Mechanism of dihydrogen cleavage by high-valent metal oxo compounds: Experimental and computational studies

Article abstract of DOI:10.1021/ic010639j

The oxidation of dihydrogen by metal tetraoxo compounds was investigated. Kinetic measurements of the oxidations of H₂ by MnO₄^- and RuO₄, performed by UV-vis spectroscopy, showed these reactions to be quite rapid at 25 °C (k₁ ≈ (3-6) x 10^-2 M^-1 s^-1). Rates measured for H₂ oxidation by MnO₄^- in aqueous solution (using KMnO₄) and in chlorobenzene (using ⁿBu₄NMnO₄) revealed only a minor solvent effect on the reaction rate. Substantial kinetic isotope effects [(k_H2/k_D2 = 3.8(2) (MnO₄^- aq), 4.5(5) (MnO₄^-, C₆H₅C1 soln) and 1.8(6) (RuO_4, CCl₄ soln)] indicated that H-H bond cleavage is rate determining and that the mechanism of dihydrogen cleavage is likely similar in aqueous and organic solutions. Third-row transition-metal oxo compounds, such as OsO₄, ReO₄^-, and MeReO₃, were found to be completely unreactive toward H₂. Experiments were performed to probe for a catalytic hydrogen/deuterium exchange between D₂ and H₂0 as possible evidence of dihydrogen σ-complex intermediates, but no H/D exchange was observed in the presence of various metal oxo compounds at various pH values. In addition, no inhibition of RuO₄-catalyzed hydrocarbon oxidation by H₂ was observed. On the basis of the available evidence, a concerted mechanism for the cleavage of H2 by metal tetraoxo compounds is proposed. Theoretical models were developed for pertinent MnO₄^- + H₂ transition states using density functional theory in order to differentiate between concerted [2 + 2] and [3 + 2] scissions of H₂. The density functional theory calculations strongly favor the [3 + 2] mechanism and show that the H₂ cleavage shares some mechanistic features with related hydrocarbon oxidation reactions. The calculated activation energy for the [3 + 2] pathway (ΔH^? = 15.4 kcal mol-¹) is within 2 kcal mol-¹ of the experimental value.

Full text of DOI:10.1021/ic010639j

6272
Inorg. Chem. 2001, 40, 6272-6280
Mechanism of Dihydrogen Cleavage by High-Valent Metal Oxo Compounds: Experimental
and Computational Studies
James P. Collman,*,† LeGrande M. Slaughter,† Todd A. Eberspacher,†
Thomas Strassner,‡ and John I. Brauman†
Department of Chemistry, Stanford University, Stanford, California 94305-5080, and
Institut fu¨r Anorganische Chemie, Technische Universita¨t Mu¨nchen, Lichtenbergstrasse 4,
D-85747 Garching, Germany
ReceiVed June 18, 2001
The oxidation of dihydrogen by metal tetraoxo compounds was investigated. Kinetic measurements of the oxidations
-
of H2 by MnO4 and RuO4, performed by UV-vis spectroscopy, showed these reactions to be quite rapid at 25
°C (k1 ≈ (3-6) × 10-2 M-1 s-1). Rates measured for H2 oxidation by MnO4- in aqueous solution (using KMnO4)
n
and in chlorobenzene (using Bu4NMnO4) revealed only a minor solvent effect on the reaction rate. Substantial
kinetic isotope effects [(kH /kD ) 3.8(2) (MnO4-, aq), 4.5(5) (MnO4-, C6H5Cl soln), and 1.8(6) (RuO4, CCl4
2
2
soln)] indicated that H-H bond cleavage is rate determining and that the mechanism of dihydrogen cleavage is
-
likely similar in aqueous and organic solutions. Third-row transition-metal oxo compounds, such as OsO4, ReO4
,
and MeReO3, were found to be completely unreactive toward H2. Experiments were performed to probe for a
catalytic hydrogen/deuterium exchange between D2 and H2O as possible evidence of dihydrogen σ-complex
intermediates, but no H/D exchange was observed in the presence of various metal oxo compounds at various pH
values. In addition, no inhibition of RuO4-catalyzed hydrocarbon oxidation by H2 was observed. On the basis of
the available evidence, a concerted mechanism for the cleavage of H2 by metal tetraoxo compounds is proposed.
Theoretical models were developed for pertinent MnO4- + H2 transition states using density functional theory in
order to differentiate between concerted [2 + 2] and [3 + 2] scissions of H2. The density functional theory
calculations strongly favor the [3 + 2] mechanism and show that the H2 cleavage shares some mechanistic features
with related hydrocarbon oxidation reactions. The calculated activation energy for the [3 + 2] pathway (∆H‡ )
15.4 kcal mol-1) is within 2 kcal mol-1 of the experimental value.
Introduction
evidence has persisted, raising the possibility that polar (i.e.,
hydride transfer) or concerted mechanisms may be operating
Metal oxo functionalities are known or believed to be the
active oxidants in a broad range of transition-metal-based
systems capable of oxidizing carbon-hydrogen bonds, including
biological monooxygenases such as the cytochromes P-450,1
synthetic metalloporphyrin oxidation catalysts,2 heterogeneous
metal oxide catalysts,3 and homogeneous inorganic oxidants
such as permanganate and chromic acid.4,5 However, despite a
wealth of mechanistic studies on the various systems, no
consistent mechanistic scheme unifying the modes of reactivity
of these related classes of oxidants has emerged. The initial
abstraction of a hydrogen radical has often been proposed as
the key step in the oxidation of a C-H bond,5-8 but conflicting
instead of or in addition to the radical pathway. Studies of the
cytochrome P-450 oxidation systems are illustrative of these
mechanistic ambiguities. The long-accepted “rebound mecha-
nism”,6 in which hydrogen atom abstraction by a ferryl moiety
is followed by the collapse of the carbon radical to give a
hydroxylated product, has been called into question by New-
comb and co-workers in a series of fast-radical clock studies.9
These experiments suggested the plausibility of a “nonsychro-
nous concerted mechanism”, in which the carbon-based radical
has a lifetime of only ∼70 fs and is best viewed as a fleeting
component of the transition state ensemble rather than a true
radical intermediate. In addition, evidence was found for a
competing pathway involving carbocation intermediates.9d More
recently, Shaik and co-workers have proposed that the mecha-
nistic problems are best explained by considering two distinct
reactive states of the iron oxo (one low-spin and one high-
spin),10 while Coon and co-workers have argued for the
existence of multiple active oxidants in P-450 oxygenations,
including iron peroxo and iron hydroperoxo species in addition
† Stanford University.
‡ Technische Universita¨t Mu¨nchen.
(1) (a) Cytochrome P-450: Structure, Mechanism, and Biochemistry; Ortiz
de Montellano, P. R., Ed.; Plenum: New York, 1995. (b) Sono, M.;
Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. ReV. 1996, 96,
2841-2887.
(2) Meunier, B. Chem. ReV. 1992, 92, 1411-1456.
(3) Structure-ActiVity and SelectiVity Relationships in Heterogeneous
Catalysis; Grasselli, R. K., Sleight, A. W., Eds.; Elsevier: New York,
1991.
(4) Oxidation in Organic Chemistry; Wiberg, K. B., Ed.; Academic
Press: New York, 1965; Part A.
(5) (a) Wiberg, K. B.; Evans, R. J. Tetrahedron 1960, 8, 313. (b) Wiberg,
K. B.; Foster, G. Chem. Ind. (London) 1961, 108. (c) Wiberg, K. B.;
Foster, G. J. Am. Chem. Soc. 1961, 83, 423.
(7) Brauman, J. I.; Pandell, A. J. J. Am. Chem. Soc. 1970, 92, 329-335.
(8) Wiberg, K. B.; Fox, A. S. J. Am. Chem. Soc. 1963, 85, 3487-3491.
(9) (a) Newcomb, M.; Le Tadic-Biadatti, M.-H.; Chestney, D. L.; Roberts,
E. S.; Hollenberg, P. F. J. Am. Chem. Soc. 1995, 117, 12085-12091.
(b) Toy, P. H.; Newcomb, M.; Hollenberg, P. F. J. Am. Chem. Soc.
1998, 120, 7719-7729. (c) Toy, P. H.; Dhanabalasingam, B.;
Newcomb, M.; Hanna, I. H.; Hollenberg, P. F. J. Org. Chem. 1997,
62, 9114-9122.
(6) (a) Groves, J. T.; McClusky, G. A. J. Am. Chem. Soc. 1976, 98, 859-
861. (b) Groves, J. T. J. Chem. Educ. 1985, 62, 928-931.
