Substituent-induced reactivity in quinonoid-bridged dinuclear complexes: Comparison between the ruthenium and osmium systems

Article abstract of DOI:10.1021/om300939x

The ligand 2,5-bis[2-(methylthio)anilino]-1,4-benzoquinone (L) was used in its doubly deprotonated form to synthesize the complex [{Cl(η⁶- Cym)Os}₂(μ-η²:η²-L_-2H)] (1; Cym = p-cymene = 1-isopropyl-4-methylbenzene). Spectroscopic characterization and elemental analysis confirms the presence of the chloride ligands in 1, which indirectly shows that the bridging ligand L_-2H acts in a bis-bidentate fashion in 1, with the thioether substituents on the bridge remaining uncoordinated. Abstraction of the chloride ligands in 1 by AgBF₄ in CH₃CN leads not only to the release of those chloride ligands but also to a simultaneous substituent-induced release of Cym with the bridging ligand changing its coordination mode to bis-tridentate. In the resulting complex [{(CH₃CN)₃Os}₂(μ- η³:η³-L_-2H)]²⁺ (2 ²⁺), the thioether groups of L_-2H are now coordinated to the osmium centers with the bridging ligand coordinating to the metal center in a bis-meridional form. The coordination mode of L_-2H in 2 ²⁺ was confirmed by single-crystal X-ray diffraction data. A structural analysis of 2²⁺ reveals localization of double bonds within the upper and lower parts of the bridging ligand in comparison to bond distances in the free ligand. Additionally, the binding of the bridge to the osmium centers is seen to occur through O^- and neutral imine-type N donors. The complexes 1 and 2²⁺ were investigated by cyclic voltammetry and UV-vis-near-IR and EPR spectroelectrochemistry. This combined approach was used to unravel the redox-active nature of the ligand L_-2H, to determine the sites of electron transfer (ligand radical versus mixed valency), and to compare the present systems with their ruthenium analogues 3 and 4²⁺ (Schweinfurth, D.Inorg. Chem. 2011, 50, 1150). The effect of replacing ruthenium by its higher homologue osmium on the reactivity and the electrochemical and spectroscopic properties were explored, and the differences were deciphered by taking into account the intrinsic dissimilarities between the two homologues. The usefulness of incorporating additional donor substituents on potentially bridging quinonoid ligands was probed in this work.

Full text of DOI:10.1021/om300939x

Article
pubs.acs.org/Organometallics
Substituent-Induced Reactivity in Quinonoid-Bridged Dinuclear
Complexes: Comparison between the Ruthenium and Osmium
Systems
Michael G. Sommer,†,§ David Schweinfurth,‡,§ Fritz Weisser,‡ Stephan Hohloch,‡ and Biprajit Sarkar*,‡
†Institut fur Anorganische Chemie, Universitat Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany
̈
̈
̈
‡Institut fur Chemie und Biochemie, Anorganische Chemie, Fabeckstraße 34-36, D-14195 Berlin, Germany
S
* Supporting Information
ABSTRACT: The ligand 2,5-bis[2-(methylthio)anilino]-1,4-
benzoquinone (L) was used in its doubly deprotonated form
to synthesize the complex [{Cl(η6-Cym)Os}2(μ-η2:η2-L‑2H)]
(1; Cym = p-cymene = 1-isopropyl-4-methylbenzene).
Spectroscopic characterization and elemental analysis confirms
the presence of the chloride ligands in 1, which indirectly
shows that the bridging ligand L‑2H acts in a bis-bidentate
fashion in 1, with the thioether substituents on the bridge
remaining uncoordinated. Abstraction of the chloride ligands
in 1 by AgBF4 in CH3CN leads not only to the release of those
chloride ligands but also to a simultaneous substituent-induced
release of Cym with the bridging ligand changing its
coordination mode to bis-tridentate. In the resulting complex
[{(CH3CN)3Os}2(μ-η3:η3-L‑2H)]2+ (22+), the thioether groups of L‑2H are now coordinated to the osmium centers with the
bridging ligand coordinating to the metal center in a bis-meridional form. The coordination mode of L‑2H in 22+ was confirmed
by single-crystal X-ray diffraction data. A structural analysis of 22+ reveals localization of double bonds within the “upper” and
“lower” parts of the bridging ligand in comparison to bond distances in the free ligand. Additionally, the binding of the bridge to
the osmium centers is seen to occur through O− and neutral imine-type N donors. The complexes 1 and 22+ were investigated by
cyclic voltammetry and UV−vis−near-IR and EPR spectroelectrochemistry. This combined approach was used to unravel the
redox-active nature of the ligand L‑2H, to determine the sites of electron transfer (ligand radical versus mixed valency), and to
compare the present systems with their ruthenium analogues 3 and 42+ (Schweinfurth, D. et al. Inorg. Chem. 2011, 50, 1150).
The effect of replacing ruthenium by its higher homologue osmium on the reactivity and the electrochemical and spectroscopic
properties were explored, and the differences were deciphered by taking into account the intrinsic dissimilarities between the two
homologues. The usefulness of incorporating additional donor substituents on potentially bridging quinonoid ligands was probed
in this work.
that, we can generate new ligands where the properties of the
metal complexes can be easily varied by using the handle “R” on
the [NR] groups. Using this concept, we have reported on
synthetic strategies for generating a host of “meta”5a and
“para”5b,c forms of ligands derived from dhbq, where two of the
[O] groups have been replaced by [NR] groups (Scheme 1).
