither the oxo ligands on the rhenium nor with water
molecules.
[H2O2] (refer to eqs 7, 8, and 16). For each of 15 experiments
a constant pH was maintained using HClO4 (pH < 3), or one
of a series of buffers, H3PO4/KH2PO4 (pH 2.9-3.8), HOAc/
NaOAc (pH 3.8-5.8), or KH2PO4/K2HPO4 (pH 5.8-6.8). A
constant ionic strength of 0.050 M was maintained with lithium
perchlorate.
CH3ReO3 + H18O18O- 9?8 ReO318O- + CH318OH (19)
CH3ReO2(η2-18O2) + HO- 9?8 ReO218O2- + CH3OH (20)
The pH profile (log kψ versus pH), shown in Figure 5, is
linear with a slope of 0.95 ( 0.01. The inset shows the fitting
of kψ directly to the expression kψ ) k′/[H3O+]. The value of
The perrhenate ions from catalyst decomposition were isolated
as the potassium salt. A 15 mg (60 µmol) sample of MTO
was dissolved in 0.60 mL of H2O16 containing 6 mg (60 µmol)
of KHCO3. A 120 µL sample of 1.5% H218O2 (90% 18O
enrichment) was added and the mixture allowed to stir for a
few minutes to ensure complete decomposition. The presence
of bicarbonate is necessary to neutralize the produced perrhenic
k′ is (9.76 ( 0.06) × 10-7 mol L-1 s-1, and therefore kA
)
3.05 × 109 L mol-1 s-1 and kMTO ) 2.03 × 108 L mol-1 s-1
at µ ) 0.050 M. For now, we simply take this to be the basis
on which the decomposition of the catalyst in Scheme 1 is
written with a direct dependence upon [OH-]. The molecular
basis for this chemistry will be presented in the Discussion.
(16) Murmann, R. K. J. Phys. Chem. 1967, 71, 974.

DeactiVation of Methylrhenium Trioxide-Peroxide Catalysts
J. Am. Chem. Soc., Vol. 118, No. 21, 1996 4971
Scheme 2
acid, since the perrhenate oxygen exchange reaction is very rapid
at acidic pHs.16
The solvent was removed under vacuum and the resulting
KReO4 collected. The KReO4 was dissolved in methanol and
the mass of the anion determined. A similar experiment was
performed by labeling the MTO oxygens with H218O and using
unlabeled H216O2.
Figure 7. Pseudo-first-order rate constants in alkaline solution for the
decomposition of MTO to CH4 and ReO4-. Conditions: 0.20 mM MTO,
0.0050-0.70 M OH-, and µ ) 1.0 M (LiClO4) at 25.0 °C.
The major perrhenate species detected by electrospray mass
spectrometry corresponded to the fully exchanged anion, that
is, Re18O4- in the case of using H218O/H216O2 and Re16O4- when
H216O/H218O2 are used. Evidently, the oxygen exchange
-
between the produced ReO4 and water16 made these labeling
experiments inconclusive with regards to resolving the true
pathway of decomposition.
Decomposition of MTO Itself. MTO is known to decom-
pose to CH4 and ReO4- in alkaline solutions, and was assumed
to proceed via a hydroxide ion adduct.17,18 We monitored the
1
rate of eq 21 by both H NMR and UV-vis methods.
-
CH3ReO3 + OH- f CH4 + ReO4
(21)
Figure 8. Spectra of A and B obtained from SPECFIT fitting shown
alongside that of MTO in the range 280-420 nm.
Addition of NaOH to 50 mM MTO in D2O resulted in an
immediate color change from clear to yellow, followed by gas
evolution over a period of time. The NMR peak of MTO (δ
2.5 ppm) shifted to 1.9 ppm, and then disappeared after several
hours. The UV-vis spectrum of 0.1 mM MTO changed
immediately with 30 mM NaOH. The new yellow compound
displays an absorption maximum at 340 nm (ꢀ ) 1500 L mol-1
cm-1). This lends support to the existence of an intermediate,
as presented in Scheme 2.
