Di- and Trinuclear Ruthenium and Osmium Bis(2-pyridyl) Ketone Oximate Derivatives

Article abstract of DOI:10.1002/ejic.200300341

Treatment of [Ru₃(CO)₁₂] with bis(2-pyridyl) ketone oxime (Hdpko) in refluxing THF leads to a separable mixture of [Ru ₃(μ,η³-dpko-N, N,O)₂(CO)₈] (1) and [Ru₂(μ,η³-dpko-N,N,O)₂(CO) ₄] (2). In both complexes, two Ru atoms are doubly bridged by two dpko ligands, which are attached to a Ru atom through the oximate O atom while chelating the other Ru atom through the N atoms of a pyridyl group and the oximate fragment. While the Ru-Ru distance of the bridged edge of complex 1 is very long [3.5388(9) A], that of complex 2 is very short [2.620(1) A]. The acetonitrile complexes [M₃(CO)₁₀(MeCN) ₂] (M = Ru, Os) react with Hdpko in THF at room temperature to give [M₃(μ-H)(μ,η³-dpko-N,N,O)(CO)₉] [M = Ru (3) Os (4)], in which the dpko ligand behaves in the same way as in 1 and 2. The thermal reaction of 3 with Hdpko leads to a mixture of 1 and 2, while an analogous treatment of complex 4 gives [Os₃(μ,η ³-dpko-N,N,O)₂(CO)₈] (5), which is isostructural with complex 1. Compounds 1 and 3 are the first examples of ruthenium clusters containing oximate ligands. These oximate complexes display low activity as DNA cleavage agents, requiring high complex concentrations, long incubation times, and the use of UV light as a trigger. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200300341

FULL PAPER
Di- and Trinuclear Ruthenium and Osmium Bis(2-pyridyl) Ketone Oximate
Derivatives
[a]
[a]
[a]
[a]
´
[c]
´
´
´
Javier A. Cabeza,* Ignacio del Rıo, Vıctor Riera, Marta Suarez, Carmen Alvarez-
[b]
[b]
[c]
´
´
Rua, Santiago Garcıa-Granda, Shih Hsien Chuang, and Jih Ru Hwu*
Keywords: Ruthenium / Osmium / Cluster compounds / Oximes / DNA cleavage
Treatment of [Ru3(CO)12] with bis(2-pyridyl) ketone oxime
(Hdpko) in refluxing THF leads to a separable mixture of
N,N,O)(CO)9] [M = Ru (3) Os (4)], in which the dpko ligand
behaves in the same way as in 1 and 2. The thermal reaction
[Ru3(µ,η3-dpko-N,N,O)2(CO)8] (1) and [Ru2(µ,η3-dpko- of 3 with Hdpko leads to a mixture of 1 and 2, while an anal-
N,N,O)2(CO)4] (2). In both complexes, two Ru atoms are
doubly bridged by two dpko ligands, which are attached to
a Ru atom through the oximate O atom while chelating the
other Ru atom through the N atoms of a pyridyl group and
the oximate fragment. While the Ru−Ru distance of the
ogous treatment of complex 4
gives [Os3(µ,η3-dpko-
N,N,O)2(CO)8] (5), which is isostructural with complex 1.
Compounds 1 and 3 are the first examples of ruthenium clus-
ters containing oximate ligands. These oximate complexes
display low activity as DNA cleavage agents, requiring high
˚
bridged edge of complex 1 is very long [3.5388(9) A], that of complex concentrations, long incubation times, and the use
˚
complex 2 is very short [2.620(1) A]. The acetonitrile com-
of UV light as a trigger.
plexes [M3(CO)10(MeCN)2] (M = Ru, Os) react with Hdpko
in THF at room temperature to give [M3(µ-H)(µ,η3-dpko-
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
rived from bis(2-pyridyl) ketone oxime (Hdpko). We de-
cided to use this oxime because, in addition to its varied
coordination possibilities, its chemistry has been little ex-
plored.[7]
Oximes and oximates are attractive ligands because they
can interact with metal atoms through either one or both
N and O atoms, acting as terminal (monodentate or chelat-
ing)[1,2] or, less frequently, as bridging ligands.[3Ϫ6] In ad-
dition, oximes can remain intact in the coordination sphere
of the metals or undergo OϪH bond cleavage to afford oxi-
mate derivatives.[1Ϫ6] In some cases they can also suffer
NϪOH bond activation to form nitrenium frag-
ments.[2b,2c,3,4] The osmium cluster-promoted OϪH vs.
NϪO bond cleavage of these ligands has been studied.[4]
Despite this rich derivative chemistry and the large number
of oximate complexes of transition metals known to
date,[1Ϫ6] it is noteworthy that no ruthenium and only a few
osmium[4Ϫ6] cluster complexes containing oximate ligands
have been reported.
Results and Discussion
Synthesis and Structural Characterization of Oximate
Complexes
Treatment of [Ru3(CO)12] with one equivalent of Hdpko,
in THF at reflux temperature, gave a complex mixture of
compounds containing some ruthenium carbonyl starting
material. When two equivalents of Hdpko were used under
the same reaction conditions, the novel trinuclear derivative
[Ru3(µ,η3-dpko-N,N,O)2(CO)8] (1) and the dioximate-
bridged binuclear complex [Ru2(µ,η3-dpko-N,N,O)2(CO)4]
(2) were obtained as the major reaction products
(Scheme 1).
Herein we report the synthesis and structural characteri-
zation of triruthenium and triosmium carbonyl clusters de-
1
[a]
´
´
´
Departamento de Quımica Organica e Inorganica, Instituto de
The H NMR spectra of both complexes display no sig-
´
´
Quımica Organometalica ‘‘Enrique Moles’’, Universidad de
Oviedo-CSIC,
nals attributable to hydride ligands, while the aromatic re-
gions reflect, in both cases, the presence of two inequivalent
pyridyl groups. Their IR spectra confirm the absence of
bridging carbonyl ligands and display no bands in the OϪH
stretching region, suggesting OϪH bond activation.