Such ligands have been used by us and others for generating
metal complexes with the goal of studying electron transfer,5,7
magnetism,5b,8 metal−metal coupling,7a−g homogeneous catal-
ysis,9 and supramolecular interactions.6c,10 Ruthenium com-
plexes with quinone and other ligands have been at the
forefront of studies on electron transfer and mixed valency.7
One interesting aspect that we observed during the course of
developing synthetic routes to such ligands is the ease of
INTRODUCTION
■
Quinones as redox-active compounds are fascinating ligands in
coordination and organometallic chemistry.1 The ability of
quinones and related ligands to exist in various redox forms
and, hence, to undergo a series of electron and charge transfer
processes has long attracted synthetic and physical inorganic
chemists to these systems.2 Recent developments have involved
the use of quinones as electron reservoirs in their metal
complexes to perform small-molecule activation and catalysis.3
The prototype of a quinone ligand capable of bridging two
metal centers is 1,4-dihydroxy-2,5-benzoquinone (dhbq;
Scheme 1). This all oxygen donor containing ligand has been
extensively used in coordination chemistry, and its metal
complexes show many interesting properties.4 We5 and others6
have argued in recent years that the tuning of steric and
electronic properties of ligands such as dhbq is best achieved by
using the [NR] for [O] isoelectronic relationship. On doing
Received: October 6, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Modification of Potentially Bridging Quinones: from [O,O,O,O] through [O,N,O,N] to a Ligand with Additional
Donors on the Nitrogen Substituents
Scheme 2. Synthesis of Diosmium Complexes 1 and 2[BPh4]2
introducing an additional donor group in the handle “R” in
[NR].5c Such a group can be a donor atom capable of
coordinating on demand during a chemical process and, hence,
acting as a hemilabile group.11 Reports of potentially bridging
quinone ligands with additional donating groups on the
substituents remain rare in the literature.6c,9e In the present
study we have used the ligand L (L = 2,5-bis[2-(methylthio)-
anilino]-1,4-benzoquinone) to generate the complexes [{Cl(η6-
Cym)Os}2(μ-η2:η2-L‑2H)] (1) and [{(CH3CN)3Os}2(μ-η3:η3-
us.5c Complex 1 was synthesized by reacting [{(Cym)-
ClOs}2(μ-Cl)2]13 with L in the presence of a base at room
temperature (Scheme 2). This reaction worked in a
straightforward manner, with no complicated purification
steps required, and resulted in a product yield of 85%
(Experimental Section). The yield and the reaction conditions
are very similar to those we recently reported for 3.5c
1
In the H NMR spectrum of 1, the signals corresponding to
the alkyl and the aryl protons of the Cym ligand are found in
the expected region (Experimental Section). Additionally, the
signal corresponding to the methyl groups of SMe appears at
2.58 ppm. This is very similar to the value of 2.45 ppm found
for the same protons in the free ligand.5c The C−H ring proton
of the substituted quinone ring appears at 4.93 ppm; the
corresponding value for the ruthenium analogue 3 is 4.78
ppm.5c In the mass spectrum a peak corresponding to [1 −
2Cl−]2+ was observed. Elemental analysis data match perfectly
with the formulation of 1. All these data point to the presence
of the chlorides as well as the Cym ligands in 1. Thus, it can be
concluded that the ligand L‑2H acts in a bis-bidentate fashion in
1, coordinating to the osmium center through the oxygen and
nitrogen atoms. The complex 1 can in principle exist in the syn
or anti form (it is the orientations of the Cym and chloride
ligands that define syn or anti). Despite several attempts, we
were not able to grow crystals suitable for single-crystal X-ray
diffraction analysis. Hence, we cannot comment on the
configuration around the metal centers in 1 with certainty.
However, we note that, in the ruthenium analogue 3, the
molecule with the chloride ligands in syn positions had
L
‑2H)]2+ (22+). By using the same ligand, we were able to show
that the initially formed complex [{Cl(η6-Cym)Ru}2(μ-η2:η2-
‑2H)] (3; Cym = p-cymene = 1-isopropyl-4-methylbenzene)
L
releases its Cym ligands as a result of chloride abstraction and
thioether coordination in an unprecedented reaction, resulting
in the formation of [{(CH3CN)3Ru}2(μ-η3:η3-L‑2H)]2+ (42+).5c
We discuss here the role that the thioether group plays in
displacing the Cym ligand on going from 1 to 22+ and compare
the reaction times and reactivity patterns of the ruthenium
5c
complexes 3 and 42+ with those of their osmium-containing
homologues 1 and 22+. Investigations of the complexes through
single-crystal X-ray diffraction studies, electrochemistry, and
UV−vis−near-IR and EPR spectroelectrochemistry will be
reported. The complexes will be discussed with respect to their
reactivity and structural, electrochemical, and spectroelectro-
chemical properties with focus on the effects brought about by
replacing ruthenium with osmium. It should be noted here that
reports on dinculear complexes of osmium with bridging
ligands in general, and quinone bridges in particular, are rare in
the literature in comparison to their ruthenium counterparts.12
Reasons for this are the often tricky synthesis of osmium
complexes in comparison to their ruthenium analogues. In this
report we present the synthesis and detailed structural,
electrochemical, and spectroscopic characterization of two
well-defined dinuclear osmium complexes.
preferentially crystallized.5c In the present case for 1, the H
1
NMR spectrum clearly showed the formation of a single isomer
(Figure S1, Supporting Information).
We next tried to use AgBF4 to abstract the chloride ions from
1. Such a procedure has been used by several groups to
generate solvent-coordinated (substitution of Cl− with solvent
molecules), quinone-bridged, dinuclear complexes with d6
metal centers, which are reacted with additional bridging
ligand, leading to solvent displacement and formation of
RESULTS AND DISCUSSION
■
Synthesis, Characterization, and Crystal Structure of
22+. The ligand L was prepared by a route recently reported by
B
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
molecular squares.4f,10 All such ligands usually contain no
additional donor atoms other than those directly bound to the
quinone ring (usually acting as a bis-bidentate ligand). In the
equivalent to each other. The Os1−S1 distance of 2.283(1) Å
lies in the range of an authentic bond. The S atom of the SMe
group is thus coordinated to the osmium center in 22+. The
bridging ligand binds to each of the osmium centers in a bis-
tridentate form with the O, N, and S atoms at each metal center
coordinating in a meridional (mer) fashion. The remaining
three coordination sites are taken up by the acetonitrile
molecules. A look at the bond lengths within the quinone
ligand in 22+ and a comparison of these with the bond lengths
1
present case, H NMR analyses of the product formed after
chloride abstraction showed no trace of signals corresponding
to the Cym protons (Figure S2, Supporting Information). The
SMe protons, which appear at 2.45 and 2.58 ppm, respectively,
for L5c and 1, show up at 2.87 ppm for this new product. The
C−H quinone ring proton is seen at 6.92 ppm. Additionally,
two sets of signals (axial and equatorial) corresponding to
coordinated acetonitrile are observed between 2.42 and 2.44
ppm and at 2.71 ppm. These data point to the release of Cym
during the chloride abstraction process, and the coordination of
the thioether groups (uncoordinated in 1) and of acetonitrile
molecules to the metal centers. Mass spectral analysis of the
product also confirmed the formation of a Cym-free compound
(Experimental Section). Having established the similar nature
of the ruthenium and osmium compounds, a look at reaction
conditions points to some differences in the two cases. Whereas
the conversion of 3 to 42+ required only 3 h of reflux in
acetonitrile and led to 85% product formation,5c the reaction of
1 to give the new product 22+ required 19 h of reflux in
acetonitrile and delivered the product in only 22% yield
(Experimental Section). Osmium is known to form much
stronger bonds to ligands in comparison to ruthenium.12 This
would explain the need for much longer reaction times in the
present case in comparison to what we have reported for the
ruthenium compounds. Such drastic conditions probably also
lead to the generation of more byproducts, which decrease
product yields. Similar problems with the synthesis of osmium
compounds are known in the literature.
5c
of the free ligand and of 42+ show some interesting trends
(Table 1). The C1−O1 bond length of 1.309(6) Å in 22+ is
Table 1. Bond Lengths (in Å)
L
42+
22+
C1−C2
C2−C3
C1−C3
C1−O1
C3−N1
M−O1
M−N1
M−S1
M−N100
M−N200
M−N300
1.432(2)
1.365(2)
1.524(2)
1.233(2)
1.344(2)
1.383(7)
1.392(9)
1.50(1)
1.385(7)
1.400(7)
1.488(6)
1.309(6)
1.360(6)
2.078(4)
2.009(4)
2.283(1)
2.012(5)
2.022(4)
1.999(5)
1.288(8)
1.335(6)
2.076(4)
2.017(5)
2.289(2)
2.013(5)
2.051(6)
2.005(5)
longer than that in the free ligand L at 1.233(2) Å. The C1−C2
bond at 1.385(7) Å is, however, shorter than 1.432(2) Å for the
same bond in L. The C2−C3 bond lengths for 22+ and L are
1.400(7) and 1.365(2) Å, respectively, showing a reasonable
elongation of this bond on metal coordination. Thus, whereas
in the free ligand L the bonds are largely delocalized in the
“upper” and “lower” parts of the molecule, metal coordination
seems to localize these bonds. Thus, the coordination of the
ligand is best described through an O− type donor and an imine
type neutral nitrogen N donor. The short C1−C2 bond next to
the C−O bond and the slightly longer C2−C3 bond next to the
C−N bond would support this hypothesis. Coordination
through an O− donor and a neutral N donor was also observed
for 42+ and for other complexes containing similar ligands. The
more electronegative O atom apparently is better suited to
stabilize the negative charge.5b,c,7c,g The C1−C3 bond which
connects the “upper” and “lower” parts of the quinone ligand
remains a long single bond in L, 22+, and 42+ (Table 1).
