Redox-responsive rhodocenium [O,O]-, [N,O]-, [N,N]-, and [N,C,N]-metalloligands

Article abstract of DOI:10.1002/ejic.201100226

As we have recently shown, the doubly substituted [1,2-O,O]H, [1,2-O,N]H, [1,2-N,N]H, and [1,3-N,N]H pentafulvenes are convenient N/O-functionalized cyclopentadienide (Cp) precursors. Deprotonation of the pentafulvenes by potassium hydride followed by rea

Full text of DOI:10.1002/ejic.201100226

FULL PAPER
DOI: 10.1002/ejic.201100226
Redox-Responsive Rhodocenium [O,O]-, [N,O]-, [N,N]-, and
[N,C,N]-Metalloligands
Daniela Eisenstecken,[a] Barbara Enk,[a] Holger Kopacka,[a] Klaus Wurst,[a]
Thomas Müller,[b] Florian Pevny,[c] Rainer F. Winter,*[c] and Benno Bildstein*[a]
Keywords: Rhodium / Metallocenes / N ligands / N,O ligands / Electrochemistry
As we have recently shown, the doubly substituted [1,2-
erties of these novel metallocene metalloligands in solution
O,O]H, [1,2-O,N]H, [1,2-N,N]H, and [1,3-N,N]H pentafulv- and in the solid state were studied by NMR spectroscopy (1H,
enes are convenient N/O-functionalized cyclopentadienide
(Cp) precursors. Deprotonation of the pentafulvenes by po-
tassium hydride followed by reaction with [Cp*RhCl2]2 or
[Cp*IrCl2]2 gave access to the first functionalized rhodocen-
ium and iridocenium salts that contain two acyl and/or imid-
oyl substituents. These air-stable compounds represent inter-
esting bis(acyl/imidoyl) or mixed acyl/imidoyl metalloligand
systems that combine an axially shielding, electrochemically
active metallocene moiety with directly attached, conjugat-
ing oxygen and/or nitrogen donor sites. The structural prop-
13C, 103Rh) and by single-crystal structure analysis. The elec-
trochemical investigations on complexes 5a, 6b, and 7a
showed the effect of the appended functional groups on the
potential and on the reversibility of the rhodocenium/rhodo-
cene reduction and provided evidence for the dimerization
that followed the reduction. Further redox processes that
were due to the heteroatom functions included the stepwise
reduction of the 1,2-diketo entity of 5a and the oxidation of
the 1,2-diimino moiety of 7a.
plications usually require metallocenes with appended func-
tional groups, and hence the design of task-specific func-
tionalized metallocenes is a worthwhile endeavour in syn-
thetic organometallic chemistry. The predominance of the
ferrocene compounds in this area is due to their ready avail-
ability by a wide range of synthetic protocols, the fully re-
versible ferrocene/ferrocenium redox couple, the biocom-
patibility of the nontoxic ferrocene moiety, and the low cost
of iron in comparison to the expensive platinum group met-
als, Ru, Os, Rh, and Ir.
Introduction
In comparison to the fully developed chemistry of ferro-
cene and its derivatives, there are surprisingly few reports
in the literature of the isoelectronic cobaltocenium, rhodo-
cenium, and iridocenium compounds. The parent, unsubsti-
tuted, air-stable group 9 metallocenium salts (Cp)2M+X–
(M = CoIII [1] RhIII [2] IrIII [2a,3]
, ; X = monovalent anion)
,
were synthesized shortly after the discovery of ferrocene,[4]
but only scattered studies dealt with such sandwich com-
pounds, as can be inferred from the scarce number of re-
ported compounds within the last 6 decades (Co: Ͼ 2000,
Rh: Ͼ 80, Ir: Ͼ 10; for comparison: Fe: Ͼ 65000).[5] In
general, the importance of the metallocenes (predominantly
ferrocenes) stems from their useful properties as the steering
ligands in (stereoselective) catalysis, in redox sensing, in
supramolecular chemistry, as components of NLO materi-
als, and in biochemical or medicinal applications. These ap-
In recent work[6] we developed an efficient modular route
to pentafulvenes that bear 2/3-acyl/imidoyl-6-hydroxy/
amino substituents. These fulvenes are new ambidentate li-
gands that are capable of forming either κ2-O/N metal com-
plexes or η5-cyclopentadienyl sandwich complexes with the
acyl/imidoyl substituents, depending on the hardness of the
metal electrophile. With “arenophilic”, electron-rich, soft,
late transition metal moieties like Cp*Ru+, the correspond-
ing disubstituted pentamethylruthenocenes[6] are conve-
niently accessible. This raised the question as to whether
this chemistry could be extended to the isoelectronic
[a] Institute of General, Inorganic and Theoretical Chemistry,
Faculty of Chemistry and Pharmacy, University of Innsbruck,
Innrain 52a, 6020 Innsbruck, Austria
group 9 metal sandwich complexes of CoIII, RhIII, and IrIII
.
