Oxidation of C-H bonds by [(bpy)2(py)RuIVO] 2+ occurs by hydrogen atom abstraction

Article abstract of DOI:10.1021/ja035276w

Anaerobic oxidations of 9,10-dihydroanthracene (DHA), xanthene, and fluorene by [(bpy)₂(py)Ru^IVO]²⁺ in acetonitrile solution give mixtures of products including oxygenated and non-oxygenated compounds. The products include those formed by organic radical dimerization, such as 9,9′-bixanthene, as well as by oxygen-atom transfer (e.g., xanthone). The kinetics of these reactions have been measured. The kinetic isotope effect for oxidation of DHA vs DHA-d₄ gives k _H/k_D ≥ 35 ± 1. The data indicate a mechanism of initial hydrogen-atom abstraction forming radicals that dimerize, disproportionate and are trapped by the oxidant. This mechanism also appears to apply to the oxidations of toluene, ethylbenzene, cumene, indene, and cyclohexene. The rate constants for H-atom abstraction from these substrates correlate well with the strength of the C-H bond that is cleaved. Rate constants for abstraction from DHA and toluene also correlate with those for oxygen radicals and other oxidants. The rate constant for H-atom transfer from toluene to [(bpy)₂(py)Ru^IVO]²⁺ appears to be close to that predicted by the Marcus cross relation, using a tentative rate constant for hydrogen atom self-exchange between [(bpy)₂(py)Ru ^IIIOH]²⁺ and [(bpy)₂(py)Ru^IVO] ²⁺.

Full text of DOI:10.1021/ja035276w

Published on Web 08/02/2003
IV
2+
2
Oxidation of C-H Bonds by [(bpy) (py)Ru O] Occurs by
Hydrogen Atom Abstraction
Jasmine R. Bryant and James M. Mayer*
Contribution from the Department of Chemistry, Campus Box 351700, UniVersity of Washington,
Seattle, Washington 98195-1700
Received March 21, 2003; E-mail: mayer@chem.washington.edu.
2
Abstract: Anaerobic oxidations of 9,10-dihydroanthracene (DHA), xanthene, and fluorene by [(bpy) (py)-
IV
2+
Ru O] in acetonitrile solution give mixtures of products including oxygenated and non-oxygenated
compounds. The products include those formed by organic radical dimerization, such as 9,9′-bixanthene,
as well as by oxygen-atom transfer (e.g., xanthone). The kinetics of these reactions have been measured.
4 H D
The kinetic isotope effect for oxidation of DHA vs DHA-d gives k /k g 35 ( 1. The data indicate a
mechanism of initial hydrogen-atom abstraction forming radicals that dimerize, disproportionate and are
trapped by the oxidant. This mechanism also appears to apply to the oxidations of toluene, ethylbenzene,
cumene, indene, and cyclohexene. The rate constants for H-atom abstraction from these substrates correlate
well with the strength of the C-H bond that is cleaved. Rate constants for abstraction from DHA and toluene
also correlate with those for oxygen radicals and other oxidants. The rate constant for H-atom transfer
IV
2+
from toluene to [(bpy)
2
(py)Ru O] appears to be close to that predicted by the Marcus cross relation,
III
2+
using a tentative rate constant for hydrogen atom self-exchange between [(bpy)
2 2
(py)Ru OH] and [(bpy) -
IV
2+
(
py)Ru O] .
IV
2+
2+
Introduction
Ru O] (RuO ) are powerful and versatile oxidants. The
2
+
properties of RuO and its related hydroxo and aquo deriva-
tives, [(bpy)2(py)Ru OH] (RuOH ) and [(bpy)2(py)Ru -
OH2] (RuOH22+), have been described in detail.20 RuO
has been reported to oxidize substrates by a variety of mech-
anisms, including electron transfer, proton-coupled electron
transfer, hydrogen atom abstraction, hydride abstraction,
and oxygen atom transfer (including epoxidation).1
Selective oxidations of organic compounds, including oxida-
tions of C-H bonds, are of fundamental, technological, and
biochemical interest. Ruthenium-oxo compounds have received
particular attention in this area.
years, Meyer and co-workers,3
that ruthenium-oxo-polypyridyl complexes such as [(bpy)2(py)-
III
2+
2+
II
2
+
2+
1
2
-22
Over the past twenty-five
4
-20
21,22
and others,
have shown
5-7
8,9
10-13
4-18
The
(
1) (a) Olah, G. A.; Moln a´ r, AÄ . Hydrocarbon Chemistry; Wiley: New York,
995; pp 375ff. (b) Shilov, A. E.; Shul′pin, G. B. ActiVation and Catalytic
(17) (a) Stultz, L. K.; Binstead, R. A.; Reynolds, M. S.; Meyer, T. J. J. Am.
Chem. Soc. 1995, 117, 2520-2532. (b) The rate constant for oxidation of
1
-1
-1
Reactions of Saturated Hydrocarbons in the Presence of Metal Complexes;
Kluwer: Boston, 2000. (c) Biomimetic Oxidations Catalyzed by Transition
Metal Complexes; Meunier, B., Ed.; Imperial College Press: London, 2000.
2) Reviews: (a) Giffith, W. P. Chem. Soc. ReV. 1992, 179-185. (b) Che,
C.-M.; Yam, V. W. W. AdV. Trans. Met. Chem. 1996, 1, 209-237. (c)
Courtney, J. L. In Organic Syntheses By Oxidation With Metal Compounds;
Mijs, W. J., de Jonge, C. R. H. I., Eds.; Plenum Press: New York, 1986;
indene to indenone is reported here as 5.74 ( 0.4 M
s , including a
2
+
stoichiometric factor of 2 to account for the consumption of 2 RuO per
indanone.
(
(18) Stultz, L. K.; Huynh, M. H. V.; Binstead, R. A.; Curry, M.; Meyer, T. J.
J. Am. Chem. Soc. 2000, 122, 5984-5996.
(19) (a) Curry, M.; Huynh, M. H. V.; Stultz, L. K.; Binstead, R. A.; Meyer, T.
J., personal communication. (b) Bryant, J. R.; Mayer, J. M., manuscript in
preparation.
4
45-467.
(
(
(
(
(
3) (a) Meyer, T. J. J. Electrochem. Soc. 1984, 131, 221C. (b) Moyer, B. A.;
(20) (a) Moyer, B. A.; Meyer, T. J. Inorg. Chem. 1981, 20, 436-444. (b) Dobson,
J. C.; Helms, J. H.; Doppelt, P.; Sullivan, B. P.; Hatfield, W. E.; Meyer, T.
J. Inorg. Chem. 1989, 28, 2200-2204. (c) The potentials vs SCE in ref
20a have been converted to NHE by adding 0.24 V.
Meyer, T. J. J. Am. Chem. Soc. 1978, 100, 3601-3603.
4) Lebeau, E. L.; Binstead, R. A.; Meyer, T. J. J. Am. Chem. Soc. 2001, 123,
1
0 535-10 544.
5) Binstead, R. A.; McGuire, M. E.; Dovletoglou, A.; Seok, W. K.; Roecker,
L. E.; Meyer, T. J.J. Am. Chem. Soc. 1992, 114, 173-186.
(21) Representative papers and leading references: (a) Yamada, H.; Hurst, J.
K. J. Am. Chem. Soc. 2000, 122, 5303-5311. (b) Farrer, B. T.; Thorp, H.
H. Inorg. Chem. 1999, 38, 2497-2502. (c) Muller, J. G.; Acquaye, J. H.;
Takeuchi, K. J. Inorg. Chem. 1992, 31, 4552-4557. (d) Gerli, A.; Reedijk,
J.; Lakin, M. T.; Spek, A. L.; Inorg. Chem. 1995, 34, 1836-1843. (e)
Goldstein, A. S.; Beer, R. H.; Drago, R. S. J. Am. Chem. Soc. 1994, 116,
2424-2429. (f) Dengel, A. C.; El-Hendawy, A. M.; Griffith, W. P.;
O’Mahoney, C. A.; Williams, D. J. J. Chem. Soc., Dalton Trans. 1990,
737. (g) Kojima, T. Chem. Lett. 1996, 121-122. (g) Hua, X.; Lappin, A.
G. Inorg. Chem. 1995, 34, 992-994. (h) Lau, T. C.; Kochi, J. K. J. Chem.
Soc., Chem. Commun. 1987, 798-799. (i) Huynh, M. H. V.; Witham, L.
M.; Lasker, J. M.; Wetzler, M.; Mort, B.; Jameson, D. L.; White, P. S.;
Takeuchi, K. J.; J. Am. Chem. Soc. 2003, 308-309.
(22) cf., (a) Che, C.-M.; Yu, W.-Y.; Chan, P.-M.; Cheng, W.-C.; Peng, S.-M.;
Lau, K.-C.; Li, W.-K.; J. Am. Chem. Soc. 2000, 122, 11 380-11 392. (b)
Che, C.-M.; Cheng, K.-W.; Chan, M. C. W.; Lau, T.-C.; Mak, C.-K. J.
Org. Chem. 2000, 65, 7996-8000. (c) Che, C.-M.; Ho, C.; Lau, T. C. J.