The quinone ring and the O and N donor atoms are largely
planar. However, the two thioether groups are displaced out of
the plane with a dihedral angle of 24.5(2)° between the
quinone OCCN and NCCS planes in 22+. The corresponding
dihedral angle for 42+ is 26.3(2)°.5c The intramolecular Os−Os
distance in 22+ is 7.8501(8) Å. The corresponding distance in
the ruthenium complex 42+ is 7.8533(8) Å.5c
Final proof of the composition of the product 22+, formed
after chloride abstraction from 1, came from a structure
determination through single-crystal X-ray diffraction analysis.
−
−
Anion exchange (BPh4 for BF4 ) was necessary to obtain
good-quality single crystals for this compound (Experimental
Section). 2[BPh4]2 crystallizes with two toluene solvate
molecules in the triclinic P1 space group (Table S1, Supporting
̅
Information). One of the toluene solvent molecules is highly
disordered at the measurement temperature of 100 K.
Each of the osmium centers in 22+ is coordinated in a
distorted-octahedral environment by the O, N, and S donors of
the ligand L‑2H and three acetonitrile molecules (Figure 1). The
distortion is imposed by the chelating nature of the ligand L‑2H
.
The molecule has a local center of inversion, making the halves
Comments on the Release of Cym from 1 To Form
22+. The processes of chloride abstraction, thioether and
acetonitrile coordination, and Cym release from 1 to produce
22+ seem to be strongly correlated. Reports by us5c and
others4f,10 have proven that it is possible to abstract chloride
atoms from related complexes that contain no additional donor
atoms in the bridging quinones and generate solvent
-substituted complexes that still have Cym as ligands. The
detection of the [1 − 2Cl−]2+ peak in the mass spectrometric
experiments with 1 shows that a simple chloride abstraction is
not enough for Cym release and thioether coordination in the
Figure 1. ORTEP plot of 2[BPh4]2·2C7H8. Ellipsoids are drawn at the
50% probability level. Toluene and tetraphenylborate molecules have
been omitted for clarity.
C
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
gas phase. Treatment of 1 with 2 equiv of Ag+ triggers a
sequence of consecutive events. Chloride abstraction from
osmium increases the Lewis acidity of the metal ions, which can
be met by binding of the dangling thioether donors. The Cym
ligand bound to osmium in a η6 fashion requires that the other
three donor atoms bind to the metal center in a facial (fac)
form.14 This, however, requires a large amount of bending
around various bonds with partial double-bond character. The
bridging ligand is not flexible enough to allow for the
corresponding distortion. Thus, the energetically feasible way
for the ligand to bind with all of its O, N, and S donors to the
metal center is in a mer form. However, once the bridging
quinone binds to the metal center in a mer form, the
coordination of Cym is no longer possible. The Cym ligands
are thus released from the molecule with concomitant
coordination of the corresponding number of acetonitrile
donors.
Table 2. Electrochemical Potentials of the Complexes
b
Ox2
Ox1
Red1
Red2
compd
E1/2
0.37
0.77
E1/2
Kc
E1/2
E1/2
22+
−0.11
0.36
1.3 × 108
7.9 × 106
−1.29
−1.15
−1.73
n.o.
c
d
42+,
a
Half-wave potentials in V from cyclic voltammetry in CH2Cl2/0.1 M
Bu4NPF6 at 298 K. Scan rate: 100 mV/s. Decamethylferrocene/
decamethylferrocenium was used as an internal standard, and the
potentials were referenced against ferrocene/ferrocenium by using
literature values.26 Kc = 10ΔE/59, ΔE = E1/2 − E1/2Ox1. Data from
b
c
Ox2
d
ref 5c, corrected with appropriate referencing. n.o. = not observed.
appropriate plot is shown in Figure 3. The osmium complex 22+
displays two reversible oxidation and two reversible reduction
processes in CH2Cl2/0.1 M Bu4NPF6 (Table 2 and Figure 3).
The necessity of such a donating solvent is clear from mass
spectrometric experiments of 1, where despite chloride
abstraction and a possible increase in the Lewis acidity of the
metal center, no Cym release occurs in the gas phase. Attempts
to perform the reaction of 1 with a Ag+ source in
dichloromethane/toluene in order to generate a coordinatively
unsaturated metal complex where each osmium center would
have 16 valence electrons were not successful.
Electrochemistry. The presence of a redox-active bridging
ligand, together with two redox-active metal centers, makes the
complexes reported herein ideal candidates for probing their
redox properties. The complex 1 shows a reversible oxidation at
0.37 V and an irreversible reduction step at −1.51 V (Figure 2)
in CH2Cl2/0.1 M Bu4NPF6.
Figure 3. Cyclic voltammograms of 42+ (top) and 22+ (bottom) in
CH2Cl2/0.1 M Bu4NPF6 at 295 K.
A comparison between the potentials of 22+ and 42+
measured under identical conditions shows some interesting
trends. A look at Table 2 shows that all potentials for 22+
containing the higher homologue osmium are cathodically
shifted compared to 42+. This is to be expected because of the
better π-donor ability of osmium in comparison to
ruthenium.17 However, the shifts of the potentials for the
individual redox processes are not uniform. Thus, whereas the
first oxidation potential for 22+ is −0.11 V, the corresponding
value for 42+ is 0.36 V; a look at the first reduction potentials
shows the values for 22+ and 42+ at −1.29 and −1.15 V,
respectively. This large shift for the oxidation potentials and
only a small change in the reduction potentials on moving from
22+ to 42+ are the first indications of greater metal contribution
to the oxidation processes in comparison to the reduction
processes (see below).
The difference in half-wave potentials of the two oxidation
processes for 22+ is larger than that for 42+. This larger potential
difference translates into a comproportionation constant, Kc,18
about 2 orders of magnitude higher for 22+ in comparison to
42+ (Table 2). Since electrochemical measurements for both the
complexes were carried out in the same solvent and electrolyte
under identical conditions, the difference in the thermodynamic
stabilities of the one-electron-oxidized species can be attributed
Figure 2. Cyclic voltammogram of 1 in CH2Cl2/0.1 M Bu4PF6 at 295
K.