Fax: +43-512-507-2934
E-mail: benno.bildstein@uibk.ac.at
In this contribution, we report on the first doubly func-
tionalized pentamethylrhodocenium salts, which are based
on our recent account on the isoelectronic ruthenocene
compounds.[6] In addition, we have shown in one example
that the corresponding iridocenium compounds may be
synthesized analogously. As mentioned above, only a few
functionalized rhodocenium compounds are known. Struc-
[b] Institute of Organic Chemistry, Faculty of Chemistry and
Pharmacy, University of Innsbruck,
Innrain 52a, 6020 Innsbruck, Austria
[c] Department of Chemistry, University of Konstanz,
Universitätsstraße 10, 78457 Konstanz, Germany
Fax: +49-7531-88-3136
E-mail: rainer.winter@uni-konstanz.de
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201100226.
2958
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 2958–2966

Redox-Responsive Rhodocenium N/O-Metalloligands
turally all of them contain only one substituent per cyclo- in combination with the strongly electron-donating penta-
pentadienyl ligand that gives rise to symmetrical 1,1Ј-disub- methylcyclopentadienide coligand.
stituted metallocene cations.[7] In contrast, these new penta-
methylrhodocenium compounds bear acyl and/or imidoyl
substituents in the 1,2- or 1,3-positions, thereby making
useful redox-active metallocene metalloligands accessible
that complement the ubiquitous ferrocene compounds. The
synthesis as well as the spectroscopic, structural, and elec-
trochemical properties of these compounds are reported
and discussed.
In contrast to the failure with CoIII, successful reactions
were achieved with RhIII and IrIII (Scheme 1). The depro-
tonation of the difunctionalized pentafulvenes[6] 1–4b in a
tetrahydrofuran solution by using excess potassium hydride
(KH) cleanly afforded the intermediate disubstituted cyclo-
pentadienide potassium salts. In our experience, KH is a
convenient, non-nucleophilic, selective base that enables
clean and complete deprotonations in organometallic chem-
istry. It is a heterogeneous reaction, and the dihydrogen
evolution that occurs during the deprotonation shifts the
equilibrium to the product side, and the observation of gas
evolution is a convenient way to monitor the course of the
reaction. After the deprotonation is complete, the excess
KH is easily removed by filtration with the exclusion of air.
The subsequent transmetalation with [Cp*RhCl2]2 or, in
one case, with [Cp*IrCl2]2 at ambient temperature over the
course of 3–5 d afforded the desired metallocenes 5a–8b in
30–98% isolated yield as yellow to red-brown, air-stable
materials. We focused on the rhodocenium salts 5a, 6a–6d,
7a,b, and 8a,b. The preparation of the iridocenium chloride
5b was performed mainly to prove the applicability of this
synthetic protocol to the 5d-metal Ir. We refrained from
synthesizing the other iridocenium salts due to the highly
expensive Ir starting materials, but it is generally accepted
that RhIII and IrIII show analogous chemical behavior.
The pentamethylmetallocenium salts 5a,b, 7a,b, and 8a,b
Results and Discussion
Synthesis
Firstly, we attempted the synthesis of 1,2-dibenzoyl-
1Ј,2Ј,3Ј,4Ј,5Ј-pentamethylcobaltocenium iodide by the reac-
tion of deprotonated 2-benzoyl-6-hydroxy-6-phenylpenta-
fulvene[8] with [Cp*CoI2]2.[9] No reaction occurred, which
indicated the insufficient stability of the desired cobaltocen-
ium species. Electronically, 1,2-diacylcyclopentadienide is
an acceptor-substituted cyclopentadienide that forms stable
sandwich metallocenes only if the second Cp ligand is able
to compensate for the loss of electron density with the com-
bination of the strong Cp–metal bonding. As in the case of
FeII, where no such metallocenes have been observed,[10] the
Cp–metal bonds for CoIII do not seem to be strong enough
to support these acceptor-substituted cyclopentadienides.
Therefore, we concluded that only the 4d (Ru,[6] Rh) and contain two identical acyl or imidoyl substituents on the
5d (Os,[11] Ir) d6-metal ions give stable compounds of this second Cp ligand and are therefore achiral, mixed-ligand,
type due to their intrinsically stronger carbon–metal bonds symmetrical metallocenes. In contrast, the 1-acyl-2-imidoyl-
Scheme 1. Synthesis of the doubly functionalized pentamethylrhodocenium/iridocenium salts 5a–8b.
