Thermal and Photochemical Reactions of Methylrhenium Diperoxide: Formation of Methyl Hydroperoxide in Acetonitrile

Article abstract of DOI:10.1021/ic970650c

Compared to the system in aqueous solution, the equilibration reactions in acetonitrile between MTO and the methylrhenium peroxides CH₃ReO₂(η²-O₂) (A) and CH₃ReO(η²-O₂)₂(H₂O) (B) are slower but more favored thermodynamically. In CH₃CN, small concentrations of water facilitate the formation of A (especially) and B. These species decompose to methyl hydroperoxide and perrhenic acid in CD₃CN, rather than to methanol and perrhenic acid as in aqueous solution. The proposed mechanism involves the intramolecular migration of the methyl group to a peroxo oxygen, followed by hydrolysis, and it is facilitated by photolysis. The potential use of B as photocatalyst does not seem promising, however.

Full text of DOI:10.1021/ic970650c

Inorg. Chem. 1997, 36, 5069-5075
5069
Thermal and Photochemical Reactions of Methylrhenium Diperoxide: Formation of Methyl
Hydroperoxide in Acetonitrile
Wei-Dong Wang and James H. Espenson*
Department of Chemistry, Iowa State University, Ames, Iowa 50011
ReceiVed May 30, 1997X
Compared to the system in aqueous solution, the equilibration reactions in acetonitrile between MTO and the
methylrhenium peroxides CH3ReO2(η2-O2) (A) and CH3ReO(η2-O2)2(H2O) (B) are slower but more favored
thermodynamically. In CH3CN, small concentrations of water facilitate the formation of A (especially) and B.
These species decompose to methyl hydroperoxide and perrhenic acid in CD3CN, rather than to methanol and
perrhenic acid as in aqueous solution. The proposed mechanism involves the intramolecular migration of the
methyl group to a peroxo oxygen, followed by hydrolysis, and it is facilitated by photolysis. The potential use
of B as photocatalyst does not seem promising, however.
Introduction
an example, illustrates the versatility of this catalytic oxo-transfer
system. Kinetic studies have revealed a common mechanism
for those oxygen atom transfer reactions which involves the
nucleophilic attack by a given substrate X on a peroxide oxygen
of A and B.6,12,14
Transition metal peroxo complexes, particularly those con-
taining d0 metals such as TiIV, VV, CrVI, MoVI, and WVI, play
an important role in catalytic oxo-transfer reactions.1,2 Recently,
attention has especially been devoted to another d0 compound,
methylrhenium trioxide (CH3ReO3, abbreviated as MTO); with
H2O2 this system exhibits high reactivity and catalytic activity.
MTO has been shown to activate H2O2 by formation of two
peroxo compounds, CH3ReO2(η2-O2) (A) and CH3ReO(η2-O2)2-
(H2O) (B), eq 1.3,4
The thermal reactions of transition metal peroxides have been
studied extensively with only limited reports of photochemical
reactions. The photochemistry of organometallic oxides and
peroxides containing d0 transition metals remains in a formative
stage.16-19 It has been shown that LnCoIII(µ-η1:η1-O2)CoIIILn
complexes undergo intramolecular ligand (peroxo) to metal
charge-transfer reactions upon photolysis, resulting in dissocia-
tion and decomposition.20 Photolysis of complexes of the type
LnM(η2-O2)1-2 usually generates molecular oxygen, in certain
cases in its electronically excited singlet state.19-24 Although
certain photochemical studies of MTO and B have ap-
peared,16,17,19 we have investigated the photochemistry of B,
which absorbs at 300-450 nm. Studies of the thermal equi-
librium and decomposition of the MTO-H2O2 system in
acetonitrile and of the photochemical reactions of B have been
carried out to learn the reactions occurring and to define their
mechanisms.
The coordinated peroxide groups can transfer an oxygen atom
to a wide range of substrates, such as alkenes,5,6 halides,7,8
organonitrogen species,9-11 phosphines,12 and several classes
of sulfur-containing compounds.13-15 Equation 2, using A as
X Abstract published in AdVance ACS Abstracts, October 1, 1997.
(1) Kitajima, N.; Akita, M.; Moro-oka, Y. Transition metal peroxides and
related compounds. In Organic Peroxides; John Wiley & Sons:
Chichester, 1992; p 535.
(2) Conte, V.; Di Furia, F.; Modena, G. Transition metal catalyzed
oxidation. The role of peroxometal complexes. In Organic Peroxides;
John Wiley & Sons: Chichester, 1992; p 559.
(3) Herrmann, W. A.; Fischer, R. W.; Scherer, W.; Rauch, M. U. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1157.
(4) Yamazaki, S.; Espenson, J. H.; Huston, P. Inorg. Chem. 1993, 32,
4683.
(5) Herrmann, W. A.; Fischer, R. W.; Marz, D. W. Angew. Chem., Int.
Ed. Engl. 1991, 30, 1638.
(6) Al-Ajlouni, A. M.; Espenson, J. H. J. Am. Chem. Soc. 1995, 117,
9243.
(7) Espenson, J. H.; Pestovsky, O.; Houston, P.; Staudt, S. J. Am. Chem.
Soc. 1994, 116, 2869.
(8) Hansen, P. J.; Espenson, J. H. Inorg. Chem. 1995, 34, 5839.