For comparison, 3, which contains ruthenium centers instead
of osmium, showed only ill-defined, irreversible processes
under identical conditions. Quinoidally bridged arene−
ruthenium complexes are known to show irreversible electro-
chemical responses.15 Reasons for this are the labile
ruthenium−chloride bonds, which are highly susceptible to
cleavage following redox processes. For 1, which contains the
higher homologue osmium, the metal−ligand bonds are
intrinsically stronger; this is a well-known effect on moving
down in a group in the periodic table. A consequence of such
robust bonds is the reversibility of the oxidation step in 1. The
reduction is irreversible, probably because of chloride
dissociation after reduction of complex 1. Such chloride
splitting following reduction has been observed before for
arene-containing complexes of a d6 metal center with redox-
active ligands.16
For the cymene-free ruthenium complex 42+ we had made an
error in referencing in our earlier report.5c The correct
potentials for this complex are given in Table 2, and an
D
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
to electronic differences between the two complexes.
Considering a superexchange mechanism, valence exchange
between the metal centers would take place via the π* orbital of
the bridging ligand. Stronger binding of the Os(II) and Os(III)
ions to the bridging ligand enforce donor−bridge−acceptor
interactions and stabilize the mixed-valence state to a greater
degree than in the case of ruthenium.17
UV−Vis−Near-IR and EPR Spectroelectrochemistry.
The Ru−arene complex 3 displays an absorption band at 526
nm (Table 3), which is tentatively assigned to a Ru (dπ) to L‑2H
cm−1. Since 1 does not display any other oxidation waves
within the dichloromethane solvent window, it is not possible
to probe if this band might disappear after a second oxidation.
The results of DFT calculations (detailed below) nevertheless
provide evidence for strong electronic coupling. The in situ
electrochemically generated one-electron-oxidized form 11+
turned out to be EPR silent down to liquid-nitrogen
temperatures. This can be taken as an evidence of a metal-
centered oxidation, because a heavy 5d element such as
osmium, with its large spin−orbit coupling constant (3000
cm−1 for OsIII),20 would lead to fast relaxation pathways and
broadening of EPR lines beyond detection.21 By putting
together the results described above, 1+ can be tentatively
assigned as a mixed-valence species. The experimental Δν1/2
value for the band at 1275 nm is 964 cm−1. On using the Hush
formulation √(2310νmax),22 the calculated Δν1/2 value turns
out to be 4256 cm−1. The larger calculated bandwidth at half-
height in comparison to the experimental value indicates that 1+
belongs to the strongly coupled class III mixed-valent systems
in the Robin and Day classification scheme. Class III character,
despite the weak intensity of near-IR bands, has previously been
observed for other systems.12a
a
Table 3. UV−Vis−Near-IR Data of the Complexes
compd
λmax/nm (ε/M−1 cm−1
)
1+
393 (1550), 590 (1330), 731 (2800), 1275 (470)
394 (1580), 582 (2950)
1
24+
23+
22+
2+
3
519 (6800), 682 (12900), 908 sh, 1657 (1090)
441 (4630), 792 (15200), 1251 (1030), 1568 (770), 1983 (910)
400 (6420), 721 (10900), 793 (11500), 884 (10700), 1129 sh
400 (6190), 549 (4790), 927 (6960), 1133 sh
327 (12400), 433 sh, 526 (20100)
b
44+
528, 760
b
43+
390, 524, 716, 830 sh, 1287, 1674 sh
369, 661 sh, 707
b
42+,
42+ displays a main MLCT band at 707 nm (Table 3 and
Figure 5). On moving to the osmium analogue 22+, the MLCT
band(s) is shifted to lower energy (Table 3 and Figure 6), as
has been observed for 1 in comparison to 3. The MLCT band
is split into three different bands for both complexes. The effect
b
4+,
327, 429, 558, 737 sh, 858, 964 sh
a
Measurements in CH2Cl2/0.1 M Bu4NPF6 (OTTLE spectroelec-
b
trochemistry). Extinction coefficients could not be determined
accurately because of the poor solubility of the compound.
(π*) metal to ligand charge transfer (MLCT) transition. On
moving to the osmium analogue 1, this band is shifted to 582
nm (Table 3 and Figure 4). The shift of the MLCT band of 1
Figure 4. Changes in the UV−vis−near-IR spectrum of 1 in CH2Cl2/
0.1 M Bu4NPF6 during the first oxidation.
to lower energy in comparison to 3 can be rationalized by the
better π-donor properties of Os(II) in comparison to Ru(II),
which would lead to a destabilization of the metal-based orbitals
in 1 in comparison to 3.17 Both 1 and 3 show other bands at
higher energy (Table 3) that are intraligand in origin. The
completely irreversible nature of the cyclic voltammogram of 3
precluded the spectroscopic characterization of the other redox
states of 3.
On one-electron oxidation of 1 to 1+ using an optically
transparent thin-layer electrochemical (OTTLE) cell,19 the
MLCT band at 582 nm loses intensity, and a new band appears
at 731 nm (Figure 4 and Table 3). This band is assigned to a
ligand to metal charge transfer (LMCT) transition. A redox-
active ligand such as L‑2H is expected to have filled orbitals close
in energy to the metal-based orbitals, and hence this assignment
seems plausible. In addition, 1+ shows a band in the near-IR
region at 1275 nm with an extinction coefficient of 470 M−1
Figure 5. Changes in the UV−vis−near-IR spectrum of 42+ in
CH2Cl2/0.1 M Bu4NPF6 during the (a) first oxidation (the inset shows
changes in the region 1000−2250 nm), (b) second oxidation, and (c)
first reduction.
E
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
equation (√(2310νmax),22 the calculated Δν1/2value is 4236
cm−1, indicating that 43+ belongs to the strongly coupled class
III mixed-valent species [Ru(2.5)−Ru(2.5)].
When the osmium complex is oxidized to 23+, several bands
appear in the visible and near-IR region. For a mixed-valent
formulation, apart from the MLCT or LMCT bands, several
interconfiguration and intervalence charge transfer bands are
possible and allowed for osmium complexes, as has been
pointed out in an excellent review article.23 Of the various
bands in the near-IR region, that at 1251 nm (Table 3 and
Figure 6) disappears on further oxidation to 24+. 23+ is EPR
silent at room temperature. At 110 K, 23+ shows an EPR signal
with an average g value of 2.140 and a reasonably large g
anisotropy of 0.440 (Table 4 and Figure 7). Both the average g
Figure 6. Changes in the UV−vis−near-IR spectrum of 22+ in
CH2Cl2/0.1 M Bu4NPF6 during the (a) first oxidation (the inset shows
changes in the region 1000−2250 nm), (b) second oxidation, and (c)
first reduction.
