Activation of molecular oxygen by a dioxygenase pathway by a ruthenium bis-bipyridine compound with a proximal selenium site

Article abstract of DOI:10.1021/ja9047027

A ruthenium(II) bipyridine complex with proximal phenylselenium tethers, [Ru](H₂O)₂, reacted intramolecularly with O₂ in a protic slightly acidic solvent, 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), to yield an O-O bond cleaved product, [Ru](O)₂, with formation of two Ru-O-Se moieties. This stable compound was isolated, and its structure was determined by X-ray diffraction. The identification of the compound in solution was confirmed by ESI-MS and the ¹H NMR with the associated Curie plot that showed that [Ru](O)₂ was paramagnetic. The magnetic susceptibility was 2.8 μB by Evan's method suggesting a ground state triplet or biradical. DFT calculations, however, predicted a ground state singlet and an oxidized Se atom. Further it was shown that [Ru](O)₂ is a potent oxygen transfer species of both O₂-derived atoms to triphenylphosphine and a nucleophilic alkene such as 2,3-dimethyl-2-butene in both HFIP and acetonitrile. UV-vis spectroscopy combined with the measured stoichiometry of PPh₃:O₂ = 2 in a catalytic oxidation of PPh₃ suggests a dioxygenase type activation of O ₂ with structural identification of the O-O bond cleavage reaction step, formation of [Ru](O)₂ as an intermediate, and the proof that [Ru](O)₂ is a donor of both oxygen atoms.

Full text of DOI:10.1021/ja9047027

Published on Web 12/16/2009
Activation of Molecular Oxygen by a Dioxygenase Pathway by
a Ruthenium Bis-bipyridine Compound with a Proximal
Selenium Site
Alexander Laskavy,† Linda J. W. Shimon,‡ Leonid Konstantinovski,‡ Mark A. Iron,‡
and Ronny Neumann*,†
Department of Organic Chemistry and Chemical Research Support Unit, Weizmann Institute of
Science, RehoVot, Israel 76100
Received June 9, 2009; E-mail: Ronny.Neumann@weizmann.ac.il
Abstract: A ruthenium(II) bipyridine complex with proximal phenylselenium tethers, [Ru](H2O)2, reacted
intramolecularly with O2 in a protic slightly acidic solvent, 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), to yield
an O-O bond cleaved product, [Ru](O)2, with formation of two Ru-O-Se moieties. This stable compound
was isolated, and its structure was determined by X-ray diffraction. The identification of the compound in
solution was confirmed by ESI-MS and the 1H NMR with the associated Curie plot that showed that [Ru](O)2
was paramagnetic. The magnetic susceptibility was 2.8 µB by Evan’s method suggesting a ground state
triplet or biradical. DFT calculations, however, predicted a ground state singlet and an oxidized Se atom.
Further it was shown that [Ru](O)2 is a potent oxygen transfer species of both O2-derived atoms to
triphenylphosphine and a nucleophilic alkene such as 2,3-dimethyl-2-butene in both HFIP and acetonitrile.
UV-vis spectroscopy combined with the measured stoichiometry of PPh3:O2 ) ∼2 in a catalytic oxidation
of PPh3 suggests a dioxygenase type activation of O2 with structural identification of the O-O bond cleavage
reaction step, formation of [Ru](O)2 as an intermediate, and the proof that [Ru](O)2 is a donor of both
oxygen atoms.
1. Introduction
tion is attained through use of monooxygenase type enzymes
where reducing agents and protons (coreactants) supply the
energy for the cleavage of the O-O bond that leads to formation
of water and reactive high-valent metal species such as iron-
oxo species.6
The literature shows that there are only a few synthetic
catalytic systems that use O2 only, i.e. without a coreagent, for
hydrocarbon oxidation. (i) Notable is the gas phase process for
ethene epoxidation and related epoxidations of substrates without
allylic hydrogen atoms.7 (ii) The formation of OsO4 and its use
for syn alkene dihydroxylation was described.8 (iii) Sterically
hindered ruthenium porphyrins were shown to catalyze alkene
epoxidation,9 and a similar mechanism has been postulated by
us for O2 activation with ruthenium substituted polyoxometa-
lates.10 (iv) In related research, an iron(IV)-oxo species that
oxidized triphenylphosphine has been reported to be formed by
The activation of molecular oxygen is of fundamental
importance and presents an important, difficult, scientific
challenge in the context of oxidation catalysis that also may
have consequences in the area of “green chemistry”.1 The
chemistry of ground state O2 is mechanistically complicated.2
Thus, O2 ubiquitously reacts with hydrocarbons by a free radical
auotoxidation mechanism. The initial hydroperoxide is typically
unstable under reaction conditions, and often nonselective
product formation is observed.3 Some metal complexes can
mediate electron transfer or oxidase type oxidation of organic
substrates with or without concomitant oxygen transfer where
O2 is used to reoxidize the catalyst.4,5 The activation of O2 in
the “biological world” for hydrocarbon (alkane, alkene) oxida-
† Department of Organic Chemistry.
‡ Chemical Research Support Unit.
(1) (a) Punniyamurty, T.; Velusamy, S.; Iqbal, J. Chem. ReV. 2005, 105,
2329. (b) Mallat, T.; Baiker, A. Chem. ReV. 2004, 104, 3037.
(2) (a) Modern Oxidation Methods; Backvall, J. E., Ed.; Wiley-VCH:
Weinheim, Germany, 2004. (b) Simandi, L. I. AdVances in Catalytic
ActiVation of Dioxygen by Metal Complexes; Kluwer Academic:
Boston, USA, 2003. (c) Sawyer, D. T. Oxygen Chemistry; Oxford
University Press: New York, 1991.