The trinuclear nature of 1 was clearly indicated by its
microanalysis and FAB MS spectrum, which also show the
presence of eight CO ligands and the incorporation of two
33071 Oviedo, Spain
Fax: (internat.) ϩ 34-985/103446
E-mail: jac@sauron.quimica.uniovi.es
[b]
[c]
´
´
´
Departamento de Quımica Fısica y Analıtica, Universidad de
Oviedo,
33071 Oviedo, Spain
Department of Chemistry, National Tsing Hua University,
Hsinchu 30013, Taiwan, Republic of China
E-mail: jrhwu@mx.nthu.edu.tw
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165
DOI: 10.1002/ejic.200300341
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4159

J. A. Cabeza, J. R. Hwu et al.
FULL PAPER
presents a non-crystallographic twofold axis. Such a sym-
metry was also indicated by its 13C{1H} NMR spectrum,
which displays only four carbonyl resonances. The length
˚
of the bridged edge [Ru(1)ϪRu(2) ϭ 3.5388(9) A] indicates
the absence of a metalϪmetal bond, as expected for a 50-
electron trinuclear cluster.[8]
The 13C{1H} NMR spectrum of 2 shows only two signals
attributable to terminal carbonyl ligands. Its binuclear
character and the incorporation of one dpko ligand per Ru
atom were pointed out by its microanalysis and FAB MS.
Its structure was confirmed by an X-ray diffraction study
(Figure 2, Table 2). The ligand arrangement in compound
2 is analogous to that described above for complex 1, but
the Ru(CO)4 fragment present in 1 is now missing. The
˚
Ru(1)ϪRu(2) distance in 2, 2.620(1) A, is in the range ex-
pected for a single metalϪmetal bond, in accordance with
the 34-electron count for the complex.[8] It is noteworthy
that the oximate NO fragment is able to bridge two metal
atoms which are either close to each other (complex 2) or
very far apart (complex 1).
Scheme 1. Synthesis of compounds 1Ϫ3
oximate ligands to the cluster shell. The structure of 1 was
established by X-ray diffraction methods (Figure 1,
Table 1). In addition to eight terminal CO ligands, the com-
plex contains two oximate ligands spanning the same edge
of the trimetallic core, in a head-to-tail arrangement,
through both the N and O atoms. Each dpko ligand is also
attached through a pyridine N-atom to one of the metal
atoms of the bridged edge, in such a way that the complex
Figure 2. Molecular structure of compound 2 (50% probability le-
vel ellipsoids); H atoms omitted for clarity
˚
Table 2. Selected interatomic distances (A) in complex 2
Ru(1)ϪRu(2)
Ru(1)ϪN(2)
Ru(2)ϪN(5)
Ru(2)ϪO(1)
N(4)ϪO(2)
2.620(1)
2.209(4)
2.176(4)
2.135(3)
1.334(4)
Ru(1)ϪN(1)
Ru(2)ϪN(4)
Ru(1)ϪO(2)
N(1)ϪO(1)
2.061(3)
2.060(4)
2.162(4)
1.341(4)
Figure 1. Molecular structure of compound 1 (50% probability le-
vel ellipsoids); H atoms omitted for clarity
IR monitoring of the reaction of [Ru3(CO)12] with
Hdpko showed that an intermediate, probably arising from
the incorporation of only one equivalent of Hdpko to the
cluster, is formed prior to compounds 1 and 2. Its isolation
was not possible under the thermal conditions applied,
since it readily reacted with more Hdpko to give 1 and 2 as
the two major final products. Fortunately, we managed to
isolate this intermediate working at room temperature,
using the reactive complex [Ru3(CO)10(MeCN)2][9] as start-
ing material. The product was characterized as the novel
˚
Table 1. Selected interatomic distances (A) in complex 1
Ru(1)ϪRu(2)
Ru(2)ϪRu(3)
Ru(1)ϪN(2)
Ru(2)ϪN(5)
Ru(2)ϪO(1)
N(4)ϪO(2)
3.5388(9)
2.8170(9)
2.173(7)
2.154(7)
2.124(5)
1.363(8)
Ru(1)ϪRu(3)
Ru(1)ϪN(1)
Ru(2)ϪN(4)
Ru(1)ϪO(2)
N(1)ϪO(1)
2.8137(9)
2.131(6)
2.115(6)
2.131(5)
1.350(8)
4160
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165

Di- and Trinuclear Ruthenium and Osmium Bis(2-pyridyl) Ketone Oximate Derivatives
FULL PAPER
trinuclear derivative [Ru3(µ-H)(µ,η3-dpko-N,N,O)(CO)9]
(3; Scheme 1).
The IR spectrum of 3 strongly suggests OϪH bond scis-
sion. This was further supported by its 1H NMR spectrum,
which shows a singlet resonance at δ ϭ Ϫ18.85 ppm at-
tributable to a hydride ligand, while the inequivalence of
two pyridine rings is evident in the aromatic region. The
incorporation of one equivalent of the oxime to the trinu-
clear framework and the presence of nine carbonyl ligands
were confirmed by its microanalysis and FAB MS.
The structure proposed for 3, and its role as an inter-
mediate in the formation of 1 and 2 from [Ru3(CO)12], were
further corroborated by treating 3 with one equivalent of
the oxime in THF at reflux temperature. This reaction led
to complexes 1 and 2, along with a small amount of
[Ru3(CO)12].
To the best of our knowledge, 1 and 3 are the first ru-
thenium cluster complexes containing oximate ligands.