Eur. J. Inorg. Chem. 2011, 2958–2966
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2959

R. F. Winter, B. Bildstein et al.
FULL PAPER
1Ј,2Ј,3Ј,4Ј,5Ј-pentamethylrhodocenium salts 6a–6d are race- N (6a) or intermolecular N–H–N (8a,b) hydrogen bonds in
mic, planar-chiral metallocenes due to their unsymmetrical, the presence of water during the workup. In our opinion,
enantiotopic O,N-disubstituted Cp ligand. For the sake of this possibility is less likely, because the trifluoromethylated
simplicity, only one of the two enantiomers is shown in rhodocenium salts are significantly weaker Brønsted bases
Scheme 1. It is likely that optical resolution of these racemic than the 1,2-bis(imidoyl)rhodocenium chlorides 7a,b, in
mixtures could have been achieved by fractional crystalli- which a strong intramolecular N–H–N hydrogen bond is to
zation of the diasteromeric salts containing anions from the be expected, as analogously demonstrated by the corre-
chiral pool. Rhodocenium salt 6b was the only compound sponding ruthenocene compounds, where such a species has
in this series that had tetraphenylborate as the counterion been unambiguously characterized by a single-crystal struc-
due to the superior workup and its purification properties ture analysis.[6b]
(see Experimental Section). The bis(imidoyl)pentamethyl-
1
In addition to standard H and 13C NMR spectroscopy,
rhodocenium chlorides 7a,b and 8a,b both have the same 103Rh NMR spectroscopy[12] is, in principle, applicable for
type of substituents but have a different substitution the characterization of 5a–8b, with the exception of the iri-
pattern. If a selective ortho-metalation at C-2 of the 1,3- docenium salt 5b. Although the spin 1/2 nuclide 103Rh is
bis(imidoyl)-Cp ligand is possible, then 8a,b are precursors 100% abundant, its detection by NMR spectroscopy suffers
to tridentate, monoanionic [N,C,N]-pincer ligands. 7a,b, on from its very weak magnetic strength with concomitant
the other hand, are bidentate, neutral [N,N]-metalloligands. very low sensitivity in comparison to other spin 1/2 nuclei.
This makes 103Rh one of the most difficult spin 1/2 metal
nuclei to measure by transition metal NMR spectroscopy.
However, state of the art FT-NMR spectroscopy enables
Spectroscopic and Structural Properties
the direct measurement of 103Rh within a reasonable data
In solution, the diagnostic NMR properties of the penta- acquisition time. Because our inhouse NMR spectrometer
methylmetallocenium salts 5a–8b are the strong singlet sig- is unsuitable for such 103Rh NMR measurements, one se-
nals of the five isochronic methyl groups of their Cp* ligand lected example, 5a, was measured at an external NMR facil-
that is detected at δ ≈ 2.0 (1H) and 10.0 ppm (13C). The acyl ity (Bruker Analytical Services, Vienna, Austria). A 103Rh
and imidoyl functionalities are evident from their 13C NMR chemical shift of δ = –28.5 ppm was observed [vs.
signals that are observed at δ = 188–194 (C=O) and 161– Rh(acac)3 as the external standard], and no other signal
169 ppm (C=N). The resolved scalar metal–carbon cou- was detected within the extremely large 103Rh shift range of
plings in the 13C NMR spectra (d, J13C-103Rh = 4–8 Hz) over 12000 ppm. This observed chemical shift is extremely
1
between the central rhodium atom and the carbon atoms deshielded in comparison to the unsubstituted, parent rho-
of both of the Cp rings allowed the chemical shifts of these docenium salt, Cp2Rh+Cl– (δ = –10147 ppm).[13,14] A cer-
carbon atoms to be assigned, except for the iridocenium tain degree of deshielding for 5a was expected due to its
compound 5b. The substitution pattern of the 1,2- and 1,3- two benzoyl acceptor substituents, but such a large differ-
bis(imidoyl)rhodocenium salts 7a,b and 8a,b is evident from ence (over 10000 ppm) cannot be explained by simple sub-
1
the multiplicity of the H NMR signals of the substituted stituent effects. At present we are unable to explain this be-
Cp ring hydrogen atoms (7a,b: t and d; 8a,b: s and s).
havior, but it would seem that 103Rh NMR spectroscopy is
In addition to these diagnostic NMR properties, the vari- of limited value for simple structure-related correlations
ous alkyl/aryl substituents of 5a–8b gave rise to their ex- based on the limited data available.