Chem. Soc., Dalton Trans. 1991, 1901. (d) Che, C.-M.; Tang, W. T.; Wong,
K.-Y.; Li, C.-K. J. Chem. Soc., Dalton Trans. 1991, 3277.
6) Binstead, R. A.; Stultz, L. K.; Meyer, T. J. Inorg. Chem. 1995, 34, 546-
5
51.
7) Trammell, S. A.; Wimbish, J. C.; Odobel, F.; Gallagher, L. A.; Narula, P.
M.; Meyer, T. J. J. Am. Chem. Soc. 2001, 120, 13 248-13 249.
(
(
8) Binstead, R. A.; Meyer, T. J. J. Am. Chem. Soc. 1987, 109, 3287-3297.
9) (a) Gilbert, J.; Roecker, L.; Meyer, T. J. Inorg. Chem. 1987, 26, 1126-
1
1
132. (b) Gilbert, J. A.; Gersten, S. W.; Meyer, T. J. J. Am. Chem. Soc.
982, 104, 6872-6873.
(
(
(
(
(
(
10) Thompson, M. S.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 5070-5076.
11) Roecker, L. E.; Meyer, T. J. J. Am. Chem. Soc. 1986, 108, 4066-4073.
12) Roecker, L. E.; Meyer, T. J. J. Am. Chem. Soc. 1987, 109, 746-754.
13) Thompson, M. S.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 4106-4115.
14) Seok, W. K.; Meyer, T. J. J. Am. Chem. Soc. 1988, 110, 7358-7367.
15) (a) Roecker, L.; Dobson, J. C.; Vining, W. J.; Meyer, T. J. Inorg. Chem.
1
987, 26, 779-781. (b) Moyer, B. A.; Sipe, K.; Meyer, T. J. Inorg. Chem.
981, 20, 1475-1480.
1
(
16) Seok, W. K.; Dobson, J. C.; Meyer, T. J. Inorg. Chem. 1988, 27, 3-5.
10.1021/ja035276w CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 10351-10361
9
10351

A R T I C L E S
Scheme 1. Thermochemical Data in the [(bpy)2(py)Ru O]
Bryant and Mayer
IV
2+
Table 1. Yields of Products Formed from the Oxidation of DHA by
4,20
2+a
System (E° vs NHE, pKa Values
)
RuO
[
RuO2+]:[DHA]
[anthrone]
[anthraquinone]
[anthracene]
total (%)
1
1
1
1
2
4
:1
:2
:4
:10
:4
0.06 (12%)
0.11 (22%)
0.09 (18%)
0.09 (18%)
0.17 (17%)
0.30 (15%)
0.06 (24%)
0.05 (20%)
0.03 (12%)
b
0.11 (22%)
0.30 (30%)
0.39 (39%)
0.55 (55%)
0.70 (70%)
0.87 (87%)
1.10 (55%)
1.33 (33%)
75
97
100
105
94
:4
78
a
Reagent amounts and yields in mM (accurate to ( ∼10%); DHA )
9
,10-dihydroanthracene. Values in parentheses are the percentages of the
total RuO oxidative equivalents consumed in the formation of this product.
2+
b
<0.01 (<4%).
Reported here are studies of the oxidations of 9,10-dihy-
droanthracene (DHA), xanthene, fluorene, and related substrates
reactions studied have involved the oxidation of O-H bonds
2+
9
12,13
5
by RuO . Both product and kinetic studies support a common
mechanism of hydrogen atom transfer followed by competitive
radical dimerization and radical trapping by the oxidant. There
is a clear correlation between the rate constant for H-atom
abstraction and the C-H bond dissociation energy (BDE) of
the organic substrate, both for the compounds studied here and
for the oxidations of other benzylic and allylic substrates. In
(H2O2, alcohols,
and hydroquinone ) as well as oxidations
of C-H bonds. The oxidations of allylic C-H bonds in indene
and cyclohexene have been proposed to occur by direct oxygen-
1
6,18
atom insertion.
Oxidations of cumene, ethylbenzene, and
toluene were suggested to occur by a nucleophile-assisted
hydride transfer mechanism,10 although this conclusion is not
1
9
supported by recent results in our lab and in the Meyer lab.
2
+
addition, the hydrogen-atom self-exchange rate for RuO
+
In this report, we present evidence that all of the C-H bond
oxidations occur by a mechanism of initial hydrogen atom
transfer from the benzylic or allylic C-H bond to the oxo group
2+
1
RuOH has been probed by dynamic H NMR line-broadening,
to apply the Marcus cross relation to these reactions.
2
+
of RuO .
Results
Work in our lab and elsewhere has shown that, in general,
the ability of a metal complex to abstract a hydrogen atom from
a substrate correlates with the thermodynamic driving force for
Oxidations of Organic Substrates and Product Charac-
terization. The oxidations of DHA, xanthene, and fluorene by
2
+
RuO were carried out anaerobically at room temperature in
2
3-25
2+
H-atom transfer.
This correlation is a subset of the more
dry acetonitrile, typically with 1-2 mM RuO and 1-10 mM
general result that H-atom transfer reactions typically follow
the Marcus cross relation, identifying the importance of both
driving force and intrinsic barriers.26 For a reaction X + HY
f HX + Y, the ∆H° is the difference between the bond
dissociation energies of HX and HY. The H-X bond strength
can be determined from the redox potential of X and the pKa
substrate. Reactions are complete within minutes and a mixture
of organic products is observed for each reaction. Organic
products were characterized by GC/MS and quantitated using
calibrated GC/FID response factors with naphthalene as an
internal standard. Yields are reproducible in samples analyzed
after extended periods of time (6-24 h). The inorganic products
were characterized by UV-vis and/or NMR spectroscopies.
DHA is oxidized within seconds to give a mixture of
anthracene, anthrone, and anthraquinone, with anthracene as the
major product (eq 1, Table 1). The product distributions are
dependent on initial concentrations of oxidant and substrate (see
below). To determine the source of oxygen in the oxygenated
+
of HX (or the E° for HX and the pKa of HX ), using
2
7-30
2+
thermochemical cycles.
The enthalpies for RuO + H f
2
+
2+
2+
RuOH and RuOH + H f RuOH2 in aqueous solution
can be calculated via this approach (Scheme 1) using the
2
0
potential and pKa data reported by Meyer et al. In this
IV
III
somewhat unusual case, the Ru /Ru reduction potential at
pH 0 corresponds to a one electron-two proton process forming
RuOH23 , whose pKa must be subtracted to get the thermody-
18
products, reactions were carried out with 10 mM added H2 O.
+
2+
18
It has been shown that RuO does not exchange with H2
O
3
1
15b
namics for one electron-one proton (hydrogen atom) transfer.
in water over the time scale of these reactions. Anthraquinone
shows 2% O incorporation by GC/MS from the intensity of
the (M + 2) vs M ions in comparison to reactions in which
no labeled water was added. No incorporation into anthrone
was detected.
1
8
(
23) (a) Mayer, J. M. Acc. Chem. Res. 1998, 31, 441-450. (b) Roth, J. P.; Mayer,
J. M. Inorg. Chem. 1999, 38, 2760-2761. (c) Larsen, A. S.; Wang, K.;
Lockwood, M. A.; Rice, G. L.; Won, T.-J.; Lovell, S.; Sad ´ı lek, M.; Tureek,
F.; Mayer, J. M. J. Am. Chem. Soc. 2002, 124, 10 112-10 123.
+
+
(24) See, for example, (a) Bakac, A. J. Am. Chem. Soc. 2000, 122, 1092-1097.
(
b) Strassner, T.; Houk, K. N. J. Am. Chem. Soc. 2000, 122, 7821-7822.
(
25) For an exception to the correlation of H-atom transfer rate with bond
strength, see: Goldsmith, C. R.; Jonas, R. T.; Stack, T. D. P. J. Am. Chem.
Soc. 2002, 124, 83-96.
(
26) (a) Roth, J. P.; Yoder, J. C.; Won, T.-J.; Mayer, J. M. Science 2001, 294,
2
2
524-2526. (b) Roth, J. P.; Lovell, S.; Mayer, J. M. J. Am. Chem. Soc.
000, 122, 5486-5498.
(
27) (a) Bordwell, F. G.; Cheng, J.-P.; Ji, G.-Z.; Satish, A. V.; Zhang, X. J.
Am. Chem. Soc. 1991, 113, 9790-9795. (b) Bordwell, F. G.; Cheng, J.-P.;
Harrelson, J. A., Jr. J. Am. Chem. Soc. 1988, 110, 1229-1231. (c) Zhang,
X.-M.; Bordwell, F. G. J. Am. Chem. Soc. 1992, 114, 7458-7462.
28) (a) Parker, V. D.; Handoo, K. L.; Roness, F.; Tilset, M. J. Am. Chem. Soc.
(
1
7
1
991, 113, 7493-7498. (b) Parker, V. D. J. Am. Chem. Soc. 1992, 114,
458-7462; and correction: Parker, V. D. J. Am. Chem. Soc. 1993, 115,
201.
Xanthene reacts with RuO2 within seconds to give a mixture
of xanthone, the major product, and 9,9′-bixanthene (eq 2, Table
2). Oxidation of fluorene proceeds similarly, about an order of
+
(
29) Gardner, K. A.; Kuehnert, L. L.; Mayer, J. M. Inorg. Chem. 1997, 36,
069-2078.