The dependence of the experimental reaction rate upon [OH-]
is shown in Figure 7. The experimental conditions were as
follows: 0.20 mM MTO, 0.0050-0.70 M NaOH, and µ ) 1.0
M (LiClO4) at 25.0 °C. The kinetic scheme gives the rate law
Figure 9. Repetitive spectral changes at 60 min intervals for the
formation of A from 1.0 mM MTO and 0.010 M H2O2 in THF at 25.0
( 0.2 °C.
(k5Kw + k6Ka)[OH-]
absorbance readings were fit globally using the program
SPECFIT.19 The known spectra of MTO and H2O2 were entered
and held constant. This fitting gives K1 ) 16.1 ( 0.2 L mol-1
and K2 ) 132 ( 2 L mol-1. The calculated spectra of A and
B are shown in Figure 8.
V )
[Re]T
(22)
Kw + Ka[OH-]
A nonlinear least squares fit gives Ka ) (1.2 ( 0.2) × 10-12
mol L-1 and (k5Kw + k6Ka) ) 2.7 × 10-16 mol L-1 s-1. The
matter cannot be resolved further on the basis of kinetic data
unless one assumes that only the k5 path is important (giving k5
) (2.7 ( 0.3) × 10-2 L mol-1 s-1), or the obverse, in which
case k6 ) (2.23 ( 0.06) × 10-4 s-1. The value of Ka for MTO
at 25 °C and µ ) 1.0 M was confirmed by a spectrophotometric
titration of MTO and NaOH; measured absorbances at 360 and
Timed repetitive scan spectra were also taken at a fixed
[H2O2] ) 10 mM and [MTO] ) 1.0 mM in THF. The overlaid
spectra, Figure 9, feature an isosbestic point at 290 nm,
indicating that no intermediates attain an appreciable concentra-
tion. In THF, A has a maximum at 305 nm (ꢀ ) 730 L mol-1
cm-1 compared to a literature value15 of ꢀ ) 600 L mol-1 cm-1).
390 nm yield an average Ka of 2.2 × 10-12 mol L-1
.
Addition of 0.47 M H2O increased the rate at which A formed.
The decomposition of the catalyst is also accelerated by water.
Equilibrium Constants for Peroxide Coordination. The
equilibrium constants for the formation of A and B in aqueous
solution were determined at 25.0 °C. These are the values of
K1 and K2 from Scheme 1. The solutions contained 1.0 M
HClO4, 1.0 M LiClO4, 0.74 mM MTO, and variable [H2O2],
0.0071-0.22 M. The spectra of 15 such solutions were recorded
in the range 280-420 nm at intervals of 0.2 nm. The
Repetitive scan spectra for the formation of B from MTO
via A show the expected two stages, Figure 10. The concentra-
tions were [MTO] ) 1.0 mM and [H2O2] ) 0.20 M. Following
formation of A, the repetitive spectra show an isosbestic point
for A and B at 340 nm (ꢀ ) 900 L mol-1 cm-1). B itself has
a maximum at 360 nm (ꢀ ) 1.10 × 103 L mol-1 cm-1). This
(17) Herrmann, W. A. Angew. Chem., Int. Ed. Engl. 1988, 27, 1297.
(18) Herrmann, W. A.; Kuchler, J. G.; Weichselbaumer, G.; Herdtweck,
E.; Kiprof, P. J. Organomet. Chem. 1989, 372, 351.
(19) Binstead, R. A.; Zuberbu¨hler, A. D. Specfit; Spectrum Software
Associates, P.O. Box 4494, Chapel Hill, NC 27515.

4972 J. Am. Chem. Soc., Vol. 118, No. 21, 1996
Abu-Omar et al.
