2,5-Diphenyl-3,4-bis(2-pyridyl)cyclopenta-2,4-dien-1-one as a redox-active chelating ligand

Article abstract of DOI:10.1002/chem.200400465

2,5-Diphenyl-3,4-bis(2-pyridyl)cyclopenta-2,4-dien-1-one (1), a close relative of tetraphenylcyclopentadienone, is a new ligand platform for use in redox switches and sensors. Compound 1 acts as a molecular electrochemical sensor towards a range of divale

Full text of DOI:10.1002/chem.200400465

FULL PAPER
2
,5-Diphenyl-3,4-bis(2-pyridyl)cyclopenta-2,4-dien-1-one as a Redox-Active
Chelating Ligand
[
a]
[a]
[b]
[c]
Ulrich Siemeling,* Imke Scheppelmann, Jürgen Heinze, Beate Neumann,
[
c]
[c]
Anja Stammler, and Hans-Georg Stammler
Abstract: 2,5-Diphenyl-3,4-bis(2-pyri-
dyl)cyclopenta-2,4-dien-1-one (1),
able two-wave behaviour. It forms che-
magnitude upon one-electron reduc-
tion. The bite angle of 1 is close to 908
in these complexes. The attachment of
the 14-valence-electron Cp*Co frag-
a
lates of the type [(1)MX ], whose sta-
2
close relative of tetraphenylcyclopenta-
dienone, is a new ligand platform for
use in redox switches and sensors.
Compound 1 acts as a molecular elec-
trochemical sensor towards a range of
divalent metal ions and exhibits favour-
bility is enhanced by five orders of
ment to the cyclopentadienone
p
Keywords: chelates electro-
·
system reduces the bite angle and thus
modulates the binding characteristics
of 1.
chemistry ligands · redox
·
N
switches · UV/Vis spectroscopy
[
6]
Introduction
versibly to the corresponding radical monoanion. The co-
ordination chemistry of 1 was completely unexplored prior
to our work. Compound 1 combines two well-known ligand
platforms, namely, oligopyridine and cylopentadienone. A
priori, it may therefore act as a bidentate N,N ligand in
[1]
Redox-active ligands are of great current interest. They
can be used inter alia as molecular electrochemical sensors
[2]
for the detection of metal cations. The vast majority of
such systems contain a redox-active group whose oxidation
leads to thermodynamic destabilisation of the metal com-
plex due to electrostatic interactions. Among such ligands,
4
Werner-type coordination chemistry and/or as an h ligand
in organometallic chemistry. Among such h -cyclopentadie-
a cyclopentadienyl-
cobalt(i) fragment are known to show reversible electro-
chemical behaviour. Therefore, we envisaged that both 1
and its cyclopentadienylcobalt(i) complexes could be used as
molecular electrochemical sensors for metals coordinating
to the pyridyl nitrogen atoms.
4
none complexes, those containing
[3]
ferrocene derivatives play a prominent role. Ligands which
can be reduced are less common, most of them being de-
rived from quinone. For electrostatic reasons, ligand reduc-
tion generally leads to a thermodynamic stabilisation of
metal complexes, and particularly large enhancements of
[6]
[2]
6
complex-formation constants of up to about 10 have been
[4]
reported for selected quinone-based systems. In this con-
text we have investigated 2,5-diphenyl-3,4-bis(2-pyridyl)cy-
Results and Discussion
[5]
clopenta-2,4-dien-1-one (1). This compound is a close rela-
tive of tetraphenylcyclopentadienone, which exhibits
redox-active cyclopentadienone core that can be reduced re-
a
Synthesis and characterisation of compounds
Organometallic complexes: The reaction of
1
with
[
CpCo(CO) ] in refluxing p-xylene afforded the cyclopenta-
2
4
[
a] Prof. U. Siemeling, Dr. I. Scheppelmann
Science Department and Center for Interdisciplinary
Nanostructure Science and Technology (CINSaT)
University of Kassel
dienone complex [CpCo(h -1)] in 89% yield. Less forcing
conditions were required for the reaction of with
Cp*Co(C H ) ], which proceeded cleanly and swiftly at
1
[
2
4 2
4
3
4109 Kassel (Germany)
608C in toluene and furnished [Cp*Co(h -1)] in 85% yield
(Scheme 1).
Fax : (+49)561-804-4777
E-mail: siemeling@uni-kassel.de
Both compounds precipitated from the reaction mixture
as red, air-stable, microcrystalline solids which are soluble in
polar solvents such as acetonitrile, ethanol and dichlorome-
thane. Crystals suitable for single-crystal X-ray structure
analysis were obtained from chloroform in each case. We
[
b] Prof. J. Heinze
Department of Chemistry, University of Freiburg
7
9104 Freiburg (Germany)
[
c] B. Neumann, A. Stammler, Dr. H.-G. Stammler
Department of Chemistry, University of Bielefeld
3
3501 Bielefeld (Germany)
also determined the structure of
1
for comparison
Chem. Eur. J. 2004, 10, 5661 – 5670
DOI: 10.1002/chem.200400465
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5661

FULL PAPER
Table 1. Cyclopentadienone fold angle a [8] and selected bond
lengths [pm] for ligand 1 and its cyclopentadienylcobalt(i) derivatives.
4
4
1
[Cp*Co(h -1)]
[CpCo(h -1)]
C1-C2
C1ꢀC5
C2ꢀC3
C3ꢀC4
C4ꢀC5
C1ꢀO1
151.8(3)
151.9(3)
134.8(3)
152.9(3)
135.3(3)
120.9(3)
1.7
147.7(3)
147.6(3)
144.3(3)
144.2(3)
143.6(3)
124.7(2)
7.2
145.8(7)
147.4(6)
145.3(7)
145.3(6)
142.3(7)
124.6(6)
12.7
a
4
4
Scheme 1. Synthesis of [CpCo(h -1)] and [Cp*Co(h -1)]. Py=2-pyridyl.
tially indistinguishable bond lengths between the four C
atoms coordinated to the Co centre in each case. The CO
carbon atoms of [CpCo(h -1)] and [Cp*Co(h -1)] point away
from the metal atom, with folding angles a of the five-mem-
bered ring of 12.7 and 7.28, respectively. This angle has a
value of only 1.78 in uncoordinated 1.
4
4
4
4
(
Figure 1). [CpCo(h -1)] and [Cp*Co(h -1)] exhibit sand-
wich-type structures, and that of the latter is shown in
Figure 2.
Delocalisation of electron density from the cyclopentadie-
nylcobalt(i) fragment into the LUMO of 1 leads to a signifi-
[7]
cant elongation of the carbon–oxygen bond of about 4 pm.
The weakening of this bond is also consistent with IR spec-
4
4
troscopic data (n : 1 1715, [CpCo(h -1)] 1592, [Cp*Co(h -
1
CO
ꢀ
1
[8]
)] 1586 cm ). This transfer of electron density is known
to lead to an increase in nucleophilicity of the oxygen atom,
which is reflected by reactions with suitable electrophiles to
[7a]
give cobaltocenium species.
4
We briefly investigated [Cp*Co(h -1)] as a representative
case in this context, because we were interested to see
whether selective reactions at the oxygen atom could be
achieved with electrophiles in the presence of the pyridyl N
atoms, which act as competing nucleophiles. This indeed
Figure 1. Molecular structure of 1 in the crystal.
proved possible with acetyl chloride, which afforded
5
[
Cp*Co(h -1COMe)]PF as a yellow solid after anion meta-
6
thesis (Scheme 2). With the softer electrophile methyl tri-
4
Figure 2. Molecular structure of [Cp*Co(h -1)] in the crystal.
4
Scheme 2. Acetylation and complete methylation of [Cp*Co(h -1)]. Py=
2
-pyridyl.
In each case, the aromatic rings attached to the cyclopen-
tadienone moiety are arranged in an axially chiral paddle-
wheel fashion. Both species crystallise as racemic com-
pounds. The aromatic substituents rotate freely on the NMR
timescale at room temperature in solution. Metal coordina-
tion leads to noticeable changes in the bond parameters of
the cyclopentadienone moiety (Table 1), which are typical of
such species. The alternating carbon–carbon bond lengths
in the diene unit of the five-membered ring of 1 are affected
substantially by metal coordination, which results in essen-
flate (1 equiv), no selective O-alkylation was observed,
which can be explained in terms of the HSAB principle.