30) Mayer, J. M. In ref 1c, pp 1-43.
2
(
1
0352 J. AM. CHEM. SOC. VOL. 125, NO. 34, 2003
9

Oxidation of C−H Bonds
A R T I C L E S
observed yield accounts for 71% of the RuO2 oxidative
+
Table 2. Yields of Products Formed from the Oxidations of
Xanthene and Fluorene (YH2) by RuO
2+a
equivalents consumed. The ruthenium product of this reaction
[
RuO2+]:[YH
2
]
[YdO]
2
[(YH) ]
total (%)
2+
2+
appears to be only RuNCMe . No RuOH2 is observed by
UV-vis, consistent with the time scale of the reaction being
longer than that for solvolysis.
xanthene: 1:1
0.36 (72%)
0.35 (70%)
0.31 (62%)
0.38 (76%)
0.66 (66%)
1.31 (66%)
0.37 (74%)
0.33 (66%)
0.31 (62%)
0.33 (66%)
0.63 (66%)
1.25 (63%)
0.013 (1.3%)
0.019 (1.9%)
0.025 (2.5%)
0.036 (3.6%)
0.045 (2.3%)
0.045 (1.1%)
0.011 (1.1%)
0.014 (1.4%)
0.017 (1.7%)
0.023 (2.3%)
0.027 (1.4%)
0.027 (0.7%)
73
72
65
80
68
67
75
67
64
68
67
64
1
1
1
2
4
:2
:4
:10
:4
:4
In the oxidations of DHA, xanthene and fluorene, the ratios
2
+
of organic products depend on the initial RuO and substrate
concentrations, as shown in Tables 1 and 2. In each table, the
percentage yields are given in terms of oxidative equivalents.
This value is the concentration of the organic product times
the number of RuO equivalents required to produce it,
dividing by the initial concentration of RuO . In the oxidation
fluorene: 1:1
1
1
1
2
4
:2
:4
:10
:4
:4
2
+
2+
2
+
of DHA, for instance, 1 equiv of RuO is consumed for each
a
anthracene, 2 for anthrone, and 4 for anthraquinone. The final
columns give the percentage of RuO oxidative equivalents
Reagent amounts and yields in mM ((10%). Values in parentheses
2
+
2+
are the percentages of the total RuO oxidative equivalents consumed in
the formation of this product. YdO is xanthone or fluorenone; (YH)2 )
bixanthene or bifluorene.
accounted for by the products listed. Less than quantitative total
yields here and elsewhere may be due in part to the slow
2
+
decomposition of RuO during storage and in solution, or to
the formation of products such as carboxylic acids which would
not have been observed with the GC conditions used. In the
DHA oxidations, the percentage yield of anthraquinone de-
magnitude slower than xanthene. This reaction produces fluo-
renone, bifluorene and a very small amount of bifluorenylidene,
with fluorenone as the major product (eq 3, Table 2). Oxidations
in the presence of 10 mM H2 O give essentially unlabeled
ketones: 2% O in the xanthone and <1% 18O in the
18
2+
18
creases with increasing [DHA] at constant [RuO ] (the first
four entries in Table 1). Conversely, comparing the three entries
with 4 mM DHA, the yield of anthraquinone increases with
fluorenone.
2
+
increasing [RuO ].
2
+
In the reactions of RuO with xanthene and fluorene, the
products account for 64-80% of the ruthenium oxidative
equivalents consumed (Table 2). The remaining RuO2 may
be consumed by further oxidation of bixanthene or bifluorene,
whose tertiary benzylic C-H bonds should be quite reactive.
A trace of bifluorenylidene is observed in the fluorene oxida-
tions. The coupled products, bixanthene and bifluorene, are
formed in low yields of 1-3% (estimated errors (∼30% of
these values). The yields of these products are larger with lower
ruthenium and higher substrate concentrations, as will be
discussed below using a kinetic model.
+
2
+
Kinetics. The kinetics of the reactions of RuO with DHA,
anthracene, xanthene, fluorene, indene, cyclohexene, and toluene
were measured by UV-vis spectroscopy at 24 °C in acetonitrile
under a nitrogen atmosphere. Reactions with cumene and
1
9
ethylbenzene will be described elsewhere. The organic sub-
strate was always in large excess (pseudo-first-order conditions).
Reactions with DHA, xanthene, fluorene, indene, and cyclo-
hexene were fast enough to require measurements with a
stopped-flow instrument. Anthracene kinetics were measured
on a diode array UV-vis spectrophotometer over 30 min;
reactions with toluene required several hours to reach comple-
tion, even with 2500-fold excess substrate (0.5 M). RuO2 is
known to decompose slowly in acetonitrile (over 24 h or more),6
but not appreciably during the time scale of kinetic measure-
ments.
The ruthenium products of these reactions appear to be a
mixture of mostly RuOH2 and the known acetonitrile complex
2+
II
2+
2+ 6
[
(bpy)2(py)Ru (NCMe)] (RuNCMe ). The mixture quan-
2+
titatively converts over time to RuNCMe as the aquo complex
-
3
-1
6
undergoes solvolysis (ksolv ) 1.66 × 10 s ; t1/2 ) 7 min).
+
After several hours, 100% conversion of RuO2+ to RuNCMe2+
is observed by UV-vis spectroscopy (λmax 440 nm, ꢀ ) 8160
-1
-1 6
M
cm ).
Anthrone is oxidized by RuO2 within seconds to form
anthraquinone (quantitatively by GC/FID) and a mixture of
+
Kinetics traces were fitted with SPECFIT,32 a global analysis
software package. The rapid reactions with DHA, xanthene and
indene could be fit with a biexponential model A f B f C.
RuOH22 and RuNCMe
+
2+
(by UV-vis). The analogous
oxidation of anthracene is much slower, requiring 30 min at
2
+
2
+
The slower reaction of RuO
with fluorene requires a
room temperature (vs 5 s for DHA). At 2 mM RuO and 4
mM anthracene, 0.47 mM anthraquinone is produced. Conver-
sion of anthracene to anthraquinone is a six-electron oxidation,
triexponential model, A f B f C f D. SPECFIT calculates
spectra for A, B, C, and D, which can then be interpreted using
the known spectral properties in this system (see Figures 1-3,
Supporting Information Figure S1). The spectra calculated for
requiring 3 equiv of RuO2 per quinone formed. Thus, the
+
(
31) In this case, BDE ) 23.06 E° - 1.37 pK
a
- C where E° (vs NHE) and
29,30
pK are measured in water and C ) -57 ( 2.
a
(32) Spectrum Software Associates, Marlborough, MA.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003 10353

A R T I C L E S
Bryant and Mayer
Figure 1. Spectra of RuO +, RuOH2+ (spectra obtained from a 1:1 mixture
2
2
+
2+
2+
of RuO , and RuOH2 , contains a slight amount of RuOH2 ),
2
+
2+
RuOH2 , RuNCMe
.
B, C, and D all appear to be due to mixtures of ruthenium
species, and some include absorbances due to organic products.
For the reactions that occur more slowly than the rate of
solvolysis of RuOH22+, the observed final product appears to
2+
be predominantly RuNCMe . When the kinetics are complete
in less than a few minutes, solvolysis is not a significant reaction
and the final product C (or D for fluorene) appears to be a
II
II
2+
mixture of Ru species, [(bpy)2(py)Ru (L)] where L ) H2O,
II
MeCN, or an organic product such as xanthone. Ru -bpy
complexes have a characteristic MLCT absorbance with ꢀ ≈
-
1
-1
2+
8
000 M
cm , as in RuOH2
(λmax ) 468 nm) and
2
+
6
RuNCMe (λmax ) 440 nm). The A f B f C f D model
is needed for the reaction of fluorene at least in part because
2
+
the C f D step includes the solvolysis of any RuOH2 that is
present in C. The kobs for the C f D rate depends on reagent
-
5
-1
-3 -1
concentrations, varying from 5.8 × 10
s
to 7.8 × 10 s ,
and is similar in magnitude to the rate constant for solvolysis,
-
3
-1 6
1
.66 × 10 s .
The calculated spectra for B are consistent with mixtures of
RuIII and Ru species. At the stage of the reaction when B is at
II
a maximum, the xanthene reaction mixture contains a substantial
II
II
amount of Ru (Figure 3). Ru species make a much smaller
III
contribution to B in the DHA reaction (Figure 2). The Ru
2+
components of B likely include both RuOH and some [(bpy)2-
III
2+
(
py)Ru OR] species, where R is derived from the substrate
Figure 2. (a) Spectral changes observed for the reaction of 0.2 mM RuO2
with 6 mM DHA over 30 s. (b) Single wavelength trace of the absorbance
at 385.5 nm with SPECFIT fit (green) showing the initial rising A f B
phase followed by the decreasing B f C phase. (c) Calculated spectra
returned by SPECFIT for A, B, and C.
+
(
e.g., fluorenyl). Similar conclusions have been made by Meyer
17-19a
et al. in related systems.