(6) (a) Cytochrome P-450, 2nd ed.; Ortiz de Montellano, P., Ed.; Marcel
Dekker: New York, 1995. (b) Sono, M.; Roach, M. P.; Coulter, E. D.;
Dawson, J. H. Chem. ReV. 1996, 96, 2841. (c) Costas, M.; Mehn,
M. P.; Jensen, R.; Que, L., Jr. Chem. ReV. 2004, 104, 939. (d) Tshuva,
E. Y.; Lippard, S. J. Chem. ReV. 2004, 104, 987. (e) Que, L., Jr.;
Tolman, W. B. Angew. Chem., Int. Ed. 2002, 41, 1114. (f) Solomon,
E. I.; Brunold, T. C.; Davis, M. I.; Kemsley, J. N.; Lee, S. K.; Lehnert,
N.; Neese, F.; Skulan, A. J.; Yang, Y. S.; Zhou, J. Chem. ReV. 2000,
100, 235.
(3) (a) Walling, C. Struct. Energ. React. Chem. Ser. 1995, 2, 24–65. (b)
Partenheimer, W. Catal. Today 1995, 23, 69. (c) Mayo, F. R. Acc.
Chem. Res. 1968, 1, 193. (d) Hermans, I.; Peeters, J.; Jacobs, P. A.
Top. Catal. 2008, 50, 124.
(7) (a) Kilty, P. A.; Sachtler, W. M. H. Catal. ReV. 1974, 10, 1. (b) Barteau,
M. A. Top. Catal. 2003, 22, 3. (c) Monnier, J. R.; Peters, K. T.;
Hartley, G. W. J. Catal. 2004, 225, 374.
(4) (a) Stahl, S. S. Angew. Chem., Int. Ed. 2004, 43, 3400. (b) Feringa,
B. L. In Transition Metals for Organic Chemistry Volume 2; Beller,
M., Bolm, C., Eds.; Wiley-VCH: 1998; p 307.
(8) (a) Do¨bler, C.; Mehltretter, G.; Beller, M. Angew. Chem., Int. Ed.
1999, 38, 3026. (b) Do¨bler, C.; Mehltretter, G.; Sundermeier, U.;
Beller, M. J. Am. Chem. Soc. 2000, 122, 10289. (c) Do¨bler, C.;
Mehltretter, G.; Sundermeier, U.; Beller, M. J. Organomet. Chem.
2001, 621, 70–76.
(5) (a) Neumann, R. Prog. Inorg. Chem. 1998, 47, 317. (b) Kozhevnikov,
I. V. Catalysts for Fine Chemical Synthesis, Volume 2; Catalysis by
Polyoxometalates; Wiley: Weinheim, Germany, 2002.
9
10.1021/ja9047027  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 517–523 517

A R T I C L E S
Laskavy et al.
Scheme 1. O2 Activation on Ruthenium-Selenium Complexes
addition of O2 to an iron(II)-cyclam complex.11 In both the
ruthenium and iron catalyzed systems,9–11 the mechanisms for
formation of the reactive metal-oxo intermediates and the O-O
bond cleavage step have not been structurally established
although the O2 activation step was followed by spectroscopy.
Herein we describe a dioxygenase type catalyst for O2
activation that enables cleavage of the O-O bond and the
observation of the O-O bond cleavage via isolation of the
reactive metal-oxo intermediate. To this end we have prepared
2,2-bipyridine ligands with arylselenium tethers. The concept
for O2 activation and subsequent oxygen transfer is presented
in Scheme 1.
This compound was designed as such because of (a) its R-cis
conformation, (b) the proximity of the adjacent oxophilic
selenium, (c) the significant steric hindrance around the transi-
tion metal center, and (d) the possibility for intramolecular O2
activation. The R-cis conformation is advantageous because it
can enable the binding of O2 at the metal center and lead, after
cleavage of the oxygen-oxygen bond, to a cis-dioxo species
that are known as effective oxygen donors. The proximal
selenium center is favorable for first aiding in the cleavage of
the oxygen-oxygen bond by introduction of an oxophilic
element and second for stabilization of the metal-oxo intermedi-
ate formed. The steric hindrance around the transition metal
center is important to avoid formation of intermolecular
µ-peroxo species that often disproportionate to µ-oxo species
that are “dead-end” thermodynamically stable compounds. The
intramolecular O2 activation is advantageous relative to the
previously reported intermolecular O2 activation because it will
allow heterogenization of the catalyst in the future, which will
be important for epoxidation of important gaseous substrates
such as propene.
Figure 1. ORTEP drawing (50% probability) of [Ru](ACN)2 (Ru, magenta;
Se, green; N, blue; C, black; solvent molecules, counteranions, and hydrogen
atoms are not shown for clarity).
Table 1. Key Interatomic Distances and Angles for [Ru](ACN)2
Bond Lengths, Å
Bond Angles, deg
Ru1 N1 2.080(5) N1 Ru1 N6 177.2(2) N2 Ru1 N4 103.9(2)
Ru1 N2 2.108(5) N4 Ru1 N5 171.8(2) N2 Ru1 N6 100.6(2)
Ru1 N3 2.075(5) N1 Ru1 N4
Ru1 N4 2.080(5) N4 Ru1 N6
Ru1 N5 2.033(6) N6 Ru1 N5
Ru1 N6 2.042(5) N5 Ru1 N1
97.4(2) N2 Ru1 N5
85.4(2) N3 Ru1 N1 101.2(2)
86.9(2) N3 Ru1 N4
90.4(2) N3 Ru1 N5
80.2(2)
79.0(2)
96.9(2)
79.4(2)
N2 Ru1 N3 177.1(2) N3 Ru1 N6
N2 Ru1 N1 78.7(2)
bipy)2(L)2](X)2 ([Ru](L)2), L ) H2O or CH3CN; X ) coun-
teranion. The X-ray structure of [Ru](ACN)2 indeed shows the
expected slightly distorted octahedral complex with an R-cis
conformation, Figure 1 and Table 1. The six Ru-N distances
are 2.033, 2.042, 2.075, 2.080, 2.080, and 2.108 Å. The distances
between the Ru and Se atoms are in the range of 3.593 to 3.628
Å respectively; there are no bonding interactions between the
Ru and Se atoms.
