Surface-assisted synthesis and behavior of dimetallic mixed-metal complexes [M2Cl2(μ-Cl)4(CO)6M′(L) 2] (M = Ru, Os; M′ = Fe, Co; L = CH3CH 2OH, H2O)

Article abstract of DOI:10.1002/ejic.200700341

The reductive carbonylation of ruthenium and osmium halides in the presence of cobalt and iron surfaces was studied. In these surface-assisted reactions the metal surface plays an active role in releasing metal ions, which can be used for the in situ synthesis of mixed-metal compounds. A linear, dimetallic, chlorido-bridged, trinuclear complex [Ru₂Cl₂(μ-Cl) ₄(CO)₆Co(CH₃CH₂OH)₂] was obtained by the interaction between reducing RuCl₃ and the cobalt surface. In the reaction the RuCl₃ is reduced with carbon monoxide in the presence of a solid cobalt surface in ethanol solution. During the reduction of the RuCl₃ the cobalt surface was simultaneously corroded, releasing cobalt cations. In addition to [Ru₂Cl ₂(μ-Cl)₄(CO)₆Co(CH₃CH ₂OH)₂], the reaction also produced other trinuclear complexes and mononuclear products, such as a bent and linear dimetallic, chlorido-bridged complex with the aqua ligand [Ru₂Cl ₂(u-Cl)₄(CO)₆Co(H₂O)₂] and ionic [RuCl₃(CO)₃]₂[Co(H₂O) ₆], The carbon monoxide reduction of Rul₃ produced the ionic complex [RuI₃(CO)₃]₂[M′(H ₂O)₆] (M′ = Fe, Co) in the presence of iron, stainless steel or cobalt. Iodido-bridged trinuclear mixed-metal complexes were not observed. The surface-assisted process proved to be useful with other metal combinations as well. When RuCl₃ was replaced with OsCl₃, a variety of trinuclear mixed-metal complexes, such as [Os₂Cl ₂(u-Cl)₄(CO)₆Co(CH₃CH ₂OH)₂] and ionic [OsCl₃(CO)₃] ₂[M(H₂O)₆] (M = Fe, Co), were obtained, depending on the metal surface used. Due the lability of the chlorido-bridged trinuclear complexes, the possible decomposition steps of [M₂Cl ₂(μ-Cl)₄(CO)₆M′(CH₃CH ₂OH)₂] (M = Ru, Os; M′ = CO, Fe) were studied computationally, using DFT methods. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejic.200700341

FULL PAPER
DOI: 10.1002/ejic.200700341
Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal
Complexes [M2Cl2(µ-Cl)4(CO)6MЈ(L)2] (M = Ru, Os; MЈ = Fe, Co; L =
CH3CH2OH, H2O)
Minna Jakonen,[a] Pipsa Hirva,[a] Taina Nivajärvi,[a] Mirja Kallinen,[a] and Matti Haukka*[a]
Keywords: Ruthenium / Osmium / Cobalt / Iron / Synthesis design
The reductive carbonylation of ruthenium and osmium ha-
lides in the presence of cobalt and iron surfaces was studied.
In these surface-assisted reactions the metal surface plays an
active role in releasing metal ions, which can be used for the
in situ synthesis of mixed-metal compounds. A linear, dimet-
allic, chlorido-bridged, trinuclear complex [Ru2Cl2(µ-Cl)4-
(CO)6Co(CH3CH2OH)2] was obtained by the interaction be-
tween reducing RuCl3 and the cobalt surface. In the reaction
the RuCl3 is reduced with carbon monoxide in the presence
of a solid cobalt surface in ethanol solution. During the re-
duction of the RuCl3 the cobalt surface was simultaneously
corroded, releasing cobalt cations. In addition to [Ru2Cl2(µ-
Cl)4(CO)6Co(CH3CH2OH)2], the reaction also produced other
trinuclear complexes and mononuclear products, such as a
bent and linear dimetallic, chlorido-bridged complex with
the aqua ligand [Ru2Cl2(µ-Cl)4(CO)6Co(H2O)2] and ionic
[RuCl3(CO)3]2[Co(H2O)6]. The carbon monoxide reduction of
RuI3 produced the ionic complex [RuI3(CO)3]2[MЈ(H2O)6] (MЈ
= Fe, Co) in the presence of iron, stainless steel or cobalt.
Iodido-bridged trinuclear mixed-metal complexes were not
observed. The surface-assisted process proved to be useful
with other metal combinations as well. When RuCl3 was re-
placed with OsCl3, a variety of trinuclear mixed-metal com-
plexes, such as [Os2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2] and
ionic [OsCl3(CO)3]2[M(H2O)6] (M = Fe, Co), were obtained,
depending on the metal surface used. Due the lability of the
chlorido-bridged trinuclear complexes, the possible decom-
position steps of [M2Cl2(µ-Cl)4(CO)6MЈ(CH3CH2OH)2] (M =
Ru, Os; MЈ = CO, Fe) were studied computationally, using
DFT methods.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Although it is known that solid metal surfaces can also
participate in the formation of the products without exter-
nal current, this phenomenon has not been widely used in
the synthesis of organometallic compounds. In fact, car-
bonylmetal chemistry started to develop via the corrosion
products formed under harsh conditions.[8] Mond was the
first to report the formation of toxic Ni(CO)4 in nickel pipes
carrying carbon monoxide gas.[8] In our earlier study, we
suggested that the release of the metal from a metal surface
by corrosion, without the use of an externally applied cur-
rent, could be used for the synthesis of other carbonyl-con-
taining metal complexes. Such an approach was applied in
the synthesis of mixed-metal trinuclear Ru-Fe-Ru com-
plexes, which were prepared by reducing RuCl3 with carbon
monoxide in the presence of an iron-containing metal sur-
face.[9] In this synthesis method, ruthenium(III) salt and the
iron-containing metal surface form a metal pair suitable for
redox processing, which releases the less noble metal ions
from the surface. The metal cations released can then be
used in the in situ synthesis of the trinuclear mixed-
metal complexes [Ru2Cl2(µ-Cl)4(CO)6Fe(L)2] (L = H2O,
Introduction
Sacrificial anodes, most commonly used in corrosion
protection,[1] have a variety of applications in synthetic
chemistry[2] and in biochemistry.[3] The sacrificial anode do-
nates metal cations in a reaction solution when current is
applied to the anode material by an external power source.
These released cations can then be used to form organome-
tallic[4] or inorganic complexes,[5] or inorganic nanomater-
ial.[6] Nanomaterials prepared by means of sacrificial an-
odes have been used, for example, in catalytic applica-
tions[6a] and as the inhibitors in biochemical applications.[6b]
Sacrificial anodes are also useful in the applications of or-
ganic chemistry:[7] in the polymerisation of organic mono-
mers,[7c] in the preparation of organic compounds and also
in electrocarboxylation reactions.[7d,7e] There also exists an
example of the electrosynthesis of the Grignard reagent
using sacrificial magnesium.[7g] The typical sacrificial anode
materials in electrosynthesis are magnesium, zinc, cadmium
and aluminum, but examples involving cobalt, copper,
nickel, palladium, silver, tin and iron are also available.[4–7]
CH3CH2OH) and
a mixture of mononuclear ions
[RuCl3(CO)3]2[Fe(H2O)6]. The potential difference between
the metal surface and the reducing metal salt determines
the rate of dissolution of the metal cations from the metal
surface into the reaction system.