pected signals and coupling patterns (see Experimental Sec-
The positive-mode FAB mass spectrometry of the penta-
tion). Whereas the NMR spectra of the non-fluorinated methylrhodocenium/iridocenium salts 5a–8b gave strong
rhodocenium/iridocenium salts are well resolved, less clean signals for their corresponding cations, which thereby fur-
and slightly broadened spectra are observed for the trifluor- ther proved the identity of these salts. In addition, the diag-
omethylated compounds 6c, 8a, and 8b. We postulate that nostic stretching vibrations of the carbonyl (ν
= 1644–
˜
C=O
the most likely reason for this behavior is the increased re- 1686 cm–1) and imine (νC=N = 1615–1646 cm–1) functionali-
˜
activity of the trifluoromethyl-substituted acyl (6c) and ties were observed in the IR spectra (see Experimental Sec-
imidoyl (8a,b) functionalities towards water, which stabi- tion).
lizes, to some minor extent, their hemiketal (6a/H2O) and
The solid-state structures are available for three represen-
hemiaminal (8a,b/H2O) forms, respectively. It is well known tatives of the rhodocenium salts. Figures 1, 2, and 3 show
that such species can be detected by NMR spectroscopy the molecular structures of the cations of 5a, 6b, and 7a,
and/or isolated if electron-withdrawing groups are pres- respectively; their counterions and the solvent molecules are
ent.[11] In our case one trifluoromethyl substituent and a neglected for clarity. Overall, regular pentamethylmetallo-
cationic rhodocene moiety are present. However, the cene sandwich structures were observed with one of the
attempts to shift this equilibrium completely to the hy- acyl/imidoyl substituents roughly coplanar with the substi-
drated or non-hydrated form by treatment with diluted tuted Cp ring, whereas the second acyl/imidoyl substituent
aqueous acid or base were unsuccessful, and no better re- is approximately orthogonal to the Cp plane. The orthogo-
solved NMR spectra were obtained. A second possible nal substituents exert some steric hindrance on the upper
reason for the broadened spectra of compounds 6c, 8a, and Cp* ring, which results in a slight deviation from planarity
8b could be the partial formation of intramolecular O–H– of the Cp and Cp* planes [5a: 3.4(1)°; 6b: 1.8(2)°; 7a:
2960
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 2958–2966

Redox-Responsive Rhodocenium N/O-Metalloligands
8.9(6)°]. The average distances from the Rh center to the As shown recently in one case of the isoelectronic [N,N]
Cp* carbon atoms are shorter by approximately 3 pm in ruthenocenes,[6b] intramolecular chelate formation is pos-
comparison to those of the Cp carbon atoms. This reflects sible.
the higher electron density of the Cp* ligand and the
slightly reduced metal–ligand bonding of the acceptor-sub-
stituted Cp ligand. In the structures with imidoyl groups
(6b, 7a), the N-aryl groups are strongly tilted with respect
to the C=N π-plane, which is due to steric hindrance by
their 2,6-substituents. At first sight one might think that
these [O,O], [N,O], and [N,N] metalloligands are incapable
of forming chelate metal complexes due to their strongly
tilted donor groups; however, we attribute their different
orientation mainly to the crystal-packing forces and ex-
pected the normal coordination chemistry for these ligands.
Figure 3. Molecular structure of the cation of 7a; the solvent mo-
lecules (CH2Cl2) are omitted for clarity. Selected bond lengths [Å]:
C(11)–N(1) 1.278(8), C(24)–N(2) 1.280(8).
Electrochemical Properties
A possible advantage of the metallocene metalloligands
over conventional diketone, imino ketone, or diimine che-
lates is the redox activity associated with the metallocene
scaffolds. As an initial test of the redox properties of these
systems, one representative from each of the complex series
5, 6, and 7, namely 5a, 6b, and 7a, was chosen for the elec-
trochemical studies. The parent rhodocene, like other
metallocenes of the heavier 4d and 5d congeners of the
group 6–9 metals, is a chemically reactive species and exists
Figure 1. Molecular structure of the cation of 5a; the solvent mo-
only at low temperatures in a true sandwich structure. The
warming of the parent rhodocene caused rapid dimerization
lecules (H2O) are omitted for clarity. Selected bond lengths [Å]:
C(11)–O(1) 1.213(2), C(18)–O(2) 1.220(2).
to a bis(η5-cyclopentadienyl)(μ-η4:η4-dihydrofulvalene)di-
rhodium structure where both rhodium atoms attain their
18 valence electron configuration. This dimerization can be
reversed at high temperatures by means of scission of the
dihydrofulvalene linkage.[15] The high chemical reactivity of
the rhodocenes was also indicated by the peculiarities ob-
served in the electrochemical experiments on the reduction
of the rhodocenium cations. The parent rhodocenium was
reduced at –1.81 V in CH3CN/Bu4NPF6 to give rhodocene,
the half-life of which was estimated to be 2 s at room tem-
perature from the ratio of the reverse and forward peak
currents. A small anodic peak at –0.9 V on the reverse scan
indicated the formation of the fulvalene-bridged dimer,
which was prepared and identified by bulk electrolysis.[16]
The replacement by one or two of the Cp rings with steri-
cally more demanding and more electron-rich Cp* ligands
(Cp* = η5-C5Me5) caused a shift in the first reduction po-
tential to –2.03 or –2.38 V in THF/Bu4NPF6.[17] The poten-
tials of the original references were transformed to the
Cp2Fe0/+ scale by the conversion factors given by Geiger
Figure 2. Molecular structure of the cation of 6b; the solvent mo-
lecules (CH2Cl2) are omitted for clarity. Selected bond lengths [Å]:
and Connelly.[18] The corresponding methylated rhodocenes
C(11)–N(1) 1.272(3), C(24)–O(1) 1.208(3).