10.1021/ic010639j CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/26/2001

Dihydrogen Cleavage by Metal Oxo Compounds
Inorganic Chemistry, Vol. 40, No. 24, 2001 6273
to ferryl intermediates.11 Recent studies by our group12 and
others13 on synthetic metalloporphyrin analogues of P-450 have
also indicated that alternatives to the rebound mechanism must
be considered. It is clear that a mechanistic understanding of
metal oxo mediated oxidations remains incomplete and that any
new evidence which adds to the current body of knowledge
will be helpful.
strongest among saturated hydrocarbons (∼105 kcal mol-1 ),21
and this would seem to preclude hydrogen atom abstraction as
a mechanistic possibility. We have undertaken an examination
of the reactions of permanganate and related metal oxo
compounds with hydrogen in more detail in order to gain
insights into the modes of reactivity of these oxidants. Herein
are presented our experimental findings, including the first
measurements of kinetic isotope effects (KIEs) for dihydrogen
cleavage by metal oxo moieties and the first observations of
dihydrogen oxidation by RuO4. We also present a theoretical
analysis which, together with our experimental results, suggests
that a concerted bond cleavage is the most plausible first step
in the oxidation of H2 by metal tetraoxo compounds.
Inorganic oxidants such as permanganate (MnO4-) and
ruthenium tetroxide (RuO4) are appealing models for the
reactivity of metal oxo species. These well-defined homoge-
neous oxidants are capable of reacting with a wide variety of
functional groups, including the C-H bonds of saturated and
unsaturated hydrocarbons.14-18 The metal tetraoxo compounds
differ from the iron oxo intermediate in cytochrome P-450 in
that they are closed-shell species, whereas the latter is known
to possess a high-spin ground state; however, these compounds
are still reasonable models for P-450 reactivity given that recent
theoretical studies predict a lowest energy pathway for alkane
hydroxylation which involves a low-spin state of the P-450 ferryl
moiety.10 Indeed, the reactivity patterns in these systems are
quite similar to those observed in the P-450 enzymes and models
(i.e., they exhibit radical-like selectivity in the oxidations of
C-H bonds (3° > 2° > 1°)), yet these reactions occur
predominantly with retention of configuration.4,17-19 Mayer and
co-workers have recently published thorough studies of benzylic
oxidations by permanganate15 and CrO2Cl216 which showed that
radical mechanisms appear to be operative in organic solvents,
but a shift to a polar mechanism may occur upon changing to
an aqueous solution.15a By contrast, a concerted C-H cleavage
has been proposed for RuO4-mediated hydroxylations of hy-
drocarbons on the basis of mechanistic investigations.18c-e
We were intrigued by decades-old reports of dihydrogen
oxidation by permanganate20 because we realized that this
reaction has interesting implications for hydrocarbon oxidations
by high-valent metal oxo species. The H-H bond of dihydrogen
is approximately as strong as the C-H bond of methane, the
Experimental Section
General Considerations. Organic solvents (chlorobenzene and CCl4)
were rigorously purified by the following procedure:22 solvents were
shaken with concentrated H2SO4 until the washings were colorless, and
then they were washed with H2O, washed with aqueous NaHCO3, dried
over CaCl2, and distilled from P2O5. The solvents were stored over
activated 4 Å molecular sieves and were passed through a column of
activated basic alumina immediately prior to use. Deionized water was
purified by boiling with potassium permanganate for 24 h followed by
distillation, and the purified water was stored under argon.
KMnO4 (Aldrich, 99%), KReO4 (Strem, 99.9%), MeReO3 (Strem),
OsO4 (Strem, 99.95%), and RuCl3‚xH2O (Platina Laboratories, Inc.,
n
41.3% Ru/wt) were used as obtained. Bu4NMnO4 was prepared by a
published procedure23 and recrystallized from CH2Cl2/Et2O. This
material was stored at -20 °C in a sealed vial which was warmed to
room temperature before opening. Warning: nBu4NMnO4 has been
reported to explode spontaneously23 and should be handled carefully.
H2 (Air Products, 99.995%) and D2 (CIL, 99.9%) were used without
additional purification.
Gas chromatographic analyses were performed on a Hewlett-Packard
5890 instrument equipped with FID and TCD detectors. UV-vis spectra
were recorded on a Hewlett-Packard 8452A diode array spectrometer.
For kinetic experiments, the sample temperature was maintained at 25
°C by the use of a thermostated cell holder connected to a Melsungen
Thermomix 1441 water circulator. A Varian XL-400 spectrometer was
used for NMR analyses.
(10) (a) Shaik, S.; Filatov, M.; Schro¨der, D.; Schwartz, H. Chem.sEur. J.
1998, 4, 193-199. (b) Ogliaro, F.; Harris, N.; Cohen, S.; Filatov, M.;
de Visser, S. P.; Shaik, S. J. Am. Chem. Soc. 2000, 122, 8977-8989.
(c) Schro¨der, D.; Shaik, S.; Schwarz, H. Acc. Chem. Res. 2000, 33,
139-145.
Kinetic Measurements: MnO4- + H2. An approximately 30 mM
n
solution of KMnO4 in purified H2O or Bu4NMnO4 in C6H5Cl was
prepared using an accurately measured sample of the permanganate
salt, and an aliquot of this solution was diluted to 5 mL to give a 0.3
mM solution. This solution was placed in an apparatus consisting of a
quartz cuvette connected to a 100 mL round-bottom flask and sealed
with a Teflon stopcock. The solution was degassed on a vacuum line
with stirring for ∼20 s at room temperature and then subjected to three
freeze-pump-thaw degas cycles at -78 °C. After the solution was
warmed to room temperature, the apparatus was opened to bubbling
H2 or D2 at 1 atm while the solution was stirred for 1 min. The cuvette
portion of the apparatus containing the solution was inserted into the
thermostated cell holder of the spectrometer, and the temperature was
allowed to equilibrate for 5 min before data collection commenced.
UV-vis spectra were acquired at intervals of 180 s for 8-36 h,
depending on the reaction rate. Three kinetic trials were performed for
each substrate (H2 and D2). After the final spectra were recorded,
solutions were analyzed using an iodometric UV-vis procedure24 to
determine the oxidation state of manganese.
(11) (a) Vaz, A. D. N.; Pernecky, S. J.; Raner, G. M.; Coon, M. J. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 4644-4648. (b) Toy, P. H.;
Newcomb, M.; Coon, M. J.; Vaz, A. D. N. J. Am. Chem. Soc. 1998,
120, 9718-9719. (c) Newcomb, M.; Shen, R.; Choi, S.-Y.; Toy, P.
H.; Hollenberg, P. F.; Vaz, A. D. N.; Coon, M. J. J. Am. Chem. Soc.
2000, 122, 2677-2686.
(12) (a) Collman, J. P.; Chien, A. S.; Eberspacher, T. A.; Brauman, J. I. J.
Am. Chem. Soc. 1998, 120, 425-426. (b) Collman, J. P.; Chien, A.
S.; Eberspacher, T. A.; Brauman, J. I. J. Am. Chem. Soc. 2000, 122,
11098-11100.
(13) (a) Nam, W.; Lim, M. H.; Lee, H. J.; Kim, C. J. Am. Chem. Soc.
2000, 122, 6641-6647. (b) Nam, W.; Lim, M. H.; Moon, S. K.; Kim,
C. J. Am. Chem. Soc. 2000, 122, 10805-10809.
(14) Stewart, R.; Spitzer, U. A. Can. J. Chem. 1978, 56, 1273-1279.
(15) (a) Gardner, K. A.; Mayer, J. M. Science 1995, 269, 1849-1851. (b)
Gardner, K. A.; Kuehnert, L. L.; Mayer, J. M. Inorg. Chem. 1997,
36, 2069-2078. (c) Mayer, J. M. Acc. Chem. Res. 1998, 31, 441-
450.
(16) (a) Cook, G. K.; Mayer, J. M. J. Am. Chem. Soc. 1994, 116, 1855-
1868. (b) Cook, G. K.; Mayer, J. M. J. Am. Chem. Soc. 1995, 117,
7139-7156.
(17) (a) Tenaglia, A.; Terranova, E.; Waegell, B. Tetrahedron Lett. 1989,
30, 5271-5274. (b) Tenaglia, A.; Terranova, E.; Waegell, B. J. Chem.
Soc., Chem. Commun. 1990, 1344-1345.
(20) (a) Just, G.; Kauko, Y. Z. Phys. Chem. 1911, 76, 601-640. (b) Hein,
F.; Daniel, W.; Schwedler, H. Z. Anorg. Chem. 1937, 233, 161-177.
(c) Webster, A. H.; Halpern, J. Trans. Faraday Soc. 1957, 53, 51-
60. (d) Halpern, J. AdV. Catal. 1957, 9, 302-311.
(21) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994, 98,
2744-2765 and references therein.
(22) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals; Pergamon: New York, 1988.
(18) (a) Bakke, J. M.; Lundquist, M. Acta Chem. Scand., Ser. B 1986, 40,
430-433. (b) Bakke, J. M.; Brænden, J. E. Acta Chem. Scand. 1991,
45, 418-423. (c) Bakke, J. M.; Bethell, D. Acta Chem. Scand. 1992,
46, 644-649. (d) Bakke, J. M.; Frøhaug, A. E. Acta Chem. Scand.