[a] University of Joensuu, Department of Chemistry,
P. O. Box 111, 80101, Joensuu, Finland
E-mail: matti.haukka@joensuu.fi
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3497

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
In the current paper we have extended the use of sacrifi- sure to carbon monoxide tends to reduce these species
cial metal surfaces from Ru-Fe systems into the other metal eventually to the highly stable Ru3(CO)12 cluster.[15,17] Like
combinations. The goal of the study was to examine RuCl3, the CO reduction of OsCl3 is commonly carried out
whether complexes such as [Ru2Cl2(µ-Cl)4(CO)6Fe(L)2], in alcohol solutions either under reflux or in a pressurized
shown in Figure 1, could be synthesized using the metal vessel under elevated CO pressure. Such reactions again
systems Ru-Co, Os-Fe and Os-Co. In addition, the influ- produce a variety of halide-containing carbonyl products
ence of the halide on the reaction progress was studied and, finally, Os3(CO)12.[18–20] A different type of surfaces
using RuCl3 and RuI3 as the reducing components.
has been used to assist the reduction process. For example,
it has been found that reductive carbonylation of silica-sup-
ported MCl3 (M = Ru,Os) produces several halide-contain-
ing carbonylmetal(II) compounds as well as the M3(CO)12
cluster.[21]
If metal surfaces are used, this opens up possibilities for
synthesizing mixed-metal compounds. For example, if the
reductive carbonylation of ruthenium trichloride is carried
out in the presence of an iron-containing surface, it acts as
a sacrificial source of iron. The active participation of the
metal surface permits the formation of mixed-metal prod-
ucts such as trinuclear [Ru2Cl2(µ-Cl)2(CO)6Fe(L)2] (L =
H2O or CH3CH2OH) and also the ionic monomeric
[RuCl3(CO)3]2[Fe(H2O)6],[6] where the cations and anions
are bound together by hydrogen bonds. It is believed that
during the reduction of Ru3+, iron from the surface is simul-
Figure 1. Schematic picture of the chlorido-bridged trinuclear ru-
thenium-iron complex.[6]
Results and Discussion
The reductive carbonylation of ruthenium(III) trichlo- taneously oxidized and released, which leads to the forma-
ride is typically performed in alcohol solutions either under tion of labile trinuclear dimetallic chlorido-bridged
reflux or in high-pressure carbon monoxide. The reduction compounds [Ru2Cl2(µ-Cl)2(CO)6Fe(L)2] (L = H2O or
produces complexes with partially or fully reduced ruthe- CH3CH2OH) with Ru2+ and Fe2+ centres. The same ap-
nium centers.[10–15] Normally, the final reduction product proach can be applied to other metal combinations using
is the zero-valent ruthenium cluster Ru3(CO)12 but several RuCl3 or OsCl3 as the reducing component and iron or co-
compounds containing carbonyl and halide ligands such as balt metal as the oxidizing component. The complexes syn-
[RuCl2(CO)3]2 can also be obtained.[11,14,16] Further expo- thesized in this work are summarized in Figure 2.
Figure 2. Schematic drawing of the main products involved in the reductive carbonylation reaction of MX3 in the presence of a sacrificial
metal surface.
3498
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508

Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal Complexes
FULL PAPER
Corrosion of the metal plate could be detected with all hydrogen bonding between the hydroxy hydrogen atom of
of the metals used as an oxidizing surface. Both iron and the ethanol ligand and the terminal chlorido ligand of the
cobalt surfaces and also stainless steel do corrode. In the ruthenium moiety. The pink crystals of [Ru2Cl2(µ-Cl)4(CO)6-
reactions with iron or stainless steel, the initially black reac- Co(CH3CH2OH)2] could be isolated from the solid product,
tion solution turned greenish-yellow during the carbon- which was obtained after recrystallization from dichloro-
ylative reduction of MCl3. Green or a greenish color was methane. The yield of complex 1 was ca. 10–15%. Complex
typical for all reactions involving an iron-containing sur- 1 could be obtained both from high-pressure reactions in
face, indicating the release of iron ions. When stainless steel an autoclave and from low-pressure reactions under reflux.
was used as the source of iron ions, the type of corrosion A reasonable yield of the complex 1 permitted further
produced was pitting, which is typical of metals and metal analysis. In order to verify the results, the elemental content
alloys with a passive oxide layer on their surface.[22] Fairly of the product was analyzed using both elemental analysis
uniform corrosion was observed when pure metal, iron or and EDS (Energy Dispersive Spectroscopy). The latter con-
cobalt, was used. In a typical reaction with cobalt surface, firmed the presence of cobalt and ruthenium. A pale violet
the reductive carbonylation of RuCl3 produced a blue solu- trinuclear complex with aqua ligand, [Ru2Cl2(µ-Cl)4(CO)6-
tion. Again, the metal surface was clearly worn down. A Co(H2O)2] (2), could also be isolated from the product, al-
blue coloration indicates the release of cobalt ions into the though the formation of complex 1 seemed to be preferred.
reaction solution.
The yield of complex 2 remained rather low at ca. 8%.
Again, the structural features of [Ru2Cl2(µ-Cl)4(CO)6-
Co(H2O)2] were very similar to [Ru2Cl2(µ-Cl)4(CO)6Fe-
(H2O)2].[6] Two different structures were found: linear 2a,
presented in Figure 4, and bent 2b, presented in Figure 5.
In both structures the intramolecular hydrogen bonds be-
tween the coordinated water molecule and the terminal
chlorido ligand further support the trinuclear arrangement
of the metal atoms. It was found that the formation of the
bent type of isomer 2b was preferred over linear 2a when
the cobalt plate was replaced by cobalt powder. Such an
observation may suggest that the morphology of the surface
plays an important role in the formation of trinuclear
mixed-metal products.
Formation of Ruthenium-Cobalt Mixed-Metal Complexes
The carbonylative reduction of RuCl3 in the presence of
a cobalt plate consistently produced a mixture of various
mixed-metal complexes and mononuclear metal complexes
with carbonyl and halide ligands. The ratio of the different
products varied with each reaction. Because the yields of
the individual products were rather low, single-crystal X-ray
diffraction was used as the principal method of analysis. In
order to crystallize the products, the reaction solutions were
concentrated to dryness and the solid products were recrys-
tallized from dichloromethane by means of slow concentra-
tion or by adding hexane to the CH2Cl2 solution. In the
case of RuCl3 reduction with a cobalt surface, the crystalli-
zation produced a mixture of pale pink crystals and color-
less crystals. X-ray analysis of the pink crystals revealed tri-
nuclear [Ru2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2], as shown
in Figure 3.
Figure 4. Thermal ellipsoid plot (50% probability level) of
OsCl2(CO)3(H2O) (2a). Hydrogen bonds [Å, °]: O4–H4 0.98,
H4···Cl3 2.30, O4···Cl3 3.242(3), O4–H4···Cl2(A) 161.8.
Figure 3. Thermal ellipsoid plot (50% probability level) of
[Ru2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2] (1). Hydrogen bond [Å, °]:
O4–H4O 0.95, H4O···Cl3(A) 2.15, O4···Cl3(A) 3.073(2), O4–
H4O···Cl3(A) 164.6. Symmetry transformations used to generate
equivalent atoms: (A) –x + 1, –y, –z + 1.
Figure 5. Thermal ellipsoid plot (50% probability level) of the bent
Structurally,
[Ru2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2]
[Ru2Cl2(µ-Cl)4(CO)6Co(H2O)2] (2b), crystallized in the triclinic
closely resembles [Ru2Cl2(µ-Cl)4(CO)6Fe(CH3CH2OH)2], as
¯
space group P1. Hydrogen bond [Å, °]: O7–H7A 0.79, H7A···O2
reported earlier.[6] The trinuclear structure is supported by 2.59, O7···O2 3.187(3), O7–H7A···O2 133.2.