2. Results and Discussion
The 6,6′-bis-phenylselenyl-2,2′-bipyridine ligand was pre-
pared by the CsOH catalyzed alkylation12 of 6,6′-dibromo-2,2′-
bipyridine with diphenydiselenide. The Ru(II) compound was
then prepared along the lines of a literature preparation13 to yield
the required Ru(II) compound with arylselenium tethers and two
labile L ) H2O or CH3CN ligands [Ru(6,6′-(SePh)2-2,2′-
The analogous aqua complex [Ru](H2O)2, was extensively
1
characterized by HR-ESI-MS; IR; and H, 13C, 15N, and 77Se
NMR spectroscopy (see Figures S5-S11). The ESI-MS showed
the expected molecular ion cluster (without anions) of peaks at
m/z ) 1069 and a cluster of peaks at m/z ) 517 assignable to
1
[M - (2 × H2O)]/2. The H, 13C, 15N, and 77Se NMR spectra
(9) (a) Groves, J. T.; Quinn, R. J. Am. Chem. Soc. 1985, 107, 5790. (b)
Groves, J. T.; Ahn, K. H. Inorg. Chem. 1987, 26, 3831. (c) James,
B. R.; Pacheco, A.; Rettig, S. J.; Thorburn, I. S.; Ball, R. G.; Ibers,
J. A. J. Mol. Catal. 1987, 41, 147.
all showed that the compound is nonsymmetric. For example,
four peaks were observed in the 15N NMR for the N atoms of
bipy at δ ) 237.21, 249.89, 253.11, and 256.72 ppm. Similarly,
four peaks were observed in the 77Se NMR at δ ) 491.42,
493.40, 501.69, and 502.15 ppm.
(10) (a) Neumann, R.; Dahan, M. Nature 1997, 388, 353. (b) Neumann,
R.; Dahan, M. J. Am. Chem. Soc. 1998, 120, 11969.
(11) Kim, S. O.; Sastri, C. V.; Seo, M. S.; Kim, J.; Nam, W. J. Am. Chem.
Soc. 2005, 127, 4178.
The Ru(II) compound [Ru](H2O)2 was dissolved in 1,1,1,3,3,3-
hexafluoro-iso-propanol (HFIP) and reacted with O2 at RT for
48 h. It should be noted that such a slightly acidic solvent was
(12) Varala, R.; Ramu, E.; Adapa, S. R. Bull. Chem. Soc. Jpn. 2006, 79,
140.
(13) Collin, J. P.; Sauvage, J. P. Inorg. Chem. 1986, 25, 135.
9
518 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Activation of O2 by RuII Bipyridine Complex
A R T I C L E S
Figure 2. ORTEP drawing (50% probability) of [Ru](O)2 (Ru, magenta;
Se, green; N, blue; O, red; C, black; solvent molecules, counteranions, and
hydrogen atoms are not shown for clarity).
Figure 3. High resolution ESI MS of [Ru](O)2.
Table 2. Key Bond Lengths and Angles for [Ru](O)2
Figure 4. 1H NMR of [Ru](O)2 at 216 K.
tetrahedral angle. Therefore, the Ru-O and Se-O bond lengths
and the Ru-O-Se bond angles are consistent with an oxygen
atom bound to and bridging between the Ru and Se atoms. The
O-O distance is 2.851 Å clearly indicating that the O2 molecule
has been cleaved.16
The high resolution positive ion mode ESI-mass spectra,
Figure 3, of [Ru](O)2 shows the expected cluster of peaks at
m/z ) 533 which is attributable to the half mass of the molecular
1
peaks (without anion and solvent). The H NMR of [Ru](O)2
taken at 216 K clearly shows that [Ru](O)2 is paramagnetic
with peaks appearing at chemical shifts ranging from -32.8 to
26.6 ppm, Figure 4. Plots of the 1H NMR spectra as a function
of temperature and then plots of the chemical shifts as a function
of temperature (Figures S12-13) show a linear behavior
consistent with a Curie paramagnetic species. Not surprisingly
[Ru](O)2 was EPR silent down to 120 K because, as the NMR
suggests, the relaxation times are very fast. Also, measurement
needed for this reaction; acetonitrile gave no product. From this
reaction we were able to isolate and crystallize the intended
[Ru](O)2 compound whose X-ray diffraction derived structure
is shown in Figure 2. Important bond lengths and angles are
listed in Table 2. The structure shows the presence of two
oxygen atoms between the ruthenium and selenium sites. The
Se-O bond lengths (1.723-1.733 Å) are longer than that
observed for SeO (1.634 Å)14 but very similar to the bond length
for SeO- (1.726 Å)15 as may be expected for a bridging
situation. The Ru-O bond lengths of ∼2.09 Å are also longer
than those expected for a terminal Ru-O bond (1.5-1.8 Å).
The Ru-O-Se bond angles are slightly larger than the
of the magnetic moment by the Evan’s method yielded µeff
2.8 µB which is roughly consistent with two unpaired spins.
)
(16) Sterically hindered RuII(2,9-dimethyl-1,10-phenanthroline)2(H2O)2 has
been reported (UV-vis spectra only) to yield dioxo species upon
reaction with H2O2. In HFIP the reaction of RuII(2,9-dimethyl-1,10-
phenanthroline)2(H2O)2 with O2 showed no change in the UV-vis
spectrum and no formation of Ru-oxo compounds. (a) Goldstein, A. S.;
Beer, R. H.; Drago, R. S. J. Am. Chem. Soc. 1994, 116, 2424. (b)
Goldstein, A. S.; Drago, R. S. J. Chem. Soc., Chem. Commun. 1991,
21.
(14) Verma, K. K.; Reddy, S. P. J. Mol. Spectrosc. 1977, 67, 360.