(9) Zhu, Z.; Espenson, J. H. J. Org. Chem. 1995, 60, 1326.
(10) Murray, R. W.; Iyanar, K.; Chen, J.; Waring, J. T. Tetrahedron Lett.
1996, 37, 805.
(16) Kunkely, H.; Turk, T.; Teixeira, C.; Bellefon, C. M.; Herrmann, W.
A.; Vogler, A. Organometallics 1991, 10, 2090.
(17) Herrmann, W. A.; Ku¨hn, F. E.; Fiedler, D. A.; Mattner, M. R.;
Geisberger, M. R.; Kunkely, H.; Vogler, A.; Steenken, S. Organo-
metallics 1995, 14, 5377.
(18) Kunkely, H.; Vogler, A. J. Photochem. Photobiol. A: Chem. 1996,
94, 135.
(19) Hatzopoulos, I.; Brauer, H.-D.; Geisberger, M.; Herrmann, W. A. J.
Organomet. Chem. 1996, 520, 201.
(20) Ledon, H.; Bonnet, M. J. Chem. Soc., Chem. Commun. 1979, 702.
(21) Vogler, A.; Kunkely, H. J. Am. Chem. Soc. 1981, 103, 6222.
(22) Seip, M.; Brauer, H.-D. J. Photochem. Photobiol. A: Chem. 1993,
76, 1.
(11) Goti, A.; Nannelli, L. Tetrahedron Lett. 1996, 37, 6025.
(12) Abu-Omar, M. M.; Espenson, J. H. J. Am. Chem. Soc. 1995, 117,
272.
(13) Huston, P.; Espenson, J. H.; Bakac, A. Inorg. Chem. 1993, 32, 4517.
(14) Vassell, K. A.; Espenson, J. H. Inorg. Chem. 1994, 33, 5491.
(15) Adam, W.; Mitchell, C. M.; Saha-Moller, C. R. Tetrahedron 1994,
50, 13121.
(23) Hoshino, M.; Yamamoto, K.; Lillis, J. P.; Chijimatsu, T.; Uzawa, J.
Inorg. Chem. 1993, 32, 5002.
(24) Kwong, D. W. J.; Chan, O. Y.; Wong, R. N. S.; Musser, S. M.; Vaca,
L.; Chan, S. L. Inorg. Chem. 1997, 36, 1276.
S0020-1669(97)00650-2 CCC: $14.00 © 1997 American Chemical Society

5070 Inorganic Chemistry, Vol. 36, No. 22, 1997
Wang and Espenson
Experimental Section
Materials. Methylrhenium trioxide, CH3ReO3 (Aldrich), H2O2
(30%, Fisher, standardized by iodometric titration), CF3COOH (Ald-
rich), HClO4 (Fisher), PPh3 (Aldrich), (CH3O)2SO2 (Kodak), KOH
(Fisher), CH3CN (Fisher), CD3CN (CIL), and CD2Cl2 (CIL) were used
as received. The compound CH3OReO3 was prepared according to
literature procedure.25 Its 1H-NMR resonance in CD3CN at room
temperature was broad. The identity of CH3OReO3 was further assured
by converting it to a more thermally stable chelated species by adding
(CH3)2NCH2CH2N(CH3)2 to CH3OReO3 in THF. High-purity H2O was
obtained by passing laboratory distilled water through a Millipore-Q
water purification system.
Kinetic Studies. Kinetic data for the formation of A and B in CH3-
CN were collected by following the absorbance changes at 320 or 360
nm, the maxima for A (ꢀ320 ) 700 L mol-1 cm-1) and B (ꢀ360 ) 1200
L mol-1 cm-1). Shimadzu UV-2101PC or UV-3101PC spectropho-
tometers equipped with a thermostated cell holder were used. The
temperature was controlled at 25.0 ( 0.2 °C, unless otherwise indicated.
The reaction occurs in two well-separated stages (kf > ks), which
allowed us to use a biexponential function, eq 3, to fit the biphasic
kinetic traces. In eq 3, kf and ks represent the pseudo-first-order rate
Figure 1. Absorbance-time profiles of the CH3ReO3 (1.0 mM)-H2O2
system in CH3CN containing 2.0 M H2O at 25 °C. Data are shown
with 36 mM H2O2 at 320 nm, the maximum for A, and with 100 mM
H2O2 at 360 nm, the maximum for B.
the yellow color of B had faded, argon was passed through the solution
and then into basic pyrogallol.27 A similar experiment was performed
for the MTO-H2O2 solution in the dark. A UV-vis spectrophotometer
was used to quantify the relative amount of O2 produced. To verify
this method, argon was passed through an air- and an oxygen-saturated
CH3CN solution and then into the pyrogallol. Attempts to quantitatively
measure O2 with an oxygen electrode were not successful owing to
the nature of the solvent and the oxidizing strength of MTO-H2O2.
Product Analysis. The reaction products were characterized by the
use of a Bruker DRX-400 spectrometer. 1H and 31P{1H} NMR spectra
were recorded at 400.13 and 161.98 MHz, respectively. 1H chemical
shifts were measured relative to residual 1H resonance in the deuterated
solvents, CHCl3 (δ ) 7.26 ppm), CDHCl2 (δ ) 5.32 ppm), CD2HCN
(δ ) 1.94 ppm). The 1H-NMR spectra of most of the organic products
in this study could be compared with the 1H-NMR spectra of authentic
samples.