proved no less reactive than their unmethylated parent, but
Eur. J. Inorg. Chem. 2011, 2958–2966
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2961

R. F. Winter, B. Bildstein et al.
FULL PAPER
they followed different reaction pathways. Cp*CpRh
rapidly transformed to the “mixed” dihydrofulvalene
CpRh(μ-η4:η4-C5Me5–C5H5)RhCp*. The dimerization of
Cp*2Rh was, however, hampered by the permethylation of
both of the rings such that the diene complex Cp*(η4-
C5Me5H)Rh formed by hydrogen abstraction from the sol-
vent. The rhodocene derived from the indenyl derivative,
Cp*(η5-C9H7)Rh+, nevertheless proved to be completely
stable towards dimerization and hydrogen abstraction,
possibly because of the facile “indenyl shift”, and gave
Cp*(η3-C9H7)Rh with a 17 valence electron configuration
at the metal atom.
The 1,2-diketo-, 1,2-ketoimine-, and 1,2-diimine-func-
tionalized complexes 5a, 6b, and 7a followed the same gene-
ral behavior as the other Cp*CpRh+ cations, but the dif-
ferent substituents had major effects on the redox potentials
and reactivities of the resulting rhodocenes. All of these
rhodocenium ions underwent a primary reduction process
in CH2Cl2/Bu4NPF6 (0.1 m). Its one-electron nature was as-
certained by comparison of the forward peak currents to
those of the ferrocene or decamethylferrocene standard
(Table 1).
Figure 4. Cyclic voltammetry of complex (a) 5a (at r.t.), (b) 6b (at
r.t.), and (c) 7a (at –70 °C, 0.1 m NBu4PF4/CH2Cl2) at v = 0.4 V/s.
Table 1. Redox potentials [V] of the rhodocenium salts 5a, 6b, and
7a.
than that obtained for 5a (peak C in Figure 4b, c; see
Table 1). It was clear from the hold-scan experiments, which
had different delays starting from potentials that were more
negative than the rhodocenium/rhodocene couple, and from
the comparative cathodic scans, where the sweep was
clipped before traversing the reduction wave, that this peak
is due to a product formed from the rhodocene. The scans
that were performed at various concentrations of the rho-
docenium ions, but under otherwise identical conditions,
showed that when the analyte concentration was increased
+/0
E1/2
Ep,follow
Other waves
5a
6b
7a
–1.59
–1.805
–1.80
–0.96
–1.20
–1.14, –1.23
–2.04, –2.29[a], 0.73[b]
–0.36[c]
0.73[b], 0.99[d]
[a] Half-wave potentials for the stepwise reduction of the 1,2-diketo
functionality at v = 0.1 V/s, T = –70 °C. [b] Peak potential for the
chloride oxidation. [c] Peak potential of the Ph4B– oxidation. [d]
Half-wave potential for the 1,2-diimine oxidation at v = 0.1 V/s, T
= –70 °C.
Not surprisingly, the diketo-functionalized 5a displayed the chemical reversibility decreased, which was measured
the least negative reduction potential of –1.59 V (–1.60 V at by the peak current ratios of the reoxidation peak B and
–70 °C) and also the highest degree of chemical reversibility the forward reduction peak A, and the amount of follow
for the +/0 wave (see Figure 4a). Thus, at room temperature, product that formed increased, which was measured by the
a small cathodic peak at –0.96 V (peak C in Figure 4a) is ratio of the current of peak C to that of peak A. This
seen at the intermediate sweep rates (v) but not at the low pointed to a reaction of higher order in rhodocene concen-
v (where the follow product, which is not present in the bulk tration, which is in line with the formation of a dimer.
solution, escaped by means of diffusive losses), or at the Peak D, which was only observed in the case of 7a (the most
high v (where the follow reaction was suppressed), or at any bulky of all of the rhodocenium ions), did not seem to show
sweep rate at –70 °C. The substitution of one or both of the concentration dependence and may thus be due to the diene
keto functions by imino groups caused a distinct shift in complex that forms as a result of hydrogen abstraction (see
E1/2 to more negative values of about –1.80 V (Table 1). Al- Figures S1–S6 of the Supporting Information).