The kinetic models yield first-order rate constants kobs for
each of the reaction phases. Plots of kobs(A f B) vs [substrate]
are linear for DHA, xanthene, and fluorene (Figure 4), giving
second-order rate constants, k ) 125 ( 8 M
(
In each case, kobs for the second phase (B f C) is not linearly
related to substrate concentration, with values varying from
0
0
4
-
1 -1
s
(DHA), 577
for all three substrates, plots of logkobs vs log[substrate] are linear
with slopes of 0.90-0.97 and intercepts equal to log k (Figure
S2). Thus, the rate-limiting step of the A f B phase involves
-
1
-1
(xanthene) and 21.9 ( 1.8 M s-1 (fluorene).
-1
13 M
s
2+
the bimolecular reaction of RuO and substrate; subsequent
-
1
-1
.027 to 0.79 s for DHA, 0.19-1.66 s for xanthene, and
reactions that form the mixture of species that make up B are
-1
.014-0.065 s for fluorene. The linearity of the data in Figure
and the negligible intercepts indicate that the A f B phase
2+
fast on this time scale. Rates of oxidation of xanthene by RuO
were investigated from 278 to 318 K, yielding the activation
parameters ∆H ) 4.9 ( 1.0 kcal mol and ∆S ) -30 ( 5
(the Eyring plot is deposited as Figure S3). These
values are similar in magnitude to those reported for other C-H
follows the simple bimolecular rate law in eq 4. Furthermore,
q
-1
q
-1
-1
cal mol
K
2+
Rate ) kAfB[RuO ][substrate]
(4)
10354 J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003

Oxidation of C−H Bonds
A R T I C L E S
Figure 4. Plot of kobs for A f B vs substrate concentration for reactions
of RuO2 with xanthene (b), DHA (9), and fluorene ([). Data are from
three separate runs at each concentration.
+
Table 3. Properties of Organic Substrates and Their Rate
Constants for Oxidation by RuO2
+
BDE(C−H)a
k (M- s-1
1
)
(kcal mol-1
)
IE (eV)
b
pK d
a
substrate
2
xanthene
DHA
indene
5.77 × 10
1.25 × 10
1.08 × 10
2.19 × 10
9.2 × 10
3.3 × 10
2.2 × 10
6.4 × 10
2.7 × 10
75.5
78
7.65
n.a.
41.0
41.1
20.1
33.6
2
1
78.9
80
8.14
7.91
8.95
8.73
8.77
8.83
7.44
1
fluorene
cyclohexene
-
1
81.6e
c
-2
-2
-3
e
cumene
84.8
ethylbenzenec
toluene
87
90
e
Figure 3. (a) Spectral changes observed for the reaction of 0.2 mM RuO2+
with 20 mM xanthene over 3 s (1000 scans/s). (b) Calculated spectra
returned by SPECFIT for A, B, and C.
f
-
1
g
anthracene
∼111
a
From refs 27; errors (∼2 kcal mol-1. b Reference 41 c Reference 19b.
d
42
Values measured in DMSO and extrapolated to CH3CN using the
43 e
2+
equation pKa(C-H) (CH3CN) ) 11 + pKa(C-H) (DMSO).
Reference 44.
bond oxidations by RuO . For example, the oxidation of
f
Reference 45. g DFT calculation.
46
cyclohexene to 2-cyclohexen-1-one has activation parameters
q
-1
q
of ∆H ) 7.4 ( 0.1 kcal mol and ∆S ) -34 ( 3.4 cal
To check for the possible intermediacy of radicals in these
reactions, the kinetics of xanthene oxidation by RuO were
-
1
-1 18
mol K .
2+
The kinetics of anthracene and anthrone oxidation by RuO2+
were measured to determine their roles as possible intermediates
in the oxidation of DHA. Spectral changes for the anthracene
reactions can be fit to a single exponential (A f B model),
with RuNCMe2 as the only product (Figure S4). Monitoring
also examined in O -sparged MeCN solutions. Measurements
2
were carried out in the stopped flow with 0.2 mM RuO2 and
+
20-60 mM xanthene. Data were analyzed as described above
-
1
and gave the first-order rate kxan,O2(A f B) ) 340 ( 30 M
+
-1
s . This is a decrease in rate by almost half from the reaction
rate under N of k ) 577 ( 13 M s . This is consistent
2+
-1 -1
at 440 nm (λmax for RuNCMe ) gives kanth ) 0.273 ( 0.01
2
xan
-1
-1
2+
M
s . The oxidation of anthrone by RuO is quite rapids
with the involvement of carbon radicals that are trapped by O2.
complete within the mixing delay of the stopped flow instru-
ment. Assuming a first-order dependence on anthrone concen-
tration and that three half-lives of the reaction are complete in
The resulting peroxyl and perhaps oxyl radicals could then
2
+
2+
reoxidize RuOH or RuOH2 , which would be observed as
a slower rate of decay of oxidized ruthenium complexes.
The rates of oxidation of toluene, ethylbenzene, cyclohexene,
and indene were similarly measured and analyzed with an A
f B f C model. The values for the A f B phase are given in
Table 3 (see below). The rate constant found for toluene
5
-1 -1
3
ms, kanthrone is calculated to be greater than ∼10 M s .
On the basis of these rate constants, the oxidation of DHA to
anthrone and anthraquinone cannot proceed via oxidation of
anthracene. A SPECFIT simulation of a DHA oxidation in which
anthrone and anthraquinone are produced solely from anthracene
predicts <1 µM of oxygenated products. The simulations also
show that anthraquinone must be produced competitively with
anthrone. The addition of the second oxygen to anthrone may
occur when the anthrone is bound to Ru (during the B f C
phase of the reaction), while there is still RuO2 present.
-
3
-1 -1
oxidation, ktol ) (6.4 ( 0.6) × 10
M
s , is more than 2
orders of magnitude faster than the previously reported value
-5
-1 -1
10
of 5 × 10
M
s
(measured aerobically). Potential origins
of this discrepancy, and that some of the results reported in ref
10 have been difficult to reproduce in Meyer’s labs and in our
+
19
own, will be discussed elsewhere. The oxidation of toluene
J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003 10355

A R T I C L E S
Bryant and Mayer
Figure 6. Plot of π∆W vs [RuOH(D)] for the hydrogen-atom (O, -) and
deuterium-atom (b, - -) self-exchange reactions (two runs each).
Self-Exchange Rates. Because the Marcus cross relation has
been reported to hold for H-atom transfer reactions of this type,
we have attempted to measure the rate of degenerate hydrogen-
2
+
2+
atom self-exchange between RuO and RuOH (eq 5) by
3
4
dynamic NMR methods. This is a challenging measurement
because RuOH is not stable in acetonitrile, undergoing
2
+
6
,35
disproportionation (eq 6)
and the resulting aquo complex
2
+
solvolyzes to RuNCMe as noted above. We have used this
instability to prepare the solutions for the self-exchange study,
as follows.
Figure 5. (a) Calculated concentrations over time for an NMR tube self-
2
+
exchange reaction based on initial concentrations of 11.2 mM RuO and
kHSE
2
+
1
2+
RuO2+ + RuOH2+
2+
2+
5
.8 mM RuOH2 . (b) Stacked plot of H NMR spectra from a RuO /
RuOH self-exchange experiment showing the sharpening of the RuO
resonance at δ53.5 from t ) 0 to 2.5 h.
*
8 *RuOH + RuO
(5)
(6)
2
+
2+
kf ) 4.07 × 103 M-1 s-1
kr ) 80.6 M-1
2
+
2+
2+
RuO + RuOH2
{
} 2RuOH
-1
s
gives benzaldehydesand possibly benzoic acid, though it was
not observed by GC/FID or GC/MS. Indene and cyclohexene
are oxidized on the stopped flow time scale (seconds) at kind )
2
+
In CD3CN solution, addition of 0.5 equiv RuOH2 to 1 equiv
of RuO forms a solution of roughly equimolar RuO and
2
+
2+
-
1
-1
-1 -1
1
0.8 ( 1 M
s
and kcyh ) 0.92 ( 0.12 M
s
to give
2
+
2+
RuOH , with a small amount of RuOH2 (Keq ) 50 for eq
indenone and cyclohexenone as the sole product in each case
by GC/MS). The rate constants are the same, within the error
limits, as the values reported by Meyer and co-workers.
Oxidation of ethylbenzene gives exclusively acetophenone, with
2
+
6
). The presence of RuOH results in the broadening of the
(
2
+
proton resonances for the pyridyl ligands of RuO because of
the self-exchange reaction (eq 5). The small concentration of
RuOH2 slowly undergoes solvolysis in CD3CN, causing the
equilibrium to shift and the amount of RuOH to decrease
Figure 5a). As a result, a range of concentrations of RuOH2
occur over time in only one NMR tube. As the concentration
1
8
2
+
-
2
-1 -1
kPhEt ) 2.2 × 10
M
s .
2
+
Deuterium isotope effects for the oxidation of DHA versus
DHA-d4 and toluene vs toluene-d8 were measured over a range
of concentrations and analyzed as above. Reaction of DHA-d4
98-99% deuterium enrichment as determined by H NMR)
+
(
2
+
2+
of RuOH decreases, the line widths of RuO sharpen
because the rate of self-exchange decreases (Figure 5b).