1994, 48, 160-163. (e) Bakke, J. M.; Frøhaug, A. E. J. Phys. Org.
Chem. 1996, 9 (7), 507-513.
(23) Karaman, H.; Barton, R. J.; Robertson, B. E.; Lee, D. G. J. Org. Chem.
1984, 49, 4509-4518.
(19) Groves, J. T.; Nemo, T. E. J. Am. Chem. Soc. 1983, 105, 6243-6248.

6274 Inorganic Chemistry, Vol. 40, No. 24, 2001
Collman et al.
-
Data Analysis: MnO4 + H2. Spectra were analyzed following a
H/D Exchange Experiments. (a) Permanganate. A solution of 40
mg of KMnO4 was prepared in 5 mL of deionized H2O or an appropriate
phosphate or carbonate buffer solution. The solution was placed in a
50 mL Schlenk flask fitted with a Suba-Seal septum and then degassed
at room temperature and again at -78 °C. The solution was warmed,
and 1 atm of D2 was admitted. The mixture was stirred for several
days, and 10 µL samples of headspace gas were periodically withdrawn
by micropipet for GC analysis. A 1/8 in. GC column packed with MnCl2-
coated alumina and maintained at -196 °C was used to analyze for
H2, HD, and D2;27 the column was connected to a heated cupric oxide
column which converted hydrogen isotopologues to water vapor for
TCD analysis.
(b) RuO4. A catalytic biphasic RuO4 system was prepared in a 5
mL Schlenk flask by dissolving 8 mg of RuCl3‚xH2O and 200 mg of
NaIO4 in 2 mL of H2O and then adding 1 mL of CH3CN and 2 mL of
CCl4. The mixture was degassed, and D2 was admitted; GC analysis
was performed as described previously.
procedure described by Gardner et al.15b Absorbances were measured
at 526 nm (a MnO4- absorption maximum) and at 494 nm (a reference
wavelength where the colloidal MnO2 product absorbs strongly but
MnO4- absorbs weakly), and these were used to calculate [MnO4-] at
each point in time. Modeling of the scattering and absorption curve of
MnO2 was required in order to obtain the ratio of effective extinction
coefficients for MnO2 at the two different wavelengths; this was done
separately for each kinetic run by performing a third-order polynomial
fit of the absorbance versus wavelength data at ∼2.5 half-lives, with
-
the region displaying residual MnO4 absorption (480-600 nm)
omitted. Pseudo-first-order rate constants were calculated by nonlinear
least-squares fitting of the [MnO4-] versus time data to the exponential
form of the first-order rate law. Use of a different pair of wavelengths
(548 and 450 nm) gave rates which were identical within experimental
error. For the reaction in C6H5Cl, significant acceleration of the rate of
-
MnO4 decay is seen at long reaction times, so only initial data were
used to compute rate constants (30 min for H2 and 50 min for D2).
These rates were corrected for the background solvent oxidation reaction
(<9%). For the reaction in H2O, no such acceleration was apparent,
and the entire data range collected (∼2.5 half-lives) was used in the
rate calculation. Reported rates are average values for three separate
runs for each substrate (H2 and D2), with uncertainties given at the
95% confidence level.25
Data were also analyzed using an alternative procedure based on
the method of initial rates.26 In this method, absorbance versus time
data from the early stages of the reaction (30-50 min) were modeled
using nth-order polynomial fits, and rate constants were calculated from
averaged second coefficients (a1) for n ) 4 and 5. This method gave
rates which differed by <10% for the reaction in C6H5Cl but by 30-
60% for the reaction in H2O. In the latter case, the method appeared to
suffer from interference by the MnO2 product early in the reaction
because of the greater effective absorbance of MnO2 at low wavelengths
in aqueous versus organic solutions. In addition, calculated uncertainties
were large (e.g., 23% for D2 in C6H5Cl and 120% for D2 in H2O), and
thus, only rates calculated by the previously described method are
reported.
(c) Other Oxidants. The 50 mM aqueous solutions were prepared
in 50 mL Schlenk flasks as described for permanganate, and all of the
other procedures were the same.
NMR Tube Experiments. Into a 5 mm NMR tube fitted with a
poly(tetrafluoroethylene) valve was placed 8-10 mg of a transition-
metal oxo compound. D2O (∼0.7 mL) was added, the tube was
evacuated on a vacuum line while immersed in a cold water bath at
∼3 °C, and 1 atm of H2 was admitted. The tube was sealed, stored at
1
25 °C, and monitored periodically by H NMR.
Kinetics of Adamantane Oxidation by RuO4. Reactions were
performed using a modification of the catalytic conditions developed
by Carlsen et al.18c,28 Into each of two 100 mL Schlenk flasks were
introduced 2.0 g of NaIO4, 18 mL of H2O, 10 mL of CH3CN, and 20
mL of an 80 mM solution of adamantane in CCl4, and the mixtures
were stirred. The flasks were fitted with Suba-Seal septa and degassed;
one flask was filled with Ar and the other with H2. A total of 2.0 mL
of a 0.16 M solution of RuCl3 was injected into each flask by syringe,
and within a few minutes, brown RuCl3 had been converted to yellow
RuO4 as monitored visually. The reaction was monitored using the
following procedure. A 100 µL aliquot of the organic (lower) layer
Kinetic Measurements: RuO4 + H2. A solution of RuO4 in CCl4
was prepared as follows. To a ∼0.58 mM solution of RuCl3 in 5 mL
of H2O were added 200 mg of NaIO4 and 5 mL of CCl4. This biphasic
mixture was stoppered and stirred vigorously for ∼2 h until the aqueous
phase was colorless and the CCl4 phase was clear yellow. The CCl4
phase was then extracted by pipet and transferred to the sealable UV
cell apparatus. UV-vis analysis indicated that >98% of RuCl3 had
been converted to RuO4 and extracted into the CCl4 phase. Even with
rigorously purified CCl4, a small amount of RuO4 decomposition
(∼10%) was observed to occur over ∼18 h in the absence of substrate,
apparently due to a trace impurity. Thus, solutions were incubated at
room temperature for 24 h under N2 prior to the addition of H2 or D2.
Solutions were then degassed and exposed to H2 and D2 as described
i
was withdrawn by syringe and mixed with 3 mL of 3:1 PrOH/H2O,
which reduced RuO4 to RuO2. This mixture was extracted twice with
1 mL of Et2O containing 1 mM tert-butyl benzene (external standard),
and this extract was analyzed by GC. The two reactions were monitored
in tandem over 48 h, and headspace gases were replaced periodically.
Rates were calculated by fitting the adamantane concentration over time
to a first-order exponential decay.
Computational Details. The density functional/Hartree-Fock hybrid
model B3LYP,29,30 as implemented in Gaussian 98,31 was used
throughout this study together with the triple-ú basis set 6-311+G-
(d,p). This methodology has proven successful for the evaluation of
flat potential energy surfaces (e.g., in examining hydrogen atom
-
for the MnO4 experiments, and UV-vis data collection was started.
(27) (a) Yasumori, I.; Ohno, S. Bull. Chem. Soc. Jpn. 1966, 39, 1302-
1306. (b) Collman, J. P.; Wagenknecht, P. S.; Lewis, N. S. J. Am.
Chem. Soc. 1992, 114, 5665-5673.
Four separate kinetic runs were performed for H2 and six for D2;
reported rate constants are average values, with errors given at the 95%
confidence level.25
(28) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J. Org.
Chem. 1981, 46, 3936-3938.
Data Analysis: RuO4 + H2. In the early stages of the reaction,
absorbances at 310, 384, and 396 nm increased somewhat. This behavior
was attributed to an initial buildup of colloidal RuO2‚xH2O (vide infra),
but after this initial period (∼2 h), solid RuO2 particles began settling
to the bottom of the cuvette, and the absorbances began uniformly
decreasing. The absorbance at 310 nm was used to calculate rates
because it was least affected by this phenomenon, and plots of Abs-
(310 nm) versus t after the initial period were modeled well by nonlinear
least-squares fitting to a first-order exponential decay. Typically, data
in the ranges of 130-300 min (H2) or 160-350 min (D2) were used.
(29) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B: Condens. Matter 1988,
37, 785-789.
(30) (a) Becke, A. D. J. Chem. Phys. 1992, 97, 9173-9177. (b) Becke, A.
D. J. Chem. Phys. 1993, 98, 1372-1377. (c) Becke, A. D. J. Chem.
Phys. 1993, 98, 5648-5652.
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,
V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R.
L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian98, Revision A.7; Gaussian,
Inc.: Pittsburgh, PA, 1998.
(24) Lee, D. G.; Perez-Benito, J. F. J. Org. Chem. 1988, 53, 5725-5728.