(15) Coe, J. V.; Snodgrass, J. T.; Freidhoff, C. B.; McHugh, K. M.; Bowen,
K. H. J. Chem. Phys. 1986, 84, 618.
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 519

A R T I C L E S
Laskavy et al.
Table 4. Selected Natural Bond Orbital Results for [Ru](O)2
Ru(83) Se(84) Se(85)
NPA Core
35.987
27.997
27.998
Valence 7.313
Rydberg 0.101
4.472
0.052
32.520
1.480
5.523
0.033
33.554
0.446
Total
43.402
0.598
Charge
Natural Electron [core] 4d7.055s0.26 [core]4s1.644p2.83 [core]4s1.724p3.81
Configuration
and a poorer agreement for a triplet state. Likewise, there is
reasonable agreement in the analogous bond distances in
[Ru](ACN)2 (see Table S1).
The measured singlet-triplet gap at the M06/SDB-cc-pVDZ//
M06-L/SDB-cc-pVDZ/DBFS level of theory is 1.05 eV. (For
comparison, at the M06-L/SDB-cc-pVDZ/DBFS level, the gap
is 1.07 eV.) To check whether M06 has a tendency to over- or
underestimate this gap, the vertical singlet-triplet gap (i.e., the
difference between the singlet and triplet energies calculated
for the singlet geometry) was calculated for a number of
different exchange-correlation functionals such as B3LYP, PBE,
and PBE0, Table S2. All the functionals predict a lower energy
for the singlet state. Accounting for differences in zero-point
vibrational energies (ZPVE) at the M06-L/SDB-cc-pVDZ/DBFS
level also led to a lower energy singlet state. It should, however,
be noted that these calculations are at 0 K while the measure-
ments were carried out above 200 K.
An additional question that was probed by the DFT calcula-
tions is with regards to the formal oxidation states of the
selenium and ruthenium atoms in [Ru](O)2. The NBO analysis
shows that significant positive charge is localized on the
selenium atoms. Table 4 shows the natural population analysis
for Se84 (bound to O, similar results are obtained for Se86),
Se85 (not bound to O, similar results are obtained for Se87),
and Ru83. On comparison of the two selenium atoms, it was
observed that there is clearly less electron density on Se84/
Se86 than on Se85/Se87. Note that there is no real physical
meaning to the absolute value of charge calculated. One can
note that the results obtained in [Ru](O)2 for Se85 (not bound
to O) are very similar to those obtained for the selenium atoms
in [Ru](ACN)2, Table S3, where these atoms are undisputedly
SeII. Thus, if one accepts that Se85/Se87 is SeII, as it is bound
to two aryl rings, then Se84/Se86 would formally have a higher
oxidation state, e.g. SeIII or SeIV.
Although the experimental results indicate a triplet state for
[Ru](O)2 from which we hypothesize low spin RuIV and two
SeIII centers, that is SeIII-O-RuIV-O-SeIII, the DFT calcula-
tions indicate a singlet ground state with a more likely
SeIV-O- · · ·RuII · · ·-O-SeIV configuration. Relevant to this
dichotomy is also the observation that, as will be shown below,
[Ru](O)2 is an oxygen donor. This typically implies a higher
valent ruthenium-oxo active species.
The UV-vis spectra of [Ru](H2O)2 and [Ru](O)2 are very
similar and relatively featureless. We were, however, interested
in the evolution of the O2 cleaved species that was more easily
observable by measurement of difference UV-vis spectra,
Figure 6. The difference spectra showed an isosbestic point at
297 nm indicating a clean bimolecular transformation of
[Ru](H2O)2 plus O2 to [Ru](O)2 on the measured time scale.
The reactivity of [Ru](O)2 was tested under a variety of
conditions. First, in stoichiometric reactions 3 µmol of [Ru](O)2
were reacted with 30 µmol of PPh3 in 1.5 mL of HFIP under
Ar. After 2 h at RT a 19.7% conversion (5.9 µmol) of PPh3 to
Figure 5. Region where differences were observed in the IR spectra of
[Ru](16O)2 (blue) and [Ru](18O)2 (red).
Table 3. Selected Bond Distances (Å) in [Ru](O)2 (Atomic Labels
Figure S15
Refer to the Calculated Structures Depicted in
)
[Ru]O2
Bond
Singlet
Triplet
X-ray
O(81)-O(82)
O(81)-Se(84)
O(82)-Se(86)
O(82)-Ru(83)
O(81)-Ru(83)
N(77)-Ru(83)a
N(78)-Ru(83)b
N(79)-Ru(83)a
N(80)-Ru(83)b
Se(84)-C(42)
Se(84)-C(53)
Se(86)-C(46)
Se(86)-C(65)
2.837
1.739
1.739
2.134
2.134
2.020
2.108
2.020
2.108
1.965
1.928
1.964
1.928
3.223
1.755
1.731
2.215
2.078
2.137
2.177
2.010
2.376
1.954
1.929
1.983
1.936
2.851
1.733
1.723
2.095
2.092
2.004
1.999
2.040
2.099
1.950
1.926
1.942
1.910
a Mutually trans pyridine rings. b Pyridine ring trans to O.
The 1H NMR spectra and magnetic susceptibility measurements
clearly suggest that [Ru](O)2 is not diamagnetic.
The IR spectra of [Ru](H2O)2 and [Ru](O)2 were also
compared (Figure S14). Most interesting in this context is the
IR assignment of the Ru-O-Se moiety that was identified by
comparison of the IR spectra of [Ru](O)2 and [Ru](18O)2, the
later being prepared from [Ru](H2O)2 and 18O2. Thus the peaks
at 557 and 517 cm-1 can be assigned to the Ru-O-Se and
Ru-18O-Se vibrations, respectively, Figure 5. The isotopically
shifted peak in the IR spectrum at 517 cm-1 is close to the
expected absorption at 525 cm-1
.