[9]
Instead, according to NMR spectroscopic results, a mixture
of mono-, di- and trimethylated species was obtained togeth-
4
er with unconsumed [Cp*Co(h -1)]. Trimethylation was
[7]
easily accomplished with an excess of methyl triflate to
5
afford [Cp*Co(h -1Me )](CF SO ) as
a
yellow solid
3
3
3 3
(Scheme 2). Attempts to remove selectively two methyl
5662
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5661 – 5670

Redox-Active Chelating Ligands
5661 – 5670
groups from this product by reaction with two equivalents
of 4-dimethylaminopyridine (DMAP) failed, and a mixture
of methylated species again resulted. Attempts were made
to achieve selective O-alkylation by steric means by utilising
trityl chloride as alkylating agent, since trityl chloride in the
presence of pyridine is commonly used for the protection of
primary alcohols. However, no reaction was observed even
after several weeks in dichloromethane solution.
lengths are identical within experimental error and indistin-
guishable from those within the Cp* ring. The carbon–
oxygen bond elongates considerably from 124.7(2) to
140.6(9) and 133.9(2) pm upon acylation and methylation,
4
4
respectively. In contrast to [CpCo(h -1)], [Cp*Co(h -1)] and
5
+
[Cp*Co(h -1COMe)] , which show regular sandwich-type
structures with essentially coplanar coordinated p decks,
5
3+
[Cp*Co(h -1Me )] exhibits a slightly bent structure with
3
The formation of cobaltocenium species by O-acylation or
O-alkylation is already indicated by the characteristic yellow
colour of the respective product and is further supported by
cobalt–carbon bond lengths ranging from 206.02(17) to
212.46(17) pm and a ring tilt angle of 9.28.
[10]
13
NMR spectroscopic data. The C NMR signal of the CO
Werner-type complexes: The reaction of 1 with metal diha-
4
carbon atom of [Cp*Co(h -1)] at 150 ppm experiences a pro-
lides MX in hot ethanol afforded chelate complexes of the
2
nounced upfield shift upon acetylation (Dd=36.7 ppm) or
methylation (Dd=22.1 ppm) of the oxygen atom.
type [(1)MX ] (MX =PdCl , PtCl , ZnCl , HgCl , HgBr ).
2
2
2
2
2
2
2
With the exception of the palladium species, the structures
of all these complexes were determined by single-crystal X-
ray diffraction. The metal atom is tetracoordinate in each
The results of single-crystal X-ray structure analyses,
5
which were performed for [Cp*Co(h -1COMe)]BPh
4
5
8
(
Figure 3) and [Cp*Co(h -1Me )](CF SO ) (Figure 4) fur-
case. The coordination environment is square-planar for d -
3
3
3 3
ther corroborate these findings.
configurated Pt and therefore presumably also for Pd,
whereas it is distorted tetrahedral for the other three com-
pounds, in which M is d -configurated. As a representative
example, the structure of the zinc complex is shown in
1
0
Figure 5. ZnCl coordination has no noticeable effect on the
2
5
Figure 3. Molecular structure of the cation of [Cp*Co(h -1COMe)]BPh
4
in the crystal.
Figure 5. Molecular structure of [(1)ZnCl ] in the crystal.
2
bond lengths of 1, and the same holds true for the other
MX complexes. The seven-membered chelate ring has bond
2
angles at C3 and C4 which deviate only slightly from the
corresponding angles in the uncoordinated ligand; the aver-
age values are 1218 for 1 and 1278 for the zinc complex.
This effect is even smaller for the other MX complexes.
2
With CdCl the chloro-bridged dimer [{(1)Cd(m-Cl)Cl} ] was
2
2
obtained, which contains two pentacoordinate Cd atoms
(
Figure 6). Their coordination is best described as distorted
square-pyramidal. The CdꢀCl(apical) vectors deviate from
the ideal pyramidal orientation by about 88. The CdꢀCl(api-
cal) distances are 241.29(8) and 240.42(8) pm. These are
considerably shorter than the corresponding distances of the
bridging basal Cl atoms, which range from 256.33(8) to
5
Figure 4. Molecular
structure
of
the
cation
of
[Cp*Co(h -
2
59.81(7) pm.
1
Me )](CF SO in the crystal.
3
3
3 3
)
It is instructive to compare the structures of [(1)MX2]
with those of unchelated analogues of the type [(py) MX ]
2
2
(
py=pyridine). For MX =ZnCl , the two structures are
2 2
In contrast to the cyclopentadienone complexes
CpCo(h -1)] and [Cp*Co(h -1)], the functionalised five-
similar, with pseudotetrahedral zinc coordination. However,
the N-Zn-N angle is 106.3(2)8 in the unchelated complex,
4
4
[11]
[
membered rings are essentially planar. Their CꢀC bond
whereas it is only 95.42(5)8 in the chelate [(1)ZnCl ]. This is
2
Chem. Eur. J. 2004, 10, 5661 – 5670
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5663

U. Siemeling et al.
FULL PAPER
The differences in the coordination behaviour of ZnCl2,
CdCl and HgX towards 1, which lead to tetracoordinate
2
2
mononuclear complexes in the case of zinc and mercury, but
to a chloro-bridged dimer exhibiting pentacoordination in
the case of cadmium, appear counterintuitive, since they
show no consistent relation to the relevant ionic radii. Es-
sentially the same behaviour was observed by Lockhart
[14]
et al. for a bidentate bis(benzimidazole) ligand.
In their
case, however, the dimeric cadmium complex is trigonal-bi-
pyramidal with strongly asymmetric chloro bridges. The
degree of covalency of the MꢀX bond increases in the order
Zn<Cd<Hg and Cl<Br, reflecting relative valence orbital
2
Figure 6. Molecular structure of [(1)Cd(m-Cl)Cl] in the crystal.
[15]
binding energies.
The intricate relationship between va-
lence orbital binding energies and structures in compounds
of zinc, cadmium and mercury was investigated by Tossell
[16]
due to the rigid ligand framework of 1. In the two mercury
compounds, the N-M-N angle is even smaller, namely,
and Vaughan, and our experimental findings are compati-
ble with their analysis.
7
8.30(10)8 for [(1)HgCl ] and 79.08(12)8 for [(1)HgBr ],
From the structures of the Group 10 metal complexes de-
scribed above it is evident that the bite-angle range of 1 lies
around 908, which is ideal for square-planar coordination. In
2
2
which is due to the larger size of the coordinated metal
atom, which leads to much longer MꢀN bonds (ca. 206 pm
for M=Zn versus ca. 241 pm for M=Hg). The same holds
for the cadmium complex, the CdꢀN bonds of which range
fact, the N-Pt-N angle in [(1)PtCl ] is 87.40(12)8, and it is ac-
2
companied by a Cl-Pt-Cl angle of 93.39(3)8. The Pt bond
from 232.8(2) to 236.4(2) pm, which results in N-Cd-N
angles of 79.85(8) and 77.12(8)8. The pronounced ligand-in-
duced decrease in the N-Zn-N angle does not lead to a con-
comitant increase of the Cl-Zn-Cl angle, which is
lengths and angles are almost identical to those observed for
[17]
cis-[(py) PtCl ].