(25) Mendham, J.; Denney, R. C.; Barnes, J. D.; Thomas, M. Vogel’s
Textbook of QuantitatiVe Chemical Analysis; Prentice Hall: New York,
2000; pp 127-128.
(26) (a) Chandler, W. D.; Lee, E. J.; Lee, D. G. J. Chem. Educ. 1987, 64,
878-881. (b) Laidler, K. J. Chemical Kinetics; McGraw-Hill:
New York, 1965; pp 15-17.

Dihydrogen Cleavage by Metal Oxo Compounds
Inorganic Chemistry, Vol. 40, No. 24, 2001 6275
abstraction by MnO4 from methane and toluene).32 All of the
Table 1. Rates of H2 and D2 Oxidation by MnO4
-
- a
geometries were fully optimized, and frequency calculations ensured
that they corresponded to either minima (NIMAG 0) or transition states
(NIMAG 1) on the potential energy surface. All of the energies listed
are zero-point energy (ZPE) corrected and are given in kcal mol-1
relative to the reactants. All of the values are unscaled. The stabilities
of wave functions were checked, and unrestricted calculations ensured
that no radical intermediates needed to be considered. Standard criteria
were employed for self-consistent field (SCF) and convergence, with
the maximum number of SCF cycles set to 250.
sub-
solvent strate
kobs
[H2/D2]
(mM)b
0.781 6.0(2) × 10-2
k2
(s-1
)
(M-1 s-1
)
KIEc
H2O
H2
D2
H2
D2
4.7(2) × 10-5
3.8(2)
1.27(6) × 10-5 0.808 1.57(8) × 10-2
C6H5Cl
7.6(8) × 10-5
2.52
3.0(3) × 10-2
4.5(5)
1.69(8) × 10-5 2.52d
6.7(3) × 10-3
a All of the data are reported for 25 °C and 1 atm of H2 or D2.
b Solubility data taken from ref 22. c Refers to kH /kD . d No solubility
data available for D2 in C6H5Cl.
2
2
The calculated vibrational frequencies for transition states were used
to compute KIEs according to Bigeleisen’s treatment.33 Plots were
created using PLUTON.34
The rates of H2 and D2 oxidation by permanganate in both
solvents are presented in Table 1. The second-order rate
constants (k2), obtained by correcting for the solubility of the
gaseous substrate in the respective solvents,37,38 reveal a minor
solvent effect, with the reaction proceeding approximately twice
as fast in H2O as in C6H5Cl. The rate data correspond to
activation energies (∆G‡) of 19.1 kcal mol-1 for the aqueous
reaction and 19.5 kcal mol-1 for the reaction in C6H5Cl. The
KIEs measured for the oxidation of D2 versus H2 are substantial
and similar in magnitude for the two solvent systems studied:
kH /kD ) 3.8(2) in H2O and 4.5(5) in C6H5Cl. The rates
measured in C6H5Cl are reported as initial rates because the
rate of disappearance of MnO4 was observed to increase as
the reaction progressed. This may be due to the formation of
solvent oxidation products or possibly oxidation products of the
nBu4N+ counterion,39 which are susceptible to further oxidation
by MnO4-. Similar rate accelerations have been reported in the
Results
Kinetic Studies and Isotope Effects. Reaction rates and KIEs
for the oxidation of H2 by metal oxo compounds were examined
in this study as a means of gaining mechanistic insights into
the H2 cleavage reaction. Kinetic data for the reaction of
permanganate with H2 in water have been previously published,20c
but no KIEs were reported. KIE measurements can provide
information about the nature of the rate-determining step of a
reaction and may allow different mechanisms to be distinguished
by comparing KIEs for similar reactions.35 Changes in the
magnitude of a KIE upon switching solvents have been used to
2
2
-
argue for a shift in mechanism in the oxidation of toluene by
- 15a
MnO4
;
thus, we examined the KIE for H2 oxidation by
-
MnO4 in both aqueous and organic solutions.
The kinetics of the reactions of H2 and D2 with dilute
solutions of MnO4- (0.3 mM) were measured at 25 °C by UV-
vis spectroscopy. For the aqueous reactions, KMnO4 was used,
and the water was rigorously purified to ensure that no
oxidizable contaminants were present. No decrease in the UV-
vis absorbance of aqueous KMnO4 solutions over time was
detected in the absence of H2. The tetrabutylammonium salt
nBu4NMnO423 was employed in order to solubilize permanganate
in organic solvents. Organic solvents with sufficient polarity to
dissolve nBu4NMnO4 are generally prone to oxidation; however,
chlorobenzene (C6H5Cl) was found to be a suitable solvent for
these reactions because it is easily purified and reacts slowly
with MnO4- relative to the reactions being studied (rates <9%
of substrate oxidation).
n
oxidations of benzylic C-H bonds by Bu4NMnO4 in organic
solvents.15b The reactions in H2O solvent showed clean pseudo-
first-order behavior with no apparent acceleration over the entire
range of data collected (∼2.5 half-lives), so all of the data were
used in the rate computation. Repeated measurements showed
that the rates were reproducible, with maximum deviations from
the mean of 3.1% in H2O and 6.1% in C6H5Cl.
The reactivities of second- and third-row metal oxo species
toward H2 were also explored. The following systems did not
react with hydrogen over several days at room temperature:
KReO4 in H2O, MeReO3 in H2O, MeReO3 in C6H6, OsO4 in
H2O, and OsO4 in CCl4.
However, RuO4 was found to readily oxidize H2. Dilute CCl4
solutions of RuO4 (∼0.5 mM) darkened to green upon initial
exposure to H2, and black RuO2 eventually began precipitating
from the solution. Efforts to measure the rate of this reaction
were complicated in the early stages of the reaction by the
apparent formation of reduced ruthenium species, which ab-
sorbed more strongly than RuO4 in the 280-500 nm range of
the UV-vis spectrum (see Experimental Section). The observed
spectral changes and the green color of the solutions match the
known properties of colloidal RuO2‚xH2O.40 Within ∼2 h, the
absorption and scattering spectrum of the colloid had disap-
peared, and solid RuO2 appeared to be the sole reduction
-
In both H2O and C6H5Cl solutions, the product of MnO4
reduction is colloidal MnO2, which exhibits a diagnostic
absorption and scattering spectrum.36 Isosbestic points were
observed in the UV-vis spectra during kinetic runs in both
-
solvents, indicating that MnO4 and MnO2 are the only
manganese-containing species present in significant concentra-
tions. Oxidation state determinations for solutions near comple-
tion of the reaction (see Experimental Section) showed the
average oxidation state of manganese to be 3.8(3). Thus, the
balanced reaction for H2 oxidation by permanganate may be
represented by eq 1. For the reaction in C6H5Cl solution, the
H2O and [nBu4N+][OH-] produced are probably associated with
the MnO2 colloid.15b For the aqueous reaction, the use of a pH
) 7 buffer resulted in no significant change in the rate compared
with that of the unbuffered reaction, so the buildup of [OH-]
did not appear to affect the kinetic measurements.
(35) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic
Molecules; Wiley: New York, 1980.
(36) Perez-Benito, J. F.; Arias, C. J. Colloid Interface Sci. 1992, 152, 70-
84.
(37) IUPAC Solubility Data Series: Hydrogen and Deuterium; Young,
C. L., Ed.; Pergamon Press: New York, 1981; Vol. 5/6, p 2 (H2/
H2O), pp 278-279 (D2/H2O), pp 240-241 (H2/C6H5Cl), p 160 (H2/
C6H6), p 287 (D2/C6H6), p 238 (H2/CCl4), p 290 (D2/CCl4).
(38) No solubility data are available for D2 in C6H5Cl. If D2 is assumed to
be 2.5% more soluble than H2, as is the case in C6H6,37 then the KIE
2MnO4- + 3H2 f 2MnO2 + 2H2O + 2OH-
(1)
-
for the oxidation of H2/D2 by MnO4 in C6H5Cl increases to 4.6(5).
-
(32) Strassner, T.; Houk, K. N. J. Am. Chem. Soc. 2000, 122, 7821-7822.
(33) Bigeleisen, J. J. Chem. Phys. 1949, 17, 675-678.
(34) Spek, A. L. PLUTON; available from http://www.cryst.chem.uu.nl/
platon.
(39) PhCOOH has been observed as the product of the reaction of MnO4
with Bu3(PhCH2)N+.15b
n
(40) Mills, A.; McMurray, N. J. Chem. Soc., Faraday Trans. 1 1988, 84,
379-390.

6276 Inorganic Chemistry, Vol. 40, No. 24, 2001
Collman et al.
product; thus, rates were determined from the data collected
beginning at 2.2-2.7 h into the reaction. The pseudo-first-order
rate of H2 oxidation by RuO4 was measured to be 8.2(17) ×
10-5 s-1. The rather poor reproducibility of the measured rates
for four separate trials is reflected in the large uncertainty. The
reproducibility of the D2 oxidation rates was even worse, but
an average of six trials gives a rate of 4.6(12) × 10-5 s-1. The
variation in rates may be due to the presence of varying amounts
of water in the reaction mixtures or to heterogeneous effects
related to the RuO2 product, and these effects are more
pronounced in the slower D2 reaction.