Indeed, one of the messages we have presented previously
is that the rate constants for the oxygen-transfer reactions of
the two intermediates A and B are nearly the same. Following,
for example, are values of k3 and k4 for a few typical oxygen-
accepting substrates that suffice to make the point:2-4,6
substrate
k3/(L mol-1 s-1
)
k4/(L mol-1 s-1
)
C6H5SCH3
Br-
PPh3
2.65 × 103
3.5 × 102
7.3 × 105
1.00
0.97 × 103
1.9 × 102
21.6 × 105
0.70
Figure 10. Repetitive spectral changes at 5 min intervals for the
formation of B in THF from MTO via A. The first spectrum does not
coincide with the isosbestic point, since the MTO to A conversion is
still in progress. Conditions: [MTO] ) 1.0 mM, [H2O2] ) 0.20 M,
and 25.0 ( 0.2 °C.
PhCHdCMe2
Clearly, A and B exhibit comparable catalytic reactivities for
a given substrate. That is not to say, of course, that a
comparable concentration of product is produced from A and
from B in any of these transformations. The contribution of
each catalytic transfer step is defined not only by k3 versus k4,
but also by the concentration of hydrogen peroxide and by the
rate constants that characterize the formation of the active
intermediates (k1, k-1, k2, and k-2). A combination of these
quantities determines the proportion of each peroxorhenium
species present in the system and its conversion to product. The
extent of the two O atom transfer processes is determined by
the peroxide concentration and all six rate constants, not simply
k3 and k4. Moreover, the thermodynamic binding constants K1
and K2 are of less relevance than the constituent rate constants,
since the peroxide binding steps are not instantaneous under
catalytic conditions, but are kinetically competitive.
Scheme 3
is close to values we have found in earlier work,20 but
considerably larger than one literature value15 of 700 L mol-1
cm-1
.
One means of gaining insight into the decompositions, where
A and B behave so differently, is to examine what would happen
were the role of each to be interchanged in the respective
processes. For example, were B to release methanol upon attack
by hydroxide ion, the chemical equation would be
Discussion
The results on catalyst deactivation can be interpreted by
means of a relatively small number of chemical reactions. They
are summarized in Scheme 1, which presents all of the reactions
that are of importance in the MTO-H2O2 system. The essence
of the reaction scheme is as follows: MTO and H2O2 form two
species that contain rhenium and peroxide in 1:1 and 1:2 ratios,
A and B. Under the conditions used, the two peroxide-binding
reactions can be treated as rapid prior equilibria, even though
in catalytic systems they are kinetic steps.
CH3Re(O)(O2)2 + OH- N CH3OH + [Re(O)3(O2)]- (23)
This reaction offers an unappealing prospect, however, in
contrast with the corresponding one for A in Scheme 1, since
the trioxo(peroxo)rhenate(VII) ion is not a known or stable
-
species. Indeed, we have found that mixtures of ReO4 and
The irreversible destruction of the catalyst to give methanol
and perrhenate proceeds via compound A or an analogous
rhenium hydroperoxide complex, [CH3Re(η1-OOH)(O3)]-. Com-
pound B, on the other hand, releases molecular oxygen, thereby
restoring MTO, the parent form of the catalyst. That is to say,
decomposition of B does not really destroy the catalyst, although
it does decompose a corresponding amount of peroxide.
Neglected Reactions. The reactions shown in Scheme 1
suffice to account for all the observations. Other chemical
equations can be written; at face value, they might appear to
be as plausible as the ones given, yet they do not participate.
For example, the answer is “no” to the question of whether B,
like A, will release methanol, and “no” again to the question of
whether A, like B, will release oxygen. Put more precisely,
the two alternatives can be demonstrated to be of negligible
importance compared to the ones found. The finding that A
and B undergo distinctive reactions is intriguing, because, when
it comes to their role in catalysis, A and B act as virtual stand-
ins for one another. We have shown that, with few excep-
tions,5,21 each peroxorhenium compound transfers a peroxidic
oxygen to a given substrate with comparable kinetic efficiency.
To illustrate this point, refer to Scheme 3, which gives the
essence of the catalytic cycle.
H2O2 show no indication of peroxo complex formation, whereas
the corresponding combination of CH3ReO3 and H2O2 readily
results in the formation of A. So our contention is that reaction
23 is not important because it would give rise to a thermody-
namically unfavorable product.