2
2
Finally, the capacity of the organometallic complex
4
[Cp*Co(h -1)] to form Werner-type complexes was tested
[11]
1
22.172(17)8 in the chelate and 120.9(1)8 in [(py) ZnCl ],
with the metal dihalides mentioned above. These reactions
2
2
in which repulsive interactions between the two chlorine
atoms prevent a closer approach.
failed in the case of cadmium and mercury. With zinc chlo-
4
ride, however, the dimetallic chelate [Cp*Co(h -1)ZnCl ]
2
Cadmium and mercury have considerably larger ionic
radii than zinc, and this allows higher coordination numbers
for analogous complexes. Thus, the pyridine complexes for
could be obtained. Adownfield shift of the signal of the pyr-
idyl a-protons of about 0.6 ppm was observed in the
1
H NMR spectrum upon metal complexation. The zinc atom
[12]
[13]
M=CdCl2 and HgCl2 are coordination polymers of the
type [(py) M(m-Cl) ] , which contain infinite chains com-
is coordinated in an exo fashion with respect to the Cp*Co
fragment (Figure 7). Metal coordination has no significant
2
2 ¥
4
posed of hexacoordinate chloro-bridged metal atoms bear-
ing two pyridine ligands in trans orientation. The architec-
ture of the chelate ligand 1 is incompatible with such a
linear N-M-N arrangement and results in a different struc-
ture type for cadmium and mercury. We note that the Hgꢀ
effects on the bond lengths of [Cp*Co(h -1)]. The angles at
the pyridyl-bearing C atoms in the seven-membered chelate
ring are slightly smaller than the corresponding angles in
4
4
[Cp*Co(h -1)]; the average values are 1278 for [Cp*Co(h -
1)] and 1218 for the zinc complex. We note that for ligand 1
Cl bond lengths of [(1)HgCl ] are significantly smaller than
coordination of ZnCl has the opposite effect and leads to
2
2
the CdꢀCl(apical) distances (ca. 236 versus ca. 241 pm) in
an increase of these angles from 121 to 1278. This behaviour
may be indicative of additional ring strain in the chelate
[
(1)Cd(m-Cl)Cl] , although the ionic radius of tetracoordi-
2
II
II
nate Hg is slightly larger than that of pentacoordinate Cd
(
110 vs 109 pm). In contrast, the corresponding MꢀN bond
lengths (vide supra) follow the trend of the ionic radii. Just
the opposite behaviour has been reported for the polymeric
pyridine complexes of CdCl and HgCl and has been inter-
2
2
preted in terms of a higher degree of metal–nitrogen p
backbonding in the case of mercury.
In contrast to its cadmium analogue [(py) Cd(m-Br) ] and
2
2 ¥
its chloro analogue [(py) Hg(m-Cl) ] , [(py) HgBr ] is mono-
2
2
¥
2
2
[13]
meric and contains tetracoordinate mercury, the bonding
parameters of which are similar to those of [(1)HgBr ]. The
2
HgꢀN and HgꢀBr bond lengths are about 239 and 248 pm,
respectively, for [(py) HgBr ] and about 241 and 249 pm, re-
2
2
spectively, for [(1)HgBr ]. The N-Hg-N angle is 90.7(7)8 in
2
[
(py) HgBr ], and 79.08(12)8 in [(1)HgBr ]; the Br-Hg-Br
2
2
2
angles are 141.2(1)8 for the former, and 137.255(17)8 for the
latter.
4
Figure 7. Molecular structure of [Cp*Co(h –1)ZnCl
2
] in the crystal.
5664
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5661 – 5670

Redox-Active Chelating Ligands
5661 – 5670
4
[
Cp*Co(h -1)ZnCl ] with respect to [(1)ZnCl ]. The bond
plexes showed quasireversible electrochemical behaviour
(Table 2).
2
2
4
parameters of the zinc atom of [Cp*Co(h -1)ZnCl ] are simi-
2
lar to those found for [(1)ZnCl ]; a noticeable difference,
2
however, is the comparatively smaller N-Zn-N (89.34(6)
versus 95.42(5)8) and Cl-Zn-Cl angles (113.621(19) versus
Table 2. Half-wave potentials [V] (vs ferrocenium/ferrocene) for the re-
duction processes exhibited by 1 and its zinc, palladium and platinum
complexes.
122.172(17)8) of the organometallic species. This effect must
be due to the presence of the Cp*Co moiety. This unit
causes shortening of the bond between the two pyridyl-bear-
ing carbon atoms of about 9 pm in 1 (vide supra) and there-
fore directly influences the bite angle of the two pyridyl
goups attached to these atoms. The bond angles at zinc devi-
ate strongly from those observed for [(py) ZnCl ] (N-Zn-N
1
[(1)ZnCl
2
]
[(1)PdCl
2
]
2
[(1)PtCl ]
[
a]
E
DE1/2
E
DE1/2
1/2 (CH
2
3
Cl
2
)
ꢀ1.37
ꢀ1.03
0.34
ꢀ1.05
ꢀ1.00
[
b]
0.32
0.37
[a]
1/2 (CH
CN)
ꢀ1.23
ꢀ0.93
[b]
0.30
2
2
ꢀ1
[
0
a] Scan rate 0.10 Vs , 0.1m [NBu
4
]PF
6
supporting electrolyte, DE
p
ꢁ
106.3(2), Cl-Zn-Cl 120.9(1)8, vide supra) and are indicative
.10 V. [b] Shift of the half-wave potential with respect to that of 1 in the
of ring strain, which conceivably leads to destabilisation of
same solvent.
the complex. In fact, further investigations showed that the
4
complex-formation constant of [Cp*Co(h -1)ZnCl ] in ace-
2
tonitrile is lower by one order of magnitude than that of the
For the sensing of nonpolarisable ions, the shift of the
half-wave potential is governed by Coulomb interactions,
and therefore large values can be expected only when the
ion-binding site is in close proximity to the redox-active
unit. This is particularly important for the highly polar sol-
vents generally used for the investigation of ionic species,
since there is an inverse dependence of DE ’ on the dielec-
tric constant of the medium, as is shown by Equation (1) (n:
number of electrons transferred during the redox process, F:
less strained [(1)ZnCl ] (vide infra). In the case of cadmium
2
4
and mercury, the ligand [Cp*Co(h -1)] would cause even
more acute bond angles than those observed in the case of
1
, for which chelates of the type [(1)MX ] exhibit N-M-N
2
angles that are already smaller than 808 for M=Cd and Hg.
Obviously, complex formation does not occur under these
circumstances.
0
II
The ionic radii of square-planar tetracoordinate Pd
II
(
78 pm) and Pt centres (74 pm) are practically identical
with that of tetracoordinate Zn (74 pm). Thus, complex for-
mation occurred with PdCl and PtCl , as evidenced by
Faraday constant, N : Avogadro constant, Qguest: charge of
A
II
the guest ion, DQsensor: charge change of the sensor caused
by the redox process, e : vacuum permittivity, e: dielectric
2
2
0
1
H NMR spectroscopic monitoring of the reaction in deuter-
constant of the solvent, d: distance between the redox-active
ated methanol. Again, a downfield shift of the signal of the
pyridyl a-protons of about 0.6 ppm was observed. The prod-
ucts proved to be only sparingly soluble in common organic
solvents and undergo light-induced decomposition with for-
centre and the point charge of the guest).
0
0
nFjDE j ¼ N
Q
guestDQsensor=4 pe
ed
ð1Þ
A
0
mation of elemental palladium and platinum, respectively.
In the present system, the shift of the half-wave potential
induced by metal complexation is rather large in each case.
As expected, the effect is more pronounced in dichlorome-
thane (e=8.93) than in the more polar acetonitrile (e=
4
Only [Cp*Co(h -1)PdCl ] could be obtained in pure form.
The cobaltocenium derivative [Cp*Co(h -1COMe)]
2
5
+
proved to be unable to form complexes with the metal diha-
lides investigated, which is probably due to repulsive elec-
trostatic effects.
[19]
36.64).
The DE1/2 values are similar for the three com-
plexes, since the metal centres have similar ionic radii (vide
supra) and identical ligands.
4
Electrochemistry: Cyclic voltammetry revealed that com-
pound 1 is reversibly reduced to its radical anion. The half-
wave potential is ꢀ1.37 V in CH Cl and ꢀ1.23 V in CH CN
The metal complexes of the type [Cp*Co(h -1)MCl ]
2
show irreversible electrochemical behaviour for M=Zn and
Pd. In the case of zinc, a reduction wave at E =ꢀ1.34 V is
2
2
3
p
[18]
(
vs ferrocenium/ferrocene). Coordination of a cyclopenta-
observed. However, the oxidative process leads to a broad
wave at about ꢀ1.00 V, which is probably due to the forma-
tion of elemental zinc. The palladium complex shows irre-
dienylcobalt(i) fragment to the p system of 1 leads to a
rather dramatic electrochemical effect, which is based on
the fact that electron density is delocalised from the organo-
versible reduction at E =ꢀ1.43 V, whereupon a reversible
p
metallic moiety into the LUMO of
1
(vide supra).
redox wave at E =ꢀ2.12 V is observed. This is close to,
1
/2
4
4
4
[CpCo(h -1)] and [Cp*Co(h -1)] undergo reversible reduc-
but not identical with, the redox potential of [Cp*Co(h -1)]
(ꢀ1.97 V). The nature of the species responsible for this
tion at ꢀ1.89 and ꢀ1.97 V, respectively, in acetonitrile; the
more electron-rich Cp* derivative is harder to reduce.