Some diruthenium() complexes structurally related to 2,
namely [Ru2(µ,η2-ONCR1R2-O,N)2(CO)4L2] (L ϭ intact
oxime or other neutral two-electron donor ligands), have
been reported.[10] It has also been reported that the reaction
of [Ru3(CO)12] with 1-nitrous-2-naphthol gives a triruthen-
ium cluster structurally related to complex 1, which also
contains two NO bridges.[11]
Figure 3. Molecular structure of compound 4 (50% probability le-
vel ellipsoids); H atoms (except the hydride) omitted for clarity
˚
Table 3. Selected interatomic distances (A) in complex 4
Os(1)ϪOs(2)
Os(2)ϪOs(3)
Os(1)ϪN(2)
N(1)ϪO(1)
2.861(5)
2.855(2)
2.078(9)
1.31(1)
Os(1)ϪOs(3)
Os(1)ϪN(1)
Os(2)ϪO(1)
2.876(1)
2.062(9)
2.170(7)
The triosmium acetonitrile complex [Os3(CO)10-
(MeCN)2][12] reacts with Hdpko in THF at room tempera-
ture to give the monohydrido derivative [Os3(µ-H)(µ,η3-
dpko-N,N,O)(CO)9] (4; Scheme 2). The structure of this
compound was established by X-ray diffraction methods
(Figure 3, Table 3). The cluster consists of a nearly equi-
lateral triangle of Os atoms in which two metal atoms are
attached to a dpko ligand in the same way as that found
previously for compounds 1 and 2. A hydride ligand spans
the same OsϪOs edge as the oximate NO fragment. The
cluster shell is completed by nine CO ligands. Therefore, the
cluster is a closed-shell, 48-electron trinuclear species.[8] The
spectroscopic data of 4 are similar to those of the ru-
thenium complex 3, indicating that both compounds are
isostructural.
Therefore, the oxime Hdpko reacts with the triruthenium
bis(acetonitrile) cluster in the same way as its triosmium
analogue, giving isostructural products. However, this is not
always the case for other ligands, such as bis(2-pyridyl)-
amine,[13] 2-amino-7,8-benzoquinoline,[14] or 2,2Ј-diamino-
1,1Ј-binaphthalene.[15]
Bearing in mind that the triruthenium complex 3 is an
intermediate in the formation of 1, we reasoned that com-
plex 4 could be a suitable precursor of a triosmium cluster
isostructural with complex 1. Indeed, treatment of complex
4 with Hdpko in toluene at reflux temperature allowed the
isolation of [Os3(µ,η3-dpko-N,N,O)2(CO)8] (5). The struc-
ture depicted for this complex in Scheme 2, which is similar
to that of the triruthenium complex 1, was determined by
X-ray diffraction methods (Figure 4, Table 4). Accordingly,
its IR and NMR spectroscopic data are also similar to
those of complex 1.
Very few triosmium clusters containing oximate ligands
have been reported prior to this work. They contain oxi-
mate fragments that span two metal atoms through the N
and O atoms.[4,5] During the preparation of this manuscript,
some triosmium clusters related to compounds 4 and 5, de-
rived from diphenyl ketone oxime and phenyl (2-pyridyl)
ketone oxime, have been reported.[6]
DNA-Cleavage Ability of the Oximate Complexes
Although most metal complex-mediated DNA cleavage
reactions use mononuclear complexes as promoters,[16,17]
a
few interesting results have been reported using bi-[18] and
trinuclear[19] complexes. Prompted by the fact that prior to
Scheme 2. Synthesis of compounds 4 and 5
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4161

J. A. Cabeza, J. R. Hwu et al.
FULL PAPER
tios of (form II)/(form I) of 1.78, 1.63 and 2.57 were found
(see lanes 5, 9, and 13). The cleaving ability of 1 and 2 was
found to be pH dependent and is greater under slightly
acidic conditions, cleaving DNA at pH 5.0 at concen-
trations of 500 n (1) and 1.0 µ (2), with ratios of (form
II)/(form I) of 1.27 and 1.08, respectively (lanes 7 and 11).
Figure 6. Electrophoresis separation of supercoiled circular ϕX174
RFI DNA (form I) and relaxed circular DNA (form II) after treat-
ment of form I DNA with the osmium complexes 4 and 5; lanes 1,
4, 7 were run in the dark, the remaining ones involved UV light
irradiation; lane 1, DNA alone (dark, pH ϭ 7.0); lane 2, DNA
alone (pH ϭ 7.0); lane 3, DNA alone (pH ϭ 5.0); lane 4, 4
(1000 µ, pH ϭ 7.0, dark); lane 5, 4 (1000 µ, pH ϭ 7.0); lane 6,
4 (1000 µ, pH ϭ 5.0); lane 7, 5 (1000 µ, pH ϭ 7.0, dark); lane
8, 5 (1000 µ, pH ϭ 7.0); lane 9, 5 (1000 µ, pH ϭ 5.0); lane 10,
5 (1.0 µ, pH ϭ 5.0)
Figure 4. Molecular structure of compound 5 (50% probability le-
vel ellipsoids); H atoms omitted for clarity
For the osmium oximate complexes 4 and 5 (Figure 6), it
was found that they can also cleave DNA upon UV acti-
vation. The ratios of (form II)/(form I) were 1.13 and 1.44
for 4 and 5 (1000 µ), respectively, in pH 7.0 buffer (see
lanes 5 and 8). Complex 5 also cleaved DNA at pH 5.0 in
concentrations of 1.0 µ to give a (form II)/(form I) ratio
of 1.04 (see lane 10).
A comparison of the cleavage activity of the pairs of iso-
structural clusters 1, 5 and 3, 4 indicates that the ruthenium
clusters (1 and 3) are more active than the osmium ones (4
and 5). In addition, the binuclear complex 2 is less active
than the trinuclear ones 1 and 3.
Although the cleaving ability of the Ru cluster complexes
1 and 3 is higher than that reported for other triruthenium
carbonyl clusters,[19c] their activity is very low when com-
pared to that of many mono- and binuclear compounds
previously studied.[17,18] The high complex concentrations
needed (millimolar range), along with long incubation times
(8 h) and UV activation, hamper the use of these complexes
for experiments in vivo.