Considering that for the starting complex, [Ru](H2O)2, the
oxidation states of the Ru and Se sites are both formally 2+, a
four electron oxidation with O2 indicates the possibility of three
most likely electronic configurations for [Ru](O)2: (i)
SeII · · ·O-RuVI-O· · ·SeII, (ii) SeIV-O- · · ·RuII · · ·-O-SeIV, or
(iii) SeIII-O-RuIV-O-SeIII. Since both ruthenium and selenium
invariably take on low spin configurations, a working hypothesis
based on the experimental evidence for the assignment of the
oxidation states in [Ru](O)2 is that the compound has a low
spin RuIV and two SeIII centers. The calculated (DFT) structure
of [Ru](O)2 is depicted in Figure S15 that includes also the
atomic numbering scheme used for the calculated structure.
Selected bond lengths for both the singlet and triplet states along
with their corresponding experimental X-ray crystallography
measured equivalents are listed in Table 3. In this context, it
should be noted that there is reasonable agreement between the
calculated and measured geometries for a ground state singlet
9
520 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Activation of O2 by RuII Bipyridine Complex
A R T I C L E S
rearrangement enabled by the proximal Se sites to yield the
stable [Ru](O)2 compound, which in the presence of a reactive
substrate is a viable oxygen transfer species. Two lines of
evidence support in general terms such a dioxygenase activation
of O2. First, the observation of the formation of [Ru](O)2 from
[Ru](H2O)2 by UV-vis spectroscopy appears to indicate a clean
one-step transformation. Second, the catalytic oxidation of PPh3
with O2 catalyzed by [Ru](H2O)2 showed a requisite PPh3:O2
stoichiometry of ∼2. A third desired line of evidence would be
to react [Ru](H2O)2 with a 1:1 mixture of 16O2:18O2 and analyze
the [Ru](16O)2:[Ru](18O)2 ratio by ESI-MS; no [Ru]16O18O
should be formed. Unfortunately, this experiment failed because
the activated O atom is freely exchanged by water present in
the electrospray ionization source used to analyze [Ru](18O)2.
Additional and numerous attempts to use other MS methods,
such as MALDI-TOF-MS with various matrices and Field
Desorption-MS, did not give molecular peaks, but rather
fragmentation products (both methods) and substitution of the
oxygen atom by the matrix (MALDI-TOF).
Figure 6. Difference spectra showing the evolution of [Ru](O)2 from 1.8
µM [Ru](H2O)2 at 25 °C in HFIP blanketed by 1 bar of O2. Spectra
measured every 15 min.
PPh3dO was measured by GC proving nearly complete oxygen
transfer of both O atoms. Although as noted above [Ru](O)2 is
not formed in acetonitrile, it can be isolated and then reacted
in this solvent. For example, 3 µmol of [Ru](O)2 was reacted
with 30 µmol of PPh3 in 1.5 mL of acetonitrile under Ar. After
8 h at RT a 19.3% conversion (5.8 µmol) of PPh3 to PPh3dO
was measured by GC proving nearly complete, albeit slower,
oxygen transfer of both O atoms. This reactivity was verified
by (a) reaction of PPh3 with 18O labeled [Ru](18O)2 yielding
PPh3d18O (>95% 18O labeled) and (b) measurement of the ESI-
MS of the reaction mixture with time that showed the disap-
pearance of [Ru](O)2 (m/z ) 533) concomitant with the
appearanceof[Ru](O)(H2O)(m/z)525)andthen[Ru](H2O)2(m/
z ) 517). At the end of the reaction mostly [Ru](H2O)2 was
observed along with a small amount of [Ru](O)(H2O). Simi-
larly, [Ru](O)2 is reactive toward a nucleophilic alkene such a
2,3-dimethyl-2-butene in both HFIP and acetonitrile. Thus, 7.4
µmol of [Ru](O)2 and 42 µmol of 2,3-dimethyl-2-butene were
reacted in 1.5 mL of solvent at 23 °C, under Ar. In acetonitrile,
the reaction yielded, after 12 h, 15.3 µmol (96%) of 2,3-
dimethyl-2-butene oxide. In HFIP, the reaction yielded, after
4 h, 15.1 µmol (94%) of 3,3-dimethyl-2-butanone (pinacolone)
and traces of 2,3-dimethyl-2,3-butanediol (pinacol). Apparently
the acidity of the HFIP solvent leads to the hydrolysis of the
initial epoxide formed and its subsequent pinacol rearrangement
to the final product. This was verified by reacting 2,3-dimethyl-
2-butene oxide in HFIP. Second, a catalytic oxidation of PPh3
with O2 catalyzed by [Ru](H2O)2 (600 µmol of PPh3, 30 µmol
of [Ru](H2O)2, 8.8 mL of HFIP, 26 °C under 1 bar of O2 in a
gas buret) showed a PPh3:O2 stoichiometry of 1.98 indicating
that a dioxygenase pathway to PPh3dO indeed is operating.
One should emphasize that the data available to date do not
exclude the possibility of a bimolecular mechanism for the
activation of O2, for example via a [Ru]O2[Ru] intermediate.
Considering, however, the steric bulk around the ruthenium
metal center, we are suggesting an intramolecular O2 activation
mechanism as an initial working hypothesis.
3. Conclusions
Synthesis of an octahedral ruthenium(II) compound based on
a 2,2′-bipyridyl ligand with proximal arylselenium tethers at
the 6,6′ position led to a compound, [Ru](H2O)2, with two labile
aqua ligands at the R-cis positions. Crystals for solid state
identification, Figure 1, were available by preparation of the
analogous [Ru](ACN)2. Reaction of [Ru](H2O)2 with O2
showed by UV-vis, Figure 6, a one-step transformation to a
ruthenium dioxo species, [Ru](O)2, that was identified in the
solid state by X-ray crystallography, Figure 2, and showed the
formation of two Ru-O-Se moieties that were also identified
by IR spectroscopy, Figure 5. In solution, HR-ESI-MS, Figure
3, showed the requisite m/z (z ) 2) cluster of molecular peaks.