The redox behaviour of 1 is also rather sensitive to metal
coordination at the pyridyl groups. An investigation of the
redox wave therefore remains unclear. The purity of the
4
platinum complex [Cp*Co(h -1)PtCl ], although not satisfac-
2
tory for analytical purposes, was sufficient for investigations
chelates of type [(1)MX ] revealed that for M=Hg irreversi-
by cyclic voltammetry. Aquasireversible reduction at E1/2 =
2
ble reduction occurred at about ꢀ0.75 V and led to deposi-
tion of elemental mercury on the electrode. The dimeric
cadmium complex also showed irreversible behaviour, with
a broad reduction peak at about ꢀ1.05 V, which may be due
to the formation of monomeric fragments. All other com-
ꢀ1.75 V was observed for this compound. The DE value
1
/2
4
of 0.22 V with respect to [Cp*Co(h -1)] is smaller than that
of 0.37 V observed for 1 and [(1)PtCl ], even when it is
2
taken into account that the latter value was determined in
the less polar solvent dichloromethane.
Chem. Eur. J. 2004, 10, 5661 – 5670
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5665

U. Siemeling et al.
FULL PAPER
Determination of complex-formation constants: The redox
and coordination equilibria in the square scheme shown in
Scheme 3 are the basis for the use of 1 as a molecular elec-
trochemical sensor. We were able to elucidate the thermo-
dynamics involved by a combination of UV/Vis spectroscop-
ic and electrochemical studies.
Figure 9. Absorbance at l=297 nm during the titration of 1 with zinc
chloride (portionwise addition of 2 equiv) in acetonitrile.
wavelength, by using an established procedure (see Experi-
mental Section). In this case, the wavelength of choice is
2
97 nm, since here the largest changes in absorbance occur
during the titration experiment. The extinction coefficient of
at l=297 nm, as determined at 298 K from UV/Vis spec-
tra of this compound by using the Lambert–Beer law (A=
1
ꢀ1
ꢀ1
ecd), is 9130m cm . The corresponding value for the zinc
complex can be approximated from the titration experiment,
under the assumption that it is the only absorbing species in
the presence of an excess of zinc chloride (see Figure 9).
This procedure leads to a value of 14900m cm . The fol-
lowing fitting procedure affords values of 8997m cm and
Scheme 3. Scheme for the binding equilibria of 1 and zinc chloride.
ꢀ
1
ꢀ1
ꢀ
1
ꢀ1
AUV/Vis titration experiment was performed with 1 and
ꢀ
1
ꢀ1
ZnCl in acetonitrile solution (Figure 8). The absorption due
14927m cm for e297 of 1 and [(1)ZnCl ], respectively, and
2
2
6
ꢀ1
to the ligand 1 (lmax =255 nm) decreases gradually, whereas
yields a value of (1.9ꢂ0.2)·10 m for K at 298 K. This cor-
1
responds to a pK value of ꢀ6.276ꢂ0.046.
1
The complexation of zinc chloride by 1 in acetonitrile so-
lution was also monitored electrochemically in a titration
experiment using square-wave voltammetry (Figure 10).
2
Figure 8. Effect of ZnCl on the UV/Vis spectrum of 1 in acetonitrile so-
lution (portionwise addition of 1 equiv).
that of the corresponding zinc complex (lmax =297 nm) in-
creases gradually upon stepwise addition of zinc chloride.
The presence of isosbestic points proves that a simple equili-
brium between just two absorbing species is operative here.
Monitoring the absorbance at a constant wavelength shows
that the zinc containing species is a 1:1 complex (Figure 9),
in accord with the crystallographically characterised chelate
Figure 10. Effect of zinc chloride (portionwise addition of 1 equiv) on the
square-wave voltammogram of 1 (acetonitrile, 0.1m [NBu ][PF ]).
4 6
Addition of a small amount of ZnCl (<10 mol%) leads
2
to the appearance of a new redox wave in the square-wave
voltammogram of 1, cathodically shifted by 0.30ꢂ0.01 V,
which corresponds to an enormous increase of the metal-
binding constant by five orders of magnitude. The two-wave
behaviour observed is characteristic for systems showing
[
(1)ZnCl ] (vide supra).
2
The complex-formation constant K for [(1)ZnCl ] can be
1
2
obtained from the UV/Vis data, provided the extinction co-
efficients of the two absorbing species are known for a given
5666
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5661 – 5670

Redox-Active Chelating Ligands
5661 – 5670
very large complex-formation constants, typically on the
behaviour. In contrast to quinone-based systems, the binding
characteristics of 1 can easily be modulated by attachment
of a suitable organometallic fragment to the redox-active
6
ꢀ1 [2b]
order of 10 m , which is nicely compatible with the value
of ꢀ6.276ꢂ0.046 determined for pK . The pK value for the
1
2
2b]
[
4
reduced species in Scheme 3, as estimated
from K and
moiety. The complex-formation constant of [Cp*Co(h -
1
the electrochemical potential shift of 0.30ꢂ0.01 V, is
1)ZnCl ] is approximately ten times smaller than that of
2
ꢀ
11.35ꢂ0.15 at 298 K.
[(1)ZnCl ].
2
AUV/Vis titration experiment was also performed for
Compound 1 is a new and convenient ligand platform for
use in redox switches and sensors, since cyclopentadienone
derivatives have not been used before in this context. Its
versatility as a ligand is enhanced by the fact that its binding
characteristics can be influenced through secondary interac-
tions involving coordination of the cyclopentadienone p
system. We envisage that fine-tuning of these effects will be
possible by using 14-valence-electron complex fragments
with different stereoelectronic properties. Furthermore, it is
conceivable that attaching other donor sets such as S,S and
O,O to the cyclopentadienone ring may give rise to further
interesting redox-active ligands. Work with the correspond-
4
the complexation of zinc chloride by [Cp*Co(h -1)] in aceto-
nitrile (Figure 11).
[20]
[20,21]
ing 2-thienyl- and 2-furyl-substitued
in progress.
analogues of 1 is
Experimental Section
4
Figure 11. Effect of ZnCl
acetonitrile solution (portionwise addition of 1 equiv).
2
on the UV/Vis spectrum of [Cp*Co(h –1)] in
General: [Cp*Co(C
2
H
4
)
2
] was prepared according to a published proce-
[
22]
dure.
All other compounds were commercially available. Synthetic
work involving air-sensitive compounds was performed under an atmos-
phere of dry argon or nitrogen by using standard Schlenk techniques or a
conventional glove box. Solvents and reagents were appropriately dried
and purified. NMR: Bruker Avance DRX 500 and Varian Unity INOVA
Again, the presence of an isosbestic point shows that a
simple equilibrium between just two absorbing species is in-
volved. By a procedure analogous to that described above
1
500 (500.13 MHz for H); MS: Finnigan MAT 8430 (EI) or PE Biosys-
tems Voyager System 1161 (ESI; MALDI-TOF, 2,5-dihydroxybenzoic
acid calibration matrix); IR: Bruker Vector 22 and Mattson Genesis FT-
IR; UV/Vis: Perkin Elmer Lambda 40, an absorbance correction taking
into account the decreasing analyte concentration during the course of
the titration experiments was performed; elemental analyses: Beller
(Gçttingen), H. Kolbe (Mülheim an der Ruhr) and microanalytical labo-
ratory of the University of Bielefeld.
for [(1)ZnCl ], the complex-formation constant of
2
4
[
(
Cp*Co(h -1)ZnCl ] in acetonitrile was determined to be
2
5
ꢀ1
1.3ꢂ0.2)”10 m at 298 K, with extinction coefficients at
ꢀ1 ꢀ1 4
l=266 nm of 24700 and 23200m cm for [Cp*Co(h -1)]
and [Cp*Co(h -1)ZnCl ], respectively. This result shows that
4
2
Determination of complex-formation constants: For the complex-forma-
tion equilibrium L+MQLM the expression for the equilibrium constant
the presence of the Cp*Co fragment leads to a pronounced
destabilisation of the chelate by approximately one order of
magnitude, for which the main reason is the unfavourable
change in bite angle induced by the organometallic moiety
0
0
L M
M
is K=c c /cLM (c=equilibrium concentration). When cL and c are the
starting concentrations used, the respective equilibrium concentrations
0
L
0
M
are c
L
=c ꢀcLM and c
M
=c ꢀcLM. Substituting these into the expression
0
L
0
M
for K gives K=(c ꢀcLM)(c ꢀcLM)/cLM. From this quadratic equation it
(
vide supra).
follows that the unknown equilibrium concentration of the complex is
0
0
M
0
0
M
2
0 0 0.5
L M
c
LM =(c +c +K)/2ꢀ[(c +c +K) /4ꢀc c ] . The absorbance at a given
L
L
wavelength l is described by the Lambert–Beer law: A/d=ec. In the
Conclusion
present system, two absorbing species are present, namely, L and LM.