˚
Table 4. Selected interatomic distances (A) in complex 5
Os(1)ϪOs(2)
Os(2)ϪOs(3)
Os(1)ϪN(2)
Os(2)ϪN(5)
Os(2)ϪO(1)
N(4)ϪO(2)
3.5903(5)
2.8274(5)
2.130(7)
2.148(7)
2.124(6)
1.346(9)
Os(1)ϪOs(3)
Os(1)ϪN(1)
Os(2)ϪN(4)
Os(1)ϪO(2)
N(1)ϪO(1)
2.8318(6)
2.125(7)
2.145(8)
2.130(6)
1.359(9)
this work no metal complexes of oximes had been studied
as DNA cleaving agents, the ruthenium and osmium oxi-
mate complexes 1Ϫ5 were tested by studying the nicking of
supercoiled circular ϕX174 RFI DNA (form I; 50 µ/base
pair) into relaxed circular DNA (form II).
Concluding Remarks
Figure 5. Electrophoresis separation of supercoiled circular ϕX174
RFI DNA (form I) and relaxed circular DNA (form II) after treat-
ment of form I DNA with the ruthenium complexes 1Ϫ3; lanes 1,
4, 8, 12 were run in the dark, the remaining ones involved UV light
irradiation; lane 1, DNA alone (dark, pH ϭ 7.0); lane 2, DNA
alone (pH ϭ 7.0); lane 3, DNA alone (pH ϭ 5.0); lane 4, 1
(1000 µ, pH ϭ 7.0, dark); lane 5, 1 (1000 µ, pH ϭ 7.0); lane 6,
1 (1000 µ, pH ϭ 5.0); lane 7, 1 (500 n, pH ϭ 5.0); lane 8, 2
(1000 µ, pH ϭ 7.0, dark); lane 9, 2 (1000 µ, pH ϭ 7.0); lane 10,
2 (1000 µ, pH ϭ 5.0); lane 11, 2 (1.0 µ, pH ϭ 5.0); lane 12, 3
(1000 µ, pH ϭ 7.0, dark); lane 13, 3 (1000 µ, pH ϭ 7.0); lane
14, 3 (1000 µ, pH ϭ 5.0)
By using the oxime Hdpko as a ligand precursor, we have
prepared one diruthenium, two triruthenium and two trios-
mium carbonyl complexes, all containing the oximate li-
gand dpko η3-coordinated to two metal atoms through the
N and O atoms of the oximate fragment and the N atom
of the pyridyl group. Compounds 1 and 3 are the first ru-
thenium cluster complexes derived from oximes.
The first DNA cleaving studies using oximate metal com-
Figure 5 shows the results obtained with the ruthenium plexes are described. Unfortunately, the observed activity is
complexes. After UV irradiation of complexes 1Ϫ3 very low when compared to those of many mono- and bi-
(1000 µ) and form I DNA, in pH 7.0 buffer for 8.0 h, ra- nuclear compounds previously studied.
4162
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165

Di- and Trinuclear Ruthenium and Osmium Bis(2-pyridyl) Ketone Oximate Derivatives
FULL PAPER
top of the column. FABϩ MS: m/z ϭ 756 [Mϩ]. C20H9N3O10Ru3
(754.55): calcd. C 31.84, H 1.20, N 5.57; found C 31.78, H 1.35, N
5.52. IR (CH2Cl2): ν(CO) ϭ 2087 (m), 2047 (s), 2021 (vs), 2000 (s),
Experimental Section
General Synthetic and Characterization Data: Solvents were dried
over sodium diphenyl ketyl (THF, Et2O, hydrocarbons) or CaH2
(dichloromethane) and distilled under nitrogen prior to use. The
reactions were carried out under nitrogen, using Schlenk-vacuum
line techniques, and were routinely monitored by solution IR spec-
troscopy (carbonyl stretching region) and spot TLC (silica gel).
[Os3(CO)10(MeCN)2] was prepared as described previously;[12] the
remaining reagents were purchased from commercial suppliers. IR:
PerkinϪElmer Paragon 1000 FT. NMR: Bruker AC-200 and DPX-
300, room temperature, TMS as internal standard. Microanalyses:
PerkinϪElmer 2400. MS: VG Autospec double-focussing mass
spectrometer operating in the FABϩ mode; ions were produced
with a standard Csϩ gun at ca. 30 kV; 3-nitrobenzyl alcohol (NBA)
was used as matrix; data given refer to the most abundant molecu-
lar ion isotopomer.
1978 (sh), 1959 (sh) cmϪ1 1H NMR (CDCl3): δ ϭ 8.75 (d, J ϭ
.
5.6 Hz, 1 H), 8.72 (dd, J ϭ 5.2, 1.2 Hz, 1 H), 8.17 (br. d, J ϭ
8.8 Hz, 1 H), 8.11 (d, J ϭ 7.8 Hz, 1 H), 7.85 (td, J ϭ 7.8, 1.6 Hz,
1 H), 7.71 (td, J ϭ 8.8, 1.6 Hz, 1 H), 7.35 (td, J ϭ 5.2, 1.2 Hz, 1
H), 7.20 (td, J ϭ 5.2, 1.2 Hz, 1 H), Ϫ18.88 (s, µ-H) ppm.
[Os3(µ-H)(µ,η3-dpko-N,N,O)(CO)9]
(4):
Hdpko
(85 mg,
0.427 mmol) was added to a solution of [Os3(CO)10(MeCN)2]
(270 mg, 0.290 mmol) in THF (25 mL). The solution was stirred at
room temperature for 2 h. The color changed from lemon yellow
to red. The solvent was removed under reduced pressure, the resi-
due was dissolved in ca. 2 mL of toluene and this solution was
placed onto a silica gel column (2 ϫ 10 cm) packed in hexanes.