though complexes 6b and 7a have very similar redox poten-
At more negative potentials, complex 5a showed two
tials, they dramatically differ in the stability of the corre- more reduction waves in addition to the rhodocenium/rho-
sponding rhodocene. Thus, whereas the rhodocene associ- docene couple. Owing to the rapid degradation of these
ated with 6b is only slightly more reactive than that formed higher reduced forms and the somewhat sluggish electron-
from 5a, the rhodocene associated with 7a is much more transfer kinetics that led to severe wave broadening at the
reactive so that cooling to –70 °C and sweep rates of higher sweep rates, these waves were best observed at –70 °C
Ͼ 0.2 V/s were required in order to observe the anodic and at low sweep rates (see Figure 5). They are most proba-
counterpeak of the 7a/7a– couple (see Figure 4b, c).
bly associated with the stepwise reduction of the 1,2-diketo-
As for 5a and other rhodocenium ions, traversing the re- cyclopentadienide to the keto enolate and then the 1,2-di-
duction wave also caused one (6b) or two (7a) new anodic methylenediolate forms, possibly with a hapticity change of
peaks to appear on the backward scan. The potential of the reduced cyclopentadienide ligands as sketched in
these anodic peaks was more than 200 mV more negative Scheme 2. No further reductions were observed for 6b and
2962
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 2958–2966

Redox-Responsive Rhodocenium N/O-Metalloligands
7a within the potential window of the CH2Cl2/Bu4NPF6 ral imino ketones that are new chiral [N,O]-ligand systems
supporting electrolyte. During the oxidative scans the com- and are useful precursors for higher denticity ligand sys-
plex salts 5a, 6b, and 7a also displayed the typical irrevers- tems by condensation of their acyl functionality, were pre-
ible oxidation waves of the corresponding counterions with pared. Thirdly, the 1,2-bis(imidoyl) derivatives, which are
the anodic peak of the Cl–/Cl2 couple near +0.73 V and the interesting diimine ligands whose steric protection can be
peak of the Ph4B– Ǟ Ph3B + 1/2 biphenyl process at tuned through the appended N-aryl substituents, were pro-
+0.36 V. Only in the case of 7a, the most electron-rich me- duced. And finally, the 1,3-bis(imidoyl) derivatives, which
tallocene metalloligand, was an additional oxidation wave are regioisomers of the former family of compounds and
observed. Whereas the process was completely irreversible are interesting precursors to the novel metallocene [N,C,N]-
at room temperature, it developed into a quasireversible fea- pincer ligand systems, were prepared.
ture at E1/2 = +0.99 V at –70 °C (Figure S7 of the Support-
These rhodocenium salts behave as regular metallocenes
ing Information). It was thus attributed to the oxidation of with the expected spectroscopic and structural properties as
the bis(imine) functionality.
was shown by 1H and 13C NMR spectroscopy, IR spec-
troscopy, and FAB mass spectrometry. Due to the presence
of the spin 1/2 and 100% abundant 103Rh nucleus, diagnos-
tic 1J(13C-103Rh) couplings were observed in their 13C
NMR spectra. The single-crystal structure analyses of the
three representatives showed more or less regular metallo-
cenes with strongly tilted acyl/imidoyl substituents. The
electrochemical investigations showed that the rhodocen-
ium ions may be reduced to the corresponding rhodocenes
and that the reduction potentials and the chemical reactiv-
ity of the thus formed rhodocenes depend on the substitu-
ents. We expect that the electrochemical properties are fur-
ther modified by metal coordination of the donor functions
and that redox chemistry can be used to alter the electron-
donating properties of these metallocene metalloligands.
In future work we plan to exploit the coordination chem-
istry of these new N/O-ligand systems. We are especially
interested in the steric and electronic influence of the di-
rectly attached pentamethylrhodocenium moiety on the re-
activity, physical properties, and catalytic applications of
the metal complexes of these new metalloligands.
Figure 5. Cyclic voltammetry of complex 5a (0.1 m Bu4NPF4/
CH2Cl2, –70 °C) at v = 0.1 V/s.
Scheme 2. Ligand-centered reductions of complex 5a.