For a similar reason, we rationalize the lack of oxygen
evolution from A:
CH3Re(O)2(O2) N CH3ReO2 + O2
(24)
This event, were it to occur, would produce a rhenium(V)
product. While this substance is known, oxygen atom abstrac-
tion from MTO takes a species with a very avid affinity for
oxygen (e.g., PPh3 and H3PO2);22,23 in such cases CH3ReO2 can
be formed slowly.23-25 It seems unlikely, however, that
conversion of peroxide to dioxygen can bring about this
transformation. As to the parallel event that B undergoes (refer
to the kB step in Scheme 1), the bis(peroxo) species upon oxygen
release generates a stable rhenium(VII) product, MTO itself.
That is to say, one peroxide anion of B is released as elemental
oxygen, and the other is converted into a pair of O2- ions that
are automatically incorporated into the newly-formed MTO. This
(22) Holm, R. H.; Donahue, J. P. Polyhedron 1993, 12, 571-589.
(23) Abu-Omar, M. M.; Espenson, J. H. Inorg. Chem. 1995, 34, 6239.
(24) Herrmann, W. A.; Roesky, P. W.; Wang, M.; Scherer, W. Orga-
nometallics 1994, 13, 4531.
(20) Yamazaki, S.; Espenson, J. H.; Huston, P. Inorg. Chem. 1993, 32,
4683.
(21) One exception appears to be the cobalt thiolate complex (en)2Co-
(SCH2CH2NH2)2+, for which k3 . k4.
(25) Zhu, Z.; Espenson, J. H. J. Mol. Catal. 1995, 103, 87.

DeactiVation of Methylrhenium Trioxide-Peroxide Catalysts
J. Am. Chem. Soc., Vol. 118, No. 21, 1996 4973
is how the more energetic Re(V) is avoided when B evolves
oxygen. If A, however, were to disproportionate H2O2 like B
does, it would necessarily require a bimolecular pathway, which
would be unlikely considering the low concentrations of A at
equilibrium.
The process for oxygen release has just been described in
somewhat simplistic terms, with the goal in mind of rationalizing
the absence of two decomposition pathways. In fact, there are
some very fundamental matters that need to be addressed about
the reactions that do take place. The timing of the events for
oxygen release from B and the chemical mechanism by which
MTO is released are examples. The balance of the discussion
focuses on these events and on plausible mechanisms by which
methanol and perrhenate are released.
implausible formulation, save for the fact that eq 26 does not
seem to represent particularly well the molecular changes that
might lead to methanol.
Returning to the first formulation, in which OH- attacks
nucleophilically at the methyl group, either we envisage that
this step leads directly to the products, in which case the requisite
electronic rearrangement of the η2-peroxide group is presumed
to be relatively rapid, eq 27, or we invoke a sequential process
in which a first-formed peroxorhenium(V) intermediate yields
the final products after internal electron shifts, eq 28.
Oxygen Evolution from B. The evolution of oxygen upon
decomposition of B to MTO is one of the reactions shown in
Scheme 1. In practice the yield of O2 is rather low compared
to [B]0, the reason being the relative values of k′A (k′MTO) and
k′B. The decomposition of B occurs relatively slowly. The
electrochemical determination of O2 buildup showed that [O2]∞
increased linearly with [H2O2]0. This substantiates that B is
the source of oxygen, as given by the kinetic scheme and eq 18
once the peroxide dependence of fA (or fMTO) is taken into
account. The experimental rate constants obtained for O2 release
at various concentrations of [H2O2] decreased hyperbolically
as [H2O2] increased; see Figure 4. This agrees with eq 13 and
with the spectrophotometric kinetic measurements, Figure 3.
Precedents for oxygen release from η2-peroxides are not very
common. A bis(η2-peroxo)(porphyrin)molybdenum(VI) is pho-
tochemically converted to oxygen and the cis-(dioxo)(porphy-
rin)molybdenum(VI).9 Although the mechanism has not been
established, an intermediate MoIV-peroxo complex has been
proposed.26 This is analogous to the mode of decomposition
by which we propose B reverts to the parent MTO.