Elution with hexanes afforded some [Os3(CO)12]. Elution with hex-
anes/dichloromethane (1:1) afforded two bands. The first band
(blue) was very weak and was not investigated. The second band
(red) afforded complex 4 (125 mg, 42%). FABϩ MS: m/z ϭ 1023
[Mϩ]. C20H9N3O10Os3 (1021.90): calcd. C 23.51, H 0.89, N 4.11;
found C 23.62, H 0.94, N 4.02. IR (CH2Cl2): ν(CO) ϭ 2091 (m),
[Ru3(µ,η3-dpko-N,N,O)2(CO)8]
(1)
and
[Ru2(µ,η3-dpko-
N,N,O)2(CO)4] (2): solution of [Ru3(CO)12] (104 mg,
A
0.163 mmol) and Hdpko (66 mg, 0.331 mmol) in THF (20 mL) was
heated under reflux for 20 min. The color changed from orange to
black. The solvent was concentrated under reduced pressure to ca.
2 mL and transferred onto TLC silica plates. Dichloromethane/
THF (95:5) eluted four bands. The first two were very weak and
were discarded. The third (orange) and fourth (pink) bands af-
forded compounds 1 (21 mg, 14%) and 2 (14 mg, 12%), respectively.
A black residue remained uneluted at the base line.
1
2051 (m), 2014 (vs), 1994 (m), 1964 (w), 1945 (w) cmϪ1. H NMR
(CD2Cl2): δ ϭ 9.04 (br. d, J ϭ 5.9 Hz, 1 H), 8.73 (ddd, J ϭ 4.7,
1.6, 0.8 Hz, 1 H), 8.40 (br. d, J ϭ 8.4 Hz, 1 H), 8.12 (dt, J ϭ 7.7,
1.2 Hz, 1 H), 7.87 (td, J ϭ 7.7, 1.8 Hz, 1 H), 7.82 (ddd, J ϭ 8.4,
7.7, 1.2 Hz, 1 H), 7.40 (ddd, J ϭ 7.5, 4.7, 1.2 Hz, 1 H), 7.30 (ddd,
J ϭ 7.5, 5.9, 1.2 Hz, 1 H), Ϫ18.14 (s, µ-H) ppm. 13C{1H} NMR
(DEPT, CD2Cl2): δ ϭ 187.7, 186.7, 184.4, 181.0, 179.9, 177.8,
177.7, 175.0, 168.4 (COs); 154.8 (CϭN); 154.7 (CH), 153.7 (Cipso),
149.3 (CH), 148.2 (Cipso), 136.7 (CH), 136.5 (CH), 127.6 (CH),
126.0 (CH), 124.6 (CH), 124.4 (CH) ppm.
1: FABϩ MS: m/z ϭ 925 [Mϩ]. C30H16N6O10Ru3 (923.74): calcd. C
39.01, H 1.75, N 9.09; found C 38.78, H 1.85, N 8.92. IR (CH2Cl2):
ν(CO) ϭ 2069 (m), 2007 (vs), 1990 (m), 1935 (w) cmϪ1. H NMR
1
([D6]acetone): δ ϭ 8.84 (ddd, J ϭ 5.4, 1.7, 0.8 Hz, 1 H), 8.42 (ddd,
J ϭ 4.8, 1.7, 0.8 Hz, 1 H), 8.00 (td, J ϭ 8.0, 1.7 Hz, 1 H), 7.54
(ddd, J ϭ 8.0, 5.4, 1.1 Hz, 1 H), 7.52 (td, J ϭ 7.7, 1.7 Hz, 1 H),
7.26 (ddd, J ϭ 7.7, 4.8, 1.1 Hz, 1 H), 7.21 (ddd, J ϭ 8.0, 1.1, 0.8 Hz,
1 H), 7.05 (ddd, J ϭ 7.7, 1.1, 0.8 Hz, 1 H) ppm. 13C{1H} NMR
(DEPT, CD2Cl2): δ ϭ 206.1, 206.0, 202.6, 194.0 (COs); 160.3 (Cϭ
N); 156.2 (Cipso), 152.3 (CH), 150.3 (Cipso), 149.8 (CH), 137.8 (CH),
136.1 (CH), 126.4 (CH), 124.8 (CH), 123.9 (CH), 123.6 (CH) ppm.
2: FABϩ MS: m/z ϭ 712 [Mϩ]. C26H16N6O6Ru2 (710.59): calcd. C
43.94, H 2.27, N 11.83; found C 43.83, H 2.42, N 11.74. IR
[Os3(µ,η3-dpko-N,N,O)2(CO)8] (5): A solution of Hdpko (16 mg,
0.080 mmol) and complex 4 (80 mg, 0.078 mmol) in toluene
(10 mL) was heated to reflux temperature for 75 min. The color
changed from red to black. The solvent was partially removed un-
der reduced pressure (to ca. 2 mL) and this solution was placed
onto a silica gel column (2 ϫ 10 cm) packed in hexanes. Elution
with hexanes afforded some [Os3(CO)12]. Elution with hexanes/di-
chloromethane (1:2) afforded two weak bands, which were not in-
vestigated. Dichloromethane/THF (20:1) eluted a red band, which
afforded complex 5 (18 mg, 19%). A black residue remained un-
eluted at the top of the column. FABϩ MS: m/z ϭ 1192 [Mϩ].
C30H16N6O10Os3 (1190.70): calcd. C 30.25, H 1.35, N 7.06; found
C 30.42, H 1.42, N 7.06. IR (CH2Cl2): ν(CO) ϭ 2071 (m), 1993
1
(CH2Cl2): ν(CO) ϭ 2018 (vs), 1971 (w), 1943 (s) cmϪ1. H NMR
([D6]acetone): δ ϭ 9.04 (ddd, J ϭ 5.2, 1.6, 0.8 Hz, 1 H), 8.65 (dt,
J ϭ 4.8, 1.6 Hz, 1 H), 7.91 (td, J ϭ 8.3, 1.6 Hz, 1 H), 7.79Ϫ7.73
(m, 3 H), 7.47 (ddd, J ϭ 7.6, 5.2, 1.6 Hz, 1 H), 7.34Ϫ7.30 (m, 1
H) ppm. 13C{1H} NMR (DEPT, CD2Cl2): δ ϭ 211.9, 203.9 (COs);
155.7 (CϭN); 153.7 (CH), 153.3 (Cipso), 149.8 (CH), 149.0 (Cipso),
137.5 (CH), 136.6 (CH), 127.3 (CH), 124.3 (CH), 124.2 (CH), 124.1
(CH) ppm.