Experimental Section
General Considerations: All of the reactions and manipulations of
the air- and/or moisture-sensitive compounds were carried out un-
der dry argon by using standard Schlenk techniques. The solvent
tetrahydrofuran (THF) was dried with and distilled from Na under
argon prior to use. The chemicals were commercially obtained and
used as received. The fulvenes 1–4b were prepared as published
recently.[6] The NMR spectra were recorded with a Bruker Avance
DPX 300 (300 MHz) spectrometer (1H, 13C) or with a Bruker Av-
ance III 400 (400 MHz) spectrometer (103Rh); the chemical shifts
are reported in ppm relative to Si(CH3)4 (1H, 13C) or Rh(acac)3
Conclusions
The readily available pentafulvenes that bear one acyl or
imidoyl substituent and one further tautomerized acyl or
imidoyl group are convenient precursors for doubly N/O-
substituted cyclopentadienide ligands. After deprotonation
of the ligand precursors by potassium hydride, reaction
with an arenophilic pentamethylcyclopentadienylrhodi-
um(III) moiety yielded the first rhodocenium salts with two
acyl/imidoyl functionalities on one cyclopentadienyl ligand.
It was shown in one case that analogous chemistry is pos-
sible with IrIII, starting from the respective pentamethylcy-
clopentadienyliridium(III) chloride. These air-stable salts
are new “metalloligands” that are composed of an electro-
chemically responsive pentamethylrhodocenium core and
two directly attached ketone or imine functionalities. Four
different families of compounds were prepared. Firstly, the
1,2-diacyl derivatives, which are interesting diketones for
further chemical transformations by condensation reactions
with functionalized amines, were synthesized. Secondly, the
(
103Rh).[14] The IR spectra were recorded with a THERMO Nicolet
5700 ATR-FTIR spectrometer. The melting points were measured
with a Leica Galen Kofler microscope. The mass spectra were re-
corded with a Finnegan MAT 95 mass spectrometer.
Representative Procedure for the Synthesis of 5a–8b: A Schlenk tube
was charged under argon with pentafulvene 1–4b (0.81 mmol), dry
THF (20 mL) and a stirring bar. After the solution had been cooled
to 0 °C, an excess of potassium hydride (approximately 100 mg,
2.49 mmol) was added in one portion with the exclusion of air. The
external cooling bath was removed, and the stirred mixture was
warmed to ambient temperature. During this period dihydrogen
evolution was observed with the concomitant darkening of the
1-acyl-2-imidoyl derivatives, which are racemic, planar-chi- solution, which indicated the formation of the potassium salt of
Eur. J. Inorg. Chem. 2011, 2958–2966
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2963

R. F. Winter, B. Bildstein et al.
FULL PAPER
1
the deprotonated functionalized fulvene. After the mixture had
been stirred at room temperature overnight, the excess, unreacted
potassium hydride was filtered off, and [Cp*RhCl2]2 (250 mg,
0.40 mmol) was added. The solution was stirred under argon at
room temperature for 4 d. The volatile materials were removed by
means of a vacuum line, the residue was exposed to air and then
hydrolyzed by the addition of water. The organic materials were
extracted at neutral pH with dichloromethane (3ϫ), and the or-
ganic layers were combined and dried with extraction with brine
and the addition of Na2SO4. The volatile materials were removed
with a rotary evaporator. The crude product mixture was dissolved
in dichloromethane (3 mL) and chromatographed through a short
column of silica. Elution with dichloromethane afforded the unre-
acted ligand precursor, and elution with methanol afforded the
pure product as a yellow to red-brown microcrystalline solid after
solvent removal.
13C) = 6.2 Hz, Cp], 97.9 [d, J(103Rh-13C) = 6.5 Hz, Cp], 103.5 [d,
1
1J(103Rh-13C) = 8.4 Hz, Cp*], 109.3 [d, J(103Rh-13C) = 6.8 Hz,
3
Cp], 122.1 (Ph), 123.9 (Ph), 125.9 [q, J(11B-13C) = 2.2 Hz, Ph of
tetraphenylborate], 127.0 (Ph), 127.9 (Ph), 128.4 (Ph), 128.5 (Ph),
130.6 (Ph), 135.4 (Ph), 136.3 (Ph), 147.6 (Ph), 160.9 (C=N), 164.2
[q, 1J(11B-13C) = 49.1 Hz, Ph of tetraphenylborate], 192.8 (C=O)
ppm. MS (FAB pos): m/z = 552.12 (M+ of cation). IR (ATR): ν =
˜
610 (m), 695 (s), 706 (s), 733 (s), 840 (m), 903 (m), 1019 (w), 1148
(w), 1210 (w), 1263 (m), 1343 (w), 1379 (w), 1419 (m), 1450 (m),
1472 (m), 1579 (w), 1637 (m, C=N), 1686 (m, C=O), 2852 (w), 2914
(w), 2985 (w), 3052 (w), 3092 (w) cm–1. C56H55BNORh (871.78):
calcd. C 77.15, H 6.36, N 1.61; found C 77.22, H 6.39, N 1.59.
Single-crystal structure analysis of 6b: Table 2, Figure 2.