Abst ) Abs∞ + Re-k t + âe-k t
(3)
f
s
constants for the two stages. The amplitude constants R and â are
given by26
(ꢀA - ꢀMTO)kf - (ꢀMTO - ꢀB)k
R )
s[MTO]0
ks - kf
(ꢀB - ꢀA)k
ks - kf
â )
f[MTO]0
(4)
Results
In some cases an insufficient number of data points were collected
in the first component. For them, ks was obtained by a first-order fit
of the latter portion after removing 10 half-lives of the faster component.
Then kf was obtained by inputting that ks as a constant into eq 3 for
the fit of the early stages of the data.
The linear plots of the pseudo-first-order rate constants kf or ks vs
[H2O2] have no intercept, indicating that contributions of the reverse
reactions of eq 1 are small, as a result of the equilibrium positions
favoring A and then B at the concentrations used.
Formation of A and B. As in THF,3,28 A in CH3CN has a
maximum at 320 nm. A kinetics experiment monitored at 320
nm, Figure 1, shows two well-separated stages. The first part
with increasing absorbance corresponds to the buildup of A
whereas the decreasing component results from the transforma-
tion of A to B.29 Data at 360 nm, which represents the buildup
of B, is also shown in Figure 1. The biphasic kinetics were
analyzed according to eq 3.
For the fast stage, with [MTO]0 ) 1.0 mM and the water
concentration kept constant at 2.0 M, the plots of kf and ks vs
[H2O2] define lines that pass through the origin; the slopes are
k1 ) 0.81 ( 0.04 L mol-1 s-1 and k2 ) 0.045 ( 0.002 L mol-1
s-1. Then, the water content was varied to 2.0 M, with [MTO]0
) 1.0 mM and [H2O2] ) 31 mM; kf (especially) and ks rise
steeply with [H2O], Figure 2. Certain experiments were
designed to evaluate the rate constant of the slow stage by
focusing on data collection at longer times. Such data were
therefore less well suited to evaluate the rate constant of the
fast stage, as insufficient data were obtained; thus scattered
values were obtained. The scatter was particularly pronounced
when the concentration of H2O is high, such that there were
not enough data points for the fast stage.
Equilibrium constants were determined by integrating the 1H
spectrum before and during the decomposition process. The decom-
position kinetics of B in CD3CN were obtained with 1H-NMR and UV-
vis methods. In the NMR measurements H2O2 was in g20-fold excess
over [MTO]0. The disappearance of B was also followed spectropho-
tometrically with [MTO]0 e 1 mM. With both techniques the rate
constants are the same at a given concentration of H2O2. The activation
parameters of the decomposition processes were obtained from kinetic
data in CH3CN between 25.0 and 59.3 °C.
Photolysis. To prevent the direct photolysis of H2O2 itself, light (λ
> 400 nm) from a 300 W sunlamp was obtained by using two KV-
399 filters (Schott, Germany). The temperature was controlled by the
use of a small submersible circulating pump with tap water for room-
temperature photolysis and with salt-ice water for low temperature (0
to -5 °C) photolysis. Most of the photochemical experiments were
carried out in NMR tubes. To minimize the interference from MTO
and especially from A, the MTO-H2O2-substrate solution was kept
in the dark for about 30 min to allow for all of the MTO and A to be
At fixed concentrations of hydrogen peroxide and MTO, 88
and 1.0 mM, the values of k1 and k2 were determined over the
temperature range 25-59.6 °C. The activation parameters are
as follows:
1
converted to B, as verified from the H-NMR spectrum. Then this
solution was photolyzed along with the control samples, MTO-
substrate and H2O2-substrate. Meanwhile, the thermal reaction of
MTO-H2O2-substrate was carried out in the dark. All of the samples
1
(27) Gordon, A. J.; Ford, R. A. The Chemist’s CompanionsA Handbook
of Practical Data, Techniques, and References; John Wiley & Sons:
New York, 1972.
(28) Abu-Omar, M. M.; Hansen, P. J.; Espenson, J. H. J. Am. Chem. Soc.
1996, 118, 4966.
were monitored by the use of H-NMR spectroscopy.
To determine that O2 is generated upon photolysis of B, a larger
scale (5 mL) reaction was carried out in a sealed Pyrex flask. After
(25) Edeards, P.; Wilkinson, G. J. Chem. Soc., Dalton Trans. 1984, 2695.
(26) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms; 2nd
ed.; McGraw-Hill: New York, 1995.
(29) We recognize that a rising and falling absorbance will be observed
irrespective of which rate constant is the larger; the assignment
indicated here is based on additional evidence.