(vs), 1921 (m) cmϪ1 1H NMR (CD2Cl2): δ ϭ 8.73 (dd, J ϭ 4.8,
.
0.8 Hz, 1 H), 8.33 (dd, J ϭ 4.8, 0.8 Hz, 1 H), 7.74 (td, J ϭ 8.0,
1.6 Hz, 1 H), 7.45 (td, J ϭ 8.0, 2.0 Hz, 1 H), 7.24Ϫ7.06 (m, 4
H) ppm.
[Ru3(µ-H)(µ,η3-dpko-N,N,O)(CO)9] (3): A solution of Me3NO X-ray Diffraction Studies: Suitable crystals of 1·C6H14, 2 and
(40 mg, 0.533 mmol) in acetonitrile (10 mL) was slowly added to a 5·CH2Cl2 were obtained by slow diffusion of hexanes into dichloro-
solution of [Ru3(CO)12] (150 mg, 0.235 mmol) in dichloromethane/ methane solutions of the appropriate complexes. Crystals of 4 were
acetonitrile (10:1, 20 mL) at Ϫ78 °C. Once the formation of [Ru3- obtained by slow diffusion of pentane into a solution of the com-
(CO)10(MeCN)2] was achieved (IR monitoring), Hdpko (51 mg,
0.256 mmol) was added and the solution was allowed to warm up
to room temperature. Solvents were removed under reduced press-
plex in diethyl ether. Diffraction data for 1·C6H14 and 5·CH2Cl2
were collected on a Nonius Kappa-CCD diffractometer equipped
with a 95-mm CCD camera on a κ-goniostat, using graphite-mono-
ure and the residue was redissolved in THF (25 mL), whereupon chromated Cu-Kα radiation. Diffraction data for 2 and 4 were
the color changed from brown to black. The solvent was removed
under reduced pressure, the residue was dissolved in ca. 1 mL of
toluene and this solution was placed onto a silica gel column (2
ϫ 8 cm) packed in hexanes. Elution with hexanes afforded some
unchanged [Ru3(CO)12]. Elution with dichloromethane afforded
complex 3 (40 mg, 23%). A black residue remained uneluted at the
collected on a Nonius CAD-4 diffractometer, with the ω-2θ scan
technique and a variable scan rate, using graphite-monochromated
Mo-Kα radiation, applying Lorentz and polarization corrections,
and reducing the data to F2o values. All structures were solved by
Patterson interpretation using the program DIRDIF-96.[20] Ab-
sorption corrections were applied using XABS2[21] (1·C6H14, 2, and
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4163

J. A. Cabeza, J. R. Hwu et al.
FULL PAPER
Table 5. Crystal, measurement, and refinement data for 1·C6H14, 2, 4, and 5·CH2Cl2
1·C6H14
2
4
5·CH2Cl2
Empirical formula
Formula mass
Cryst. system
Space group
C30H16N6O10Ru3·C6H14
1009.87
C26H16N6O6Ru2
710.59
monoclinic
P21/c
9.029(2)
21.231(7)
13.43(2)
90
92.31(6)
90
C20H9N3O10Os3
1021.90
monoclinic
P21/a
18.92(3)
7.566(2)
18.95(1)
90
120.05(7)
90
C30H16N6O10Os3·CH2Cl2
1276.01
triclinic
triclinic
¯
¯
P1
P1
˚
a, A
10.6768(3)
12.4667(4)
15.6330(6)
83.596(3)
79.614(2)
72.495(2)
1948.3(1)
2
10.8194(4)
12.4036(4)
15.4080(6)
85.049(2)
80.843(2)
71.770(2)
1937.4(1)
2
˚
b, A
˚
c, A
α, deg
β, deg
γ, deg
3
˚
V, A
2573(4)
4
2349(4)
4
Z
F(000)
1000
1.721
Cu-Kα (1.54180)
9.838
1400
1.835
Mo-Kα (0.71073)
1.229
1832
2.889
Mo-Kα (0.71073)
16.246
1176
2.187
Cu-Kα (1.54180)
19.960
Dcalcd, g cmϪ3
˚
Radiation (λ, A)
µ, mmϪ1
Cryst size, mm
Temp., K
θ limits, deg
0.33 ϫ 0.20 ϫ 0.13
200(2)
2.88 to 69.76
Ϫ12/12, Ϫ14/15, Ϫ13/18
9338
6866 (0.0525)
6121
XABS2
0.33 ϫ 0.10 ϫ 0.10
293(2)
1.80 to 25.98
Ϫ11/0, Ϫ26/0, Ϫ16/16
5361
5037 (0.0280)
3399
XABS2
0.27 ϫ 0.10 ϫ 0.07
293(2)
1.24 to 25.98
Ϫ23/0, Ϫ9/0, Ϫ20/23
4764
4608 (0.0415)
3415
XABS2
0.17 ϫ 0.07 ϫ 0.07
200(2)
2.91 to 69.93
Ϫ12/13, Ϫ14/15, 0/18
22303
7242 (0.0783)
5680
SORTAV
1.249/0.744
0/496
Min/max h, k, l
No. of collected rflns.
No. of unique rflns. (Rint
No. of rflns. with I Ͼ 2σ(I)
Absorption correction
Max/min transmission
No. of restraints/params.