The 1H NMR, Figure 4, and the associated Curie plots showed
that [Ru](O)2 is paramagnetic. The magnetic susceptibility by
Evans’ method was 2.8 µB. This value is close to a triplet or
biradical and suggests a formulation of oxidation states in
[Ru](O)2 as SeIII-O-RuIV-O-SeIII. An alternative formulation
is suggested by DFT calculations at 0 K which pointed toward
a ground state singlet and oxidized Se atoms, more likely a
SeIV-O- · · ·RuII · · ·-O-SeIV formulation. The dichotomy be-
tween the experimental evidence and the calculated results
remains unresolved; however, it should be stressed that [Ru](O)2
is a potent oxygen donor, capable of transferring both oxygen
atoms originally derived from O2 as demonstrated by formation
of [Ru](18O)2 from [Ru](H2O)2 and 18O2. The fact that [Ru](O)2
was indeed reactive toward nucleophiles such as triphenylphos-
phine and 2,3-dimethyl-2-butene suggests that the ruthenium
center is indeed partly in a higher oxidation state, since it would
The turnover frequency was 0.83 h-1
.
Given the results collected and presented above, one may
present a working hypothesis for the activation of O2 by
[Ru](H2O)2, Scheme 2. The reaction is initiated by coordination
of O2 to [Ru](H2O)2 via ligand exchange. In a likely proton
assisted step, a protonated Ru(III)-superoxo species undergoes
Scheme 2. Proposed Mechanism for Formation of [Ru](O)2
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 521

A R T I C L E S
Laskavy et al.
appear more likely based on literature precedence3,9 that a higher
valent Ru-oxo species would be an oxygen donor versus a
SeIV-O- · · ·RuII · · ·-O-SeIV configuration which would perhaps
be expected to be inert as an oxygen donor. Future research
will be devoted to more reactive analogues of [Ru](H2O)2 that
will be able to activate O2 and oxygenate less nucleophilic
terminal aliphatic alkenes.
Table 5. Crystal Data and Structure Refinement for Compounds
[Ru](ACN)2 and [Ru](O)2
Compound
[Ru](ACN)2
[Ru](O)2
Empirical formula
C48H38N6RuSe4 +
Na+3ClO4
C44H32N4O2RuSe4+
PF6 + Cl + 29/6CHCl3
1823.01
Formula weight
Space group
a, Å
b, Å
c, Å
1437.09
j
j
Triclinic, P 1
Rhombohedral, R 3
12.830(3)
14.283(3)
16.894(5)
90.61(2)
104.71(2)
116.42(1)
2654.2(13), 2
1.798
21.8438(4)
21.8438(4)
21.8438(4)
99.34(1)
99.34(1)
99.34(1)
9956.3(3), 1
1.824
3.136
4. Experimental Section
R, deg
4.1. Synthetic Procedures. 6,6′-Bis-phenylselenyl-2,2′-bipyri-
dine (bipySePh). 1.51 g (4.8 mmol) of 6,6′-dibromo-2,2′-bipyridine,
2.25 g (7.2 mmol) of diphenyl diselenide ((SePh)2), and 2.19 g
(13.1 mmol) of CsOH ·H2O in 14.4 mL of dry DMSO were stirred
under Ar at 110 °C for 6 h. The DMSO was distilled off under
vacuum, and the remaining solid was washed 5 times with H2O.
The ligand was purified on a silica gel column using 65/35 CH2Cl2/
n-hexane as eluent. Yield ) 67.1% (3.13 g) (white microcrystalline
product). Elemental Analysis, calculated: C, 56.67%; H, 3.46%;
N, 6.01%. Found: C, 56.49%; H, 3.39%; N, 5.96%. For further
characterization, see Supporting Information, Figures S1-S4.
ꢀ, deg
γ, deg
V (Å3), Z
dcalc (mg/cm3)
µ (mm-1
)
3.264
R1 ) 0.0580,
wR2 ) 0.1430
R [I > 2σ(I)]a
R1 ) 0.0591
wR2 ) 0.1525
R1 ) 0.1121;
wR2 ) 0.1984
R (all data)
R1 ) 0.0877;
wR2 ) 0.1542
a
2
2
R1 ) Σ|F0| - |Fc|/Σ; wR2 ) {Σ[w(F0 - Fc2)2]/Σw(F0 )2]}1/2
.
[Ru(6,6′-(SePh)2-2,2′-bipy)2(H2O)2](PF6)2 ([Ru](H2O)2). 105 mg
(0.40 mmol) of RuCl3 ·xH2O, 424 mg (0.91 mmol) of bipySePh,
and 268 mg (6.32 mmol) of LiCl were stirred in 2.6 mL of ethylene
glycol under Ar at 150 °C for 10 h. The mixture was cooled to
RT, and the preciptate formed was washed with H2O until the wash
was colorless and then dried under vacuum overnight. The
precipitate was dissolved in 40 mL of MeOH, and 300 mL of Et2O
were slowly added while stirring to yield a brown microcrystalline
precipitate. The wash with Et2O was continued until all the free
ligand was extracted and nearly pure Ru(6,6′-(SePh)2-2,2′-bipy)2Cl2
was obtained (ESI-MS: m/z - 1104.70). Yield ) 77.4% (324 mg).
Ru(6,6′-(SePh)2-2,2′-bipy)2Cl2 (324 mg) was dissolved in 2.0 mL
of acetonitrile (ACN) under Ar, and 17 mL of H2O were added
and the solution was heated at 60 °C for 3 h. The reaction mixture
was cooled to 0-5 °C. Five drops of HPF6 (65% in water) and
702 mg of NH4PF6 in 8.8 mL of H2O were added. After 2 h the
brown microcrystalline precipitate was collected and washed 4 times
with 100 mL of 0.1 M HPF6 and dried under vacuum overnight.
Yield ) 97.6% (388 mg) (brown microcrystalline product).
Elemental Analysis, calculated: C, 41.99%; H, 2.88%; N, 4.45%.