A plot of the increase in line width (π∆W) for two different
1
(
-
1 -1
gave kDHA-d4 ) 3.6 ( 0.4 M s , or kDHA/kDHA-d4 ) 35 ( 5.
-
5
-1 -1
Similarly, ktoluene-d8 ) (9.56 ( 0.18) × 10
M
s
which
2
+
resonances versus [RuOH ] is roughly linear (Figure 6)
gives ktoluene/ktoluene-d8 ) 67 ( 4. Because these values are so
large, the residual H contributes to the reactivity of the
deuterated materials.33 Taking this into account (but ignoring
secondary isotope effects), the actual kH/kD for DHA is between
2
+
2+
suggesting that RuO and RuOH undergo chemical ex-
change on the NMR time-scale. This process is a net hydrogen-
atom self-exchange as these complexes differ by a proton and
5
0 (calculated for 1% residual H) and 100 (for 2% residual H),
(33) For a discussion of the effect of residual H on observed k /k , see: Sorokin,
H
D
A.; Robert, A.; Meunier, B. J. Am. Chem. Soc. 1993, 115, 7293-7299.
34) Sandstr o¨ m, J. Dynamic NMR Spectroscopy; Academic Press: New York,
1982.
and the 0.5% residual H in the methyl group of the toluene-d8
gives an actual kH/kD ) 100. A large deuterium isotope effect
has been reported for RuO oxidizing a CH vs CD bond in
cyclohexene vs cyclohexene-d4.1
(
IV
II
2+
2
(35) For related studies of polypyridyl Ru )O + Ru (H O) reactions, see: (a)
Iordanova, N.; Hammes-Schiffer, S. J. Am. Chem. Soc. 2002, 124, 4848-
4856. (b) Farrer, B. T.; Thorp, H. H. Inorg. Chem. 1999, 38, 2497.
8
10356 J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003

Oxidation of C−H Bonds
A R T I C L E S
an electron. We have previously used this approach to measure
H-atom self-exchange rates for iron complexes.26,36 The data
are consistent for two experiments with different batches of Ru
starting materials, each measured at two different proton
resonances. The slope of the line suggests an average self-
reaction solutions causes a change in the rate of the first (Α f
B) phase of xanthene oxidation kAfB, indicating that organic
radicals are involved in the Α f B phase when RuO2 is being
consumed. The large deuterium isotope effect for the oxidation
of DHA-d (k /k g 35 ( 1) indicates that C-H bond breaking
+
4
H
D
4
-1 -1
exchange rate constant of kHSE ) (7.6 ( 0.4) × 10 M s .
The same procedure was carried out with the addition of Ru-
is an important part of the rate-limiting step. Isotope effects of
this magnitude are often attributed to tunneling and have been
2+
2+ 9b,12,13
OD2 to determine the deuterium-atom self-exchange rate. The
CD3CN used in the experiment was dried thoroughly over CaH2
and P2O5, and the NMR indicated only a small amount (<0.5
mM) of H2O present in the solvent. As a check on the isotopic
purity of the reaction mixture, the integral for H2O in the CD3-
CN reaction solution does not grow in appreciably through the
experiment indicating nearly quantitative D2O incorporation into
the starting material [solvolysis of RuO(H/D)22 by CD3CN
increases the amount of free (H/D)2O in the sample]. A plot of
the increase in line width (π∆W) for two different resonances
observed in many other oxidations by RuO ,
including
kH/kD ) 21 ( 1 for the oxidation of cyclohexene vs cyclohex-
1
8
ene-d4.
The data indicate that at least a portion of the reactions of
RuO2+ proceed by initial H-atom abstraction. A mechanism
2+
that involves solely oxygen atom insertion by RuO , and
2+
formation of radicals only by RuOH is difficult to rationalize
+
with the effect of oxygen on the rate and the higher yield of
radical coupling products at higher xanthene and fluorene
concentrations (see below). The high yield of the non-oxygen-
ated product anthracene would have to be explained by
2+
versus [Ru-OD ] is linear and gives a self-exchange rate
4
-1 -1
constant of kDSE ) (6.4 ( 0.4) × 10 M
s
(Figure 6). This
dehydration of an intermediate 9,10-dihydroanthracenol complex
gives a self-exchange isotope effect kHSE/kDSE ) 1.2 ( 0.1. This
small isotope effect raises the question of whether the NMR
line broadening really results from the self-exchange reaction
or is due to some other process. Two different resonances of
2+ 37
to anthracene and RuOH2
competitive with its oxidation
III
to anthrone. The significant formation of Ru intermediates in
the oxidation of DHA (observed in spectrum B) seems
inconsistent with a two-electron oxygen-insertion pathway, since
2
+
RuO broaden equally, consistent with an exchange process.
2+
2+
comproportionation of RuO with RuOH2 is slower than
But it is difficult to obtain direct evidence for exchange, such
C-H bond oxidation under our conditions. For example, at 40
2
+
as peak coalescence, because of the instability of RuOH .
2+
mM DHA and 0.2 mM RuO , t1/2 for the A f B phase of the
4
-1 -1
Thus, the assignment of kHSE as (7.6 ( 0.4) × 10 M
s
oxidation is 0.13 s, whereas t1/2 ) 0.85 s for reaction of
should be considered tentative (certainly an upper limit).
2+
2+
RuOH2 with 0.2 mM RuO . Xanthene is even faster.
It is possible that there is a mixed mechanism, with RuO2
+
Discussion
reacting both by direct oxygen atom insertion and by hydrogen
atom abstraction. However, there is no need to invoke direct
oxygen insertion, as all of the data can be rationalized by an
H-atom abstraction pathway as described in the following
sections. It should be noted that there are few documented cases
Mechanistic Considerations. [(bpy)2(py)Ru O] (RuO2+)
is an extensively studied material that is known to oxidize a
wide range of substrates.3 A variety of mechanisms have been
proposed for its oxidations of C-H and O-H bonds. Hydrogen-
atom abstraction and related proton-coupled electron transfer
IV
2+
-22
38
of direct O-atom insertion into C-H bonds, whereas experi-
2
+
(
PCET) pathways have been proposed for RuO oxidations
ments have increasingly shown H-atom abstraction by metal-
5
9
of O-H bonds (e.g., hydroquinone, H2O2, and the compro-
portionation/disproportionation of eq 6)6 but these have not
been the suggested mechanisms for C-H bond oxidations. In
detailed studies of allylic oxidation of cyclohexene and indene
by RuO , Meyer and co-workers considered a mechanism of
H-atom abstraction followed by rapid radical trapping but
concluded in favor of direct oxygen insertion because reactions
of 3,3,6,6-tetradeuteriocyclohexene (d4-cyclohexene) gave very
3
9
2+
2+
oxo and other complexes. RuO and RuOH are thermo-
dynamically good hydrogen atom abstractors, as they form O-H
,8
-
1
bonds of 84 and 82 ( 2 kcal mol (Scheme 1). Direct oxygen
transfer likely does occur in the oxidation of anthracene by
2+
2
+
RuO . Aromatic oxidations in general do not proceed by
hydrogen atom abstraction because the aromatic C-H bond is
2
+
too strong (see Table 3). Most likely, RuO acts as an
electrophile toward anthracene, analogous to the mechanism
1
6,18
little allylic rearrangement.
Oxidation of cyclohexene by
2
+ 14
2+
proposed for phenol oxidation by RuO . RuO is known
III
2+
Ru OH , however, was concluded to proceed by H-atom
abstraction.18 Che and co-workers have suggested an H-atom
abstraction mechanism for C-H bond oxidations by ruthenium-
(
37) (a) Rao, S. N.; More O’Ferrall, R. A.; Kelly, S. C.; Boyd, D. R.; Agarwallb,
R. J. Am. Chem. Soc. 1993, 115, 5458-5465. (b) Dey, J.; O’Donoghue,
A. C.; More O’Ferrall, R. A. J. Am. Chem. Soc. 2002, 124, 8561-8574.
38) See, for example: (a) Chen, K.; Que, L., Jr. J. Am. Chem. Soc. 2001, 123,
6327-6337. (b) Shustov, G. V.; Rauk, A. J. Org. Chem. 1998, 63, 5413-
22b,d
(VI) dioxo complexes.
(
The oxidations of DHA, xanthene, and fluorene presented
5
422 and Du, X.; Houk, K. N. J. Org. Chem. 1998, 63, 6480-6483, and
here give both oxygenated (anthrone, anthraquinone, xanthone,
fluorenone) and non-oxygenated products (anthracene, 9,9′-
bixanthene, bifluorene). The oxygen in the products does not
come from water (as shown by added H2 O experiments), so
under our anaerobic conditions, it must come from the ruthenium-
oxo group. The formation of bixanthene and bifluorene by
references therein.