GOF (on F2)
)
0.278/0.141
83/515
1.106
0.884/0.863
0/361
1.015
0.321/0.156
0/326
1.030
0.962
Final R1 [on F, I Ͼ 2σ(I)]
0.0630
0.2319
1.266/Ϫ1.487
0.0317
0.0884
0.617/Ϫ0.554
0.0384
0.1102
1.974/Ϫ2.404
0.0481
0.1291
1.986/Ϫ1.708
Final wR2 (on F2, all data)
Ϫ3
˚
Max/min residuals, e·A
4) or SORTAV[22] (5·CH2Cl2). Isotropic and full-matrix anisotropic In a typical experiment, a reaction mixture (10 µL) containing
least-squares refinements were carried out using SHELXL-97.[23]
All non H atoms were refined anisotropically. Hydrogen atom posi-
tions were geometrically calculated and refined riding on their par-
supercoiled circular ϕX174 RFI DNA stock solution (form I, 50
µ/base pair, 1.0 µL), an oximate complex (10 m, 1.0 µL), and a
phosphate buffer (0.10 , 8.0 µL) was preincubated at 37 °C in a
ent atoms, except the hydride H atom of 4, which was located by Pyrex vial and irradiated with UV light (350 nm, 32-W) under aero-
Fourier difference maps. The disordered solvent molecule of bic conditions for 8.0 h. After addition of the gel-loading buffer
1·C6H14 was treated with a mixture of constraints and restraints as (0.25% bromophenol blue, 0.25% xylene cyanol, and 30% glycerol),
described elsewhere.[24] The solvent molecule of 5·CH2Cl2 was
found disordered over two positions, with partial occupancies of
the reaction mixture was loaded onto a 1% agarose gel with ethid-
ium bromide staining. The electrophoresis tank was attached to a
0.5. The molecular plots were made with the EUCLID program power supply at a constant current of about 100 mA). The gel was
package.[25] The WINGX program system[26] was used thorough visualized by 312-nm UV transilluminator and photographed by
the structure determinations. Selected crystal, measurement, and
refinement data for the four X-ray structures are given in Table 5.
CCDC-182375 (1·C6H14), -182376 (2), -203169 (4) and -203170
(5·CH2Cl2) contain the supplementary crystallographic data for
this paper. These data can be obtained free of charge at
www.ccdc.cam.ac.uk/conts/retrieving.html [or from the Cambridge
Crystallographic Data Centre, 12, Union Road, Cambridge
CB2 1EZ, UK; Fax: (internat.) ϩ44-1223/336-033; E-mail:
deposit@ccdc.cam.ac.uk].
an FB-PDC-34 camera. Quantification of DNA cleavage activity
was performed by integration of the optical density of the spot as
a function of the band area by use of a Microtek scanner and the
NIH 1.60 image program.
Acknowledgments
This work has been supported by the Spanish MECD (grant PB98-
1555 to J. A. C.), MCYT (grants BQU2000-2019 to S. G.-G. and
BQU2002-2623 to J. A. C.), and Principado de Asturias (grant PR-
01-GE-7 to V. R.), and by the Republic of China’s National Science
Council (to J. R. H.).
DNA-Cleavage Experiments: All reactions were carried out in oven-
dried (120 °C) or autoclaved glassware and eppendorf tubes. Super-
coiled circular ϕX174 RFI DNA (molecular weight 3.50 ϫ 106,
5386 base pairs in length) was purchased from New England Biol-
abs Co. The remaining reagents were also purchased from commer-
cial suppliers. Photolytic experiments were carried out at 37 °C,
using a medium pressure mercury lamp (350 nm, 32-W). The NIH
1.60 image program, provided by Dr. R. Wayne of National Insti-
tutes of Health, U. S. A., was used for the quantitative analysis of
DNA cleavage.
[1]
A. Chakravorty, Coord. Chem. Rev. 1974, 13, 1Ϫ46.
[2] [2a]
A. R. Middleton, J. R. Thornback, G. Wilkinson, J. Chem.
[2b]
Soc., Dalton Trans. 1980, 174Ϫ179.
T. Daniel, W. Knaup,
[2c]
M. Dziallas, H. Werner, Chem. Ber. 1993, 126, 1981Ϫ1993.
H. Werner, T. Daniel, W. Knaup, O. Nürnberg, J. Organomet.
[2d]
Chem. 1993, 462, 309Ϫ318.
R. Castarlenas, M. A. Es-
4164
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165

Di- and Trinuclear Ruthenium and Osmium Bis(2-pyridyl) Ketone Oximate Derivatives
FULL PAPER
G. Roelfes, M. E. Branum, W. Li,
[16d]
´
´
´
teruelas, E. Gutierrez-Puebla, Y. Jean, A. Lledos, M. Martın,
2000, 122, 8814Ϫ8824.
´
J. Tomas, Organometallics 1999, 18, 4296Ϫ4303.
L. Que, B. L. Feringa, J. Am. Chem. Soc. 2000, 122,
[3]
[4]
[5]
[16e]
T. Carofiglio, S. Stella, C. Floriani, A. Chiesi-Villa, C.
Guastini, J. Chem. Soc., Dalton Trans. 1989, 1957Ϫ1962.
A. J. Deeming, D. W. Owen, N. I. Powell, J. Organomet. Chem.
1990, 398, 299Ϫ310.
11517Ϫ11518.
2003, 42, 2483Ϫ2485.
For recent reviews, see:
S. Dhar, A. R. Chakravarty, Inorg. Chem.
[17]
[17a]
G. Pratviel, J. Bernadou, B. Meu-
[17b]
nier, Adv. Inorg. Chem. 1997, 45, 251Ϫ312.
B. Armitage,
M. H. Chao, S. Kumaresan, Y. S. Wen, S. C. Lin, J. R. Hwu,
K.-L. Lu, Organometallics 2000, 19, 714Ϫ717.
Chem. Rev. 1998, 98, 1171Ϫ1200. [17c] Z. Guo, P. J. Sadler, An-
[17d]
gew. Chem. Int. Ed. 1999, 38, 1512Ϫ1531.