Scheme 1
The solubilities of H2 and D2 in CCl4 were corrected for (3.3
and 3.4 mM, respectively, at 1 atm/298 K),37 and second-order
rate constants of 2.5(5) × 10-2 M-1 s-1 for H2 and 1.4(4) ×
10-2 M-1 s-1 for D2 were obtained; thus, the rates were very
close to those of the corresponding MnO4- reactions. The KIE
for the reaction is 1.8(6), but this value must be considered
approximate in view of the reproducibility problems.
is also characteristic of the hydrogenase enzymes.46 Therefore,
the KMnO4 experiments were repeated at pH ) 3.0, 5.5, 7.0,
and 9.9, but still no H/D exchange was observed. The addition
of 50 mM Ag+ (as AgNO3), which is known to catalyze the
oxidation of H2 by MnO4,20b,c also did not result in detectable
H/D exchange.
H/D Exchange Experiments. In the course of these kinetic
studies, we recognized the possibility of dihydrogen σ-complex
intermediates preceding bond cleavage. Previous studies of H2
-
oxidation by MnO4 predated the discovery of H2 complexes
of transition metals,41 so this possibility had not been considered.
Such hydrogen adducts of high-valent metal oxo species are
reasonable intermediates because amine adducts of OsO4 are
well-known,42 and kinetic evidence has been found for thioether
Because dihydrogen complexes of second- and third-row
transition metals are known to be more stable than those of
first-row metals,47 the metal oxo species RuO4, OsO4, and
ReO4- as well as the related organometallic compound MeReO3
were examined for H/D exchange activity. For highly reactive
RuO4, a catalytic biphasic system18c,28 was employed (see
Experimental Section); for the other compounds, 50 mM
aqueous solutions were used as in the MnO4- experiments. No
H/D exchange was observed in these systems. In addition, D2O
solutions of OsO4, KReO4, and MeReO3 under 1 atm of H2
-
complexes of MnO4 as intermediates in the oxidation of
organic sulfides.43
As a probe for the existence of dihydrogen complexes of
metal oxo compounds, we examined several systems for
catalytic H/D exchange between H2O solvent and D2. This
reaction, which is characteristic of a number of known dihy-
drogen complexes,44 could occur via the loss of D+ from an
acidic D2 complex followed by the protonation and release of
H-D (Scheme 1). Such H/D exchange could also signify the
existence of a metastable metal hydride intermediate.45 In a
typical permanganate experiment, a 50 mM solution of KMnO4
in H2O was degassed and placed under 1 atm of D2, and samples
of gas were withdrawn periodically over several days for
analysis using a cryogenic GC column capable of resolving H2,
1
were examined by H NMR for changes in the chemical shift
or T1 of H2, which would constitute evidence of an H2 complex,
but no such changes were apparent.
RuO4-Catalyzed Oxidation under H2. As a further test for
the intermediacy of H2 complexes of metal oxos, the catalytic
hydroxylation of adamantane was examined in the presence of
-
H2. The oxidation of adamantane is known to show clean
HD, and D2.27 The disappearance of the purple color of MnO4
kinetics over several half-lives,18c-e and the presence of an
intermediate H2 complex of RuO4 in significant concentrations
would be expected to inhibit hydrocarbon oxidation because of
the sequestration of the active catalyst. Hydroxylation rates
determined by monitoring the disappearance of adamantane in
the presence of catalytic RuO4 under 1 atm of H2 and in the
absence of H2 were identical (kobs ) 9.3(5) × 10-6 s-1 at 20
mol % RuO4; see Experimental Section). As previously
reported,18c,d the major oxidation product was 1-adamantanol,
with a small amount of 2-adamantanone (∼2%) also produced.
Because RuO4 reacts with H2 as well as adamantane (vide
supra), the lack of inhibition indicates that reoxidation of reduced
ruthenium species to RuO4 is not rate-limiting in the catalytic
cycle and that an H2 σ-complex intermediate is present in
negligible concentrations or not at all.
and appearance of brown MnO2 indicated that permanganate
was reduced by a reaction with D2, but no traces of HD or H2
were detected. Some dihydrogen complexes exhibit maximum
H/D exchange rates at a particular pH,44c,d a phenomenon which
(41) (a) Kubas, G. J. Acc. Chem. Res. 1988, 21, 120-128. (b) Heinekey,
D. M.; Oldham, W. J. Chem. ReV. 1993, 93, 913-926.
(42) (a) Criegee, R.; Marchand, B.; Wannowius, H. Justus Liebigs Ann.
Chem. 1949, 550, 99. (b) Cleare, M. J.; Hydes, P. C.; Griffith, W. P.;
Wright, M. J. J. Chem. Soc., Dalton Trans. 1977, 941-944. (c)
Griffith, W. P.; Skapski, A. C.; Woode, K. A.; Wright, M. J. Inorg.
Chim. Acta 1978, 31, L413-L414. (d) Svendsen, J. S.; Marko´, I.;
Jacobsen, E. N.; Rao, C. P.; Bott, S.; Sharpless, K. B. J. Org. Chem.
1989, 54, 2263-2264. (e) Nelson, D. W.; Gypser, A.; Ho, P. T.; Kolb,
H. C.; Kondo, T.; Kwong, H.-L.; McGrath, D. V.; Rubin, A. E.;
Norrby, P.-O.; Gable, K. P.; Sharpless, K. B. J. Am. Chem. Soc. 1997,
119, 1840-1858.
(43) Xie, N.; Binstead, R. A.; Block, E.; Chandler, W. D.; Lee, D. G.;
Meyer, T. J.; Thiruvazhi, M. J. Org. Chem. 2000, 65, 1008-1015.
(44) (a) Albeniz, A. C.; Heinekey, D. M.; Crabtree, R. H. Inorg. Chem.
1991, 30, 3632-3635. (b) Kubas, G. J.; Burns, C. J.; Khalsa, G. R.
K.; Van Der Sluys, L. S.; Kiss, G.; Hoff, C. D. Organometallics 1992,
11, 3390-3404. (c) Collman, J. P.; Wagenknecht, P. S.; Hembre, R.
T.; Lewis, N. T. J. Am. Chem. Soc. 1990, 112, 1294-1295. (d)
Collman, J. P.; Wagenknecht, P. S.; Hutchison, J. E.; Lewis, N. S.;
Lopez, M. A.; Guilard, R.; L’Her, M.; Bothner-By, A. A.; Mishra, P.
K. J. Am. Chem. Soc. 1992, 114, 5654-5664.
(46) (a) Farkas, A.; Farkas, L.; Yudkin, J. Proc. R. Soc. London, Ser. A
1934, B115, 373. (b) Lespinat, P. A.; Berlier, Y.; Fauque, G.;
Czechowski, M.; Dimon, B.; LeGall, J. Biochimie 1986, 68, 55-61.
(c) Arp, D. J.; Burris, R. H. Biochim. Biophys. Acta 1982, 700, 7-15.
(d) Teixeira, M.; Fauque, G.; Moura, I.; Lespinat, P. A.; Berlier, Y.;
Prickril, B.; Peck, H. D.; Peck, H. D., Jr.; Xavier, A. V.; LeGall, J.;
Moura, J. J. G. Eur. J. Biochem. 1987, 167, 47-58.
(45) Stahl, S. S.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 1996,
118, 5961-5976.
(47) Millar, J. M.; Kastrup, R. V.; Nelchior, M. T.; Horva´th, I. T.; Hoff,
C. D.; Crabtree, R. H. J. Am. Chem. Soc. 1990, 112, 9643-9645.

Dihydrogen Cleavage by Metal Oxo Compounds
Inorganic Chemistry, Vol. 40, No. 24, 2001 6277
Scheme 2
are precluded by the KIEs and lack of solvent dependence on
rates and will not be considered here. Evaluation of the four
mechanistic possibilities described above in light of the
experimental evidence allows some of them to be discounted
as implausible.