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3499

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
The chlorido-bridged trinuclear mixed-metal complexes long enough (Ͼ 17 h), the dominating carbonylation prod-
of types 1 and 2 are labile and sensitive towards moisture. uct is zero-valent, fully carbonylated and reduced Ru3-
Previous results suggest that the exposure of [Ru2Cl2(µ-Cl)4- (CO)12. This seems to be true for both reactions, with or
(CO)6Fe(CH3CH2OH)2] to moisture leads first to ligand ex- without the presence of the solid metal surface. The solvent
change on the central metal atom and then to the formation also has an effect on the progression of the reaction and on
of [Ru2Cl2(µ-Cl)4(CO)6Fe(H2O)2]. Additional exposure to the products formed during the reduction. If the alcohol
moisture finally breaks down the chlorido bridges and pulls solvent is replaced with water, the final carbonylation prod-
the metal atoms apart, yielding ionic [Ru(CO)3Cl3]2- uct is polymeric [Ru(CO)4]n.[23] However, the presence of
[Fe(H2O)6].[6] Similar behavior was also observed with com- the metal surface tends to favor the partially reduced and
plex 1. As in the case of the RuCl3/Fe reaction, the RuCl3/ carbonylated ruthenium species. If RuCl3 was simply heated
Co reaction produced an ionic decomposition product, under reflux in ethanol and in a CO flow in the presence
[Ru(CO)3Cl3]2[Co(H2O)6] (3), with variable ratios (Fig- of an iron or cobalt plate (17 h), Ru3(CO)12 was not de-
ure 6).
tected at all. After 3 h of reaction with 20 bar of CO, the
cluster product was still not observable. Increasing the reac-
tion temperature from 125 °C to 150 °C led to the forma-
tion of Ru3(CO)12 after only 3 h of reaction with 20 bar of
carbon monoxide, but under these conditions Ru3(CO)12
was only a minor product. At 50 bar of CO, Ն 125 °C and
Ն 3 h reaction times with or without the presence of a
metal surface, the principal product was an Ru3(CO)12 clus-
ter. Cluster formation under similar conditions without
metal surface is well known and reported by Bruce et al.[24]
Nevertheless, it should be noted that partially carbonylated
species can still be found even with elevated pressures
(50 bar) and long reaction times (17–20 h), where Ru3-
(CO)12 is already the main product, including in reactions
with a metal surface. It seems likely that products with par-
tially reduced ruthenium centres [Ru2Cl2(µ-Cl)4(CO)6-
M(L)2], and [Ru(CO)3Cl3]2[M(H2O)6] (M = Fe, Co and L
= H2O, CH3CH2OH) are stable enough to interfere with
Figure 6. Thermal ellipsoid plot (50% probability level) of
[RuCl3(CO)3]2[Co(H2O)6]·2(H2O) (3). The symmetry-related water
molecule is not shown. Hydrogen bonds [Å, °]: O4–H4A 0.85,
H4A···Cl1 2.36, O4···Cl1 3.140(2), O4–H4A···Cl1 152.8, O5–H5B
0.84, H5B···Cl3 2.44, O5···Cl3 3.220(2), O5–H5B···Cl3 153.8, O6–
H6B 0.84, H6B···O99 1.88, O6···O99 2.716(2), O6–H6B···O99
173.3.
Compared to [Ru2Cl2(µ-Cl)4(CO)6Fe(CH3CH2OH)2] and the reduction of ruthenium in the zero-valent state.
[Ru2Cl2(µ-Cl)4(CO)6Fe(H2O)2], the decomposition of com- From the synthetic point of view, the formation of an
plexes 1 and 2 into the ionic product 3 seems to be slightly unpredictable mixture of products is the main setback,
slower. For example, complex 1 can be stored in air much which complicates the use of metal surfaces for synthetic
longer than the corresponding iron compound without se- purposes. The yields of the most interesting products, the
vere decomposition.
mixed-metal trinuclear [Ru2Cl2(µ-Cl)4(CO)6Co(L)2] com-
In addition to complexes 1, 2, and 3, crystallographic plexes, consistently remained rather low at around 8–15%.
analysis of the reaction products obtained after the re- This is mainly due to the lability of the trinuclear com-
duction of RuCl3 in the presence of cobalt revealed other pounds in the presence of moisture. However, compared to
types of mononuclear products as well such as neutral its Ru-Fe-Ru equivalent, [Ru2Cl2(µ-Cl)4(CO)6Co(L)2] can
RuCl2(H2O)(CO)3 and CoCl2(H2O)4, which could be iden- certainly be synthesized with better reproducibility and en-
tified by means of single-crystal X-ray diffraction. This sug- hanced yields.
gests that an alternative decomposition route for trinuclear
The reaction products 1–3, which were obtained with
Ru-Co-Ru may also be possible. The formation of corre- adequate yields, were also analysed using IR spectroscopy.
sponding neutral mononuclear species has been observed in The common feature of all these compounds is that the
RuCl3/Fe systems earlier, which means that a similar alter- arrangement of the carbonyl ligands around the ruthenium
native route also exists in the case of RuCl3/Co systems. center is very similar in all of them. The facial arrangement
However, it should be mentioned that, according to the very of the CO ligands in each compound gives a ν(CO) pattern
low yields of the neutral products, this route seems not to with two strong bands: a sharp band at higher wave-
be very feasible. Actually, neutral products such as numbers around 2140–2145 cm–1 and a broad- or double-
RuCl2(H2O)(CO)3 were found only occasionally in some of headed one at lower wavenumbers at ca. 2055–2065 cm–1.
the reaction products. Each reaction was repeated several Two pairs of ν(CO) signals were found at 2143 cm–1 and
times, and the main decomposition product was always 2077 cm–1 in the CH2Cl2 solution immediately after reac-
found to be ionic [Ru(CO)3Cl3]2[Co(H2O)6].
tion, when ethanol is evaporated from the Ru-Co reaction
The product distribution of the carbonylative reduction solution and the solid material is dissolved in CH2Cl2. Typi-
of RuCl3 depends strongly on the reaction conditions. If the cally, the higher band was sharp and the lower band much
reaction is carried out without metal surfaces and if the CO broader. The result closely resembles those found in reac-
pressure is high enough (Ͼ 20 bar) and the reaction time tions conducted with iron-containing surfaces, where the
3500
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508

Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal Complexes
FULL PAPER
corresponding signals have been reported to be at and Ru(CO)4I2 (12), which were characterized by single-
2143 cm–1 and 2078 cm–1.[6]
crystal X-ray diffraction. Structurally, compounds 4 and 5
ATR-IR (Attenuated Total Reflectance IR) measure- are isostructural with the corresponding complex 3 formed
ments were performed for the chlorido-bridged Ru-Co-Ru in the reduction reactions of RuCl3. The structures of 4, 11
complexes 1 and 2, and for ionic 3. In the IR measurements and 12 are shown in Figures 7, 8, and 9.
of the reaction solutions different complexes could not be
assigned, but with the ATR technique the complexes could
be identified. In the case of the ATR measurements two
bands were detected again, and as with the FTIR spectra
the peak at the higher wavenumber was once again strong
and rather sharp. The lower peak was double-headed and
actually a combination of two peaks. In the spectrum of
complex 1 the highest signal is at the wavenumber
2137 cm–1, while the lower band has two peaks, at 2075 and
2054 cm–1. Similar patterns can be found also in the spectra
of the compounds 2 and 3 with the wavenumbers 2135,
Figure 8. Thermal ellipsoid plot (50% probability level) of
[RuI2(CO)3]2 (11). Symmetry transformations were used to generate
equivalent atoms: (A) x, –y, z; (B) –x, –y, –z + 1.
2070, and 2054 cm–1 for 2 and 2139, 2075, and 2056 cm–1
for 3.
The Effect of the Halide in the Reducing Salt
In order to study the role of the halide in the reducing
salt, the RuCl3 was replaced with RuI3. Otherwise, the re-
ductive carbonylation reaction with cobalt or iron surfaces
was carried out under conditions similar to those used for
a typical RuCl3 reduction (20 bar of CO, 125 °C, 3 h).
Again, the release of iron and cobalt cations was observed
in the reaction. The most noticeable feature of these reac-
tions was the absence of halide-bridged trinuclear mixed-
metal products. They were detected in reduction reactions
with either of the metals (Fe, Co) used. This may indicate
that the trinuclear iodido-bridged products are less stable
than their chlorido-bridged analogues. The reaction with
RuI3 produced variable amounts of [RuI3(CO)3]2[Fe-
(H2O)6] (4), [RuI3(CO)3]2[Co(H2O)6] (5), [RuI2(CO)3]2 (11)
Figure 9. Thermal ellipsoid plot (50% probability level) of
RuI2(CO)4 (12).
Another difference between the RuI3 reactions and the
reduction reactions of RuCl3 was the strongly decreased
formation of Ru3(CO)12. In fact, this was not observed at
all even at higher pressures (50 bar) and with long reaction
times (20 h). In the reactions under 50 bar pressure of car-
bon monoxide, the reduction reaction (125 °C, 3–20 h) pro-
duced [RuI2(CO)3]2 (11), presented in Figure 8. The Ru-
I2(CO)4 (12) presented in Figure 9 was found only in the
reactions with iodine salt, while no equivalent chloride
product was observed in the RuCl3 reactions.