Reactions of Methylrhenium Diperoxide
Inorganic Chemistry, Vol. 36, No. 22, 1997 5071
Table 1. Summary of the Rate and Equilibrium Constants for Reaction 1 between CH3ReO3 and H2O2 in Different Solvents at 25 °C
in H2Oa
CH3CN:H2Oc (1:1)
in CH3OHd
261
in CH3CNe + 2.6 M H2O
k1/L mol-1 s-1
k-1/s-1
80
10
32.5
3.0
13
0.81 ( 0.04
(3.9 ( 0.22) × 10-3
209 ( 6
K1/L mol-1
7.7 (µ ) 0.10)
16.1 (µ ) 2.0)b
k2/L mol-1 s-1
k-2/s-1
5.2
1.05
0.008
136
0.045 ( 0.002
(6.8 ( 0.72) × 10-5
660 ( 64
0.04
K2/L mol-1
145 (µ ) 0.10)a
814
132 (µ ) 2.0)b
a pH ) 1.0, ref 4. b pH 0, µ ) 2.0 M, ref 28. c pH ) 1.0, ref 12. d Reference 9. e This work; k-1 and k-2 were calculated from K1 ) k1/k-1 and
K2 ) k2/k-2
.
Figure 2. The effect of H2O on the rate constants kf and ks from the
biexponential fits to the absorbance-time data. To include data at
varying peroxide concentrations, the rate constants were divided by
[H2O2].
Figure 3. A spectrophotometric titration with 0.012 M 2,4,6-collidine
solution in CH3CN. The perrhenic acid was titrated in a spent solution
that initially contained 1.0 mM CH3OReO3 and 0.088 M H2O2, and
which was diluted 8-fold before the titration.
∆Sq/J mol-1 K-1
∆Hq/kJ mol-1
k298/L mol-1 s-1
[A]
[B]
K1 )
K2 )
(6)
k1
k2
-212 ( 6
-214 ( 2
24.5 ( 1.9
29.0 ( 0.7
0.81 ( 0.04
0.045 ( 0.002
[CH3ReO3][H2O2]
[A][H2O2]
Formation of HOReO3. The photochemical and thermal
decomposition of B forms perrhenic acid, easily recognized in
the UV spectrum from the unique fine structure of ReO4-. It
was particularly evident after the excess peroxide had been
decomposed with a small amount of MnO2. The amount of
HOReO3 was determined by spectrophotometric titration with
2,4,6-collidine. Aliquots of samples first used for UV-vis and
NMR kinetics were diluted and then titrated with a collidine
solution in CH3CN. The amount of acid from the titration
corresponds to the initial MTO concentration, allowing us to
conclude that perrhenic acid is the only rhenium product. A
typical titration graph is given in Figure 3. The method was
verified with an authentic HOReO3 sample.
Formation of CH3OOH and Other Organic Products. In
addition to B and the signal at δ 3.26 ppm (CH3OH), a singlet
appeared at δ 3.75 ppm in the 1H-NMR upon photolyzing B in
CD3CN for 2 h. Unlike the photolysis reaction, the thermal
decomposition of B in CD3CN at room temperature was very
slow and the predominant product appeared at δ 3.75 ppm,
which does not correspond to CH3OH. This new species has a
13C signal at δ 64.7 ppm, compared with the CH3OH, δ 49.5
ppm. These two compounds account for >99% of the decom-
position products in that the sum of their integrals nearly equals
that of the initial CH3ReO3. Figure 4 displays these results.
Acidifying the spent solution of B with DClO4 or CF3CO2H
did not affect the ratio of these two compounds in weeks. The
δ 3.75 ppm species remained unchanged for hours after addition
of a ca. 20-fold excess of 2,4,6-collidine, showing that the
HOReO3 measured by titration does not arise from the hydroly-
sis of other rhenium products. Introducing CH3OCH3 gave rise
to a new singlet at δ 3.24 ppm. The unidentified product is
clearly not CH3OCH3. The spectroscopic data are, however,
in agreement with methyl perrhenate, CH3OReO3, or a deriva-
Equilibrium Constants. In MTO-H2O2 solutions in ac-
etonitrile-d3, the individual methylrhenium species (MTO, A,
and B) could be detected and measured with H-NMR. The
1
methyl resonance of A is shifted upfield of MTO in CD3CN,
the reverse of their positions in THF-d8 or CD3OD. In accord
with eq 1, the concentration-time values for A and B were
fitted to biexponential kinetics and [MTO] to first-order kinetics.
During the first 35 min of a reaction with [H2O2]0 ) 140
mM, [MTO]0 ) 13 mM, and [H2O] ) 2.6 M (2.1 M added and
0.5 M from hydrogen peroxide), [A], after a quite rapid increase,
then declined from 8.9 to 0.30 mM and [MTO] from 0.40 to
0.012 mM. K1 was evaluated at each point, the average being
K1 ) 209 ( 6 L mol-1. This constancy confirms that reaction
1 had attained equilibrium before the first reading; as B
decomposed, the equilibrium was maintained. At [H2O2]0 )
73 mM, K1 ) 249 ( 16 L mol-1, the same within the error. A
similar procedure was applied to [A] and [B] values at longer
times, 50-5660 min, to calculate K2 ) 660 ( 64 L mol-1. At
other peroxide concentrations the K2 values are 690 (290 mM
H2O2) and 730 L mol-1 (420 mM), seemingly the same within
the error. The equilibrium concentration of peroxide was
calculated from the mass balance relation in eq 5 that explicitly
recognizes its diminution from the formation of methyl hydro-
peroxide.