-
Hydrogen atom abstraction has been proposed for the MnO4
oxidations of benzylic carbon-hydrogen bonds on the basis of
a number of mechanistic studies.7,8,15 The strongest evidence
against a radical mechanism in the present case is simply that
the oxidation of dihydrogen proceeds so rapidly. Compared with
- 15
the oxidation of toluene by MnO4
,
the reaction of H2 with
MnO4 or RuO4 is ∼105 times faster. Because the H-H bond
of dihydrogen (BDE ) 104 kcal mol-1) is considerably stronger
than the benzylic C-H bond of toluene (BDE ) 92 kcal
mol-1),21 it is improbable that these reactions are both radical
abstractions. Toluene is a special case among hydrocarbons in
that benzylic stabilization of the radical may play a role, but
this should make hydrogen abstraction more facile rather than
slower relative to a more thermodynamically robust substrate
such as hydrogen. Experimental data for free radicals such as
•OH52,53 and halogen atoms54 show that the rates of hydrogen
atom abstraction from H2 are virtually identical (i.e., different
by less than a factor of 5) to the rates of H• abstraction from
CH4, which has a C-H bond comparable in strength to the H-H
bond (BDE ) 104.9 kcal mol-1).21 By contrast, hydrogen
abstraction from toluene by •OH occurs near the diffusion-
controlled limit, 102-103 times faster than the reaction of •OH
with H2.55 These data lead to the conclusion that bond strength
is the primary influence on the activation energy in a simple
radical abstraction rather than orbital differences between H-H
and C-H bonds. Arguments correlating H• abstraction rates with
a thermodynamic driving force, which were employed by Mayer
and co-workers to argue for a radical mechanism in the reaction
of MnO4- with toluene,15 would predict such a high activation
energy for H• abstraction from H2 that the reaction could not
occur.56
-
Discussion
Mechanism of H2 Cleavage. The kinetic studies show that
dihydrogen undergoes facile oxidation by MnO4- and RuO4 at
room temperature, with activation energies (∆G‡) of 19.1-19.5
kcal mol-1 for MnO4 and 19.6 kcal mol-1 for RuO4. These
-
reactions are faster than the oxidations of benzylic C-H bonds
by permanganate15 and the oxidations of tertiary C-H bonds
of saturated hydrocarbons by RuO4.18 The substantial KIEs
observed for the dihydrogen oxidation reactions are consistent
with a rate-determining step involving the cleavage of the H-H
bond. The similarity of the KIEs for H2 versus D2 oxidation by
MnO4- in H2O and C6H5Cl solutions as well as the lack of an
appreciable solvent effect on absolute reaction rates suggests a
common mechanism in organic and aqueous solutions. The
absence of catalytic H/D exchange in these systems and the
lack of inhibition of hydrocarbon oxidation by H2 indicate that
dihydrogen complexes, at least, acidic ones, are unlikely
intermediates.
Hydride transfer has been implicated in the MnO4- oxidations
of alcohols59 and in the aqueous MnO4- oxidation of toluene.15a
However, a polar mechanism such as this would be expected
In view of the unreactivity of third-row transition-metal oxo
compounds such as OsO4, ReO4-, and MeReO3 toward dihy-
-
drogen, it is clear that MnO4 and RuO4 possess a special
reactivity among metal tetraoxo compounds. It has been noted
by Holm and co-workers48 that greater thermodynamic stabiliza-
tion of the higher oxidation states and stronger metal oxo bonds
are two important factors contributing to the diminished
reactivity of third-row versus second-row transition-metal oxo
compounds.
(51) (a) Ostrovic, D.; Bruice, T. C. Acc. Chem. Res. 1992, 25, 314-320.
(b) Kim, T.; Mirafzal, G. A.; Liu, J.; Bauld, N. L. J. Am. Chem. Soc.
1993, 115, 7653-7664. (c) Gross, Z.; Nimri, S. J. Am. Chem. Soc.
1995, 117, 8021-8022.
(52) For gas phase where (a) k(•OH + H2) ) 3.5(2) × 106 M-1 s-1 (295
K), see: Overend, R. P.; Paraskevopoulos, G.; Cvetanovic, R. J. Can.
J. Chem. 1975, 53, 3374-3382. For (b) k(•OH + CH4) ) 3.4(3) ×
106 M-1 s-1 (293 K), see: Dunlop, J. R.; Tully, F. P. J. Phys. Chem.
1993, 97, 11148-11150.
Because the rate-determining H-H cleavage is the only
-
kinetically accessible step in the oxidation of H2 by MnO4
(53) For aqueous solution where k(•OH + H2) ) 4.2 × 107 M-1 s-1 (298
K) and k(•OH + CH4) ) 1.1 × 108 M-1 s-1 (298 K), see: Buxton,
G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys. Chem.
Ref. Data 1988, 17, 513.
and RuO4, the mechanistic discussion will focus on this bond
scission event. Four plausible mechanisms for the reaction of
H2 with a metal oxo are shown in Scheme 2: (1) the abstraction
of an H• radical, constituting a 1e- reduction of the metal, (2)
hydride transfer, resulting in a 2e- reduction of the metal, (3)
a concerted [2 + 2] addition across an MdO bond, with no
change in the oxidation state of the metal, and (4) a concerted
[3 + 2] addition involving two MdO functionalities, resulting
in a 2e- reduction of the metal. Mechanisms involving an initial
electron transfer to MnO4- followed by bond cleavage, such as
those that have been discussed for the oxidations of organic
(54) For gas phase at 298 K where k(Cl• + H2) ) 7.7 × 106 M-1 s-1
,
k(Cl• + CH4) ) 3.6 × 107 M-1 s-1, k(Br• + H2) ) 1.0 × 10-3 M-1
s-1, and k(Br• + CH4) ) 2.2 × 10-3 M-1 s-1, see: Kerr, J. A. In
Selected Elementary Reactions; Bamford, C. H., Tipper, C. F. H., Eds.;
Comprehensive Chemical Kinetics 18; Elsevier: Amsterdam, 1976;
pp 45-51.
(55) For aqueous solution where k(•OH + CH3Ph) ) 3 × 109 M-1 s-1 54
.
(56) The fitting of Mayer’s data for benzylic hydrogen abstractions by nBu4-
NMnO4 to the Polanyi relation ∆G‡ ) â + R(∆H°)57 and the
15b
-
extrapolation of the line to 24 kcal mol-1 (∆H° for MnO4 + H2 f
O3MnOH- + H•, assuming D(MnO-H) ) 80 kcal mol-1) gives k1
≈ 10-16 M-1 s-1. The actual rate constant should be even smaller
because R f 1 for very endothermic reactions.58
compounds by MnO4 ,
- 49 OsO4,50 and oxometalloporphyrins,51
(48) (a) Tucci, G. C.; Donahue, J. P.; Holm, R. H. Inorg. Chem. 1998, 37,
1602-1608. (b) Sung, K.-M.; Holm, R. H. J. Am. Chem. Soc. 2001,
123, 1931-1943.
(49) Wiberg, K. B.; Freeman, F. J. Org. Chem. 2000, 65, 573-576.
(50) Wallis, J. M.; Kochi, J. K. J. Am. Chem. Soc. 1988, 110, 8207-8223.
(57) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1936, 32, 1333-
1360.
(58) Pross, A. Theoretical and Physical Principles of Organic ReactiVity;
Wiley: New York, 1995; pp 175-182.
(59) Stewart, R. J. Am. Chem. Soc. 1957, 79, 3057-3061.

6278 Inorganic Chemistry, Vol. 40, No. 24, 2001
Collman et al.
Figure 1. B3LYP/6-311+G** transition states and products for H2
-
Figure 2. Calculated energy surfaces for [3 + 2] and [2 + 2] addition
cleavage by MnO4
.
-
of H2 to MnO4
.
to give rise to a substantial solvent effect. In the present study
of hydrogen oxidation by MnO4-, only a twofold increase in
rate was observed on changing from C6H5Cl (ꢀ ) 5.6) to the
much more polar solvent H2O (ꢀ ) 78.5).
Optimized transition states were checked using unrestricted
single-point calculations, and both the [2 + 2] and [3 + 2]
transition states were calculated to lie on singlet energy surfaces.
Vibrational analyses confirmed that each transition state pos-
sessed one imaginary frequency. No transition states leading to
hydrogen atom or hydride abstraction were found. Furthermore,
no intermediates preceding H2 cleavage (i.e., H2 σ complexes)
were located along the reaction coordinates.
A concerted mechanism thus appears to fit the experimental
evidence best. Concerted [2 + 2] addition of a C-H bond to a
2-
manganese oxo has been previously suggested for MnO4
oxidations,60 and a [3 + 2] mechanism has been proposed for
the RuO4 oxidations of alkanes.18c-e Because the [3 + 2] and
[2 + 2] addition pathways are kinetically indistinguishable, we
turned to computational modeling in order to assess which of
these mechanisms is more plausible for the oxidation of
dihydrogen.
The computed energetics of the reaction are illustrated in
Figure 2. The [3 + 2] addition is calculated to be exothermic
by 46.0 kcal mol-1, while the [2 + 2] pathway is endothermic
by 32.1 kcal mol-1. The calculated activation energies are 61.9
kcal mol-1 for [2 + 2] addition and 15.4 kcal mol-1 for [3 +
2] addition. The latter value compares favorably with the ∆H‡
of 14.0 kcal mol-1 calculated from the rate data for the oxidation
of H2 in H2O.62 Thus, the model predicts that [3 + 2] addition
is strongly favored both kinetically and thermodynamically. This
result can be rationalized by considering that the presence of
multiple oxo ligands, which are strong σ and π donors, disfavors
an increase in the coordination number of the manganese center.