0
Therefore, A/d=e
written as A/d=e
L
c
L
+eLM
c
LM. Since c
L
=c ꢀcLM, this equation can be
L
0
0
L
(c ꢀcLM)+eLM
L
c
LM =e
L
c +(eLMꢀe
L
L LM
)cLM. With c =
We have explored the coordination chemistry of 2,5-diphen-
yl-3,4-bis(2-pyridyl)cyclopenta-2,4-dien-1-one (1). This com-
pound can act as a bidentate ligand that forms stable che-
lates with a range of metal dihalides. Pseudotetrahedral,
square-planar and square-pyramidal coordination has been
observed. Compound 1 is a redox-active ligand showing
well-behaved electrochemistry. It can therefore exhibit
switchable binding. In the case of the zinc complex
0
L
0
M
0
0
2
0
0
0.5
(
e
c +c +K)/2ꢀ[(c +c +K) /4ꢀc c ] this can be written as A/d=
L
M
L M
0
L
0
L
0
L
0
M
0
L
0
M
2
L
(c ꢀcLM)+eLM
c
LM =e
L
c +(eLMꢀe
L
){(c +c +K)/2ꢀ[(c +c +K) /
0
L
0
0.5
M
4
ꢀc c ] }. The complex formation constant K can be obtained from this
equation by standard fitting procedures, given that the following values
are known in the UV/Vis titration experiment: V (sample volume at the
s
start), cLs (starting concentration of L in the sample), cMt (concentration
of M in the titre solution), v (volume of titre solution added). During the
titration, the total volume is V=V
centrations at any stage of the titration are c =c (V /V) and c =cMt(v/
s
+v. Consequently, the relevant con-
0
0
L
Ls
s
M
[
(1)ZnCl ], enhancement of the complex-formation constant
L
V). The experimental determination of e is straightforward, and eLM can
2
be obtained by extrapolation from the titration experiment, under the as-
sumption that LM is the only absorbing species in the presence of an
excess of M.
by five orders of magnitude is observed upon one-electron
reduction. This is similar to the best results obtained with
quinone-based switches. Compound 1 can be used as a mo-
lecular electrochemical sensor for the detection of divalent
metal ions in solution, since it shows favourable two-wave
4
Electrochemical investigations: Experiments involving [CpCo(h -1)],
4
4
[
Cp*Co(h -1)] and complexes of the type [Cp*Co(h -1)MCl
2
] were car-
ried out in specially constructed cells containing an internal drying
Chem. Eur. J. 2004, 10, 5661 – 5670
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5667

U. Siemeling et al.
FULL PAPER
1
column with highly activated alumina. The working electrode was a Pt
disc sealed in soft glass (1.0 mm diameter). APt wire was used as the
counterelectrode. Potentials were calibrated with ferrocene [E1/2(ferroce-
nium/ferrocene)=0.38 V versus Ag/AgCl/KCl(aq)]. The setup was con-
trolled by a Jaissle potentiostat IMP88. Material and apparatus for all
(80%). H NMR (CDCl
3
): d=1.69 (s, 15H), 2.33 (s, 3H), 6.68 (d, J=
7.8 Hz), 7.05 (m, 4H), 7.27 (m, 2H), 7.35 (“t”, 4H), 7.42 (m, 2H), 7.53
(m, 2H), 8.51 ppm (d, J=5.4 Hz); C{ H} NMR (CDCl
93.7, 94.1, 99.6, 113.3, 123.9, 125.9, 126.6, 129.1, 129.6, 130.1, 136.4, 149.0,
149.7, 168.7 ppm; MS (MALDI-TOF): m/z: 623 [Cp*Co(h -1COMe)] ;
elemental analysis (%) calcd for C39 CoF P (768.63): C 60.94, H
.72, N 3.64; found: C 60.91, H 4.72, N 3.61.
1
3
1
3
): d=8.9, 20.6,
5
+
[
23]
other electrochemical studies have been described elsewhere.
,5-Diphenyl-3,4-bis(2-pyridyl)cyclopenta-3,4-diol-1-one: The compound
H
36
N
2
6 2
O
4
2
[
5]
5
was prepared previously, but analytical data were unavailable. Pow-
dered potassium hydroxide (2.00 g, 35.6 mmol) was added to a vigorously
stirred solution of dibenzyl ketone (33.3 g, 158 mmol) and 2,2’-pyridil
[Cp*Co(h -1COMe)]BPh
in analogy to the synthesis of the hexafluorophosphate by using a saturat-
ed aqueous solution of NaBPh . It was characterised by X-ray crystallog-
raphy. Elemental analysis (%) calcd for C63 58BCoN (944.91): C
4
: This compound was obtained in 77% yield
4
(
25.1 g, 119 mmol) in ethanol (400 mL). After 1 h, the voluminous white
H
2 2
O
precipitate was filtered off with suction, washed with ethanol (3”30 mL)
80.08, H 6.19, N 2.96; found: C 79.67, H 6.25, N, 3.21.
1
5
and dried in vacuo. Yield 47.5 g (94%). H NMR (CDCl
3
): d=4.78 (s,
H), 6.98 (m, 4H), 7.11 (m, 2H), 7.22 (m, 6H), 7.36 (“d”, apparent J=
.1 Hz, 2H), 7.52 (m, 2H), 8.37 (d, J=5.0 Hz, 2H), 8.74 ppm (s, 2H);
[Cp*Co(h -1Me )](CF SO ) : Methyl triflate (10 mL) was added drop-
3
3
3 3
4
2
8
wise to [Cp*Co(h -1)] (528 mg, 0.91 mmol). The mixture was stirred for
14 h and subsequently reduced to dryness in vacuo. Ethanol (5 mL) was
added to the brownish yellow residue, and the mixture stirred for 10 min.
The yellow solid was filtered off, washed with diethyl ether (5 mL) and
1
3
1
C{ H} NMR (CDCl
3
): d=65.9, 82.1, 122.4, 123.2, 127.1, 127.7, 130.9,
ꢀ
1
1
33.3, 137.3, 146.5, 162.5, 213.3 ppm; IR (KBr): n˜ =1743 (nCO) cm ; MS
+
1
(
EI): m/z: 422 [M] ; elemental analysis (%) calcd for
422.48): C 76.76, H 5.25, N 6.63; found: C 76.92, H 5.40, N 6.62.
: The synthesis was performed by using a dehydration protocol devel-
oped by Hughes et al. for the synthesis of 1,5-di-tert-butyl-1,3-cyclopenta-
C
27
H
22
N
2
O
3
n-hexane (5 mL) and dried in vacuo. Yield 456 mg (47%). H NMR
(
(CD CN): d=1.69 (s, 15H), 3.62 (s, 6H), 4.10 (s, 3H) 7.32 (m, 4H), 7.54
3
(
“t”, 4H), 7.61 (m, 2H), 7.98 (d, J=8.0 Hz, 2H), 8.12 (m, 2H), 8.62 (d,
1
1
3
1
3
J=5.9 Hz, 2H), 8.75 ppm (“t”, 2H); C{ H} NMR (CD CN): d=9.7,
[
24]
47.6, 64.5, 89.2, 93.5, 102.8, 125.3, 127.9, 130.6, 130.7, 131.1, 132.5, 134.0,
diene,
since this gave yields superior to those reported by Eistert
5
[
5]
144.1, 149.2, 151.6 ppm; MS (MALDI-TOF): m/z: 625 [Cp*Co(h -
et al. POCl (15.0 g, 97.8 mmol) was added to a suspension of 2,5-di-
3
+
1
Me
3
)] ; elemental analysis (%) calcd for C43
H
42
N
2
CoF
9
O
10
S
3
(1072.93):
phenyl-3,4-bis(2-pyridyl)cyclopenta-3,4-diol-1-one (21.1 g, 50.0 mmol) in
pyridine (90 mL). The reaction mixture was stirred at 858C for 14 h and
was subsequently allowed to cool to room temperature. Volatile compo-
nents were removed in vacuo. The solid residue was dissolved in di-
chloromethane (500 mL). The solution was cooled to 08C and treated
with a saturated aqueous solution of sodium carbonate (100 mL). After
gas evolution had stopped, the organic layer was separated, washed with
water (2”100 mL) and dried with sodium sulfate. The solution was re-
duced to dryness in vacuo and the brownish red crude product recrystal-
C 48.14, H 3.95, N 2.61; found: C 48.17, H 4.24, N 2.91.