(39) See, for instance: (a) References 1c, 23-26, 29, 38a. (b) Cook, G. K.;
Mayer, J. M. J. Am. Chem. Soc. 1994, 116, 1855-1868 (correction, ibid.,
8
859). (c) Cook, G. K.; Mayer, J. M. J. Am. Chem. Soc. 1995, 117, 7139-
18
7156. (d) Bryant, J. R.; Taves, J. E.; Mayer, J. M. Inorg. Chem. 2002, 10,
2
769-2776. (e) Lockwood, M. A.; Blubaugh, T. J.; Collier, A. M.; Lovell,
S.; Mayer, J. M. Angew. Chem., Int. Ed. Engl. 1999, 38, 225-227. (f)
Mahapatra, S.; Halfen, J. A.; Tolman, W. B. J. Am. Chem. Soc. 1996, 118,
11 575-11 586. (g) Qin, K.; Incarvito, C. D.; Rheingold, A. L.; Theopold,
2+
RuO indicates the intermediacy of radicals in these processes
in sufficient concentrations to dimerize. Addition of O2 to the
K. H. J. Am. Chem. Soc. 2002, 124, 14 008-14 009. (h) Auclair, K.; Hu,
Z.; Little, D. M.; Ortiz de Montellano, P. R.; Groves, J. T. J. Am. Chem.
Soc. 2002, 124, 6020-6027 and references therein. (i) Reitz, J. B.; Solomon,
E. I. J. Am. Chem. Soc. 1998, 120, 11 467-11 478. (j) Fontecave, M.;
Pierre, J. L. Comptes Rendus Acad. Sci. Ser. II C 2001, 4, 531-538
“Mechanisms of Formation of Free Radicals In Biological Catalysis.”
(
36) Yoder, J. C.; Roth, J. P.; Gussenhoven, E. M.; Larsen, A. S.; Mayer, J. M.
J. Am. Chem. Soc. 2003; 125, 2629-2640.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003 10357

A R T I C L E S
Bryant and Mayer
Figure 8. Plot of the log of the rate constant for H-atom abstraction for
DHA (b, -) and toluene (O, - -) vs the strength of the X-H bond formed
by the oxidant. The straight lines are drawn through the two oxygen radical
Figure 7. Plot of log k′ vs BDE for oxidation of various substrates by
2
+
RuO
.
IV
III
points. Mn2O2 represents the value for oxidation by [(phen)2Mn (O)2Mn -
to attack other electron-rich substrates such as phosphines,
sulfides, and alkenes.15,17
3+
III
III
3+
(
phen)2] , Mn2(O)(OH) for [(phen)2Mn (O)(OH)Mn (phen)2] , and Fe-
III
2+ 23,29
2
(Hbim) for [Fe (Hbim)(H bim) ]
2
.
Reactivity vs Thermodynamic Driving Force. A hallmark
of hydrogen atom abstraction processes is that the rate constants
correlate with C-H bond dissociation energies (BDEs). A plot
Scheme 2
40
of rate constant vs BDE for all of the substrates studied here
(except anthracene) is shown in Figure 7 (data in Table 3). The
rate constants in this plot (k′H) are kAfB values that have been
statistically corrected for the number of reactive hydrogens (e.g.,
DHA, 4; xanthene, 2; toluene, 3). The k′H values have not,
however, been corrected the stoichiometries of the reaction, that
one H-atom transfer can lead to the reduction of more than one
reactions, we have found that rate constants for metal complexes
correlate well with the rate constants for the oxygen radicals
2+ 29,39b,c
molecule of RuO .
The A f B kinetic phase does
include multiple steps, as noted above, but the magnitude of
this correction is difficult to estimate. The close correlation of
t
•
•
t
s
23
BuO and ROO (R ) Bu, Bu). The oxidations of DHA and
2+
toluene by RuO fit reasonably well on these correlation lines
5
k′H with BDE is quite good over a range of ca. 10 in rate
(
Figure 8). The RuO-H2 bond strength of 84 kcal mol is
+
-1
constants. It encompasses both benzylic and allylic oxidations
and includes primary, secondary, and tertiary C-H bonds. This
correlation is strong evidence for a common hydrogen-atom
abstraction mechanism for all of the substrates. The rates of
oxidation, kAfB, do not correlate with other properties of the
substrates, such as ionization energy (IE) or acidity (pKa), as
shown in Table 3. For instance, cyclohexene has a lower IE
than toluene, but reacts more slowly. Similarly, indene and
fluorene are much more acidic than xanthene or DHA, but react
more slowly. The data indicate that the reactions do not occur
by rate-limiting initial electron or proton transfer. This is not
-
29
higher than the hydrogen-atom affinities of MnO4 ,
III
2+ 23b
3+
[
(
Fe (Hbim)(H2bim)2] , and [(phen)2Mn(µ-O)2Mn(phen)2]
-
1
23c
80, 76, and 79 kcal mol , respectively). It should be noted
that Stack and co-workers have reported an exception to this
25
correlation. The correlation is a special case of a more general
Marcus treatment (see below).
Product Formation. For xanthene and fluorene, the forma-
tion of both dimeric and oxygenated products indicates that
radical coupling competes with trapping of the radical by the
oxidant (Scheme 2). Oxidizing metal oxo compounds are known
to be good radical traps, forming oxygenated products.47 This
trapping can compete with the almost diffusion controlled radical
2
+
20a
surprising, as RuO is not basic and is a very poor electron-
transfer oxidant.4
9
-1 -1 48
dimerization or disproportionation (kd ≈ 10 M s ) because
2+
The rapid rate of DHA oxidation by RuO is consistent with
H° ) -6 ( 3 kcal mol for the hydrogen-atom transfer step.
This value is based on the 78 kcal mol C-H bond strength
in DHA and the 84 kcal mol bond strength in RuOH (Table
-1
∆
(
41) http://webbook.nist.gov/chemistry: NIST Standard Reference Database
Number 69-July 2001 Release.
-
1
(42) (a) Reference 27a. (b) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456-
-1
2+
463 and references therein.
(
43) Wayner, D. D. M.; Parker, V. D. Acc. Chem. Res. 1993, 26, 287-294.
44) Denisov, E. T.; Denisova, T. G. Handbook of Antioxidants, 2nd ed.; CRC
Press: Boca Raton, 2000; p 21-36.
3
, Scheme 1). In previous studies of H-atom abstraction
(
(
40) (a) Free Radicals; Kochi, J. K., Ed.; Wiley: New York, 1973, especially
(45) Bierbaum, V.; DePuy, C.; Davico, G.; Ellison, B. Int. J. Mass Spectrom.
Ion Phys. 1996, 156, 109-131.
(46) Barckholtz, C.; Barckholtz, T. A.; Hadad, C. M. J. Am. Chem. Soc. 1999,
121, 491-500.
Ingold, K. U. Chapter 2, Vol 1, pp 67ff and Russell, G. A. Chapter 7, Vol.
1, pp 275-331. (b) Tedder, J. M. Angew. Chem., Int. Ed. Engl. 1982, 21,
01-410. (c) See also refs 23a and 25.
4
10358 J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003

Oxidation of C−H Bonds
A R T I C L E S
Scheme 3. Kinetic Models for Oxidations of Xanthene (XH2) and
Fluorene (FlH2) by RuO
In the oxidation of DHA, trapping of the hydroanthracenyl
radical (HA ) can occur by either C-O bond formation or by
2+
•
transfer of a second hydrogen atom. The former path leads to
anthrone and anthraquinone, and the latter gives anthracene.
Anthracene can also be formed by disproportionation of two
•
9
-1
hydroanthracenyl radicals (2HA f A + DHA, kdisp ≈ 10 M
-1
48
•
s ). Hydrogen transfer from HA is facile because of the very
-
1 49
weak C-H bond strength in this radical, 43 kcal mol .
Meyer and co-workers ruled out a hydrogen atom transfer
mechanism for the allylic oxidation of cyclohexene because
reactions of 3,3,6,6-tetradeuteriocyclohexene (d4-cyclohexene)
1
8
gave very little allylic rearrangement. For this result to be
consistent with Scheme 2, the d3-cyclohexenyl radical would
2+
have to be trapped rapidly by RuOH , within the solvent cage.
Groves and co-workers had previously found 30% allylic
rearrangement in the oxidation of this d4-cyclohexene with
cytochrome P450 enzymes and with an iron-porphyrin +
Table 4. Experimental and Calculated Product Concentrations
Using the Model in Scheme 3
a
5
0
] (mM)a
(HY)
2
[YdO] (mM)a
iodosylbenzene catalytic system. They argued on this basis
[
5
1
50
RuO2+]:[YH
observedb
observedb
for a rebound mechanism analogous to Scheme 2. The
Groves result suggests that 60% of the cyclohexyl radicals have
enough lifetime to be trapped randomly at either allylic position,
but also that 40% of the oxidized products are formed regio-
specifically, presumably by very rapid trapping. Similar expla-
nations have been given for a number of C-H bond oxidations
[
2
](mM)
calculated
calculated
Xanthene: 1:1
0.013
0.019
0.025
0.036
0.045
0.045
0.011
0.014
0.017
0.023
0.027
0.027
0.010
0.010
0.021
0.043
0.017
0.020
0.006
0.014
0.026
0.053
0.027
0.025
0.36
0.35
0.31
0.38
0.66
1.31
0.37
0.33
0.31
0.33
0.63
1.25
0.38
0.38
0.38
0.36
0.76
1.55
0.49
0.49
0.47
0.45
0.97
1.97
1
1
1
2
4
:2
:4
:10
:4
:4
Fluorene: 1:1
VI
VII
by Cr and Mn reagents that occur with partial retention of
1
1
1
2
4
:2
52
configuration, starting with a report by Wiberg in 1961. These
studies thus provide precedent for trapping that is sufficiently
rapid that rearrangement is not observed. However, it is not
clear why trapping of cyclohexenyl radicals by RuOH2 would
so efficient. With other hydrocarbon substrates, cage escape
competes with trapping within the solvent cage, as indicated
by the formation of dimeric products and the effect of O2.