Found: C, 41.59%; H, 2.76%; N, 4.57%. For further characteriza-
tion, see Supporting Information, Figures S5-S11.
structures were solved by direct methods with SHELXS. Full-matrix
least-squares refinement was based on F2 with SHELX-97. 33 181
(9972 unique) reflections (Rint ) 0.078) were collected over a range
of q ) 2.55-25.68 with -15 e h e 14, -17 e k e 17, 0 e l e
20. 2θmax ) 51.36°, from a crystal of size of 0.15 × 0.15 × 0.15
mm3; the largest electron density peak 1.841 eÅ-3 and hole -1.108
dÅ-3. For [Ru](O)2: Crystal data collected at 100 K using a Bruker
Kappa ApexII CCD diffractometer with Mo KR (λ ) 0.710 73 Å)
radiation and Miracol optics. The data were processed with the
Bruker Apex2 suite, and the structure was solved with direct
methods in Bruker Apex2 Autostructure. Full-matrix least-squares
refinement was based on F2 with SHELX-97. 97 880 (12 593
unique) reflections (Rint) 0.071) were collected over a range of q
) 2.63-25.68 with -26 e h e 26, -26 e k e 25, -25 e l e 26.
2θmax ) 51.36°, from a crystal of size 0.20 × 0.05 × 0.05 mm3;
4.2.2. NMR Spectroscopy. 1H, 13C{1H}, 15N, and 77Se{1H}
NMR spectra were recorded at 298 or 216 K as noted (in CDCl3
or CD2Cl2) on a 500 MHz spectrometer operating at 500.13, 125.76,
50.70, and 95.38 MHz respectively, using 5 mm sample tubes. The
1H and 13C chemical shifts were referenced to TMS; 15N chemical
shifts were referenced to liquid NH3. 77Se chemical shifts are relative
to Me2Se (external standard selenophene, δ ) 605 ppm).17 The
pulse programs of the gsCOSY, 13C-1H gsHMQC, and 15N-1H
gsHMBC experiments were used from the Bruker software library.
The gradient HMBC was acquired using SW ) 90 ppm with 256
increments in F1 and 96 transients per increment. Long-range
coupling was 7 Hz.
largest electron density peak 2.876 eÅ-3 and hole -1.570 eÅ-3
.
[Ru(6,6′-(SePh)2-2,2′-bipy)2(ACN)2](ClO4)2 · NaClO4
([Ru-
](ACN)2) for X-ray structure determination was similarly prepared.
Thus, 50 mg (0.045 mmol) of Ru(6,6′-(SePh)2-2,2′-bipy)2Cl2 in 0.3
mL of ACN under Ar were treated with 5.0 mL of H2O at 60 °C
for 3 h. After cooling to 0-5 °C, 1.5 mL of a saturated solution of
NaClO4 in 0.1 M HClO4 was added dropwise. After 2 h, the brown
microcrystalline precipitate was washed 5 times with 15 mL of
0.1 M aq. HClO4 and dried under vacuum. Yield ) 91.4% (46
mg). Orange-brown prismatic crystals were grown from a solution
in ACN by slow addition of benzene vapors. Elemental Analysis,
calculated: C, 40.12%; H, 2.67%; N, 5.85. Found: C, 39.87%; H,
2.58%; N, 5.67%.
4.2.3. Mass Spectrometry. Low resolution spectra were taken
on an MS Micromass ZMD 4000 Mass Spectrometer equipped with
an ESI probe for electrospray analysis. High resolution spectra were
taken on a UPLC-MS Micromass Q-TOF Premier spectrometer
equipped with ESI for electrospray analysis.
4.2.4. IR Spectroscopy. The FT-IR spectra were measured on
a Nicolet 6700 FTIR; samples were prepared by evaporation of
samples of [Ru](H2O)2, [Ru](16O)2, and [Ru](18O)2 on KBr plates.
4.2.5. UV-vis Spectroscopy. UV-visible spectra were recorded
on an Agilent 89090A spectrophotometer using the following
experimental procedure: A 1.8 µM solution of [Ru](H2O)2 was
prepared by dissolving 0.573 mg (0.540 µmol) of [Ru(6,6′-(SePh)2-
2,2′-bpy)2 (H2O)2](PF6)2 in 100 µL of CHCl3 (cleaned from
peroxides by neutral alumina and degassed by He, and then
blanketed by Ar). Then 1 µL of this solution was added to 3.0 mL
of CF3-CH(OH)-CF3 (degassed by He and placed under Ar). To
[Ru(6,6′-(SePh)2-2,2′-bipy)2(O)2](PF6)Cl ·xCHCl3 ([Ru](O)2). 9.5
mg (8.9 µmol) [Ru](H2O)2 were dissolved in 2.0 mL of HFIP in
a 20 mL glass tube and placed under 1 bar of O2 and stirred for
48 h at RT. After evaporation of the solvent and drying under
vacuum, the remaining solid was dissolved in 2.0 mL of CHCl3.
Black needle-like crystals were grown by slow addition of n-hexane
vapors. Yield 88% (10.6 mg). Elemental Analysis, calculated: C,
32.18%; H, 2.04%; N, 3.07%. Found: C, 32.55%; H, 2.08%; N,
2.91%. For further characterization, see Figures 3-5 and S12-14.
4.2. Characterization. 4.2.1. X-ray Structure Determina-
tion. For [Ru](ACN)2: Crystal data were collected at 120 K using
a Nonius Kappa CCD diffractometer with Mo KR (λ ) 0.710 73
Å) radiation. The data were processed with Denzo-scalepack. The
(17) Choi, M.-G.; Angelici, R. J. J. Am. Chem. Soc. 1991, 113, 5651.