[H2O2]eq ) [H2O2]0 - [A] - 2[B] - 2[CH3OOH] (5)
The expressions in eq 6 used to calculate the K values do
not introduce explicitly the concentration of H2O, constant at
2.6 M. Table 1 summarizes the equilibrium and rate constants
obtained in different solvents. The rate constants of the reverse
reactions in eq 1 in CH3CN were calculated from the equations
k-1 ) k1/K1 and k-2 ) k2/K2.

5072 Inorganic Chemistry, Vol. 36, No. 22, 1997
Wang and Espenson
in the range 25.0-59.3 °C. The activation parameters are ∆Hq
) 50 kJ mol-1, ∆Sq ) -190 J mol-1 K-1; k298 ) 2.1 × 10-6
s-1
.
Photoinduced Methyl Migration. Irradiation (>400 nm for
5 h) of a sample initially containing 12 mM B and 0.65 M H2O2
forms 6.8 mM CH3OH (57%) and 5.2 mM CH3OOH (43%).
Further photolysis for several hours did not change the CH3-
OH/CH3OOH ratio. Under similar conditions in the dark, only
22% of B had reacted in 8 h, giving 18% CH3OOH and 4%
CH3OH. After B had completely disappeared in 8 days, nearly
96% of the organic products was CH3OOH. The methyl
hydroperoxide thus formed did not convert to CH3OH in a week
at 25 °C.
The photodecomposition of B under oxygen and argon is the
same. Adding 0.2 M CH3OH before photolysis did not affect
the rates and product distribution. Similar outcomes have been
observed for the thermal decomposition of B.
The photolysis of B in the absence of excess of H2O2 in CH2-
Cl2 at 248-365 nm was reported to give chloromethane, ReO3,
and O2 under these conditions.19 We photolyzed B (prepared
from and in equilibrium with MTO and H2O2) in CD2Cl2 and
detected CH3OH (δ 3.42 ppm) and CH3OOH (δ 3.86 ppm) by
1H-NMR.
Photolysis of B in the Presence of Substrates. The
photodecomposition of B in the presence of benzhydrol deriva-
tives was studied, but it yielded little in the way of new
reactions. Most proceeded as before, but a few new organic
products were found, especially in an oxygen atmosphere as
compared to argon. With 4-methylbenzhydrol,1H-NMR and
GC-MS showed benzaldehyde (but only a trace of 4-methyl-
benzaldehyde) and 3-methylphenol, which were comparable in
amount to a ring-oxidized product.
To test for a phenoxyl radical, 2,4,6-triphenylphenol trapping
was attempted. No identifiable products resulted. The thermal
and photochemical reactions with 1,4-dihydronaphthalene were
carried out below 0 °C so that the thermal epoxidation would
not interfere. The major photoproduct was naphthalene, but
the control (photolysis of hydrogen peroxide with 1,4-dihy-
dronaphthalene) also gave naphthalene.
Figure 4. 1H-NMR data for the conversion of CH3OOH from B in
CD3CN. Conditions: [MTO]0 ) 14 mM, [H2O2]0 ) 0.75 M, [H2O]0 )
2.6 M; 25 °C.
tive; it is ruled out, however, by its chemical properties and by
the fact that all of the rhenium was accounted for as perrhenic
acid.
Addition of excess PPh3 converted all of the δ 3.75 ppm
species into CH3OH, perhaps implicating an organic peroxide.
Though the NMR data for CH3OOCH3 and CH3OOH are not
available in the literature, their likely 1H NMR chemical shifts
appear compatible on the basis of Swern’s work on other organic
peroxides.30 A convenient published procedure was used to
prepare in solution CH3OOCH3 and CH3OOH from dimethyl
sulfate and hydrogen peroxide in the presence of potassium
hydroxide.31 The major gas phase product was CH3OOCH3,
though a small amount of CH3OCH3 was also observed. The
CD3CN solution of dimethyl peroxide showed a resonance at δ
3.76 ppm; however, a new signal at δ 3.75 ppm grew in as
CH3ReO3 and H2O2 were added. Again, the δ 3.75 ppm peak
transformed to CH3OH and the CH3OOCH3 resonance remained
the same upon adding excess PPh3. The major products of the
(CH3)2SO4-H2O2-KOH solution upon acidifying with HClO4
were CH3OOH and CH3OSO3-, with NMR signals at δ 3.75
and δ 3.60 ppm, respectively. The CH3OOH thus prepared
reacts with PPh3 to yield CH3OH and Ph3PdO on the basis of
1H and 31P{1H} NMR spectra. When MTO and H2O2 were
added to the CH3OOH solution, the resonance at δ 3.75
increased. Thus all of the evidence indicates that the decom-
position product of B seen at δ 3.75 ppm is CH3OOH.
The methyl hydroperoxide formed from B is stable at 40 °C
for at least 24 h. It slowly decomposes to HCO2H, CH2(OH)2,
and CH3OH at elevated temperature. At 60 °C, 36% of CH3-
OOH decomposed in 24 h to HCO2H (20%), CH2(OH)2 (32%),
and CH3OH (47%). The product distribution was calculated
Photoassisted Formation of O2. Molecular oxygen was
produced in the photolysis of B in acetonitrile at λ >400 nm,
just as it was in CH2Cl2 at 248-365 nm.19
Discussion
Equilibrium Constants. Consistent values of the equilibrium
constants K1 and K2 were obtained at different peroxide
concentrations throughout the course of the decomposition
process, which is much slower than the equilibration reactions
between MTO, A, and B. The kinetic studies confirmed that
also in this medium, as in others, A forms faster than B. The
decomposition of B is much slower than the formation of A
and B, allowing the decomposition system to be treated as two
fast equilibria and a rate-controlling step.