It should be noted that hydrogen atom abstraction and direct
insertion are extremes of a mechanistic continuum. Many current
theoretical treatments of C-H bond hydroxylation by metal-
oxo species show that O-H bond formation and C-H bond
cleavage occur prior to significant C-O bond formation, in part
because the light hydrogen atom moves more quickly.53 C-O
bond formation often has a very small or essentially zero
reaction barrier. Thus, it seems possible that rate constants for
a set of processes could reflect an H-atom transfer rate-limiting
step, whereas selectivities could suggest direct insertion.54
:4
:10
:4
:4
+
a
(HY)2 is 9,9′-bixanthene or bifluorene; YdO is xanthone or fluorenone.
b
From Table 2.
the oxidant is present in much higher concentration than the
radical. This model explains why the yield of dimerization
products decreases with increasing concentrations of RuO
2
+
•
2+
(
Table 2). Addition of R to the oxo group in RuO or to the
2
+
III
2+
hydroxo group in RuOH would form [(bpy)2(py)Ru OR]
II
2+
or [(bpy)2(py)Ru (HOR)] , consistent with the observed
spectra. Further oxidation of these species and solvolyses,
occurring typically in the slower kinetic phases such as B f
C, generate the ketone and ruthenium products observed. The
sensitivity of kBfC to reaction conditions reflects the changing
mixture of the processes occurring in this step.
The mechanism in Scheme 2 has been modeled for the
xanthene and fluorene reactions using SPECFIT simulations
-
1
(
49) The sum of the H-AH bond strength (78 kcal mol ) and the H-A bond
strength equals ∆H° for H A f A + 2 H (A ) anthracene). The ∆H° can
be determined from ∆H°
anthracene + H
2
-1
(
Scheme 3). In both cases, the k1 steps are taken as the measured
kAfB. The trapping rates, assumed for simplicity to be the same
f
(H) ) 52.1 kcal mol and ∆H° for DHA f
-1
2
) 17 kcal mol (Cox, J. D.; Pilcher, G. Thermochemistry
of Organic and Organometallic Compounds; Academic Press: New York,
1970).
2
+
2+
6
-1 -1
for RuO and RuOH , need to be 5 × 10 M
s
for
(
50) Groves, J. T.; Adhyam, D. V. J. Am. Chem. Soc. 1984, 106, 2177-2181.
6
-1 -1
xanthenyl radicals and 1 × 10 M
s
for fluorenyl radicals
(51) (a) Reference 39h and references therein. (b) Schoneboom, J. C.; Lin, H.;
Reuter, N.; Thiel, W.; Cohen, S.; Ogliaro, F.; Shaik, S. J. Am. Chem.
Soc.2002, 124, 8142-8151.
to account for observed product ratios. A test of the model is
that it reproduces the trends in ketone vs dimer yields on varying
the initial concentrations (Table 4). A similar model was used
in our study of oxidations by manganese µ-oxo dimers, which
(52) Oxidation of (+)-3-methylheptane by chromic acid in 91% acetic acid gave
the corresponding tertiary alcohol with 70-85% retention of configura-
tion: Wiberg, K. B.; Foster, G. J. Am. Chem. Soc. 1961, 83, 424. (b)
2 2
Oxidation of optically active PhCHDEt by CrO Cl gives optically active
3
4
-1 -1 23c
1-chloro-1-phenylpropane in 35-40% chemical yield and with about 35%
retention of configuration: Stephenson, L. M.; Egnatchik, J.; Speth, D. R.
J. Org. Chem. 1979, 44, 346-349. (c) Permanganate oxidations of
arylvaleric acids gave 30-40% retention: Brauman, J. I.; Pandell, A. J. J.
Am. Chem. Soc. 1970, 92, 329-335. See also: (d) Wiberg, K. B.; Fox, A.
S. J. Am. Chem. Soc. 1963, 85, 3487. (e) Eastman, R. H.; Quinn, R. A. J.
Am. Chem. Soc. 1960, 82, 4249.
found trapping rates of 10 -10 M s . In both that study
and this work, trapping of xanthenyl radicals is faster than that
of fluorenyl radicals, although the reason for this is not known.
(
47) (a) Steenken, S.; Neta, P. J. Am. Chem. Soc. 1982, 104, 1244-1248. (b)
Al-Sheikhly, M.; McLaughlin, W. L. Radiat. Phys. Chem. 1991, 38, 203-
(53) See, for instance: (a) Ogliaro, F.; Harris, N.; Cohen, S.; Filatov, M.; de
Visser, S. P.; Shaik, S. J. Am. Chem. Soc. 2000, 122, 8977-8989. (b) Hata,
M.; Hirano, Y.; Hoshino, T.; Tsuda, M. J. Am. Chem. Soc. 2001, 123,
6410-6416. (c) Baik, M.-H.; Gherman, B. F.; Friesner, R. A.; Lippard, S.
J. J. Am. Chem. Soc. 2002, 124, 14 608-14 615. (d) Strassner, T.; Houk,
K. N. J. Am. Chem. Soc. 2000, 122, 7821-7822.
2
11. (c) Libson, K.; Sullivan, J. C.; Mulac, W. A.; Gordon, S.; Deutsch, E.
Inorg. Chem. 1989, 28, 375-377. (d) Elliot, A. J.; Padamshi, S.; Pika, J.
Can. J. Chem. 1986, 64, 314-320. (e) References 23c, 29, 39b,c.
48) Arends, I. W. C. E.; Mulder, P.; Clark, K. B.; Wayner, D. D. M. J. Phys.
Chem. 1995, 99, 8182-8189.
(
J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003 10359

A R T I C L E S
Bryant and Mayer
+
Self-Exchange Rates and Application of the Marcus Cross
Relation. We have recently shown that rate constants for a broad
oxo group, to give RuOH2 and the organic radical. This radical
can be trapped by the oxidant or can encounter another radical
to form dimerization products (Scheme 2). Rate constants for
the initial hydrogen-atom transfer step correlate well with the
set of hydrogen atom transfer reactions XH + Y follow the
Marcus cross relation (eq 7) fairly well.2
6a,55
The correlations
5
of rate constants with bond strengths in Figures 7 and 8 are
special cases of the Marcus equation and cross relation, that
hold
strength of the C-H bond being cleaved, over a range of 10
in k. RuO reacts slightly more slowly than ROO , consistent
with its 84 kcal mol affinity for a hydrogen atom being smaller
than that of the oxygen radical (90 kcal mol ; see Figure 8).
2
+
•
-1
-
1
1
H NMR line-broadening experiments have given a tentative
kXY
)
kXXkYY
x
K
f
(7)
XY XY
2+
2+
value of the RuO /RuOH hydrogen-atom self-exchange rate
constant kHSE. Using this value, or an alternative estimate of
kHSE, the rate constant for H-atom transfer from toluene to
when the intrinsic barriers are roughly constant across the series
or vary linearly with ∆G). The intrinsic barriers appear in eq
primarily through the rate constants for degenerate self-
(
7
IV
2+
[
(bpy)2(py)Ru O] is well predicted by the Marcus cross
relation.
exchange XH + X (kXX). The factor fXY is close to unity for
reactions of moderate driving-force.55
Experimental Section
The applicability of the Marcus cross relation to the oxidation
reactions described here can be tested using the rate of toluene
General Considerations. All experiments were performed under
an N atmosphere using standard techniques unless otherwise noted.
2
Solvents (including deuterated solvents from Cambridge Isotope) were
degassed and dried according to standard procedures.5 Acetonitrile
2
+
oxidation by RuO . Benzyl radical + toluene self-exchange
-
5
-1 -1
(
kYY) occurs with a rate constant of ∼1 × 10
M
s
per
8
56
2+
hydrogen. H-transfer from PhCH3 to RuO is 6 ( 3 kcal
(
low-water brand) was purchased from Burdick and Jackson and
dispensed from a stainless steel keg plumbed directly into the drybox.
CD CN was stirred over CaH for 2 days, vacuum transferred to P
and stirred for 4 h. It was then transferred back to CaH , stirred for 30
-
1
-1
mol uphill, based on D(PhCH2-H) ) 90 kcal mol and
D(RuO-H) ) 84 kcal mol . Assuming ∆S ) 0 for the H-atom
transfer, the ∆H° corresponds to an equilibrium constant of 4
10 , with an uncertainty of about 2 orders of magnitude.