The Carbon Monoxide Reduction of OsCl3 in the Presence
of a Sacrificial Metal Surface
The carbonylative reduction of OsCl3 follows the general
lines of the reduction of RuCl3. The final reduction product
is the zero-valent cluster Os3(CO)12, especially under high
pressures.[18–21] In the current study, the reduction reactions
were performed in an alcohol solution either under reflux
with a CO stream passing through the reaction solution or
in a pressurized autoclave. In both of the reaction setups a
sacrificial metal plate (stainless steel, Fe or Co) was added
to the system. During the reduction reaction the colour of
the solution turned from black to brownish-yellow when
an iron-containing plate was used, and greenish-blue when
cobalt was used. In all of the osmium reactions, either un-
der reflux or in an autoclave, a dark brown-grey metallic
powder was formed. The powder was removed from the
solution by filtering prior to crystallization of the carbon-
Figure 7. Thermal ellipsoid plot (50% probability level) of
[RuI3(CO)3]2[Fe(H2O)6]·2(H2O) (4). The symmetry-related second
[RuI3(CO)3]– anion and the solvent water molecule are not shown.
Hydrogen bonds [Å, °]: O4–H4A 0.78, H4A···O1 3.14, O4···O1
3.534(6), O4–H4A···O1 114.2, O5–H5A 0.81, H5A···I2 2.83,
O5···I2 3.602(4), O5–H5A···I2 159.3, O5–H5B 0.79, H5B···I3 3.32,
O5···I3 3.721(4), O5–H5B···I3 176.1, O6–H6A 0.85, H6A···O(99)
1.90, O6···O99 2.752(5), O6–H6A···O99 176.1.
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3501

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
ylation products. The metallic powder material was charac- peaks at wavenumbers of 2043 cm–1 and 2024 cm–1. Com-
terized with the aid of EDS measurements, which con- pared with the ATR-IR spectra of ruthenium analogues,
firmed that the solid material consisted mainly of metallic the signals are shifted to a lower wavenumber, but the shape
osmium. The filtrate was analyzed after it had been concen- and intensity of the signals remained the same.
trated and redissolved in dichloromethane. IR analysis re-
vealed two peaks in the ν(CO) region with both Os-Fe and
Os-Co systems. In solutions containing Os-Fe, the higher
band at 2135 cm–1 was again sharp and the lower one at
2058 cm–1 was broader. A similar pattern was also found
with the Os-Co reaction solution, which produced corre-
sponding peaks at 2135 cm–1 and 2054 cm–1. In comparison
with the Ru-Fe/Co bands, the Os-Fe/Co bands shifted
towards lower wavenumbers, but the shape and relative in-
tensities of the bands remained the same in all cases.
The crystallization of the osmium-iron and osmium-co-
balt products was much more difficult than that of the ru-
thenium-iron or, in particular, of the ruthenium-cobalt
products. In consequence, the final yields of the separated
products remained very low. This also prevented further
analysis of the isolated products. Crystallization was per-
formed according to the same procedure as for the ruthe-
nium reactions. Only the mononuclear products could be
isolated and characterized from the Os-Fe reactions. Ionic
[OsCl3(CO)3]2[Fe(H2O)6] (6) is the structural equivalent of
[RuCl3(CO)3]2[Co(H2O)6] (3) and another ionic product
[OsCl3(CO)3][FeCl2(H2O)4] (7). In 7, the osmium atom is in
an oxidation state of 2+ and the iron atom is in an oxi-
dation state of 3+. However, this minor product was ob-
tained only from one reaction. The equivalent of complex
7 was observed neither in any of the Ru-Co or Ru-Fe reac-
tions nor in the Os-Co reactions. Another mononuclear re-
duction product of OsCl3 with no ruthenium counterpart
was OsCl2(CO)3(CH3CH2OH) (10). This product was also
a minor product found only from one reaction. The solid-
state structure of 10 consists of two osmium moieties bound
together with a hydrogen bond (Figure 10).
Figure 11. Thermal ellipsoid plot (50% probability level) of
[OsCl3(CO)3]2[Co(H2O)6]·2(H2O) (8a). The symmetry-related sec-
ond [OsCl3(CO)3]– anion and the second water molecule are not
shown. Hydrogen bonds [Å, °]: O4–H4B 0.99, H4B···Cl1 2.27,
O4···Cl1 3.216(3), O4–H4B···Cl(1) 159.9, O7(A)–H7(A) 0.94,
H7(A)···Cl1 2.36, O7(A)···Cl1 3.292(3), O7–H7(A)···Cl1 173.5.
An ethanol-containing trinuclear complex with chlorido
bridges, [Os2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2] (9), could
also be found in the osmium-cobalt reactions. Structurally,
9 is similar to [Ru2Cl2(µ-Cl)4(CO)6M(CH3CH2OH)2] (M =
Fe, Co) (Figure 12).
Figure 12. Thermal ellipsoid plot (50% probability level) of
[Os2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2] (9). Hydrogen bond [Å, °]:
O4–H4 0.95, H4···Cl3(A) 2.15, O4···Cl3(A) 3.072(6), O4–
H4···Cl3(A) 162.4.
The formation of trinuclear halide-bridged compounds
is not as favorable with Os/Fe and Os/Co metal pairs as it
is with the Ru/Co pair. Complex 9 was separated in some
of the reactions, but significantly more rarely than with the
ruthenium analogue 1. Although Os/Co trinuclear halide-
bridged complexes were rare and Os/Fe analogues were not
Figure 10. Thermal ellipsoid plot (50% probability level) of
OsCl2(CO)3(EtOH) (10).
From the osmium-cobalt reaction, [OsCl3(CO)3]2[Co- detected during our study, this does not necessarily mean
(H2O)6] (8) could be isolated and characterized by means that such a compound cannot be formed at all. Such a
of the single-crystal X-ray diffraction (Figure 11). ATR product may either be less stable, or the absence of these
measurements of the analysed complex 8a revealed a complexes may simply be due to the fact that the reaction
pattern of signals that was similar to those found by means setup as it now exists is unsuitable for osmium reactions. In
of the measurements of the Ru-Co complexes 1, 2 and 3. addition, it can be stated that control of the products in the
The higher band was again sharp at a wavenumber of OsCl3 reactions was undoubtedly more difficult than with
2126 cm–1, and the lower peak was a combination of two RuCl3.
3502
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508

Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal Complexes
FULL PAPER
Scheme 1. Decomposition steps of µ-Cl dimetallic trinuclear complexes.
Stability of the Trinuclear Complexes
The decomposition of [M2Cl2(µ-Cl)4(CO)6MЈ(H2O)2]
can follow two different main steps (steps II and III in
Scheme 1). The addition of 4 equiv. of H2O breaks down
the chlorido bridges and results in ionic [MCl3(CO)3]2-
[MЈ(H2O)6] or [MCl2(CO)3(H2O)] and [MЈCl2(H2O)4]. Ac-
cording to our calculations, the relative energies of the de-
composition products formed along step II are around
300 kJ/mol more favourable than for the trinuclear
chlorido-bridged complexes. The result agrees well with the
experimental results: [MCl3(CO)3]2[MЈ(H2O)6] is the fav-
oured product, although [MCl2(CO)3(H2O)] and
[MЈCl2(H2O)4] have been observed from reactions of the
ruthenium-cobalt system. It also seems that decomposition
step III is around 40 kJ/mol more favourable with iron than
with cobalt, which is also consistent with experimental ob-
servations. According to our calculations, in MЈCl2(H2O)4
the most favourable isomer is the trans(Cl) structure when
the metal atom is iron and the cis(Cl) structure when the
metal atom is cobalt, although the energy difference be-
tween the isomers is less than 5 kJ/mol for both metal chlo-
rides. There are no notable differences in the calculation
results between the Ru-Fe, Ru-Co, Os-Fe and Os-Co sys-
tems.