Figure 7. X-band EPR spectrum of in situ electrochemically generated
23+ at 110 K (top) and 2+ at 295 K (bottom) in CH2Cl2/0.1 M
Bu4NPF6.
is more prominent for the osmium complex, possibly due to a
larger vibronic coupling in the osmium complex 22+.12c,23
On one-electron oxidation to 43+, the band at 707 nm loses
intensity and a new band appears at 830 nm. This band is
assigned to a LMCT transition. 43+ also displays an absorption
in the near-IR region at 1287 nm. This band disappears on
further oxidation to 44+. As has been reported previously by
us,5c the in situ electrochemically generated species 43+ is EPR
silent at room temperature. At 110 K 43+ shows an EPR signal
with an average g value of 2.065 and a g anisotropy of 0.168
(Table 4). The deviation of the average g value from the free
electron g value of 2.0023 and the substantial g anisotropy are
an indication of spin at the metal centers.24 Taking the UV−
vis−near-IR and EPR data together, 43+ can be assigned as a
mixed-valence diruthenium species. The experimental Δν1/2
value for the band at 1287 nm is 942 cm−1; on using the Hush
value and the g anisotropy are substantially larger for the odd-
electron osmium complex 23+ in comparison to its ruthenium
counterpart 43+. These results are probably a combined effect of
more spin density on the metal centers for 23+ as well as the
higher spin−orbit coupling constant for osmium as compared
to ruthenium.20 The appearance of an EPR signal at 110 K for
23+ but not for 1+, despite both being diosmium complexes,
points to different electronic situations and a more facile
relaxation pathway for 1+ in comparison to 23+. The combined
UV−vis−near-IR and EPR investigations suggest that 23+ is a
mixed-valent species. Analysis of the band at 1251 nm reveals
an experimental Δν1/2 value of 480 cm−1; the calculated Δν1/2
value according to Hush22 is 4297 cm−1, indicating that 23+, like
1+ and 43+, belongs to the strongly coupled class III mixed-
valent species.
a
Table 4. EPR Data of the Complexes
b
d
compd
g1
g2
g3
Δg
gav
On one-electron reduction to 4+, the MLCT band is shifted
to higher energy, and a new low-energy band appears at 858
nm, which is assigned to an LMCT transition. 4+ shows a
narrow EPR signal at 295 K centered at g = 1.998, as reported
previously.5c The appearance of the signal at room temperature
and the small deviation of the g value from the free electron g
value are indications of the unpaired electron residing in a
predominantly ligand-centered orbital.24 The one-electron
c
2+
1.982
2.140
1.998
2.065
23+
ce
,
2.355
2.143
2.128
2.074
1.915
1.975
0.440
0.168
4+
e
43+
a
EPR data of in situ generated species in CH2Cl2/0.1 M Bu4NPF6.
b
c
d
Δg = g1 − g3. Isotropic g values for measurements at 295 K. For 23+
e
2
2
2
and 43+, gav = √(g1 + g2 + g3 )/3. From ref 5c.
F
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
reduction of 42+ thus leads to a reduction of the L‑2H ligand in
that molecule.
Just as for the ruthenium case, the one-electron reduction of
the osmium compound to 2+ shifts the MLCT band to higher
energy and leads to the emergence of a new low-energy band at
927 nm, which is assigned to a LMCT transition (Figure 6 and
Table 3). 2+ shows an EPR signal at 295 K centered at g =
1.982. This spectrum could be simulated with hyperfine
coupling to osmium (osmium satellites)20 of 13 G. The
appearance of an EPR signal at room temperature despite the
presence of a 5d metal center, the small deviation of the g value
from the free electron g value, and the relatively small hyperfine
coupling constant to osmium are indications of a ligand-
centered reduction.24 Thus, just as for 4+, 2+ has its spin
predominantly located on the L‑2H bridging ligand. The second
reduction of 22+ was not completely reversible on the
spectroelectrochemistry time scale and hence cannot be
discussed here.
DFT Calculations. Structure-based DFT calculations using
the functionals B3LYP was used to calculate spin density
distributions for the paramagnetic forms of the metal
complexes. A look at the spin densities calculated according
to Lowdin population analysis shows that for 1+ more than 60%
̈
spin density is located at the osmium centers (Figure 8). This
result indirectly reinforces the assignment of the near-IR band
observed for 1+ as an indication of this being a mixed-valent
species (vide supra).
Figure 9. DFT calculated spin density plot of 23+ (top; each osmium
atom accounts for 32% of the spin population) and 2+ (bottom).
CONCLUSIONS
■
In conclusion, we have presented here the synthesis and
characterization of two quinonoid-bridged diosmium com-
plexes. The arene−osmium-containing complex 1 displays a
substituent-induced reactivity on chloride abstraction, which
leads to the coordination of the thioether groups of the
bridging ligand to the osmium centers and to arene release.
This set of chemical transformations leads to the acetonitrile-
containing complex 22+, where the bridging ligand acts in a “bis-
tridentate” fashion. Comparison of 1 and 22+ with their
ruthenium analogues 3 and 42+ reveals the more robust nature
of the osmium−ligand bonds, leading to higher reaction times
and temperatures for the osmium compounds in comparison to
their ruthenium analogues. The diosmium complex 22+ displays
a higher comproportionation constant in comparison to its
diruthenium analogue. Electrochemical and UV−vis−near-IR
spectroelectrochemical investigation of 1 and 22+ shows a
predominantly metal-centered oxidation for both of these
compounds, resulting in the generation of completely
delocalized class III mixed-valence species in the odd-electron
oxidized form. Metal contribution to the singly occupied
molecular orbital (SOMO) is higher for 23+ in comparison to
the analogous ruthenium compound 43+. This is evident from
much larger g anisotropy for 23+ in comparison to 43+. The one-
electron reduction of 22+ leads to the reduction of the bridging
quinone and the generation of a ligand-radical-bridged
homovalent diosmium(II) complex. Spin density calculations
corroborate the assignments made on the basis of spectroelec-
trochemical experiments. The UV−vis−near-IR spectra of the
osmium complexes are complicated by the presence of many
Figure 8. DFT calculated spin density plot of 1+. Each osmium atom
accounts for 31% of the spin population.
For the one-electron-oxidized form 23+, a total of more than
60% spin density is localized on the osmium centers. In
contrast, for the reduced form 2+, the spin density on the
osmium centers is negligible (Figure 9). These results thus
corroborate the assignments made from spectroelectrochemis-
try measurements of 23+ being a mixed-valent species with
predominantly osmium-centered spin and 2+ being a ligand-
radical-bridged diosmium(II) species with almost exclusive
ligand-centered spin. The corresponding spin density distribu-
tions for the ruthenium complexes show slightly more than
50% spin densities on the ruthenium centers for 43+ and almost
negligible spin densities on the metal centers for 4+ (Figure S3,
Supporting Information).
G
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
under reduced pressure gave 2[BF4]2 as a green solid. Anion exchange
with excess NaBPh4 gave the pure green product (40 mg, 22%).
Crystals were obtained by slow diffusion of diethyl ether into an
acetonitrile/toluene (1/1) solution of 2[BPh4]2. Anal. Calcd for
C87H82B2N8O2Os2S2 (M + toluene): C, 60.13; H, 4.76; N, 6.45.
Found: C, 60.21; H, 5.01; N, 6.14. HRMS (ESI): calcd for
bands in the visible and near-IR region because of the
observance of forbidden transitions in case of the 5d metal
center. We note here that the use of potentially bridging
quinones that contain additional donor atoms on their nitrogen
substituents is rare in the literature. Ligand-bridged diosmium
complexes have also been much less investigated in comparison
to their diruthenium counterparts. We have presented here
examples of two new diosmium complexes by combining both
of these concepts. Additionally, the arene−osmium complex
presented here displays reversible electrochemical responses.
This is in contrast to their ruthenium counterparts known in
the literature.10 Hence, the use of arene−osmium components
might deliver rectangles and cages better suited for redox-
switching and redox-sensing studies.