Computational Studies. The reaction of H2 with MnO4- was
examined computationally using the hybrid density functional/
Hartree-Fock model B3LYP29,30 with the 6-311+G(d,p) basis
set (see Experimental Section). This level of theory, which
employs a large basis set with diffuse functions for improved
modeling of anionic species, has been used successfully in
-
studies of MnO4 reactions with alkenes61 and toluene,32
predicting activation enthalpies within 2 kcal mol-1 of the
experimental values.
-
Given the similarity of the H2 oxidation rates for MnO4 and
RuO4, it is reasonable to propose that both reactions proceed
by the same concerted mechanism. Calculations for OsO4 and
Transition states and “products” were located for the two
concerted pathways. The [2 + 2] addition transition state is
characterized by a significant degree of O-H bond forming,
with one O-H distance of 1.283 Å, and only a very weak
interaction (2.317 Å) between the second hydrogen atom and
the manganese center (Figure 1a). The manganese(VII) species
resulting from the [2 + 2] addition, [O3Mn(OH)H]- (Figure
1b), possesses a distorted trigonal-bipyramidal geometry, with
the hydride and one oxo ligand occupying axial positions
( Oax-Mn-H ) 149.4°). The manganese hydride bond (1.606
Å) is significantly lengthened compared with the sum of the
covalent radii (1.48 Å). The transition state for [3 + 2] addition
shows a lesser degree of O-H bond forming (O-H distances
1.502 Å) but features a dissociating H2 unit which is sym-
metrically placed between two oxo functionalities (Figure 1c).
The [3 + 2] addition step leads to the distorted tetrahedral
manganese(V) species [O2Mn(OH)2]- (Figure 1d), which pos-
sesses two MnOH groups with an HO-Mn-OH angle of 86.6°.
-
ReO4 identify the [3 + 2] mechanism as the energetically
favored pathway but predict significantly higher activation
energies,63 consistent with the experimentally observed lack of
reactivity of these compounds toward H2.
Parallels with the calculated energy surface for the carbon-
hydrogen activation of toluene by MnO4- are noteworthy. The
B3LYP-computed transition state for this reaction32 clearly
shows the involvement of two oxo groups, with one oxo forming
an O-H bond and the other weakly interacting with the benzylic
carbon of toluene. The transition state resembles a radical pair
yet lies on a singlet energy surface. The collapse of the radical
pair to give [O2Mn(OH)(OCH2Ph)]- (the analogue of the [3 +
-
2] product in the reaction of MnO4 with H2) occurs with no
barrier and no formation of free radicals. Experimental observa-
(62) Calculated using ∆G‡ ) 19.1 kcal mol-1 from rate data in the present
study and ∆S‡ ) -17 eu from Halpern’s Eyring plot.20d The ∆H‡ for
the reaction in C6H5Cl is calculated to be 14.4 kcal mol-1 using the
same ∆S‡ value, and Halpern’s data for the aqueous reaction gave
(60) Lee, D. G.; Chen, T. J. Am. Chem. Soc. 1993, 115, 11231-11236.
(61) (a) Houk, K. N.; Strassner, T. J. Org. Chem. 1999, 64, 800-802. (b)
Strassner, T.; Busold, M. J. Org. Chem. 2001, 66, 672-676.
∆H‡ ) 13.8 kcal mol-1
.
(63) Strassner, T., paper in preparation.

Dihydrogen Cleavage by Metal Oxo Compounds
Inorganic Chemistry, Vol. 40, No. 24, 2001 6279
tions of the partial loss of stereochemistry in benzylic oxidations
by permanganate7,8 can be explained by invoking the existence
of a separate pathway for carbon-based radical dissociation from
the singlet radical pair, which occurs at a finite rate. The toluene
oxidation transition state is obviously related to the [3 + 2]
transition state for the H2 cleavage calculated in this study; the
H2 addition transition state is “earlier” and more symmetric by
virtue of the greater exothermicity of the reaction and the
potential to form two MnO-H bonds. These differences result
in a purely concerted mechanism for the H-H activation, while
the benzylic C-H activation proceeds by a more asynchronous
mechanism resembling a hydrogen abstraction/oxygen rebound,
though the essential features of the two reactions are similar.
KIE Magnitudes. The KIEs of 3.8-4.5 for H2 versus D2
oxidation by MnO4- are substantial.64 The largest published KIE
for a related solution-phase reaction is 2.5, measured for the
hydrogenolysis of the neopentyl ligand of (C5Me5)2Th(CH2-
CMe3)(OCH2CMe3).72 Typical KIEs for the oxidative addition
of H2 to coordinatively unsaturated metal centers are smaller,
usually falling in the range 1.05-1.5.73 The reported KIE for
the gas-phase reaction of •OH radicals with H2/D2 is 2.6.74
Large primary KIEs are typically attributed to two factors:75
(1) similar degrees of bond formation and bond breakage in
the transition state and (2) the linear transfer of a hydrogen atom.
The energetically favored [3 + 2] addition transition state for
involves the transfer of two hydrogen atoms rather than just
one, so it is reasonable to question the validity of standard
rationalizations of KIE magnitudes on the basis of simple three-
atom models75 in this case.
We examined the KIE for H2/D2 activation by MnO4- using
our computational model. Calculated vibrational frequencies for
the H2 [3 + 2] addition transition state and the D2 analogue76
yielded a calculated KIE of 1.63,77 considerably lower than the
experimentally measured values. Even the use of a frequency
scale factor of 0.9, such as has been commonly employed to
make Hartree-Fock-computed frequencies more realistic,80,81
only increased the calculated KIE to 2.34, still in poor agreement
with experiment. The discrepancy could result from a systematic
error in the calculations, for example, poor modeling of low-
frequency isotopically sensitive vibrations.83 No significant
problems are apparent in the analysis of the experimental rate
data, and the use of an alternative method of rate calculation
gave similar KIE values.84 Another possible explanation for the
disparity is tunneling, which has most notably been invoked to
account for extraordinarily large KIEs of 20-50 in enzymatic
C-H activations85,86 but has also been used to explain H2
addition KIEs which were only moderately larger than those
predicted by classical models.73c Application of the standard
Wigner correction87 with the unscaled frequencies results in a
small but significant increase of the calculated KIE to 2.20. This
correction is based on an X-H-Y situation in the transition
state, and it is conceivable that the unusual X-H-H-Y
configuration in the [3 + 2] transition state might give rise to
a larger tunnel effect. However, pending the development of
-
MnO4 + H2 is relatively early (r(H-H) ) 0.876 Å) for the
computational model employed in this study and does not feature
linear hydrogen transfer. However, the reaction of interest
(64) KIEs for H2/D2 cleavages are generally significantly smaller than those
for the corresponding C-H/C-D scissions. Several examples of metal-
mediated C-H activation KIEs with magnitudes of 8-13 have been
reported.7,15a,19,65-69
(76) Experimental frequencies were used for H2 (4395 cm-1) and D2 (3118
cm-1). Moelwyn-Hughes, E. A. Physical Chemistry; Pergamon
Press: New York, 1961; p 427.
(65) A theoretical study by Abu-Hasanayn et al.70 suggested that the small
magnitudes of H2/D2 cleavage KIEs are due to the appearance of five
new isotopically sensitive vibrational modes in the transition state
relative to free H2. These vibrations produce significant inverse (<1)
EXC and ZPE factors which compensate for the large H2/D2 mass
moment of inertia (MMI) term of ∼5.7. Early transition states have
also been invoked to explain the small KIEs.71
(77) The same value is obtained using either the Bigeleisen expression33
or the Redlich-Teller product rule.78 The factored energetic contribu-
tions (KIE ) MMI × EXC × ZPE) are as follows: MMI ) 5.07,
EXC ) 0.78, ZPE ) 0.41. A previous study of equilibrium isotope
effects for hydrocarbon addition79 has shown that large substrate MMI
terms are effectively canceled by inverse EXC and ZPE factors from
“new” isotopically sensitive vibrational modes appearing upon sub-
strate binding or activation, which correspond to remnants of free
substrate rotation and translation. Accordingly, new isotopically
sensitive vibrations in the [3 + 2] H2 addition transition state account
for the entire EXC term of 0.78 and a factor of 0.11 in the ZPE term,
leaving a ZPE factor of 3.75 from the weakened H-H stretch.
(78) Schaad, L. J.; Bytautas, L.; Houk, K. N. Can. J. Chem. 1999, 77,
875-878.
(66) (a) Hjelmeland, L. M.; Aronow, L.; Trudell, J. R. Biochem. Biophys.
Res. Commun. 1977, 76, 541-549. (b) Groves, J. T.; McClusky, G.
A.; White, R. E.; Coon, M. J. Biochem. Biophys. Res. Commun. 1978,
81, 154-160. (c) Miwa, G. T.; Walsh, J. S.; Lu, A. Y. H. J. Biol.