Computational work was done to investigate the stability
of the complexes by means of their relative energies. In our
previous study we investigated[6] the decomposition path of
[Ru2Cl2(µ-Cl)4(CO)6Fe(CH3CH2OH)2]. The presence of
water launches the decomposition of the halide-bridged tri-
nuclear mixed-metal complexes. The decomposition path-
way proceeds by the replacement of the ethanol ligands
with water and the formation of [Ru2Cl2(µ-Cl)4(CO)6-
Fe(H2O)2] and its eventual decomposition into the mono-
nuclear products [RuCl3(CO)3][Fe(H2O)6], [RuCl2(CO)3-
(H2O)] and [FeCl2(H2O)4]. In the current study we have in-
vestigated a similar decomposition route for the Ru/Co, Os/
Fe and Os/Co systems by using computational DFT meth-
ods. The decomposition steps involved are summarized in
Scheme 1 and in Table 1.
Table 1. Calculated DFT energies of the decomposition routes pre-
sented in Scheme 1. Values in bold are taken from ref.[9]
M
MЈ
EI (kJ/mol) EII (kJ/mol) EIII (kJ/mol)
Ru
Ru
Os
Os
Fe
Co
Fe
–34
–31
–33
–29
–325
–287
–327
–287
67
68
92
Co
106
Conclusions
The first step in the decomposition of [M2Cl2(µ-Cl)4-
(CO)6MЈ(CH3CH2OH)2], the replacement of CH3CH2OH
The reductive carbonylation of MCl3 (M = Ru or Os) in
with H2O at the bridging metal atom (either Fe or Co), is the presence of a metal surface (Fe or Co) produces a vari-
favourable with all of the metal combinations. The reaction ety of mixed-metal products, whose components originate
energy of the ligand exchange is around –30 kJ/mol. The from the reducing MCl3 and from the oxidizing surface.
two possible structural isomers of [M2Cl2(µ-Cl)4(CO)6- The metal ions released from the oxidizing surface can be
MЈ(H2O)2], bent and linear, are energetically very similar to used for an in situ complex formation. Although this ap-
each other, which explains the co-existence of both struc- proach produces a mixture of products, it nevertheless pro-
tures. The relative energy between the bent and linear iso- vides a useful method for synthesizing trinuclear mixed-
mers for all the metal combinations is ca. 4 kJ/mol.
metal compounds of the type [M2Cl2(µ-Cl)4(CO)6MЈ(L)2]
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3503

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
oxide (98%) used in reactions was purchased from Messer and
AGA. The FTIR spectra were measured with a Nicolet Magna
IR Spectrometer 750, and ATR-IR measurements were performed
using an FTS 7000 Series DIGILAB FTIR spectrometer equipped
with a UMA 600 DIGILAB microscope. CE Instruments EA 110
CHNS-O was used in the elemental analyses. Qualitative analyses
of cobalt and osmium were conducted using an FE-SEM Hitachi
S-4800 Scanning Electron Microscope equipped with a Norman
System Six EDS.
(M = Ru, Os; MЈ = Fe, Co; L = CH3CH2OH, H2O). For
example,
a synthesis of [Ru2Cl2(µ-Cl)4(CO)6Co(CH3-
CH2OH)2] can be repeated with moderate yields. The main
problem in the production of [M2Cl2(µ-Cl)4(CO)6MЈ-
(CH3CH2OH)2] is that this can be decomposed relatively
easily by moisture into mononuclear species such as
[MCl3(CO)3]2[MЈ(H2O)6], [MCl2(CO)3(H2O)] and [MЈCl2-
(H2O)4].
Mechanistically, the reductive carbonylation of MCl3 (M
= Ru or Os) in the presence of a metal surface (Fe or Co)
follows similar reaction pathways regardless of the reducing
metal halide and oxidizing metal surface used. Further, the
decomposition of the mixed-metal trinuclear [M2Cl2(µ-Cl)4-
(CO)6MЈ(L)2] by moisture is similar in all studied cases. In
the case of [M2Cl2(µ-Cl)4(CO)6MЈ(CH3CH2OH)2], the first
step in the decomposition is the replacement of the ethanol
ligands by water at the central metal atom (Fe or Co). This
is followed by cleavage of the chlorido bridges and the for-
mation of mononuclear products. Energetically, the most
favorable decomposition products are ionic compounds of
the type [MCl3(CO)3]2[MЈ(H2O)6].
Syntheses
High-Pressure Reactions: Metal trichloride [RuCl3·x(H2O), RuI3 or
OsCl3·x(H2O)] reduction reactions were carried out in a Berghof
autoclave equipped with a PTF liner. The solid metals (Co, Fe,
stainless steel, 7ϫ7 mmϫ0.02–0.03 mm or 50 mg of cobalt pow-
der) and 100 mg of metal trichloride reagent were placed in the
liner together with 4 mL of ethanol. The autoclave was pressurized
with 10, 20 or 50 bar of carbon monoxide (in a typical reaction the
CO pressure was 20 bar), quickly heated to 125 °C, and maintained
at that temperature for 3 h. After the reaction, the autoclave was
cooled in an ice bath. The reaction solution was filtered to remove
any possible solid metallic precipitations formed during the reac-
tion. The filtrate was concentrated to dryness and the solid residue
dissolved in CH2Cl2. Crystallization was performed using the
CH2Cl2 solution by slow concentration or by adding hexane to the
CH2Cl2 solution.
The iodide salt RuI3 can be used instead of RuCl3 in the
preparation of ionic products, but no trinuclear compounds
of the type [M2I2(µ-I)4(CO)6MЈ(L)2] have been obtained so
far. This is probably due to the poorer stability of [M2I2(µ-
I)4(CO)6MЈ(L)2] compared to its chlorido equivalents.
Reactions under Reflux: Metal trichloride [RuCl3·n(H2O) or
OsCl3·3(H2O), 250 mg] was placed in a two-necked round-bottom
vessel with a metal plate (Co or Fe or stainless steel, 7ϫ7 mm).
MCl3 was diluted in 40 mL of ethanol and the reaction mixture
was refluxed in a carbon monoxide stream for 17–20 h, until the
reaction solution was yellowish with iron and blue or green with
cobalt. Any solid material formed during the reduction was filtered
off and the filtrate was placed in a 100 mL round-bottom vessel.
The solution was concentrated to dryness and the residue dissolved
in CH2Cl2. The products were crystallized as described above.
Experimental Section
Materials and Measurements: RuCl3·x(H2O) (99.9%) was pur-
chased from Alfa Aesar, OsCl3·x(H2O) from Sigma Aldrich and
RuI3 from Strem. The metal plates Co (99.9%, CO000279/3), Fe
(99.5%, FE000400/11) and stainless steel were from GoodFellow.
The cobalt powder (99.995%) was a product from Aldrich. MeOH
(p.a.) and CH2Cl2 (p.a.) were obtained from Fluka, hexane (p.a.)
from Merck, and ethanol (99.5%) from Primalco. The carbon mon-
All of the reactions produced a mixture of products with variable
ratios depending on the reaction time and conditions. In the cases
of complexes 1, 2, 3, 8 and 11, the yields were very low and varied
Table 2. Crystallographic data for 1–2.
1
2a
2b
2c
Empirical formula
Formula mass
Temperature [K]
λ [Å]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C10H12Cl6CoO8Ru2
733.97
100(2)
0.71073
monoclinic
P21/c
9.1922(3)
10.9226(3)
11.0099(4)
90
96.482(2)
90
1098.36(6)
2
2.219
2.866
C6H4Cl6CoO8Ru2
677.86
120(2)
0.71073
monoclinic
P21/n
7.0428(15)
15.489(3)
8.4505(15)
90
96.808(15)
90
915.3(3)
2
2.459
3.428
8499
C6H4Cl6CoO8Ru2
677.86
150(2)
C6H4Cl6CoO8Ru2
677.86
150(2)
0.71073
monoclinic
P21/c
10.1061(3)
6.3789(2)
27.6699(9)
90
95.5550(10)
90
1775.39(10)
4
2.536
3.535
0.71073
triclinic
¯
P1
6.3339(2)
10.4649(3)
14.3176(5)
87.315(2)
81.818(2)
74.300(2)
904.28(5)
2
γ [°]
V [Å3]
Z
ρcalcd. [Mg/m3]
µ(Mo-Kα) [mm–1]
No. of reflections
Unique reflections
Rin[at]
2.490 Mg/m3
3.470 mm–1
6593
4070
0.0256
22720
2499
0.0580
0.0245
8861
3797
0.0342
0.0294
2090
0.0564
0.0309
0.0507
R1 (I Ն2σ)
0.0289
0.0601
[b]
wR2 (IՆ2σ)
0.0571
0.0511
[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|. [b] wR2 = {Σ[w(Fo – Fc ) ]/Σ[w(Fo2)2]}1/2
.