2+
C24H22N4O2Os2S2 ([M − 4CH3CN − 2BPh4−]2+) 422.0186;
1
found 422.0191. H NMR (250 MHz, CD3CN): δ 2.42−2.44 (m,
12H, 2 CH3CN trans), 2.71 (s, 6H, CH3CN), 2.87 (s, 6H, SCH3), 6.84
3
−
(t, JH−H = 7.1 Hz, 8H, aryl-H BPh4 −), 6.92 (s, 2H, quinone-H), 6.99
(t, 3JH−H = 7.2 Hz, 16H, aryl-H BPh4 ), 7.20−7.30 (m, 18H,16 aryl-H,
BPh4 , 2 aryl-H), 7.43 (dt, JH−H = 7.9 Hz, JH−H = 1.5 Hz, 2H, aryl-
−
3
4
3
4
H), 7.73 (dd, JH−H = 7.8 Hz, JH−H = 1.3 Hz, 2H, aryl-H), 8.14 (d,
3JH−H = 8.2 Hz, 2H, aryl-H).
X-ray Crystallography. Single crystals of 2[BPh4]2 were grown by
slow diffusion of diethyl ether into an acetonitrile/toluene (1/1)
solution of 2[BPh4]2. The X-ray intensity data for 2[BPh4]2 were
collected at 100 K using a Bruker Kappa Apex II duo diffractometer.
Calculations were performed with the SHELXTL PC 5.03 and
SHELXL-97 program.27
The alert “PLAT413_ALERT_2_A Short Inter XH3 .. XHn H69A
.. H71C .. 1.63 Ang” in the CIF-Check can be explained due to the
strongly disordered toluene moiety in the structure. The refinement of
the methyl group at the toluene at two different positions results in the
short distance of the two protons.
EXPERIMENTAL SECTION
■
13
General Considerations. [Os(Cym)2Cl2]2 and the ligand L5c
were prepared according to reported procedures. All other reagents are
commercially available and were used as received. All solvents were
dried and distilled using common techniques unless otherwise
mentioned.
Instrumentation. 1H NMR spectra were recorded at 250.13 MHz
on a Bruker AC250 instrument. EPR spectra in the X band were
recorded with a Bruker System EMX instrument. Simulations of EPR
spectra were done using the Easyspin program.25 UV−vis−near-IR
absorption spectra were recorded on a J&M TIDAS spectrometer.
Cyclic voltammetry was carried out in 0.1 M Bu4NPF6 solutions using
a three-electrode configuration (glassy-carbon working electrode, Pt
counter electrode, Ag wire as pseudoreference) and PAR 273
potentiostat and function generator. The decamethylferrocene/
decamethylferrocenium couple served as internal reference,and the
values were then converted to the ferrocene/ferrocenium couple as
standard by using literature values.26 Spectroelectrochemical measure-
ments were carried out using an optically transparent thin-layer
electrochemical (OTTLE) cell consisting of platinum-grid working
and auxiliary electrodes and a silver quasi-reference electrode sealed
between optical windows.19 The cell was mounted on the J&M TIDAS
diode-array spectrometer, and the spectra were collected continuously
during the potential scan within the redox steps. Elemental analysis
was performed on a Perkin-Elmer Analyzer 240 instrument. Mass
spectrometry experiments were carried out on a Bruker Daltronics
Mictrotof-Q mass spectrometer.
CCDC 904338 contains a CIF file for 2(BPh4)2. These data can be
obtained free of charge from the Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_requests/cif.
DFT Calculations. The program package ORCA 2.9.1 was used for
all calculations.28 The geometry optimization and single-point
calculations were performed by the DFT method with BP86 and
B3LYP functionals, respectively,29 including relativistic effects in zero-
order regular approximation (ZORA).30 Convergence criteria for the
geometry optimization were set to default values (OPT), and “tight”
convergence criteria were used for SCF calculations (TIGHTSCF).
Triple-ζ valence quality basis sets (def2-TZVP) were used for all
atoms.31 The resolution of the identity approximation (RIJCOSX) was
employed32,33 with matching auxiliary basis sets.34 All spin densities
were calculated according to Lowdin population analysis.34 Spin
̈
densities were visualized via the program Molekel.35
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis. [{Cl(η6-Cym)Os}2(μ-η2:η2-L‑2H)] (1). Os2(Cym)2Cl4
(80.0 mg, 0.10 mmol) and the ligand L (38.0 mg, 0.10 mmol) were
dissolved in CH2Cl2 (20 mL) under an argon atmosphere. NEt3 (0.1
mL) was added, and the solution was stirred overnight at room
temperature. The solution was concentrated, and the product was
precipitated by addition of hexane. The compound was filtered,
washed with hexane, and dried in vacuo. The compound was obtained
as a reddish purple solid (94 mg, 85%). Anal. Calcd for
C40H44Cl2N2O2Os2S2: C, 43.66; H, 4.03; N, 2.55. Found: C, 43.50;
H, 4.03; N, 2.70. HRMS (ESI): calcd for C40H44N2O2Os2S2 ([M −
2Cl−]2+) m/z 515.1018; found 515.1043. 1H NMR (400 MHz,
A CIF file and table giving crystallographic data for 2[BPh4]2
1
and figures and tables giving H NMR spectra, spin density
plots for 4n+, and coordinates for DFT calculations. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Email: biprajit.sarkar@fu-berlin.de.
CD3CN): δ 1.09 (d, 3JH−H = 6.9 Hz, 6H, CH(CH3)2), 1.13 (d, 3JH−H
=
Author Contributions
6.9 Hz, 6H, CH(CH3)2), 1.90 (s, 6H, CH3), 2.55 (s, 6H, SCH3), 2.91
(sept, 3JH−H = 6.8 Hz, 2H, CH(CH3)2), 4.93 (s, 2H, quinone-H), 5.38
(d, 3JH−H = 5.5 Hz, 2H, arene-H), 5.65 (d, 3JH−H = 5.3 Hz, 2H, arene-
§Both of these authors contributed equally to this work.
Notes
3
3
H), 5.69 (d, JH−H = 5.4 Hz, 2H, arene-H), 5.83 (d, JH−H = 5.6 Hz,
The authors declare no competing financial interest.
2H, arene-H), 7.23 (2, 3JH−H = 4.0 Hz, 4H, aryl-H), 7.34 (m, 2H, aryl-
3
H), 7.44 (d, JH−H = 7.9 Hz, 2H, aryl-H).
[{(CH3CN)3Os}2(μ-η3:η3-L‑2H)][BPh4]2 (2[BPh4]2). Compound 1 (120
mg, 0.11 mmol) and AgBF4 (53 mg, 0.27 mmol) were dissolved in
acetonitrile (40 mL) and refluxed for 19 h under an inert atmosphere
of argon. After it was cooled to room temperature, the reaction
mixture was filtered and the solvent was removed under reduced
pressure. The resulting green solid was purified by flash column
chromatography (aluminum oxide + 5 wt % water). The green 2[BF4]2
was eluted with CH3CN/CH3OH (97/3). Evaporation of the solvent
ACKNOWLEDGMENTS
■
We are indebted to the Baden-Wurttemberg Stiftung and the
̈
Fonds der Chemischen Industrie (Chemiefondsstipendium for
D.S.) for financial support of this work. ZEDAT, FU Berlin is
acknowledged for access to the computing center. This paper is
dedicated to Prof. Pierre Braunstein on the occasion of his 65th
birthday, with our warm wishes.