General procedure for the preparation of the chelates of 1 with metal di-
halides: Equimolar amounts of 1 and the respective metal dihalide MX
2
were stirred in ethanol (10 mL per 100 mg of 1) at 708C for 14 h. The
mixture was allowed to cool to room temperature. The solid was filtered
off, washed with diethyl ether and n-hexane and dried in vacuo. The CO
1
3
carbon atom gave rise to a low-intensity C NMR signal, which some-
times could not be detected with certainty. [(1)ZnCl ] was obtained as a
red, microcrystalline solid in 72% yield. H NMR (CDCl ): d=7.23 (m,
12H), 7.58 (m, 2H), 7.70 (m, 2H), 9.06 ppm (d, J=5.5 Hz, 2H); C{ H}
NMR (CDCl ): d=125.0, 128.4, 128.6, 129.0, 129.2, 130.3, 130.9, 140.4,
48.7, 149.5, 152.8 ppm; elemental analysis (%) calcd for
C H N Cl OZn (522.76): C 62.04, H 3.47, N 5.36; found: C 62.00, H
2
1
lised from ethanol to afford the product as dark red crystals. Yield 14.3 g
3
1
13
1
(
74%). H NMR (CDCl
3
): d=7.06 (m, 2H), 7.23 (m, 12H), 7.51 (m,
1
3
1
2
1
2
H), 8.32 ppm (d, J=4.4 Hz, 2H); C{ H} NMR (CDCl
25.0, 126.7, 128.0, 128.2, 130.0, 130.2, 135.9, 149.0, 153.1, 153.2,
3
): d=122.3,
3
1
ꢀ
1
+
00.7 ppm; IR (KBr): n˜ =1715 (nCO) cm ; MS (EI): m/z: 386 [1] ; ele-
O (386.45): C 83.92, H 4.70, N
27 18
2
2
3
.44, N 5.33. [(1)Cd(m-Cl)Cl]
2
was obtained in 63% yield as a violet red,
microcrystalline solid. H NMR ([D ]acetone): d=7.21 (m, 6H), 7.29 (m,
6H), 7.40 (“t”, 2H), 7.70 (m, 2H), 8.95 ppm (m, 2H); C{ H} NMR
[D ]acetone): d=125.5, 127.8, 129.0, 129.4, 130.5, 131.0, 131.2, 140.8,
50.3, 151.4, 154.1 ppm; elemental analysis (%) calcd for
mental analysis (%) calcd for C27
18 2
H N
1
7
.25; found: C 83.93, H 4.74, N 7.23.
6
1
3
1
4
[
CpCo(h -1)]: Asolution of 1 (3.86 g, 10.0 mmol) and [CpCo(CO)
2
]
(
1
6
(
1.81 g, 10.1 mmol) in p-xylene (20 mL) was heated to reflux for 14 h.
The mixture was allowed to cool to room temperature. The dark red, mi-
C
54
H
36
N
4
Cd
2
Cl
4
O
2
(1139.54): C 56.92, H 3.18, N 4.92; found: C 56.46, H
] and [(1)HgBr ] were obtained as a violet red, mi-
crocrystalline solids in 70% yield. Their NMR spectroscopic data are vir-
crocrystalline precipitate was isolated by filtration, washed with n-hexane
3
.26, N 4.66. [(1)HgCl
2
2
1
(
3”20 mL) and dried in vacuo. Yield 4.54 g (89%). H NMR (CDCl
3
):
d=4.98 (s, 5H), 7.09 (“t”, 2H), 7.15 (d, J=7.7 Hz, 2H), 7.22 (m, 6H),
1
tually identical. H NMR (CD
2
Cl
m, 2H), 7.63 (m, 2H), 8.89 ppm (d, J=4.7 Hz, 2H); C{ H} NMR
CD Cl ): d=124.7, 126.8, 128.6, 128.8, 129.4, 129.9, 130.3, 139.4, 149.1,
50.5, 153.3, 199.5 ppm. [(1)PdCl ] was obtained as a dark red, microcrys-
talline solid in 48% yield. H NMR (CD Cl ): d=7.13 (d, J=7.8 Hz,
H), 7.34 (m, 10H), 7.39 (m, 2H), 7.62 (m, 2H), 9.07 ppm (d, J=5.1 Hz,
2
): d=7.20 (m, 6H), 7.28 (m, 6H), 7.39
7
3
1
1
.42 (“t”, 2H), 7.63 (“d”, apparent J=3.8 Hz, 4H), 8.41 ppm (d, J=
1
3
1
(
(
1
1
3
1
.4 Hz, 2H); C{ H} NMR (CDCl
3
): d=79.1, 85.9, 92.8, 122.1, 126.7,
2
2
27.1, 127.8, 128.9, 130.3, 133.1, 135.2, 149.1, 153.6 ppm; IR (KBr): n˜ =
ꢀ1
+
2
593 (nCO) cm ; MS (ESI): m/z: 511 [M] ; elemental analysis (%) calcd
CoO (510.49): C 75.29, H 4.54, N 5.49; found: C 76.04, H
1
2
2
for C32
23 2
H N
2
2
1
4
.93, N 4.96.
13
1
H); C{ H} NMR (CD
29.8, 140.3, 146.4, 153.3, 153.6, 198.9 ppm; elemental analysis (%) calcd
Cl OPd (563.60): C 57.54, H 3.22, N 4.97; found: C 56.91, H
.46, N 4.61. [(1)PtCl ] was obtained as a brown, microcrystalline solid in
9% yield. H NMR (CD Cl ): d=7.11 (d, J=7.9 Hz, 2H), 7.33 (m,
2 2
Cl ): d=126.6, 127.9, 128.1, 128.7, 128.9, 129.1,
4
[
Cp*Co(h -1)]: Asolution of 1 (3.86 g, 10.0 mmol) and [Cp*Co(C
2 4 2
H ) ]
(
2.50 g, 10.0 mmol) in toluene (100 mL) was heated to 508C for 14 h. The
for C27
3
2
H
18
N
2
2
mixture was allowed to cool to room temperature. The red, microcrystal-
2
line precipitate was isolated by filtration, washed with n-hexane (2”
1
2
2
1
13
1
2
0 mL) and dried in vacuo. Yield 4.94 g (85%). H NMR (CDCl
3
): d=
12H), 7.58 (m, 2H), 9.05 ppm (d, J=5.6 Hz, 2H); C{ H} NMR
(CD Cl ): d=125.9, 127.1, 128.4, 128.7, 128.8, 129.3, 129.6, 140.0, 147.2,
1
.45 (s, 15H), 6.95 (d, J=7.8 Hz, 2H), 7.05 (“t”, 2H), 7.14 (m, 6H), 7.35
2
2
(
2
“t”, 2H), 7.72 (“d”, apparent J=6.8 Hz, 4H), 8.45 ppm (d, J=4.3 Hz,
27 18 2 2
154.4, 154.5, 199.3 ppm; elemental analysis (%) calcd for C H N Cl OPt
(652.31): C 49.72, H 2.78, N 4.29; found: C 48.62, H 2.57, N 3.95.
1
3
1
H); C{ H} NMR (CDCl
3
): d=8.0, 77.8, 90.9, 92.6, 121.6, 126.2, 126.3,
1
27.2, 130.0, 132.6, 135.0, 148.9, 150.0, 154.9 ppm; IR (KBr): n˜ =1562
CO) cm ; MS (ESI): m/z: 581 [M] ; elemental analysis (%) calcd for
4
General procedure for the preparation of the chelates of [Cp*Co(h -1)]
ꢀ
1
+
(
n
4
with metal dichlorides: Equimolar amounts of [Cp*Co(h -1)] and the re-
2
were stirred in ethanol (10 mL per
00 mg of 1) at 708C for 14 h. The mixture was allowed to cool to room
temperature. The solid was filtered off, washed with diethyl ether and n-
hexane and dried in vacuo. The CO carbon atom gave rise to a low-inten-
sity C NMR signal, which could not be detected with certainty.