2
2 2
3504
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508

Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal Complexes
FULL PAPER
from reaction to reaction. In these cases the identification of the
products was based on single-crystal X-ray analysis.
[OsCl3(CO)3]2[Co(H2O)6] (8): C6H12Cl6CoO12Os2 (928.21): calcd.
C 7.76, H 1.30; found C 7.70, H 1.31. ATR-IR ν(CO) = 2126 (vs),
˜
2043 (vs), 2024 cm–1 (vs). Yield: 24%.
Analysis of Separated and Purified Complexes 1–3, 8 and 11
[RuI2(CO)3]2 (11): C60I4O6Ru2 (877.82): calcd. C 8.21; found C
[Ru2Cl2(µ-Cl)4(CO)6Co(CH3CH2OH)2] (1): C10H12Cl6CoO8Ru2
8.21. IR (CH Cl ): ν(CO) = 2118 (s) , 2055 (sh, s) cm–1.
˜
2
2
(733.99): calcd. C 16.36, H 1.65; found C 16.24, H 1.69. ATR-IR
ν(CO) = 2137 (vs), 2075 (vs), 2054 cm–1 (vs). Yield: 10–15%.
X-ray Crystal Structure Determinations: The crystals were im-
mersed in cryo-oil, mounted in a Nylon loop and measured at a
temperature of 100–150 K. The X-ray diffraction data was col-
lected by means of a Nonius KappaCCD diffractometer using
Mo-Kα radiation (λ = 0.71073 Å). The Denzo-Scalepack[25] or
EvalCCD[26] program packages were used for cell refinements and
data reductions. The structures were solved by direct methods or
by the heavy-atom method using SHELXS-97,[27] SIR-97,[28] or
˜
[Ru2Cl2(µ-Cl)4(CO)6Co(H2O)2]·(H2O)
(2):
C6H6Cl6CoO9Ru2
(695.90): calcd. C 10.36, H 0.87; found C 10.11, H 0.97. ATR-IR
ν(CO) = 2135 (vs), 2070 (vs), 2054 cm–1 (vs). Yield: 8%.
˜
[RuCl3(CO)3]2[Co(H2O)6]·(H2O) (3): C6H14Cl6CoO13Ru2 (767.96):
calcd. C 9.38, H 1.84; found 9.34, H 2.01. ATR-IR ν(CO) = 2139
˜
(vs), 2075 (vs), 2056 cm–1 (vs). Yield: 16%.
Table 3. Crystallographic data for 3–6.
3
4
5
6
Empirical formula
Formula mass
Temperature [K]
λ [Å]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C6H16Cl6CoO14Ru2
785.96
100(2)
C6H16FeI6O14Ru2
1331.58
100(2)
0.71073
monoclinic
P21/c
12.6166(2)
9.5667(2)
12.4005(2)
90
109.3620(10)
90
1412.08(4)
2
3.132
8.170
9461
C6H16CoI6O14Ru2
1334.66
110(2)
0.71073
monoclinic
P21/c
12.6371(4)
9.5821(3)
12.4173(5)
90
109.3740(10)
90
1418.47(9)
2
3.125
8.207
12152
C6H12Cl6FeO12Os2
925.11
120(2)
0.71073
0.71073
triclinic
triclinic
¯
¯
P1
P1
6.2044(3)
7.1225(2)
14.7970(7)
91.603(3)
98.438(2)
115.186(3)
582.33(5)
1
2.241
2.729
8390
2634
6.2890(14)
6.3520(10)
13.939(3)
87.901(7)
77.217(5)
80.630(7)
535.8(2)
1
2.867
13.293
6598
1840
γ [°]
V [Å3]
Z
ρcalcd. [Mg/m3]
µ(Mo-Kα) [mm–1]
No. of reflections
Unique reflections
Rin[at]
3177
0.0606
0.0333
0.0791
3217
0.0408
0.0242
0.0582
0.0241
0.0219
0.0536
0.0820
0.0523
0.1215
R1 (IՆ2σ)
[b]
wR2 (IՆ2σ)
2
2 2
[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|. [b] wR2 = {Σ[w(Fo – Fc ) ]/Σ[w(Fo2)2]}1/2
.
Table 4. Crystallographic data for 7–9.
7
8a
8b
9
Empirical formula
Formula mass
Temperature[K]
λ [Å]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C3H12Cl5FeO9Os
C6H16Cl6CoO14Os2
964.22
120(2)
C6H12Cl6CoO12Os2
928.19
120(2)
C10H12Cl6CoO8Os2
912.23
120(2)
0.71073
monoclinic
P21/c
9.1912(5)
10.9224(6)
11.0355(6)
90
96.174(3)
90
1101.43(10)
2
2.751
13.012
615.43
120(2)
0.71073
monoclinic
P21/n
11.7840(12)
9.4780(11)
15.1490(11)
90
108.032(6)
90
1608.9(3)
4
2.541
9.650
11930
3663
0.0474
0.0308
0.0603
0.71073
0.71073
triclinic
triclinic
¯
¯
P1
P1
6.2144(2)
7.1074(2)
14.8771(6)
91.495(2)
98.393(2)
115.275(2)
584.95(3)
1
2.737
12.277
10546
2673
6.2750(12)
6.3310(13)
13.877(3)
87.455(11)
77.067(11)
80.810(12)
530.39(19)
1
2.906
13.527
6704
1824
γ [°]
V [Å3]
Z
ρcalcd. [Mg/m3]
µ(Mo-Kα) [mm–1]
No. of reflections
Unique reflections
Rin[at]
11536
2164
0.0437
0.0319
0.0306
0.0231
0.0517
0.0788
0.0496
0.1200
R1 (IՆ2σ)
[b]
wR2 (IՆ2σ)
0.0693
2
2 2
[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|. [b] wR2 = {Σ[w(Fo – Fc ) ]/Σ[w(Fo2)2]}1/2
.
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3505

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
SIR2004[29] with a WinGX[30] graphical user interface. An empirical
absorption correction was applied to all of the data (XPREP in
SHELXTL v.6.14–1[31] or SADABS v.2.10[32]). Structural refine-
ments were carried out using SHELXL-97.[33] In 1, 2a, 2c, 3, 4, 5,
7, 8a, 9 and 10 the H2O and OH hydrogen atoms were located from
the difference Fourier map but constrained to ride on their parent
atom with Uiso = 1.5 Uiso(parent atom). In 2b the H2O hydrogen
atoms were also located from the difference Fourier map and re-
fined with fixed coordinates and fixed Uiso = 0.05. In 6 and 8b the
idealized positions of the H2O hydrogen atoms were estimated
using the HYDROGEN[34] program and they were constrained to
ride on their parent atom by means of Uiso = 1.5 Uiso(parent atom).
Other hydrogen atoms were positioned geometrically and were also
constrained to ride on their parent atoms by means of C–H 0.98–
0.99 Å and Uiso = 1.2–1.5 Uiso(parent atom). The structure of 12 is
known and has been previously published in the early 1960s by
Dahl and Wampler.[35] The crystallographic details are summarized
in Tables 2, 3, 4 and 5. The thermal ellipsoid plots and selected
bond lengths and angles for structures 1, 2a, 3, 4, 8a and 9–12 are
given in Figures 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 and Tables 6 and
7; for the other structures, see the Supporting Information. CCDC-
630761 to -630775 (1–12, respectively) contain the supplementary
Table 5. Crystallographic data for 10–12.