On the other hand, the rhenium(V)-peroxo complex (HBpz3)-
Re(O)(O2) decomposes to (HBpz3)ReO3 bimolecularly, rather
than losing O2 directly.10 The authors attribute the failure of
dioxygen release from a single site, despite a substantial driving
force, to a symmetry restriction against the interconversion of
a d2 metal peroxide to a d0 metal dioxo species. If this
symmetry-derived restriction is general to all d2 complexes, the
decomposition of B to MTO and O2 would proceed via a
concerted mechanism rather than a stepwise transformation that
involves a rhenium(V) peroxo intermediate.
(27)
(28)
The alternative route for production of methanol and perrhe-
nate ions involves the attack of HO2- on MTO. This proceeds
most likely via a rhenium-hydroperoxo complex that is
analogous to compound A, eq 29. Following its coordination
(29)
to the rhenium(VII) center, the hydroperoxide ligand becomes
electrophilically activated and more susceptible to attack by the
methyl group bound to rhenium. The outcome of this rearrange-
ment (as illustrated by eq 29) is the elimination of methanol,
leaving behind perrhenate as the final rhenium product.
Although neither kinetics nor labeling experiments were
successful in discriminating between mechanisms 27 and 28
on the one hand and mechanism 29 on the other, a comment
on which pathway would be more chemically sound based on
our present knowledge is in order at this stage. Since kA and
kMTO are close to the diffusion limit and the hydroperoxide anion
is notably more nucleophilic than hydroxide,27 one expects to
see diffusion-controlled peroxide-induced rather than hydroxide-
induced decomposition. Therefore, in light of these consider-
ations and the fact that dilute MTO solutions without H2O2
persist for days in aqueous solution, we are inclined to favor
the mechanism shown in eq 29 as the route by which the catalyst
is irreversibly deactivated.
Methanol Formation. Since the two possible pathways for
the production of methanol (Scheme 1) cannot be resolved
further on the basis of kinetics, and the labeling studies were
inconclusive due to the oxygen exchange of perrhenate, we will
discuss the molecular mechanisms by which each pathway may
proceed. The decomposition of A may occur by the direct attack
of OH- on the methyl group of A, in effect, an SN2 displace-
ment, eq 27 or 28. Alternatively, a species with OH-
coordinated to rhenium may be responsible, which is represented
by the following steps:
A precedent relating to the decomposition of MTO is the
formation of (ButO)2Mo(O)2 from the reaction of Mo(OBut)4
with molecular oxygen in dilute solutions.28 In concentrated
solution, however, the product is (ButO)4MoO, from bimolecular
oxygen activation. The dilute reaction appears to be unimo-
lecular, as in eq 30.
A
Ka
A + H2O y z A-OH- + H+
(25)
(26)
k25
-
A-OH- 98 CH3OH + ReO4
A
A
with kA ) k25Ka /Kw, which affords k25Ka ) 6.2 × 10-5 mol
L-1 s-1. If we approximate KaA as equal to KaMTO (1.2 × 10-12
mol L-1; see eq 21), then k25 ) 5 × 107) s-1. This is not an
(30)
(27) Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry, 3rd
ed.; Plenum Press: New York, 1990; Part A, p 268.
(28) Chisholm, M. H.; Folting, K.; Huffman, J. C.; Kirkpatrick, C. C.
Inorg. Chem. 1984, 23, 1021.
(26) Ledon, H. J.; Bonnet, M.; Gillard, D. J. Am. Chem. Soc. 1981, 103,
6209.

4974 J. Am. Chem. Soc., Vol. 118, No. 21, 1996
Abu-Omar et al.