[Cp*Co(h -1)ZnCl ] was obtained as a dark red, microcrystalline solid in
64% yield. H NMR (CD Cl ): d=1.52 (s, 15H), 7.23 (m, 12H), 7.57 (m,
2H), 7.85 (m, 2H), 9.16 ppm (brs, 2H); C{ H} NMR (CD CN): d=9.2,
88.4, 92.0, 97.4, 126.3, 127.7, 129.2, 129.6, 132.6, 133.0, 141.1, 149.7,
C
37
H
33
N
2
CoO (580.62): C 76.54, H 5.73, N 4.82; found: C 75.98, H 5.72,
spective metal dichloride MCl
1
N 4.77.
5
[
Cp*Co(h -1COMe)]PF
6
: Acetyl chloride (6.0 mL) was added dropwise
4
to [Cp*Co(h -1)] (181 mg, 0.31 mmol). The resultant yellow solution was
stirred for 1 h and subsequently reduced to dryness in vacuo. The remain-
ing light orange solid was dissolved in ethanol (5 mL). The yellow prod-
uct was precipitated by dropwise addition of a saturated aqueous solution
of ammonium hexafluorophosphate, filtered off, washed with diethyl
ether (10 mL) and n-hexane (10 mL) and dried in vacuo. Yield 234 mg
1
3
4
2
1
2
2
1
3
1
3
5668
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5661 – 5670

Redox-Active Chelating Ligands
5661 – 5670
1
51.4 ppm; elemental analysis (%) calcd for C37
H
33
N
2
Cl
2
CoOZn (716.91):
94.4, 125.8, 128.3, 128.5, 131.1, 131.9, 137.7, 155.4 ppm; no satisfactory el-
emental analysis could be obtained.
4
C 61.99, H 4.64, N 3.91; found: C 61.45, H 5.21, N 4.30. [Cp*Co(h -
1
)PdCl
2
] was obtained as a dark red, microcrystalline solid in 67% yield.
]methanol): d=1.57 (s, 15H), 7.20 (d, J=7.8 Hz, 2H), 7.27
m, 4H), 7.35 (m, 6H), 7.59 (m, 2H), 7.93 (m, 2H), 9.07 ppm (d, J=
4
X-ray crystallography: X-ray crystallographic data for 1, [CpCo(h -1)],
1
4
5
5
H NMR ([D
(
4
[
Cp*Co(h -1)], [Cp*Co(h -1COMe)]BPh
4 3 3 3 3
, [Cp*Co(h -1Me )](CF SO ) ,
[(1)PtCl ], [(1)ZnCl ], [(1)HgCl ], [(1)HgBr ], [{(1)Cd(m-Cl)Cl }] and
[Cp*Co(h -1)ZnCl ] are collected in Table 3. Data collection was per-
2
2
2
2
2
2
1
3
1
4
5
.3 Hz, 2H); C{ H} NMR ([D
4
]methanol): d=7.2, 84.6, 92.2, 97.4, 123.3,
1
26.3, 126.4, 129.2, 130.0, 136.1, 152.5, 153.7 ppm; elemental analysis (%)
formed with
a a Siemens P21 four-circle diffractometer for 1 and
5
calcd for C37
H
33
N
2
Cl
2
OPd (757.76): C 58.65, H 4.39, N 3.70; found: C
[Cp*Co(h -1COMe)]BPh and with a Nonius Kappa CCD diffractometer
for all other compounds. Graphite-monochromatised MoKa radiation (l=
0.71073 ꢁ) was used in each case. The structures were solved by direct
methods. Programs used were Siemens SHELXTL PLUS
4
4
5
7.74, H 4.16, N 3.56. [Cp*Co(h -1)PtCl
2
] was obtained as a dark brown,
Cl ): d=1.47 (s, 15H),
.02 (d, J=8.0 Hz, 2H), 7.27 (m, 10H), 7.40 (m, 2H), 7.68 (m, 2H),
1
microcrystalline solid in 62% yield. H NMR (CD
2
2
[
25]
7
9
and
1
3
1
[26]
2
.14 ppm (d, J=5.0 Hz, 2H); C{ H} NMR (CD
2
Cl
2
): d=9.2, 84.1, 85.2,
SHELXL 97.
Full-matrix least-squares refinement on F was carried
Table 3. X-ray crystallographic data.
4
4
5
5
Compound
1
[CpCo(h -
[Cp*Co(h -1)]
[Cp*Co(h -1COMe)]
[Cp*Co(h -
2 2 2
[(1)PtCl ]·2CH Cl
1
)]·0.5CHCl
3
[BPh
4
]
1Me
3
)](CF
3
SO
3
)
3
empirical formula
formula weight
crystal system
space group
a [ꢁ]
C
27
H
18
N
2
O
C
32.5
H
23.5Cl1.5CoN
2
O
C
580.58
monoclinic
P2
12.5750(1)
10.8040(1)
21.2930(3)
95.1240(7)
2812.57(5)
4
37
H
33CoN
2
O
C
944.85
monoclinic
P2
17.520(9)
14.882(5)
20.277(5)
112.16(5)
4896(3)
4
63
H
58BCoN
2
O
2
C
43
H
42CoF
9
N
2
O
10
S
3
29 6 2
C H22Cl N OPt
822.28
orthorhombic
386.43
monoclinic
P2 /c
19.021(6)
9.188(3)
11.532(3)
99.28(2)
1989.0(10)
4
570.14
monoclinic
P2
9.5320(1)
30.7260(5)
9.5900(1)
111.8831(8)
2606.34(6)
4
1072.90
monoclinic
P2 /c
19.2730(2)
10.7090(1)
23.9510(3)
114.8440(5)
4485.87(8)
4
1
1
1
/n
1
/c
1
P
2
1
na
20.085(4)
11.399(2)
26.266(5)
b [ꢁ]
c [ꢁ]
b [8]
V [ꢁ]
3
6014(2)
8
1.816
5.226
3184
Z
ꢀ
3
1
calcd [gcm
]
1.290
0.079
808
1.453
0.842
1172
1.371
0.644
1216
1.282
0.399
1992
1.589
0.620
2200
ꢀ
1
m [mm
]
F(000)
crystal size [mm]
q range [8]
reflections collected
independent reflections
0.5”0.4”0.3 0.24”0.05”0.04
0.26”0.15”0.04 0.3”0.2”0.1
0.23”0.21”0.07
3.0–27.5
81348
10249
0.049
0.30”0.27”0.20
2.0–30.0
16869
2.2–30.0
6053
5782
3.0–25.0
30621
8980
3.0–27.5
48577
6428
1.9–22.5
6533
6282
0.0773
3400
0
16869
R
int
0.0437
0.067
0.062
reflections with I>2s(I) 3241
restraints
6665
4940
8090
0
16070
1
0
29
0
parameters
271
1.011
0.0680/
742
1.027
0.0483/0.1142
502
1.041
0.0391/0.1034
430
1.015
0.0862/0.1961
781
1.011
0.0335/0.0828
704
0.992
0.0251/0.0580
2
GOF on F
[
a]
[b]
R1 (I>2s(I))/wR2
0.1490
largest diff. peak/
0.315/
ꢀ0.297
0.527/ꢀ0.285
0.542/ꢀ0.513
0.565/ꢀ0.461
0.495/ꢀ0.460
1.450/ꢀ1.462
ꢀ
3
hole [eꢁ
]
4
Compound
[(1)ZnCl
2
]·CHCl
OZn
3
[(1)HgCl
2
]
[(1)HgBr
2
]·0.5C
HgN
3
H
6
O
[(1)Cd(m-Cl)Cl]
Cl
2
·C
3
H
6
O
[Cp*Co(h -1)ZnCl
2
]·CH
2 2
Cl
empirical formula
formula weight
crystal system
space group
a [ꢁ]
C
28
H
19Cl
5
N
2
C
27
H
18Cl
2
HgN
2
O
C
28.5
H
21Br
2
2
O
1.5
C
57
H
42Cd
2
4
N
4
O
3
C
38
H
35Cl
4
CoN
2
OZn
642.07
monoclinic
P2 /n
6.8760(1)
21.4650(2)
18.8700(2)
96.3740(4)
2767.87(6)
4
657.92
orthorhombic
Pbca
13.0660(8)
14.8700(8)
23.5860(10)
775.88
orthorhombic
Pbcn
27.0840(1)
13.3410(1)
14.5320(2)
1197.55
monoclinic
P2 /c
12.0220(2)
21.0170(3)
20.3800(3)
94.3850(7)
5134.27(14)
4
801.78
monoclinic
P2 /n
9.0860(1)
21.9900(3)
17.2380(2)
94.6140(5)
3433.01(7)
4
1
1
1
b [ꢁ]
c [ꢁ]
b [8]
V [ꢁ]
3
4582.6(4)
8
1.907
6.974
2528
5250.81(8)
8
1.963
8.933
2944
Z
ꢀ
3
1
calcd [gcm
]
1.541
1.395
1296
1.549
1.085
2400
1.551
1.528
1640
ꢀ
1
m [mm
]
F(000)
crystal size [mm]
q range [8]
reflections collected
0.34”0.08”0.05
2.2–30.0
15732
0.16”0.14”0.09
2.7–30.0
78864
0.32”0.30”0.08
3.0–25.0
72990
0.24”0.06”0.05
3.1–27.5
88169
0.30”0.06”0.06
2.2–30.0
19109
independent reflections
8043
6676
4592
11771
9996
R
int
0.0231
0.0786
0.065
0.068
0.0329
reflections with I>2s(I)
restraints
6332
0
5344
0
4033
0
8643
0
7752
0
parameters
334
1.026
0.0308/0.0743
0.661/ꢀ0.669
298
1.083
390
1.060
0.0297/0.0829
2.141/ꢀ1.684
783
1.032
0.0346/0.0833
1.074/ꢀ0.661
429
1.020
0.0329/0.0724
0.411/ꢀ0.489
2
GOF on F
[
a]
[b]
R1 (I>2s(I))/wR2
largest diff. peak/hole [eꢁ
0.0390/0.0751
1.537/ꢀ1.804
0.5
ꢀ
3
]
2
o
2
c
2
2
o
2
[
a] R1=ꢀjjF jꢀjF jj/ꢀjF
o
c
o
j. [b] wR2={ꢀ[w(F ꢀF ) ]/ꢀ[w(F ) ]}
.