H
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(8) Margraf, G.; Kretz, T.; de Biani, F. F.; Lasi, F.; Loschi, S.; Zanello,
P.; Bats, J. W.; Wolf, J.; Removic-Langer, K.; Lang, M.; Prokofiev, A.;
Assmus, W.; Lerner, H.-W.; Wagner, M. Inorg. Chem. 2006, 45, 1277.
(9) (a) Zhang, D.; Jin, G.-X. Organometallics 2003, 22, 2851.
(b) Scheuermann, S.; Kretz, T.; Vitze, H.; Bats, J. W.; Bolte, M.;
Lerner, H. W.; Wagner, M. Chem. Eur. J. 2008, 14, 2590. (c) Hohloch,
S.; Braunstein, P.; Sarkar, B. Eur. J. Inorg. Chem. 2012, 546. (d) Taquet,
J.-P; Siri, O.; Braunstein, P.; Welter, R. Inorg. Chem. 2004, 43, 6944.
(e) Yang, Q.-Z.; Kermagoret, A.; Agostinho, M.; Siri, O.; Braunstein, P.
Organometallics 2006, 25, 5518.
(10) (a) Jia, W.-G.; Han, Y.-F.; Lin, Y.-J.; Weng, L.-H.; Jin, G.-X.
Organometallics 2009, 28, 3459. (b) Han, Y.-F.; Jia, W.-G.; Yu, W.-B.;
Jin, G.-X. Chem. Soc. Rev. 2009, 38, 3419. (c) Han, Y.-F.; Li, H.; Jin, G.-
X. Chem. Commun. 2010, 46, 6879. (d) Han, Y.-F.; Lin, Y.-J.; Jia, W.-
G.; Jin, G.-X. Organometallics 2008, 27, 4088. (e) Han, Y.-F.; Jia, W.-
G.; Lin, Y.-J.; Jin, G.-X. Organometallics 2008, 27, 5002. (f) Han, Y.-F.;
Jia, W.-G.; Lin, Y.-J.; Jin, G.-X. Angew. Chem., Int. Ed. 2009, 48, 6234.
(g) Paul, L. E. H.; Therrien, B.; Furrer, J. Inorg. Chem. 2012, 51, 1057.
(h) Mattsson, J.; Govindaswamy, P.; Renfrew, A. K.; Dyson, P. J.;
REFERENCES
■
(1) (a) Kaim, W. Inorg. Chem. 2011, 50, 9752. (b) Ward, M. D.;
McCleverty, J. A. J. Chem. Soc., Dalton Trans. 2002, 275. (c) Pierpont,
C. G. Coord. Chem. Rev. 2001, 216−217, 95. (d) Zanello, P. Coord.
Chem. Rev. 2006, 50, 2000. (e) de Bruin, B.; Hetterscheid, D. G. H.;
Koekkoek, A. J. J.; Grutzmacher, H. Prog. Inorg. Chem. 2007, 55, 247.
(f) Poddel’sky, A. I.; Cherkasov, V. K.; Abakumov, G. A. Coord. Chem.
Rev. 2009, 253, 291.
̈
(2) (a) Chaudhuri, P.; Verani, C. N.; Bill, E.; Weyhermuller, T.;
̈
Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213.
(b) Bhattacharya, S.; Gupta, P.; Basuli, F.; Pierpont, C. G. Inorg.
Chem. 2002, 41, 5810. (c) Mandal, S.; Paul, N.; Banerjee, P.; Mondal,
T. K.; Goswami, S. Dalton Trans. 2010, 39, 2717. (d) Fabre, M.;
Bonvoisin, J. J. Am. Chem. Soc. 2007, 129, 1434. (e) Sarkar, B.;
Huebner, R.; Pattacini, R.; Hartenbach, I. Dalton Trans. 2009, 4653.
(f) Deibel, N.; Schweinfurth, D.; Fiedler, J.; Zalis, S.; Sarkar, B. Dalton
Trans. 2011, 40, 9925. (g) Paretzki, A.; Das, H. S.; Weisser, F.;
Scherer, T.; Bubrin, D.; Fiedler, J.; Nycz, J. E.; Sarkar, B. Eur. J. Inorg.
Chem. 2011, 2413. (h) Maurer, J.; Linseis, M.; Sarkar, B.; Schwederski,
B.; Niemeyer, M.; Kaim, W.; Zalis, S.; Anson, C.; Zabel, M.; Winter, R.
F. J. Am. Chem. Soc. 2008, 130, 259.
Stepnicka, P.; Suss-Fink, G.; Therrien, B. Organometallics 2009, 28,
̈
4350.
(3) (a) Chirik, P. J.; Wieghardt, K. Science 2010, 327, 794.
(b) Ringenberg, M. R.; Kokatam, S. L.; Heiden, Z. M.; Rauchfuss,
T. B. J. Am. Chem. Soc. 2008, 130, 788. (c) Praneeth, V. K. K.;
Ringenberg, M. R.; Ward, T. R. Angew. Chem. 2012, 124, 10374.
(d) Dzik, W. I.; van der Vlugt, J. I.; Reek, J. N. H.; de Bruin, B. Angew.
Chem. 2011, 123, 3416. (e) Nguyen, A. I.; Blackmore, K. J.; Carter, S.
M.; Zarkesh, R. A.; Heyduk, A. F. J. Am. Chem. Soc. 2009, 131, 3307.
(f) Tsai, M. K.; Rochford, J.; Polyansky, D. E.; Wada, T.; Tanaka, K.;
Muckerman, J. T. Inorg. Chem. 2009, 48, 4372. (g) Deibel, N.;
Schweinfurth, D.; Hohloch, S.; Fiedler, J.; Sarkar, B. Chem. Commun.
2012, 48, 2388.
(4) (a) Ward, M. D. Inorg. Chem. 1996, 35, 1712. (b) Leschke, M.;
Melter, M.; Lang, H. Inorg. Chim. Acta 2003, 350, 114. (c) Kitagawa,
S.; Kawata, S. Coord. Chem. Rev. 2002, 224, 11. (d) Carbonera, C.; Dei,
A.; Letard, J.-F.; Sangregorio, C.; Sorace, L. Angew. Chem., Int. Ed.
2004, 43, 3136. (e) Miller, J. S.; Min, K. S. Angew. Chem., Int. Ed. 2009,
(11) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.
(12) (a) Kaim, W.; Lahiri, G. K. Angew. Chem., Int. Ed. 2007, 46,
1778. (b) Gupta, P.; Das, A.; Basuli, F.; Castineiras, A.; Sheldrick, W.
S.; Mayer-Figge, H.; Bhattacharya, S. Inorg. Chem. 2005, 44, 2081.
(c) Conception, J. J.; Dattelbaum, D. M.; Meyer, T. J.; Rocha, R. J.
Philos. Trans. R. Soc. A 2008, 366, 163. (d) Barrigelleti, F.; Flamigni, L.;
Guardigli, M.; Juris, A.; Beley, M.; Chodorowski-Kimmes, S.; Collin, J.-
P.; Sauvage, J.-P. Inorg. Chem. 1996, 35, 136.
(13) Werner, H.; Zenkert, K. J. Organomet. Chem. 1988, 345, 151.