Table 2. Summary of Equilibrium and Rate Constants at 25 °C
reaction
symbol
value
CH3ReO3 + H2O2 a CH3Re(O)2(O2)(A) + H2O
A + H2O2 a CH3Re(O)(O2)2(OH2)(B)
K1
16.1 ( 0.2 L mol-1
K2
132 ( 2 L mol-1
CH3ReO3 + H2O a CH3ReO3(OH)- + H+
KaMTO
k5
(1.2 ( 0.2) × 10-12 mol L-1
(2.7 ( 0.3) × 10-2 L mol-1 s-1
(2.2 ( 0.1) × 10-4 s-1
a
CH3ReO3 + OH- f ReO4- + CH4
a
CH3ReO3(OH)- f ReO4- + CH4
k6
kA
kMTO
kB
- b
A + OH- f CH3OH + ReO4
(6.2 ( 0.2) × 109 L mol-1 s-1
(4.1 ( 0.1) × 108 L mol-1 s-1
2 × 107 L mol-1 s-1
CH3ReO3 + HO2- f ReO4- + CH3OHb
B + OH- f CH3ReO3 + O2c + OH-
a The reaction represented by k5 is written in an alternative form represented by k6. b µ ) 0.010 M and 25 °C. c pH 3.21, µ ) 0.10 M, and 25
°C.
Scheme 4
Scheme 5
The spectrum of A could be obtained independently in THF,
where the formation reactions for A and B are much slower
than in water. Although the spectrum of A in water from
SPECFIT fitting does not match exactly with that measured in
THF, they are in reasonable agreement considering the proce-
dures and fitting errors.
We contend that the monoperoxorhenium compound A was
incorrectly identified in the NMR spectrum previously.15 It
seems likely to us that the reported chemical shift can be
attributed to methanol, a decomposition product.
The rates at which methanol is produced and B decomposed
are very sensitive to the solvent and pH. In both aqueous and
semiaqueous environments, these compounds are stabilized by
acid. In neat organic solvents the peroxides form more slowly
than in water, and their temporal stabilities are greatly enhanced.
For example, at pH 2 in water, the experimental rate constant
is 6.2 × 10-3 s-1 whereas in THF it is 3.1 × 10-4 s-1. As a
result B is stable in the presence of excess peroxide for days in
organic solvents, whereas 1 M acid is required to stabilize B
comparably in water.
Any mechanism that would involve deprotonation of the
methyl group is disregarded because deuterium incorporation
into MTO does not occur in D2O, Scheme 4. The lack of
hydrogen exchange for deterium in D2O deems the carbene
tautomer of MTO insignificant.
Methane Formation. The decomposition of (peroxide-free)
CH3ReO3 in alkaline solutions produces methane quantitatively.
The kinetic and spectroscopic data show that an intermediate
is “instantly” formed between MTO and hydroxide ions. This
is reasonably a coordination compound, analogous to the adducts
formed between MTO and chloride ions and nitrogen bases.29
The short sequence of subsequent events that leads to methane
formation can easily be depicted, as in eq 31.
(31)
Summary. The important reactions are shown together in
Scheme 5, and the rate and equilibrium constants are given in
Table 2.
Formation of A and B. The equilibrium constants K1 and
K2 were previously determined from absorbance readings at
selected wavelengths.6,20 The redeterminations given here were
based on a global fitting routine19 that in effect uses the
continuous spectra. This treatment not only resulted in improved
values and statistics for the equilibrium constants, but was also
able to resolve the spectrum of A. Except at the lowest peroxide
concentrations, the equilibrium situation is such that A in water
is present at relatively low levels compared to MTO and B.
Acknowledgment. This research was supported by the U.S.
Department of Energy, Office of Basic Energy Sciences,
Division of Chemical Sciences, under Contract No. W-7405-
Eng-82. P.J.H. is indebted to the administration and Board of
Trustees of Northwestern College (Iowa) for their award of
sabbatical leave. We are grateful to Kamel Harrata for the mass
spectra determinations. J.H.E. is grateful to S. N. Brown and
J. E. Bercaw for helpful comments.
(29) Herrmann, W. A.; Kuchler, J. G.; Kiprof, P.; Riede, J. J. Organomet.
Chem. 1990, 395, 55.
JA952305X