Chem. Eur. J. 2004, 10, 5661 – 5670
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5669

U. Siemeling et al.
FULL PAPER
out anisotropically for the non-hydrogen atoms. Hydrogen atoms were
included at calculated positions using a riding model, except for the non-
Chinn, Jr., M. B. Hall, Organometallics 1984, 3, 284–288; c) R. Hoff-
mann, P. Hoffmann, J. Am. Chem. Soc. 1976, 98, 598–604.
[8] This effect has long been known. See, for example: E. Weiss, W.
Hübel, J. Inorg. Nucl. Chem. 1959, 11, 42–55.
4
5
disordered
Me )](CF
hydrogen
atoms
of
[Cp*Co(h -1)],
[Cp*Co(h -
1
3
3
SO , [(1)HgBr
3
)
3
2
] and [(1)Cd(m-Cl)Cl]
2
, which were refined
isotropically. CCDC-238172–CCDC-238182 contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.uk/conts/retrieving.html (or from the Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2
[9] T.-L. Ho, Chem. Rev. 1975, 75, 1–20.
[10] P. Jutzi, U. Siemeling, A. Müller, H. Bçgge, Organometallics 1989, 8,
1744–1750.
[11] W. L. Steffen, G. J. Palenik, Acta Crystallogr. Sect. B 1976, 32, 298–
300.
1
EZ, UK; fax: (+44)1223-336-033; or deposit@ccdc.cam.uk).
[
[
[
12] K. Satoh, T. Suzuki, K. Sawada, Monatsh. Chem. 2001, 132, 1145–
1
155.
13] A. J. Canty, C. L. Raston, B. W. Skelton, A. H. White, J. Chem. Soc.
Dalton Trans. 1982, 15–18.
14] C. J. Matthews, W. Clegg, S. L. Heath, N. C. Martin, M. N. Stuart
Hill, J. C. Lockhart, Inorg. Chem. 1998, 37, 199–207.
Acknowledgement
This work was supported by the Fonds der Chemischen Industrie. We
thank Prof. Wilhelm Knoche, Dr. Jçrg Kleimann and Priv.-Doz. Dr.
Rainer Winter for helpful discussions, and Ottilie Thorwarth for electro-
chemical measurements.
[15] a) D. M. P. Mingos, Essential Trends in Inorganic Chemistry, Oxford
University Press, Oxford, 1998, Chaps. 3.5, 5.5; b) J. Huheey, E.
Keiter, R. Keiter, Anorganische Chemie, 2nd ed., de Gruyter,
Berlin, 1995, Chap. 4.
[
[
16] J. A. Tossell, D. J. Vaughan, Inorg. Chem. 1981, 20, 3333–3340.
17] P. Colamarino, P. L. Orioli, J. Chem. Soc. Dalton Trans. 1975, 1656–
1659.
[
1] Recent reviews: a) T. Hirao, Coord. Chem. Rev. 2002, 226, 81–91;
b) A. M. Allgeier, C. A. Mirkin, Angew. Chem. 1998, 110, 936–952;
Angew. Chem. Int. Ed. 1998, 37, 894–908.
[18] All potentials are referred to the ferrocenium/ferrocene couple,
whose formal electrode potential versus SCE is 0.46 V in dichloro-
[
2] Recent reviews: a) A. E. Kaifer, M. Gꢂmez-Kaifer, Supramolecular
Electrochemistry, Wiley-VCH, Weinheim, 1999, Chap. 10; b) A. E.
Kaifer, S. Mendoza in Comprehensive Supramolecular Chemistry,
Vol. 1 (Ed.: G. W. Gokel), Pergamon, Oxford, 1996, Chap. 19.
3] Selected reviews: a) I. Haiduc, F. T. Edelmann, Supramolecular Or-
ganometallic Chemistry, Wiley-VCH, Weinheim, 1999, Chap. 2.1.5;
b) P. D. Beer, Adv. Inorg. Chem. 1992, 39, 79–157; c) P. D. Beer,
Chem. Soc. Rev. 1989, 18, 409–450.
4 6
methane and 0.40 V in acetonitrile (0.1m [NBu ]PF supporting elec-
trolyte); see N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96,
877–910.
[19] CRC Handbook of Chemistry and Physics, 82nd ed. (Ed.: D. R.
Lide), CRC Press, Boca Raton, FL, 2001, Table 8-127.
[20] W. Ried, A. Gçbel, Chem. Ber. 1967, 100, 3405–3412.
[21] P. Bergmann, H. Paul, Chem. Ber. 1967, 100, 828–835.
[22] S. A. Frith, J. L. Spencer, Inorg. Synth. 1985, 23, 15–21.
[23] U. Siemeling, B. Neumann, H.-G. Stammler, O. Kuhnert, Polyhedron
1999, 18, 1815–1819.
[
[
4] See, for example, a) A. Gouloumis, R. C. Lawson, P. Vꢃzquez, L.
Echegoyen, T. Torres, Tetrahedron 2002, 58, 961–966; b) M. Gꢂmez-
Kaifer, P. A. Reddy, C. D. Gutsche, L. Echegoyen, J. Am. Chem.
Soc. 1994, 116, 3580–3587.
[24] R. P. Hughes, A. S. Kowalski, J. R. Lomprey, D. R. Neithamer, J.
Org. Chem. 1996, 61, 401–404.
[
[
[
5] B. Eistert, G. Fink, M. A. El-Chahawi, Liebigs Ann. Chem. 1967,
[25] G. M. Sheldrick, SHELXTL PLUS, Siemens Analytical Instruments,
Madison, WI, 1990.
[26] G. M. Sheldrick, SHELXL 97, University of Gçttingen, Gçttingen,
1997.
7
03, 104–108.
6] H. van Willigen, W. E. Geiger, Jr., M. D. Rausch, Inorg. Chem. 1977,
6, 581–584.
1
7] See, for example, a) R. Gleiter, R. Roers, F. Rominger, B. Nuber, I.
Received: May 12, 2004
Hyla-Kryspin, J. Organomet. Chem. 2000, 610, 80–87; b) J. W.
Published online: October 7, 2004
5670
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5661 – 5670