The values of K1 and K2 in CH3CN are similar to those
obtained in CH3OH. These values are considerably larger than
those in water, and the cooperativity effects (K2 > K1) in CH3-
CN and CH3OH are less pronounced than that in H2O. The
significant difference of equilibrium constants between organic
solvents and water may (partially) result from the omission of
the activity of water in the equilibrium constant calculation.
Nonetheless we would emphasize that such a representation of
the thermodynamics is the only reasonable course, given the
uncertainty as to whether A is the pictured “anhydride” structure
or whether it incorporates highly labile water molecule(s), not
detected by NMR.28 These options lead to an unspecified
1
on the basis of integrations of the H signals.
As mentioned above, a small amount of CH3OH was also
detected, in an amount that varied with the initial concentration
of H2O2. The higher the [H2O2]0, the less the CH3OH produced.
For example, with [MTO]0 ) 14 mM and >20-fold excess of
H2O2, <6% CH3OH was formed. However, 20% CH3OH was
found when [MTO]0 ) 14 mM and [H2O2]0 < 0.14 M. A trace
amount of hydrated formaldehyde was also observed, particu-
larly when the [H2O2]0/[MTO]0 was <10.
Kinetics of the Thermal Decomposition of B. The first-
order rate constant was determined from spectrophotometric
measurements in acetonitrile. At a given [MTO], the rate
constant was independent of the initial concentration of hydro-
gen peroxide. The rate constant was evaluated in experiments
(30) Swern, D.; Clements, A. H.; Loung, T. M. Anal. Chem. 1969, 41,
412.
(31) Davies, D. M.; Deary, M. E. J. Chem. Soc., Perkin Trans. 2 1992,
559.

Reactions of Methylrhenium Diperoxide
Inorganic Chemistry, Vol. 36, No. 22, 1997 5073
Table 2. Product Distributions from CH3ReO3-H2O2 in Aqueous
and Acetonitrile Solutions at 25 °C
Scheme 1
amount of product (%)
in CH3CN
in H2Oa
product
thermal
thermal
photochemical
CH3OH
CH3OOH
HOReO3
100
0
100
4
96
100
57
43
100
a Reference 28.
number of water molecules on one side or the other of steps 1
and 2 in eq 1.
total CH3OH was observed in the first 1H-NMR spectrum upon
addition of H2O2 to the MTO solution, consistent with this
assignment.
Solvent Effects on Kinetics. Water has a pronounced effect
on the rates of formation of the peroxorhenium compounds;
the reaction is fastest in water itself. We offer two plausible
reasons for this effect: (a) the more polar solvent component
would stabilize a polar transition state, and (b) water may also
play a specific chemical role in the reaction, possibly by assisting
proton transfer through a hydrogen-bonded intermediate. We
shall examine each in turn. Kinetics data, including isotope
effects and volumes of activation,32 have been used to study
the formation of A. It was concluded that this sequence applies:
Mechanism for the Formation of CH3OOH. The formation
of CH3OOH from B follows first-order kinetics with k ) 2 ×
10-6 s-1 at 25 °C. This rate constant is independent of the
initial concentrations of B and H2O2. Scheme 1 was proposed
to account for the observations of the formation of CH3OOH
and the unimolecular kinetic process. Coordinated water in
Scheme 1 was omitted for clarity. The suggested derivatives,
CH3OOReO(η2-O2) and CH3OOReO3, the later analogous to
the known compound CH3OReO3, could be postulated. The
literature reports that CH3OReO3 is moisture sensitive.25 Indeed,
we have shown that the addition of H2O to a CD3CN solution
of CH3OReO3 produces CH3OH promptly. It is reasonable to
propose that CH3OOReO3 hydrolyzes CH3OOH and HOReO3.
CH3OOH may instead arise from the hydrolysis of CH3OOReO-
(η2-O2). Excess H2O, introduced with 30% H2O2 used to
prepare B, is inevitably present.
The key difference between the two proposed pathways is
whether a ReV or a ReVII intermediate is involved. The
concerted pathway, a, entangles many bond-breaking and
-forming processes and seems to be unlikely, although the
oxidation state of Re remains as VII throughout. Path b
involving a ReV(d2)-ReVII(d0) conversion, CH3OOReO(η2-O2)
to CH3OOReO3 or HOReO(η2-O2) to HOReO3, is symmetry
forbidden;33 however, the formation of CH3OOH is so slow that
path b may indeed operate. The postulated thermal methyl
migration to form rhenium methyl peroxide from B is different
from the rearrangement proposed in the alkylperoxopermeth-
yltantalocene system, eq 7. The conversion of Cp*2Ta(η2-
O2)(CH3) to Cp*2Ta(O)(OCH3) at 80 °C occurs in a non-first-
order fashion and is acid-catalyzed.34 A concerted mechanism
is preferred. On the other hand, the rearrangement in Scheme
The transition state shown has bonds considerably more polar
than in MTO and H2O2, suggesting that a kinetic acceleration
arises from the general increase in polarity of the medium.