Chem. 1984, 259, 3000-3004. (d) Lindsay-Smith, J. R.; Sleath, P. R.
J. Chem. Soc., Perkin Trans. 2 1983, 621-628. (e) Jones, J. P.;
Korzekwa, K. R.; Rettie, A. E.; Trager, W. F. J. Am. Chem. Soc. 1986,
108, 7074-7078. (f) Jones, J. P.; Trager, W. F. J. Am. Chem. Soc.
1987, 109, 2171-2173. (g) Jones, J. P.; Trager, W. F. J. Am. Chem.
Soc. 1988, 110, 2018. (h) Jones, J. P.; Rettie, A. E.; Trager, W. F. J.
Med. Chem. 1990, 33, 1242-1246.
(79) Slaughter, L. M.; Wolczanski, P. T.; Klinckman, T. R.; Cundari, T.
R. J. Am. Chem. Soc. 2000, 122, 7953-7975.
(80) (a) Flock, M.; Ramek, M. Int. J. Quantum Chem. 1993, 27, 331-
336. (b) Harris, N. J. J. Phys. Chem. 1995, 99, 14689-14699. (c)
Cundari, T. R.; Raby, P. D. J. Phys. Chem. 1997, 101, 5783-5788.
(81) Frequency scale factors required to bring density functional theory
calculated frequencies into close agreement with the experiment are
generally very close to unity,82 and thus, these frequencies are typically
reported unscaled.
(67) (a) Nappa, M. J.; McKinney, R. J. Inorg. Chem. 1988, 27, 3740-
3745. (b) Sorokin, A. B.; Khenkin, A. M. J. Chem. Soc., Chem.
Commun. 1990, 45-46. (c) Sorokin, A.; Robert, A.; Meunier, B. J.
Am. Chem. Soc. 1993, 115, 7293-7299.
(68) Wayland, B. B.; Ba, S.; Sherry, A. E. J. Am. Chem. Soc. 1991, 113,
5305-5311.
(82) Rauhut, G.; Pulay, P. J. Phys. Chem. 1995, 99, 3093-3100.
(83) According to the density functional theory model, low-frequency
vibrations result in strongly inverse EXC and ZPE factors which
overcompensate for the large MMI term (see ref 77).
(69) Schaller, C. P.; Cummins, C. C.; Wolczanski, P. T. J. Am. Chem.
Soc. 1996, 118, 591-611.
(70) Abu-Hasanayn, F.; Goldman, A. S.; Krogh-Jespersen, K. J. Phys.
Chem. 1993, 97, 5890-5896.
(84) Initial rates calculated by polynomial fitting of Abs versus time data26
(see Experimental Section) gave KIEs of 4.3 for C6H5Cl solution and
3.1 for aqueous solution, though the calculated uncertainties were large
(35% and 55%, respectively).
(71) Collman, J. P.; Roper, W. R. AdV. Organomet. Chem. 1967, 7, 53-
94.
(72) Lin, Z.; Marks, T. J. J. Am. Chem. Soc. 1987, 109, 7979-7985.
(73) (a) Chock, P. B.; Halpern, J. J. Am. Chem. Soc. 1966, 88, 3511. (b)
Vaska, L.; Wernecke, F. Trans. N.Y. Acad. Sci. 1971, 33, 70. (c) Zhou,
P.; Vitale, A. A.; San Filippo, J.; Saunders, W. H. J. Am. Chem. Soc.
1985, 107, 8049-8054. (d) Bullock, R. M. In Transition Metal
Hydrides; Dedieu, A., Ed.; VCH: New York, 1992; pp 263-307.
(74) Paraskevopoulos, G.; Nip, W. S. Can. J. Chem. 1980, 58, 2146-
2149.
(75) (a) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic
Chemistry; Harper & Row: New York, 1987; pp 236-238. (b)
Carpenter, B. K. Determination of Organic Reaction Mechanisms;
Wiley: New York, 1984; pp 83-111.
(85) (a) Nesheim, J. C.; Lipscomb, J. D. Biochemistry 1996, 35, 10240-
10247. (b) Valentine, A. M.; Stahl, S. S.; Lippard, S. J. J. Am. Chem.
Soc. 1999, 121, 3876-3887.
(86) (a) Glickman, M. H.; Wiseman, J. S.; Klinman, J. P. J. Am. Chem.
Soc. 1994, 116, 793-794. (b) Glickman, M. H.; Klinman, J. P.
Biochemistry 1995, 34, 14077-14092. (c) Hwang, C.-C.; Grissom,
C. B. J. Am. Chem. Soc. 1994, 116, 795-796. (d) Lewis, E. R.;
Johansen, E.; Holman, T. R. J. Am. Chem. Soc. 1999, 121, 1395-
1396.
(87) (a) Wigner, E. Z. Phys. Chem., Abt. B. 1932, 19, 203. (b) Bell, R. P.
Tunnel Effect in Chemistry; Chapman & Hall: New York, 1980.

6280 Inorganic Chemistry, Vol. 40, No. 24, 2001
Collman et al.
-
Scheme 3
for the oxidation of H2 by MnO4 and RuO4 implicate a rate-
determining H-H cleavage, though the origins of the large KIE
magnitudes of 3.8-4.5 for the MnO4- reactions remain unclear.
Though the rates of H2 oxidation by the high-valent metal
oxos studied here are much faster than those of the correspond-
ing oxidations of C-H bonds, similarities in the B3LYP-
calculated energy surfaces for these reactions, coupled with the
success of these models in predicting activation energies, offer
hope that a unified mechanistic picture can be devised which
encompasses both H2 and hydrocarbon activations. The con-
certed [3 + 2] addition characteristic of H2 cleavage can be
transformed into a C-H activation mechanism resembling
hydrogen atom abstraction by replacing one O-H interaction
with a very weak O-C interaction in the transition state and
increasing the degree of bond forming in the other O-H
interaction. While one oxo group rather than two is presumed
to be involved in cytochrome P450-type metalloporphyrin-
catalyzed oxidations, the nonsynchronous concerted mechanism,
which has recently been proposed as an alternative to the radical
abstraction/oxygen rebound mechanism in these systems,9 is
clearly related to the asynchronous mechanism of C-H activa-
tion of toluene by MnO4-. Considered together with these
hydrocarbon activation reactions, the concerted reaction of H2
an appropriate tunneling model, it is best to consider the origin
of the unusually large measured KIEs as an open question. The
excellent agreement between the experimental and calculated
activation energies still strongly favors the [3 + 2] addition
mechanism.
Overall Reaction. Assuming that the rate-determining H-H
-
cleavage is a [3 + 2] addition, the overall reaction of MnO4
with H2 to give MnO2 and H2O may be represented by the
sequence of reactions in Scheme 3. The product of the first step
is the manganese(V) intermediate [O2Mn(OH)2]-, which could
rapidly lose water to yield MnO3-. Manganese(V) species of
this type are known to be extremely reactive,88 and this
intermediate would readily oxidize another equivalent of H2,
the solvent, or a nBu4N+ counterion.39 The latter two reactions
might lead to the formation of species which would react rapidly
with MnO4-, accounting for the rate acceleration observed in
the C6H5Cl solution. The RuO4 reaction might proceed by an
analogous series of steps, with an RuO3 intermediate rapidly
oxidizing a second equivalent of H2 to give the observed RuO2
product.
-
with MnO4 and RuO4 adds to the growing body of evidence
which suggests that a more sophisticated mechanistic description
is required for metal oxo mediated oxidations than simple,
stepwise hydrogen abstraction.
Acknowledgment. We gratefully acknowledge the NSF
(Grant No. CHE9612725) for financial support of this research.
T.S. thanks Prof. Wolfgang Hermann (TUM) for research
support and the Leibniz-Rechenzentrum Mu¨nchen for computer
time.
Conclusions
-
We have examined the oxidation of dihydrogen by MnO4
and other high-valent metal oxo compounds in order to gain
insights into the mechanism of this unusual reaction. The rapid
reaction rates and lack of significant solvent effects argue against
radical or hydride abstraction mechanisms, and computational
modeling of the MnO4 reaction supports a concerted [3 + 2]
addition of H2 involving two oxo ligands. The KIEs measured
Supporting Information Available: Representative kinetic plots
for H2 oxidation by MnO4 (in H2O and C6H5Cl) and by RuO4 (in
-
CCl4); Gaussian98 archive files for calculated H2 + MnO4- transition
states and products, containing optimized geometries and thermochemi-
cal outputs; and assignments of calculated vibrational modes for H2/
D2 + MnO4- [3 + 2] addition transition states. This material is available
free of charge via the Internet at http://pubs.acs.org.
-
(88) Zahonyi-Budo, E.; Simandi, L. I. Inorg. Chim. Acta 1996, 248, 81-
84.
IC010639J