Using the tentative RuO /RuOH self-exchange rate constant
-1
3
2
2 5
O
2
-
5
×
min, and transferred to a sealable flask prior to use. Dihydroanthracene
(DHA) was recrystallized twice from absolute ethanol. Other reagents
were purchased from Aldrich and used as received unless otherwise
2+
2+
4
-1 -1
57
kHSE ) 7.6 × 10 M
s
discussed above, and fXY ) 1,
gives kcalc(RuO2+ + PhCH3) ) 5.6 × 10
-3
M
-1 -1
s . This is
noted. Ru(bpy)
cording to literature methods.
and [Ru(bpy) (py)O](PF
literature procedures. The same procedure for the synthesis of the
2
Cl
2
and [Ru(bpy)
2
(py)Cl](PF
6
) were synthesized ac-
2
0a,59
2+
[Ru(bpy)
2
(py)(OH )](PF ) (RuOH )
2 6 2 2
remarkably close to the experimental rate constant (per hydro-
2
6 2
)
(RuO2 ) were synthesized by modified
+
-3
-1 -1
gen) of 2.1 × 10
M
s , a factor of 3 larger. As an
2+
2+
alternative estimate of the RuO /RuOH self-exchange rate,
one could use the geometric mean of the comproportionation/
disproportionation rate constants in eq 6, because these are close
2
0b
perchlorate salt [Ru(bpy)
that a saturated solution of KPF
product. A procedure similar to the one used to synthesize [Ru(bpy)
2
(py)(OH
2 4 2
)](ClO ) was followed, except
6
in water was used to precipitate the
2
-
-
1 6
to self-exchange reactions (∆G° ) (2.3 kcal mol ). Using
18
13a
(
Br
py) O](ClO
4
)
2
was used to produce [Ru(bpy)
was added in a very small amount to the dissolved RuOH
the resulting solution was purged with N for several minutes. The
2
solution was cooled and 1-2 mL of saturated KPF
to precipitate the product. Ru-OD
identical to that used for the synthesis of RuOH
O at all points in the synthesis from [Ru(bpy) (py)Cl]-
2
). Only one batch of Ru-OD
stored in a vial in the drybox between experiments. RuO , RuOH
2
(py)O](PF
6
)
2
. Liquid
2
-1 -1
2+
this mean value (k′HSE ) 5.7 × 10 M s ), kcalc(RuO
+
2+
2
2
and
-
4
-1 -1
PhCH3) ) 4.9 × 10
M
s , a factor of 4 smaller than the
experimental value. Thus, both estimates based on the cross
relation are within the uncertainty of this calculation of a little
more than an order of magnitude (primarily from uncertainties
in the bond strengths used to estimate Keq). It should be noted
that while the Marcus cross relation is valuable for hydrogen
atom transfer reactions in some ways, it does not provide an
explanation of the large isotope effects. For instance, the
6
in water was added
2+
2
was synthesized in a manner
2+
2
2
except that D O
replaced H
2
2
0b
2+
(PF
6
2
was synthesized and it was
2
+
2+
,
2+
2+
2+
RuOH
and UV-vis spectroscopies. DHA-d
member of the lab by exchange of protons with sodium dimsylate-d
2
, Ru-OD
2
, and RuNCMe were characterized by NMR
20
4
was synthesized by a previous
2
k(toluene)/k(toluene-d8) of ∼10 would require that the isotope
5
effect on toluene/benzyl radical self-exchange be an unrealistic
60
6
in DMSO-d .
4
1
0 .
1H NMR spectra were recorded on Bruker AC-200, AF-300, and
DRX-500 spectrometers at ambient temperatures and are reported in
Conclusions
(bpy)2(py)Ru O] (RuO ) oxidizes C-H bonds of alkyl-
aromatic and allylic substrates, forming oxygenated products
such as xanthone), a dehydrogenated product (anthracene), and/
or products from radical coupling (9,9′-bixanthenyl). The kinetic
and mechanistic data are consistent with a mechanism of initial
hydrogen-atom abstraction from the substrate to the ruthenium
4
ppm relative to SiMe . UV-vis spectra were recorded on a Hewlett-
IV
2+
2+
[
Packard 8453 diode array spectrophotometer and are reported as λ
max
-1
-1
(nm), (ꢀ, M cm ). GC/MS data were obtained on a Hewlett-Packard
5971 instrument equipped with a nonpolar capillary column and a mass
spectral analyzer. GC/FID spectra were obtained on a Hewlett-Packard
(
5
890 instrument equipped with a similar column. Yields of organic
products were quantified using response factors and confirmed by the
addition of authentic samples.
(
54) (a) This argument is related to one made by Groves, personal communica-
tion, ref 39h, and Brazeau, B. J.; Austin, R. N.; Tarr, C.; Groves, J. T.;
Lipscomb, J. D. J. Am. Chem. Soc. 2001, 123, 11 831-11 837. (b) Related
arguments have been in the context of dioxirane oxidations, cf., Curci, R.;
Dinoi, A.; Fusco, C.; Lillo, M. A. Tetr. Lett. 1996, 37, 249-252.
Typical Procedure for Organic Oxidations. A solution of xanthene
4 mg, 20 µmol) and RuO (5 µmol) in 5 mL MeCN turned orange
2+
(
(58) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory Chemicals,
(
55) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265-322.
56) Jackson, R. A.; O’Neill, D. W. J. Chem. Soc., Chem. Commun. 1969, 1210-
3rd ed.; Pergamon Press: New York, 1988.
(
(59) Sullivan, B. P.; Salmon, D. J.; Meyer, T. J. Inorg. Chem. 1978, 17, 3334-
1
211.
3341.
57) ln fxy ) [1/4(lnKxy)2]/[ln(kxx
k
yy/Z )]. Using the values given above, with
2
55
(60) Bailey, R. J.; Card, P. J.; Schechter, H. J. Am. Chem. Soc. 1983, 105, 6096-
(
11
Z ) 10 (ref 55), fxy ) 0.98.
6103.
10360 J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003

Oxidation of C−H Bonds
A R T I C L E S
within 5 min. 9,9′-Bixanthene was observed by GC/MS m/z 181, M+
(
Self-Exchange Measurements. The self-exchange rates for RuO2+/
at twice the retention time of xanthene), 152, 69, 39.39d Xanthone was
RuOH2+ and RuO /Ru-OD were measured by dynamic H NMR
2+
2+
1
2+
observed at m/z 196, 168, 139. Oxidation of fluorene showed bifluorene
m/z 330, 165), bifluorenylidene (m/z 328), and fluorenone (m/z 180,
52). Oxidations of DHA were monitored by GC/MS and by UV-vis
spectroscopy, where anthracene production was evident from its
characteristic spectrum in CH CN: 359 (8800), 378 (8200) after
line broadening. A solution of 4.5 mg RuO (5.6 µmol), 2.3 mg
2
+
(
1
RuOH
2
(2.9 µmol) and 1 µL (Me
3 2
Si) O (as an internal standard) in
1
0.5 mL CD
3
CN was prepared in a sealable NMR tube. H NMR spectra
2
+
were acquired every 10 min for 3-4 h. Line widths for RuO were
measured at δ 53.1 and -31.0 ppm by fitting to Lorentzian functions
using the commercially available NUTS software (Acorn NMR). Values
3
II
removal of the Ru product by passing the solution through a silica
pipet column. Anthracene was quantitated by UV-vis and by GC/FID
response factors. Anthraquinone production was also observed by GC/
2+
for ∆W were obtained by subtracting the natural line widths for RuO
in the absence of RuOH2 from the values obtained by Lorentzian
+
1
8
MS (m/z 208, 180, 152). O-labeling studies were conducted under
fitting.
18
2
the same conditions as above, but with 10mM added H O. The amount
1
8
+
+
of O incorporation was measured by comparing the M :M + 2 peak
Acknowledgment. We are grateful to Dr. Martin Sadilek and
Dr. Brian Bales for assistance with mass spectra and chroma-
tography and to Dr. T. J. Meyer and Dr. M. H. V. Huynh for
sharing results prior to publication. We acknowledge the
National Institutes of Health and the UW PRIME Fellowship
for financial support.
1
8
2
intensity ratios for products formed in the presence of H O to those
formed in the absence of H O.
2
1
8
Kinetics Studies. Stopped-flow kinetic data were obtained using
an OLIS stopped flow apparatus with a rapid scanning monochromator,
typically obtaining spectra from 350 to 650 nm every 0.001 to 0.5 s
over 1-1000 s. Kinetic data were analyzed using the global analysis
3
2
2+
software package SPECFIT. Reactions of RuO with xanthene,
2
+
Supporting Information Available: Kinetics traces and
SPECFIT calculated spectra for the oxidation of fluorene, plots
of log k vs log[x] for xanthene, DHA, and fluorene, Eyring plot
for the oxidation of xanthene, and an overlay plot and kinetic
trace for the oxidation of anthracene are included in the
Supporting Information. This material is available free of charge
via the Internet at http://pubs.acs.org.
fluorene and DHA were carried out with 0.2 mM RuO and 2-80
mM substrate in MeCN. Solutions were made up in an N -filled drybox
immediately prior to use. Oxidations of xanthene in O -saturated
solution were carried out in the stopped flow instrument using
acetonitrile sparged with O prior to mixing with the reagents. The
2
2
2
kinetics of anthracene oxidation by RuO2+ were slow enough to be
carried out with the HP 8453 diode array spectrophotometer, monitoring
every 2 s for 600 s. Solutions were made up and transferred to a cuvette
with a Teflon Kontes valve in a nitrogen-filled glovebox.
JA035276W
J. AM. CHEM. SOC.
9
VOL. 125, NO. 34, 2003 10361