Additionally, one or two water molecules may play an explicit
chemical role in which hydrogen bonding to an intervening
water molecule facilitates proton transfer in the step diagramed
above. This idea is embodied in a modified transition state:
Thus it is not surprising to see a significant H2O effect on the
kinetics since formation and cleavage of O-H bonds are
involved in the processes.
Solvent Effect on Organic Products. The aqueous decom-
position of the MTO-H2O2 system in aqueous solution produces
CH3OH and ReO4-, not from B, but from a very fast reaction
between MTO and HOO-.28 In organic media, however, the
MTO-H2O2 system is more stable. To lower the deactivation
rate of MTO-H2O2 in CD3CN to about the same as that in
aqueous solution would require that [H+] be increased to 5 M,
there being an inverse rate effect on [H+] in water. The major
organic product is CH3OOH in CD3CN and also in CDCl3, CD2-
Cl2, acetone-d6, and 2-propanol-d8. Comparison of the product
distributions in H2O and CH3CN solvents is given in Table 2.
The formation of CH3OH in CD3CN likely results from the
reaction of MTO with HOO-. The qualitative correlation
between the amount of CH3OH produced and the initial
concentrations of H2O2 supports this contention. The large
equilibrium constants and the small rate constants of the reverse
reactions in acetonitrile suggest that very little rhenium species
is present as MTO except at the beginning. About 60% of the
1 is similar to the Cp*2Ta(η2-CH2O)(H) system, eq 8.35 Instead
of C-O bond cleavage concurrent with C-H bond formation,
a stepwise process supported by the observation of an inverse
kinetic deuterium isotope effect was proposed. Conversion of
Cp*2Ta(η2-CH2O)(H) to Cp*2TaO(CH3) obeys first-order kinet-
ics with a rate constant of 3.03 × 10-6 s-1 at 140 °C, with an
equilibrium between Cp*2Ta(η2-CH2O)(H) and Cp*2Ta(OCH3).
(33) Brown, S. N.; Mayer, J. M. Inorg. Chem. 1992, 31, 4091.
(34) Van Asselt, A.; Trimmer, M. S.; Henling, L. M.; Bercaw, J. E. J. Am.
Chem. Soc. 1988, 110, 8254.
(32) Pestovsky, O.; van Eldik, R.; Huston, P.; Espenson, J. H. J. Chem.
Soc., Dalton Trans. 1995, 133.
(35) Van Asselt, A.; Burger, B. J.; Gibson, V. C.; Bercaw, J. E. J. Am.
Chem. Soc. 1986, 108, 5347.

5074 Inorganic Chemistry, Vol. 36, No. 22, 1997
Wang and Espenson
Scheme 2
O2)2‚H2O systems upon photolysis.22,37 The molecular orbital
interaction diagram of B would be similar to that of (HMPA)-
MoO(η2-O2)2(H2O). As a result, the absorption band around
360 nm for B is corresponding to a d r π* (O22-) transition.
When the pentagonal plane is designated xy, cleavage of the
Re-O (η2-O2) bond must result from the transition of the in-
plane π* (O22-) orbitals to the dxy or dx -y orbital. Whether a
2
2
•
CH3 is also formed as a result will be discussed below.
Step II consists of a methyl migration followed by an
intramolecular redox process. Unlike the thermal reaction, the
methyl rearrangement need not cleave the Re-O bond. As a
result, this process is much faster than the thermal methyl
migration.
The elimination of molecular oxygen, step III, has been well
documented in photochemical reactions involving metal peroxo
complexes. As mentioned previously, CH3ReO(η2-O2)(η1-OO•)
resulting from Re-O bond cleavage may be responsible for
the formation of CH3OOH. However, the generation of O2 may
not be from the same excited state if the O-O bond cleavage
product, CH3ReO(η2-O2)(η1-O•)2, could also be produced upon
photolysis. The electronic state of molecular oxygen was not
determined in these studies, but 3O2 was produced in CH2Cl2.19
Since MTO is also formed as each oxygen molecule is released,
decomposition of MTO to CH3OH will compete with equilibra-
tion. The formation of CH3OH and equilibration between MTO,
A, and B, shown in the top part of the scheme, resemble the
aqueous reactions. The product distribution ([CH3OH] > [CH3-
OOH]), Table 2, suggests that the rates of steps III and II would
be comparable.
•
If the mechanism involved CH3 , then CH3OOH might be
•
formed from CH3 with O2 and subsequent hydrogen-atom
abstraction. The lack of effect of molecular oxygen and the
large activation entropy (-190 J mol-1 K-1) of the decomposi-
tion processes rule out possible homolytic cleavage of the Re-
CH3 bond.