K. E. Erkkila,
[6]
[7]
J. S. Y. Wong, W. T. Wong, New J. Chem. 2002, 26, 94Ϫ104.
D. T. Odom, J. K. Barton, Chem. Rev. 1999, 99, 2777Ϫ2796.
Hdpko has heretofore been used to prepare Ag,[11a] Cu,[11a,11b]
G. Pratviel, J. Bernadou, B. Meunier, Angew. Chem. Int.
[17e]
Zn[11c] and Ni[11d] derivatives.
S. O. Sommerer, B. L.
Ed. Engl. 1995, 34, 746Ϫ769. [17f] S. J. Franklin, Cur. Op. Chem.
Biol. 2001, 5, 201Ϫ208.
[18] [18a]
[7a]
Westcott, A. J. Jircitano, K. A. Abboud, Inorg. Chim. Acta
[7b]
1995, 238, 149Ϫ153.
hedron 1999, 18, 3425Ϫ3431.
V. L. Pecoraro, Inorg. Chem. 1995, 34, 2271Ϫ2272.
Psomas, C. Dendrinou-Samara, M. Alexiou, A. Tsohos, C. P.
Raptopoulou, A. Terzis, D. P. Kessissoglou, Inorg. Chem. 1998,
37, 6556Ϫ6557.
M. A. S. Goher, F. A. Mautner, Poly-
F. OЈReilly, J. Kelly, A. Kirsch-De-Mesmaeker, Chem.
[7c]
[18b]
A. J. Stemmler, J. W. Kampf,
Commun. 1996, 1013Ϫ1014.
Biol. Inorg. Chem. 1999, 4, 593Ϫ600.
M. E. Branum, L. Que, J.
[7d]
[18c]
G.
M. E. Branum, A.
K. Tipton, S. Zhu, L. Que, J. Am. Chem. Soc. 2001, 123,
1898Ϫ1904.
[19] [19a] K. J. Humphreys, K. D. Karlin, S. E. Rokita, J. Am. Chem.
[8]
[9]
[19b]
D. M. P. Mingos, D. J. Wales, Introduction to Cluster Chemistry,
Prentice-Hall International, Englewood Cliffs, NJ, 1990.
G. A. Foulds, B. F. G. Johnson, J. Lewis, J. Organomet. Chem.
1985, 296, 147Ϫ153.
Soc. 2001, 123, 5588Ϫ5589.
C. S. Allardyce, P. J. Dyson,
J. Cluster Sci. 2001, 12, 563Ϫ569. [19c] J. R. Hwu, K.-L. Lu, S.-
F. Yu, L.-J. Yu, S. Kumaresan, K.-J. Lin, S.-C. Tsay, Photo-
chem. Photobiol. 2002, 75, 457Ϫ461.
[19d]
E. Rosenberg, F.
[10] [10a]
M. Langenbahn, H. Stoeckli-Evans, G. Süss-Fink, J. Or-
Spada, K. Sugden, B. Martin, L. Milone, R. Gobetto, A. Viale,
J. Fiedler, J. Organomet. Chem. 2003, 668, 51Ϫ58.
[10b]
ganomet. Chem. 1990, 397, 347Ϫ358.
M. Langenbahn, H.
[20]
Stoeckli-Evans, G. Süss-Fink, Helv. Chim. Acta 1991, 74,
P. T. Beurskens, G. Beurskens, W. P. Bosman, R. de Gelder, S.
´
549Ϫ553.
Garcıa-Granda, R. O. Gould, R. Israe¨l, J. M. M. Smits, The
[11]
K. K. H. Lee, W. T. Wong, J. Chem. Soc., Dalton Trans.
1997, 2987Ϫ2996.
M. J. Nicholls, M. D. Vargas, Inorg. Synth. 1990, 28, 232Ϫ235.
DIRDIF-96 Program System, Crystallography Laboratory, Uni-
versity of Nijmegen, The Netherlands, 1996.
S. Parkin, B. Moezzi, H. Hope, J. Appl. Cryst. 1995, 28, 53Ϫ56.
R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33Ϫ38.
G. M. Sheldrick, SHELXL-97, version 97Ϫ2, University of
Göttingen, Germany, 1997.
J. F. van der Maelen, Cryst. Rev. 1999, 7, 125Ϫ187.
A. L. Spek, The EUCLID Package, in Computational Crystal-
lography (Ed.: D. Sayre), Clarendon Press, Oxford, UK, 1982,
p. 528.
[12]
[13]
[21]
[22]
[23]
´
´
´
J. A. Cabeza, I. del Rıo, S. Garcıa-Granda, V. Riera, M. Sua-
rez, Organometallics 2002, 21, 2540Ϫ2543.
[14]
[15]
´
´
´
J. A. Cabeza, I. del Rıo, S. Garcıa-Granda, V. Riera, M. Sua-
rez, Organometallics 2002, 21, 5055Ϫ5059.
[24]
[25]
´
´
J. A. Cabeza, I. da Silva, I. del Rıo, S. Garcıa-Granda, V. Riera,
M. G. Sanchez-Vega, Organometallics 2003, 22, 1519Ϫ1525.
´
[16] [16a] N. Farrell, in Catalysis by Metal Complexes (Eds.: R. Ugo,
[26]
B. R. James), Kluwer Academic Publishers, Dordrecht, The
L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837Ϫ838.
[16b]
Netherlands, 1989, Vol. 11, p. 17.
J. Kim, M. F. Sistare, P.
Received June 4, 2003
Early View Article
Published Online October 10, 2003
J. Carter, H. H. Thorp, Coord. Chem. Rev. 1998, 171, 341Ϫ349.
[16c]
A. Sreedhara, J. D. Freed, J. A. Cowan, J. Am. Chem. Soc.
Eur. J. Inorg. Chem. 2003, 4159Ϫ4165
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4165