9
522 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Activation of O2 by RuII Bipyridine Complex
A R T I C L E S
observe the formation of [Ru](O)2 a quartz cuvette was placed under
1 bar of O2 at 25 °C.
of the calculation. Because of the large size of the ligands (87 atoms,
450 electrons), the use of DFBSs is essential to be able to optimize
the geometries in a reasonable amount of time. Because the use of
DFBSs precludes the use of a hybrid DFT exchange-correlation
functional, the local version of the M06 family (M06-L) was
employed.21 This functional was shown to provide performance
similar to that M06 for transitional metals.20 Truhlar recently
recommended its use in such a scheme for optimizing large
complexes followed by energy calculations with either M06 or M06-
2X as appropriate (the former in this case because of the presence
of a transition metal).27 The automatic DFBS generation algorithm
built-in to Gaussian03 was employed.
The geometries of [Ru]O2 were fully optimized as both a singlet
and a triplet. Bulk solvent effects were approximated by single-
point energy calculations using a polarizable continuum model
(PCM),28 specifically the integral equation formalism model (IEF-
PCM)28 with water as the solvent as in the experiments. Explicit
spheres were used on the hydrogen atoms.
4.2.6. Elemental Analyses. Elemental analyses (C, H, N) were
performed on a CHN elemental analyzer (FlashEA 1112, Eager
300 Software).
4.3. Oxygen Transfer Reactions. The reaction between
[Ru](O)2 and Ph3P or 2,3-dimethy-2-butene was carried out in HFIP
or acetonitrile under Ar in a 25 mL pressure tube using the amounts
given in the paper. The products and substrates were quantified
using a GLC with a flame ionization detector and a 30 m × 0.32
mm 5% phenylmethylsilicone (0.25 µm coating) capillary column
and helium carrier gas. Products identified by GC-MS. A catalytic
reaction to determine the PPh3:O2 stoichiometry was carried out in
a 25 mL flask attached to a oxygen buret. Conditions: 600 µmol of
PPh3, 30 µmol of [Ru](H2O)2, 8.8 mL of HFIP, 26 °C under 1 bar
of O2.
4.4. Computational Methods. Calculations are at the PCM-
(CH3CN)-M06/SDB-cc-pVDZ//M06-L/SDB-cc-pVDZ/BFBS level
of theory. All calculations were carried out using Gaussian 03
Revision E.0118 to which the MNGFM patch19 was applied; this
patch from the University of Minnesota adds the M06 (Vide infra)
family of DFT exchange-correlation functionals to the commercial
version. Two DFT exchange-correlation functionals were used. The
first is the new M06 functional,20 a meta-hybrid functional
containing 27% HF exchange, which was shown to have even
superior performance in the study of transition metal reactions. The
second is the local version of the M06 family (M06-L).21 This
functional was shown to provide similar performance as M06 for
transitional metals.20 With this functional, the SDB-cc-pVDZ
combines the Dunning cc-pVDZ basis set22 on the main group
elements and the Stuttgart-Dresden basis set-RECP23 on the
transition metals with an added f-type polarization exponent taken
as the geometric average of the two f-exponents given in the
appendix of ref 24. The cc-pVDZ-PP basis set-RECP combination
of Peterson et al. was used on selenium.25
For interpretive purposes, natural population analysis (NPA)
charges29 were derived from natural bond order (NBO) analyses
calculated at the M06/SDB-pc1//M06-L/SDD(d)/DFBS level of
theory. Mulliken charges30 are also provided, although the NPA
charges are generally considered to be more reliable.
Acknowledgment. The research was supported by the German-
Israeli Project Cooperation (DIP-G7.1), the Israel Science Founda-
tion, the Helen and Martin Kimmel Center for Molecular Design,
and the Bernice and Peter Cohn Catalysis Research Fund. R.N. is
the Rebecca and Israel Sieff Professor of Organic Chemistry.
Supporting Information Available: X-ray cif files and NMR,
IR, and mass spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA9047027
Density Fitting Basis Sets (DFBS), as implemented in Gaussi-
an03,26 were employed to improve the computational efficiency
(26) (a) Dunlap, B. I. J. Chem. Phys. 1983, 78, 3140. (b) Dunlap, B. I.
THEOCHEM 2000, 529, 37.
(18) Frisch, M. J. Gaussian 03, revision E.01; Gaussian, Inc.: Wallingford,
CT, 2004.
(19) Minnesota Gaussian Functional Module version 3.1; Zhao, Y.; Truhlar
D. G. with assistance from Iron, J. M. A.; Martin, M. L. University
of Minnesota, Minneapolis, MN, 2008.
(20) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215. (b)
Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
(21) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
(22) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007.
(23) Dolg, M. EffectiVe Core Potentials. In Modern Methods and Algorithms
of Quantum Chemistry; John von Neumann Institute for Computing:
Ju¨lich, 2000; Vol. 1, p 479.
(24) Martin, J. M. L.; Sundermann, A. J. Chem. Phys. 2001, 114, 3408.
(25) Peterson, K. A.; Figgen, D.; Goll, E.; Stoll, H.; Dolg, M. J. Chem.
Phys. 2003, 119, 11113.
(27) Truhlar, D. G. “2006 Schro¨dinger Medal Lecture: New Density
Functionals with Broad Applicability for Thermochemistry, Thermo-
chemical Kinetics, Noncovalent Interactions, Transition Metals, and
Spectroscopy” presented at the Eighth Triennial Congress of the World
Association of Theoretical and Computational Chemistry (WATOC),
Sydney, Australia, September 14-19, 2008.
(28) (a) Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 106, 5151. (b)
Cance`s, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032.
(c) Cossi, M.; Barone, V.; Mennucci, B.; Tomasi, J. Chem. Phys. Lett.
1998, 286, 253. (d) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V.
J. Chem. Phys. 2002, 117, 43. (e) Mennucci, B.; Cance`s, E.; Tomasi,
J. J. Phys. Chem. B 1997, 101, 10506. (f) Tomasi, J.; Mennucci, B.;
Cance`s, E. THEOCHEM 1999, 464, 211.
(29) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899.
(30) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833.
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 523