10
11
12
Empirical formula C5H6Cl2O4Os
C6I4O6Ru2
877.80
100(2)
0.71073
Monoclinic
C2/m
11.0765(4)
7.6845(3)
9.6081(4)
90
C4I2O4Ru
466.91
110(2)
0.71073
Monoclinic
C2/c
7.1062(3)
10.9172(4)
12.5486(3)
90°
92.399(3)°
90
972.66(6)
4
3.188
7.927
Formula mass
Temperature [K]
λ [Å]
Crystal system
Space group
a [Å]
391.20
120(2) K
0.71073 Å
Monoclinic
P21/n
6.0129(5)
16.5884(8)
9.7260(4)
90
b [Å]
c [Å]
α [°]
β [°]
96.584(5)
90
963.72(10)
4
2.696
13.758
7892
107.759(2)
90
778.85(5)
2
3.743
9.877
γ [°]
V [Å3]
Z
ρcalcd. [Mg/m3]
µ(Mo-Kα) [mm–1
]
No. of reflections
4707
4190
952
0.0228
0.0180
0.0336
Unique reflections 2197
Rint 0.0469
832
0.0284
0.0132
0.0282
[a]
R1 (IՆ2σ)
0.0274
0.0379
[b]
wR2 (IՆ2σ)
[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|. [b] wR2 = {Σ[w(Fo – Fc2)2]/Σ[w(Fo ) ]}1/2
.
2
2 2
Table 7. Selected bond lengths [Å] and bond angles [°] of the com-
plexes 8a and 9–12 presented in Figures 8, 9, 10, 11, and 12.
Table 6. Selected bond lengths [Å] and bond angles [°] of the com-
plexes 1, 2a, 2b, 3, 4, presented in Figures 4, 5, 6, and 7.
Bond length
8a
9
10
11
12
Bond length [Å]
1
2a
2b
3
4
[Å]
C1–M1
C2–M1
C3–M1
C4–M2
C5–M2
C6–M2
X1–M1
X2–M1
X3–M1
O4–M1
X4–M2
X5–M2
X6–M2
X1–MЈ1
X2–MЈ1
X4–MЈ1
X5–MЈ1
O4–MЈ1
O5–MЈ1
O6–MЈ1
O7–MЈ1
O8–MЈ1
1.900(3) 1.909(4)
1.916(3) 1.899(5)
1.908(3) 1.901(4)
1.916(3)
1.8994
1.904(3)
1.911(4)
1.908(3)
1.911(4)
1.907(2)
1.921(2)
1.901(2)
1.913(6)
1.908(6)
1.916(6)
C1–M1
C2–M1
C3–M1
C4–M2
C5–M2
C6–M2
X1–M1
X2–M1
X3–M1
O4–M1
X4–M2
X5–M2
X6–M2
X1–MЈ1
X2–MЈ1
X4–MЈ1
X5–MЈ1
O4–MЈ1
O5–MЈ1
O6–MЈ1
O7–MЈ1
O8–MЈ1
1.897(5)
1.902(5)
1.918(5)
1.905(10) 1.909(5)
1.896(8) 1.892(5)
1.912(9) 1.896(6)
1.907(3)
1.933(4)
1.933(4)
1.982(4)
2.4208(6) 2.4251(10) 2.4166(7) 2.4239(5) 2.7441(5)
2.4275(6) 2.4083(11) 2.4221(9) 2.4105(5) 2.7471(5)
2.4186(6) 2.4129(11) 2.4154(8) 2.4275(6) 2.7357(5)
2.4359(10) 2.436(2) 2.3908(12) 2.7292(3) 2.7332(4)
2.4329(10) 2.429(2) 2.4154(12)
2.4165(10) 2.424(2)
2.7355(4)
2.133(3)
2.4082(8)
2.4284(8)
2.4120(9)
2.4876(10)
2.4646(8)
2.4353(8)
2.4711(9)
2.4844(6)
2.5172(6)
2.514(2)
2.486(2)
2.053(3)
2.0823(16) 2.142(4)
2.0775(16) 2.077(4)
2.0988(16) 2.132(4)
2.078(3)
2.107(3)
2.080(3)
1.996(6)
2.057(3)
2.019(2)
Bond angle [°]
1
2a
2b
3
4
Bond angle [°]
8a
9
10
11
12
C1–M1–C2
C1–M1–X1
C1–M1–C3
C5–M2–C6
C3–M1–X3
C4–M2–X4
C3–M1–O4
X1–M1–X2
X4–Ru2–X5
O4–MЈ1–X1
O4–MЈ1–X2
X1–MЈ1–X2
X4–MЈ1–X5
O4–MЈ1–O5
O4–MЈ1–O6
O5–MЈ1–O6
O8–MЈ1–O7
C1(A)–M1–C1
X1(B)–M1–X1
91.45(11) 91.4(2)
92.07(14) 93.40(9)
93.4(2)
C1–M1–C2
C1–M1–X1
C1–M1–C3
C5–M2–C6
C3–M1–X3
C4–M2–X4
C3–M1–O4
X1–M1–X2
X4–Ru2–X5
O4–MЈ1–X1
O4–MЈ1–X2
X1–MЈ1–X2
X4–MЈ1–X5
O4–MЈ1–O5
O4–MЈ1–O6
O5–MЈ1–O6
O8–MЈ1–O7
C1(A)–M1–C1
X1(B)–M1–X1
92.1(2)
91.2(4)
92.3(2)
178.68(12)
92.90(16)
92.32(15)
179.15(8) 176.85(12) 178.89(11) 174.92(7) 177.5(2)
176.44(10)
177.05(13) 178.7(2)
178.36(11)
173.75(15)
84.95(2) 86.18(3)
85.28(3)
87.15(3)
90.76(2)
87.53(3)
81.22(6) 88.13(4)
90.61(5)
92.02(5)
81.77(2) 84.76(3)
82.89(3)
85.61(3)
92.010(16)
90.88(6)
92.16(7)
89.68(7)
87.40(15)
87.57(15)
89.97(14)
90.26(12)
89.58(12)
92.11(13)
87.32(10)
95.46(15)
85.632(11)
3506
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508

Surface-Assisted Synthesis and Behavior of Dimetallic Mixed-Metal Complexes
FULL PAPER
[9] M. Haukka, M. Jakonen, N. Nivajärvi, M. Kallinen, Dalton
Trans. 2006, 3212–3220.
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
[10] J. Chatt, B. L. Shaw, A. E. Field, Chem. Ind. 1964, 3466–3475,
and ref.[1] cited therein.
Computational Work: Full geometry optimization of the complexes
was performed using the DFT method with a non-local hybrid
functional B3PW91. The Stuttgart-Dresden ECP basis set was used
for the metals (Fe, Co, Ru and Os)[36] and 6-31G* for others (C,
H, O and Cl). The calculations were made with the Gaussian03
program package.[37]
[11] M. L. Berch, A. Davison, J. Inorg. Nucl. Chem. 1973, 35, 3763–
3767.
[12] C. R. Eady, P. F. Jackson, B. F. G. Johnson, J. Lewis, M. C.
Malatesta, J. Chem. Soc. Dalton Trans. 1980, 383–392.
[13] A. J. Deeming, M. Karim, Polyhedron 1991, 10, 837–840.
[14] H. A. Mirza, J. J. Vittal, R. J. Puddephatt, Inorg. Chem. 1993,
32, 1327–1332.
[15] E. Lucenti, E. Cariati, C. Dragonetti, D. Roberto, J. Or-
ganomet. Chem. 2003, 669, 44–47.
[16] M. Fauré, L. Maurette, B. Donnadieu, G. Lavigne, Angew.
Chem. Int. Ed. 1999, 38, 518–522.
Supporting Information (see footnote on the first page of this arti-
cle): Figures S1–S5: Thermal ellipsoid plots (at 50% probability
level) of 2c, 5, 6, 7, 8a and 10; Figures S6 and S7: ATR-IR spectra
of complexes 1, 2, 3 and 8.
[17] M. Fauré, C. Saccavini, G. Lavigne, Chem. Commun. 2003,
1578–1579.
[18] J. P. Collman, W. R. Roper, J. Am. Chem. Soc. 1966, 88, 3504–
3508.
Acknowledgments
[19] M. Aracama, M. A. Esteruelas, F. J. Lahoz, J. A. Lopez, U.
Meyer, L. A. Oro, H. Werner, Inorg. Chem. 1991, 30, 288–293.
[20] S. Dev, J. P. Selegue, J. Organomet. Chem. 1994, 469, 107–110.
[21] a) D. Roberto, R. Psaro, R. Ugo, Organometallics 1993, 12,
2292–2296; b) D. Roberto, R. Psaro, R. Ugo, J. Mol. Catal.