(14) Elschenbroich, C. Organometallics; Wiley-VCH: Weinheim,
Germany, 2006.
(15) Han, Y.-F.; Lin, Y.-J.; Jia, W.-G.; Jin, G.-X. Organometallics 2008,
27, 4088.
(16) Sarkar, B.; Kaim, W.; Fiedler, J.; Duboc, C. J. Am. Chem. Soc.
2004, 126, 14706.
(17) Kaim, W.; Kasack, V. Inorg. Chem. 1990, 29, 4696.
(18) (a) Creutz, C. Prog. Inorg. Chem. 1983, 30, 1. (b) Crutchley, R.
J. Adv. Inorg. Chem. 1994, 41, 273.
(19) Krejcik, M.; Danek, M.; Hartl, F. J. Electroanal. Chem. 1991, 317,
179.
(20) Weil, J. A.; Bolton, J. R.; Wertz, J. E. Electron Paramagnetic
Resonance; Wiley: New York, 1994; p 532.
(21) (a) Wanner, M.; Scheiring, T.; Kaim, W.; Slep, L. D.; Baraldo, L.
48, 262. (f) Therrien, B.; Suss-Fink, G.; Govindaswamy, P.; Renfrew,
̈
A. K.; Dyson, P. J. Angew. Chem. 2008, 120, 3833.
(5) (a) Braunstein, P.; Bubrin, D.; Sarkar, B. Inorg. Chem. 2009, 48,
2534. (b) Das, H. S.; Weisser, F.; Schweinfurth, D.; Su, C.-Y; Bogani,
L.; Fiedler, J.; Sarkar, B. Chem. Eur. J. 2010, 16, 2977. (c) Schweinfurth,
D.; Das, H. S.; Weisser, F.; Bubrin, D.; Sarkar, B. Inorg. Chem. 2011,
50, 1150.
(6) (a) Bayen, S.; Barooah, N.; Sarma, R. J.; Sen, T. K.; Karmakar, A.;
Baruah, J. B. Dyes Pigm. 2007, 75, 770. (b) Yang, Q.-Z.; Siri, O.;
Braunstein, P. Chem. Commun. 2005, 2660. (c) Yang, Q.-Z.; Siri, O.;
Braunstein, P. Chem. Eur. J. 2005, 11, 7237.
(7) (a) Kumbhakar, D.; Sarkar, B.; Maji, S.; Mobin, S.; Fiedler, J.;
Aparicio, R. J.; Kaim, W.; Lahiri, G. K. J. Am. Chem. Soc. 2008, 130,
17575. (b) Keyer, T. E.; Forster, R. J.; Jayaweera, P. M.; Coates, C. G.;
McGarvey, J. J.; Vos, J. G. Inorg. Chem. 1998, 37, 5925. (c) Kar, S.;
Sarkar, B.; Ghumaan, S.; Janardanan, D.; van Slageren, J.; Fiedler, J.;
Puranik, V. G.; Sunoj, R. B.; Kaim, W.; Lahiri, G. K. Chem. Eur. J.
2005, 11, 4901. (d) Masui, H.; Freda, A. L.; Zerner, M. C.; Lever, A. B.
P. Inorg. Chem. 2000, 39, 141. (e) Das, H. S.; Das, A. K.; Pattacini, R.;
Huebner, R.; Sarkar, B.; Braunstein, P. Chem. Commun. 2009, 4387.
(f) Cotton, F. A.; Jin, J.-Y.; Li, Z.; Murillo, C. A.; Reibenspies, J. H.
Chem. Commun. 2008, 211. (g) Weisser, F.; Huebner, R.;
Schweinfurth, D.; Sarkar, B. Chem. Eur. J. 2011, 17, 5727. (h) Paretzki,
A.; Pattacini, R.; Huebner, R.; Braunstein, P.; Sarkar, B. Chem.
Commun. 2010, 46, 1497. (i) Deibel, N.; Schweinfurth, D.; Huebner,
R.; Braunstein, P.; Sarkar, B. Dalton Trans. 2011, 40, 431. (j) Man, W.
Y.; Xia, J.-L.; Brown, N. J.; Farmer, J. D.; Yufit, D. S.; Howard, J. A. K.;
Low, P. J. Organometallics 2011, 30, 1852. (k) Xia, J. L.; Man, W. Y.;
Zhu, X.; Zhang, C.; Jin, G.-J.; Schauer, P. A.; Fox, M. A.; Yu, G.-A.;
Low, P. J.; Liu, S. H. Organometallics 2012, 31, 5321. (l) Shao, J.-Y.;
Yang, W.-W.; Yao, J.; Zhong, Y.-W. Inorg. Chem. 2012, 51, 4343.
(m) Yao, C.-J.; Yao, J.; Zhong, Y.-W. Inorg. Chem. 2011, 50, 6847.
M.; Olabe, J. A.; Zal
́
is,
̌
S.; Baerends, E. J. Inorg. Chem. 2001, 40, 5704.
(b) Waldhor, E.; Kaim, W.; Olabe, J. A.; Slep, L. D.; Fiedler, J. Inorg.
̈
Chem. 1997, 36, 2969.
(22) Hush, N. S. Prog. Inorg. Chem. 1967, 8, 391.
(23) Demadis, K. D.; Hartshorn, C. M.; Meyer, T. J. Chem. Rev.
2001, 101, 2655.
(24) Patra, S.; Sarkar, B.; Mobin, S. M.; Kaim, W.; Lahiri, G. K. Inorg.
Chem. 2003, 42, 6469.
(25) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42.
(26) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877.
(27) Sheldrick, G. M. SHELXL-97, Program for refinement of crystal
structures; University of Gottingen, Gottingen, Germany, 1997.
̈
̈
(28) Neese, F. ORCA-an Ab Initio, Density Functional and
Semiempirical Program Package, version 2.9.1; Department of molecular
theory and spectroscopy, Max Planck Institute for Bioinorganic Chemistry,
Mulheim/Ruhr, Germany, January 2012.
̈
(29) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C. T.;
Yang, W. T.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(30) van Wullen, C. J. Chem. Phys. 1998, 109, 392.
̈
(31) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
(32) (a) Baerends, E. J.; Ellis, D. E.; Ros, P. Chem. Phys. 1973, 2, 41.
(b) Dunlap, B. I.; Connolly, J. W. D.; Sabin, J. R. J. Chem. Phys. 1979,
71, 3396. (c) Vahtras, O.; Almlof, J.; Feyereisen, M. W. Chem. Phys.
Lett. 1993, 213, 514.
I
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(33) (a) Eichkorn, K.; Greutler, O. T.; Ohm, H.; Haser, M.; Ahlrichs,
R. Chem. Phys. Lett. 1995, 242, 652. (b) Eichkorn, K.; Weigend, F.;
Treutler, O.; Ahlreichs, R. Theor. Chem. Acc. 1997, 97, 119.
(34) (a) Lowdin, P. O. J. Chem. Phys. 1950, 18, 365. (b) Lowdin, P.
̈
̈
O. Adv. Quantum Chem. 1970, 5, 185.
(35) Portmann, S. Molekel, version 5.4.0.8; CSCS/UNI, Geneva,
Switzerland, 2009.
J
dx.doi.org/10.1021/om300939x | Organometallics XXXX, XXX, XXX−XXX