The rate constant for the decomposition of B to MeOOH in
acetonitrile is 2 × 10-6 s-1, nearly 150 times as large as the
rate constant for its decomposition to O2 in water, k9 ) 2.9 ×
10-4 s-1, both at 25 °C (eq 9). The solvent effect is quite large,
CH3Re(O)(O2)2OH2 f CH3ReO3 + O2 + H2O (9)
sufficient to change not only the rate constant but the entire
course of the reaction. No oxygen evolution was detected in
acetonitrile, although the method is not too sensitive. If as much
as 10% O2 would have escaped detection, then in acetonitrile
k9 < 2 × 10-7 s-1. The original issue, that methyl hydroper-
oxide is favored so much over oxygen in acetonitrile, becomes
two others: How is it that MeOOH becomes favored over
MeOH? This was addressed in the preceding paragraphs. And
why is the rate constant for O2 evolution favored so strongly in
the one solvent? The strong solvent dependence for other MTO
reactions comes to mind: the previously-discussed reaction of
B to release A and hydrogen peroxide, k-2 ) 4 × 10-2 s-1 and
7 × 10-5 s-1, which is similar to the increase in k9 (rate ratios
are 600 and 1500). The key to understanding both lies in the
role played by hydrogen bonding to stabilize the terminal O
atom or OH group of an intermediate in which the atom bonded
to rhenium is only one O atom (or OH group) of the incipient
O2 or H2O2 molecules. With this picture in mind, higher rates
in water are easily understood.
Cleavages of the Re-O (η2-O2) and Re-CH3 bonds upon
photolysis of B in CH2Cl2 with 248-365 nm light were
proposed.19 Chloromethane was observed. The energies of the
light sources were in the range of 328-482 kJ mol-1, which
are probably larger than the unknown Re-CH3 bond energy.
In our studies, the photolysis energy was lower, <299 kJ mol-1
,
but still larger than the energy of one of the strongest transition
metal carbon bonds, Ta-CH3 (DTa-C ) 261 kJ mol-1).38 The
photoproducts from B in CD3CN under Ar are the same as those
under O2, but a mechanism involving a free methyl radical still
could not be ruled out since O2 was produced upon irradiation
of B. However, to date the only detected products from the
•
reaction of CH3 with O2 in the literature are CH3OH, CH2O,
and CH3OOCH3,39 in contrast to CH3OH and CH3OOH
observed in the photolysis of B. The photoproducts from B in
CD2Cl2 at >400 nm are different from those at 248-365 nm,
suggesting that CH3• is not likely involved in the photolysis of
B at >400 nm.
Photocatalysis. One of our goals was to investigate the
possibility of using B as a photocatalyst. As indicated in
Scheme 2, the use of an oxygen-centered radical CH3ReO(η2-
O2)(η1-OO•), generated from the initial photolysis step, was
limited owing to step II. To trap CH3ReO(η2-O2)(η1-OO•), 1,4-
dihydronaphthalene was used. The results are indicative but
not conclusive on this point. Phenoxyl radicals were suggested
by the oxygen-dependent photoreactions with 2,4,6-trialkylphe-
nols. The stability of the phenoxyl radical could be the driving
force for the reaction of CH3ReO(η2-O2)(η1-OO•) with ArOH.
Rearrangement of ArO• to a carbon-centered radical and its
subsequent reaction with oxygen are known.40 The low yields
Photolysis of B. CH3OH and CH3OOH are not intercon-
verted thermally or photochemically under the experimental
conditions, meaning that they arise independently. The first
step in Scheme 2 consists of homolytic cleavage of a Re-O
bond, as proposed in many photoinduced reactions involving
metal peroxo complexes, though exceptions are known.23,36 For
example, the photocleavage of the O-O bond instead of a M-O
bond has been proposed in the photochemical reactions of
peroxotitanium(IV) porphyrin.23 The experimental results in
the case of the diperoxorhenium system could not conclusively
rule out the O-O cleavage pathway; however, the formation
of CH3OOH and theoretical calculations on the analogous
complex do favor the M-O cleavage pathway. Extended
Hu¨ckel calculations are in favor of the homolysis of a Mo-O
bond in the (TPP)Mo(η2-O2)2 and the (HMPA)MoO(η2-
(37) Ledon, H. L.; Bonnet, M. J. Am. Chem. Soc. 1981, 103, 6209.
(38) Adedeji, F. A.; Connor, J. A.; Skinner, H. A.; Galyer, L.; Wilkinson,
G. J. Chem. Soc., Chem. Commun. 1976, 159.
(39) Simon, F.-G.; Schneider, W.; Moortgat, G. K. Int. J. Chem. Kinet.
1990, 22, 791.
(40) Altwicker, E. R. Chem. ReV. 1967, 67, 475.
(36) Peterson, M. W.; Van Atta, R. B.; Richman, R. M. Inorg. Chem. 1983,
22, 1982.

Reactions of Methylrhenium Diperoxide
Inorganic Chemistry, Vol. 36, No. 22, 1997 5075
were discouraging as far as the prospects for photocatalysis go,
and the matter was not pursued further.
CH3OH in aqueous solution. Light facilitates the decomposition
processes, but B is not a likely photocatalyst given its rapid
conversion to CH3OOH.
Conclusion. Water affects the rates and equilibria of the
MTO-H2O2 system in CH3CN, more for the formation of A
than B. Although the MTO-H2O2 system is much more stable
in organic solvents than water, the generation of the real oxo-
transfer catalysts, A and B, is much slower in organic solvents
than in water. This may be a consequence of the direct
involvement of water in the mechanism of peroxide binding.
The decomposition of B produces CH3OOH in acetonitrile but
Acknowledgment. This work was supported by a grant from
the National Science Foundation (CHE-9007283). Some of the
experiments were conducted with the use of the facilities of
the Ames Laboratory.
IC970650C