1994, 86, 109–120; c) E. Lucenti, D. Roberto, R. Ugo, Organo-
metallics 2001, 20, 1725–1733; d) D. Roberto, E. Cariati, R.
Psaro, R. Ugo, Organometallics 1994, 13, 734–737; e) D. Ro-
berto, E. Cariati, R. Ugo, R. Psaro, Inorg. Chem. 1996, 35,
2311–2316; f) D. Roberto, E. Cariati, E. Lucenti, M. Respini,
R. Ugo, Organometallics 1997, 16, 4531–4539; g) C. Roveda,
E. Cariati, E. Lucenti, D. Roberto, J. Organomet. Chem. 1999,
580, 117–127.
[22] H. H. Uhlig, R. W. Revie, Corrosion and Corrosion Control, An
Introduction to Corrosion Science and Engineering, 3rd ed.,
John Wiley & Sons, New York, 1985, p. 78–87 and references
cited therein.
[23] P. Hirva, M. Haukka, M. Jakonen, T. A. Pakkanen, Inorg.
Chim. Acta 2006, 359, 853–862.
Financial support in the form of a grant (grant no. 210265) pro-
vided by the Academy of Finland is gratefully acknowledged. The
authors would also like to thank Dr. Sari Suvanto for the EDS and
ATR measurements, Ritva Romppanen for the elemental analysis
measurements, and Dr. Larisa Oresmaa for her valuable comments.
[1] a) R. Brittain, A. L. Hickman, Brit. UK Pat. Appl.
20504727 19810107, 1981; b) L. Bertolini, M. Gastaldi, M. P.
Pedeferri, E. Redaelli, Corros. Sci. 2002, 44, 1497–1513; c) S.
Paul, K. Basu, P. Mitra, Bull. Electrochem. 2005, 21, 269–273;
d) G. Glass, PCT Int. Appl. WO 2006003473, 2006.
[2] a) A. M. Vecchio-Sadus, J. Appl. Electrochem. 1993, 23, 401–
416; b) M. T. Reetz, M. Winter, R. Breinbauer, T. Thurn-Al-
brecht, W. Vogel, Chem. Eur. J. 2001, 7, 1084–1094.
[3] A. Shantaram, H. Beyenal, R. R. A. Veluchamy, Z. Lewan-
dowski, Environ. Sci. Technol. 2005, 39, 5037–5042.
[24]
M. I. Bruce, C. M. Jensen, N. L. Jones, Inorg. Synth. 1989, 26,
259–261.
[4] a) M. A. Méndez-Rojas, F. Cordova-Lozano, G. Gojon-Zor-
illa, E. González-Vergara, M. A. Quiroz, Polyhedron 1999, 18,
2651–2658; b) B. I. Kharisov, L. M. Blanco, A. D. Garnovskii,
A. S. Burlov, L. I. Kuznetsova, L. V. Korovina, D. A. Garnov-
skii, T. Dieck, Polyhedron 1998, 17, 381–389; c) J. M. Paratian,
E. Labbé, S. Sibile, J. Y. Nédélec, J. Périchon, J. Organomet.
Chem. 1995, 487, 61–64; d) J. Sánchez-Piso, A. Carcía-
Vázquez, J. Romero, M. L. Durán, A. Sousa-Pedras, E. Labis-
bal, O. R. Nascimento, Inorg. Chim. Acta 2002, 328, 111–122;
e) H. Fillon, C. Gosmini, J.-Y. Nédélec, J. Périchon, Tetrahe-
dron Lett. 2001, 42, 3843–3846.
[25]
Z. Otwinowski, W. Minor, in Methods in Enzymology, vol. 276
(“Macromolecular Crystallography”), part A (Eds.: C. W. Car-
ter, J. Sweet), Academic Press, New York, USA, 1997, pp. 307–
326.
[26]
[27]
A. J. M. Duisenberg, L. M. J. Kroon-Batenburg, A. M. M.
Schreurs, J. Appl. Crystallogr. 2003, 36, 220–229.
G. M. Sheldrick, SHELXS-97, Program for Crystal Structure
Determination, University of Göttingen, Göttingen, Germany,
1997.
[28]
[29]
A. Altomare, M. C. Burla, M. C. Camalli, C. Giacovazzo, A.
Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
M. C. Burla, M. Camalli, B. Carrozzini, G. L. Cascarano, C.
Giacovazzo, G. Polidori, R. Spagna, J. Appl. Crystallogr. 2005,
38, 381–388.
[5] J. Castro, S. Cabaleiro, P. Pérez-Lourido, J. Romero, J. A. Gar-
cia-Vásquez, A. Sousa, Z. Anorg. Allg. Chem. 2002, 628, 1210–
1217.
[6] a) M. Faticanti, N. Cioffi, S. De Rossi, N. Ditaranto, P. Porta,
L. Sabbatini, T. Bleve-Zacheo, Appl. Catal., B 2005, 60, 73–82;
b) N. Cioffi, N. Ditaranto, L. Torsi, R. A. Picca, E. De Giglio,
L. Sabbatini, L. Novello, G. Tantillo, T. Bleve-Zacheo, P. G.
Zambonin, Anal. Bioanal. Chem. 2005, 382, 1912–1918.
[7] a) E. Duñach, S. Dérien, J. Périchon, J. Organomet. Chem.
1989, 364, C33–C36; b) C. Moreau, F. Serein-Spirau, M. Bor-
deau, C. Biran, J. Organomet. Chem. 1998, 570, 147–154; c) T.-
Y. Lin, T.-C. Chou, J. Appl. Electrochem. 1999, 29, 489–496; d)
C. Gosmini, J. Y. Nédélec, J. Périchon, Tetrahedron Lett. 2000,
41, 5039–5042; e) A. A. Isse, A. Gennaro, J. Electrochem. Soc.
2002, 149, D113–D117; f) A. Palma, B. A. Frontana-Uribe, J.
Cárdenas, M. Saloma, Electrochem. Commun. 2003, 5, 455–
459; g) H. Lund, H. Svith, S. U. Pedersen, K. Daasbjerg, Elec-
trochim. Acta 2005, 51, 655–664; h) O. Scialdone, A. Galia, C.
La Rocca, G. Filardo, Electrochim. Acta 2005, 50, 3231–3242.
[8] L. Mond, C. Langer, F. Quincke, J. Chem. Soc. Trans. 1890,
57, 749–753.
[30]
[31]
L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837.
G. M. Sheldrick, SHELXTL, v. 6.14-1, Bruker Analytical X-
ray Systems, Bruker AXS, Inc., Madison, Wisconsin, USA,
2005.
[32]
[33]
G. M. Sheldrick, SADABS, Bruker Nonius scaling and absorp-
tion correction, v. 2.10, Bruker AXS, Inc., Madison, Wisconsin,
USA, 2003.
G. M. Sheldrick, SHELXL-97, Program for Crystal Structure
Refinement, University of Göttingen, Göttingen, Germany,
1997.
[34]
[35]
M. Nardelli, J. Appl. Crystallogr. 1999, 32, 563–571.
L. F. Dahl, D. F. Wampler, Acta Crystallogr. 1962, 15, 946–
950.
[36]
The basis set was obtained from the Extensible Computational
Chemistry Environment Basis Set Database, version 9/12/01,
Eur. J. Inorg. Chem. 2007, 3497–3508
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3507

M. Jakonen, P. Hirva, T. Nivajärvi, M. Kallinen, M. Haukka
FULL PAPER
as developed and distributed by the Molecular Science Com-
puting Facility, Environmental and Molecular Sciences Labo-
ratory, which is part of the Pacific Northwest Laboratory, P.
O. Box 999, Richland, Washington 99352, USA, and funded
by the U. S. Department of Energy. The Pacific Northwest
Laboratory is a multi-program laboratory operated by the Bat-
telle Memorial Institute for the U. S. Department of Energy
under contract DE-AC06-76RLO 1830. Contact David Feller
or Karen Schuchardt for further information.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morok-
uma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzew-
ski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V.
Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A.
Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian 03, re-
vision C.02, Gaussian, Inc., Wallingford, CT, 2004.
[37] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T.
Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Received: March 27, 2007
Published Online: June 5, 2007
3508
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3497–3508