1-(2-{[(2,6-Dimethylphenyl)imino](phenyl)methyl}-1Ј,2Ј,3Ј,4Ј,5Ј-penta-
methylrhodocen-1-yl)-2,2,2-trifluoroethanone Chloride (6c): Starting
materials: 1-(5-{1-[(2,6-dimethylphenyl)amino]-1-phenylmethyl-
1,2-Dibenzoyl-1Ј,2Ј,3Ј,4Ј,5Ј-pentamethylrhodocenium Chloride (5a): ene}cyclopenta-1,3-dien-1-yl)-2,2,2-trifluoroethanone (2b) (299
Starting materials: 2-benzoyl-6-hydroxy-6-phenylpentafulvene (1)
(222 mg, 0.81 mmol) and [Cp*RhCl2]2 (250 mg, 0.40 mmol). Yield:
353 mg (80%). M.p. 135 °C. 1H NMR (300 MHz, CD2Cl2, 298 K):
δ = 2.04 (s, 15 H, Cp*CH3), 6.22 (d, 2 H, CpH), 6.73 (unresolved
t, 1 H, CpH), 7.42 (t, 4 H, o-PhH), 7.56 (t, 2 H, p-PhH), 7.77 (d,
4 H, m-PhH) ppm. 13C NMR (300 MHz, CD2Cl2, 298 K): δ = 10.0
mg, 0.81 mmol) and [Cp*RhCl2]2 (250 mg, 0.40 mmol). Yield:
444 mg (86%). H NMR (300 MHz, CD2Cl2, 298 K): δ = 1.25 (br.
1
s, 3 H, CH3), 1.61 (br. s, 3 H, CH3), 2.23 (br. s, 15 H, Cp*CH3),
4.21 (q, 1 H, OH), 4.67 (br. s, 3 H, CpH), 6.70–7.07 (br. m, 6 H,
PhH), 7.51–7.71 (br. m, 2 H, PhH) ppm. 13C NMR (300 MHz,
CD2Cl2, 298 K): δ = 11.1 (Cp*CH3), 11.2 (Cp*CH3), 14.1 (CH3),
68.3 (Cp), 103.8 (br., Cp), 127.0, 128.1, 128.2, 128.3, 128.6, 128.8,
129.0, 129.4, 131.2 (Ph), 167.9 (C=N) ppm. MS (FAB pos): m/z =
1
1
(Cp*CH3), 91.7 [d, J(103Rh-13C) = 6.2 Hz, Cp], 92.6 [d, J(103Rh-
1
13C) = 6.2 Hz, Cp*], 103.8 [d, J(103Rh-13C) = 8.2 Hz, Cp], 129.0
(o-Ph), 129.2 (m-Ph), 134.4 (p-Ph), 136.3 (Ph), 188.7 (C=O) ppm.
606.08 (M+ of cation). IR (ATR): ν = 705 (m), 769 (m), 871 (m),
˜
103Rh NMR (400 MHz, CD2Cl2, 300 K): δ = –28.5 ppm. MS (FAB 952 (w), 1023 (m), 1130 (m), 1169 (s), 1214 (m), 1290 (w), 1335
pos): m/z = 511.08 (M+ of cation). IR (ATR): ν = 671 (m), 695 (s),
(w), 1375 (w), 1446 (m), 1468 (m), 1584 (m), 1606 (m), 1722 (w),
˜
716 (s), 740 (s), 839 (m), 856 (s), 931 (s), 1024 (m), 1174 (m), 1249 2856 (w), 2919 (w), 3034 (w), 3355 (w) cm–1. C32H32ClF3NORh
(s), 1338 (s), 1379 (m), 1448 (m), 1468 (w), 1575 (m), 1588 (m), (641.97): calcd. C 59.87, H 5.02, N 2.18; found C 60.01, H 5.00, N
1644 (s, C=O), 2914 (w), 3052 (w), 3350 (m) cm–1. C29H28ClO2Rh 2.15.
(546.90): calcd. C 63.69, H 5.16; found C 63.90, H 5.18. Single-
crystal structure analysis: Table 2, Figure 1.
1-(2-{[(2,6-Diisopropylphenyl)imino](phenyl)methyl}-1Ј,2Ј,3Ј,4Ј,5Ј-
pentamethylrhodocen-1-yl)ethanone Chloride (6d): Starting materi-
1,2-Dibenzoyl-1Ј,2Ј,3Ј,4Ј,5Ј-pentamethyliridocenium Chloride (5b):
Starting materials: 2-benzoyl-6-hydroxy-6-phenylpentafulvene (1)
(362 mg, 1.32 mmol) and [Cp*IrCl2]2 (500 mg, 0.63 mmol). Yield:
213 mg (30%). 1H NMR (300 MHz, CDCl3, 298 K): δ = 2.07 (s,
15 H, Cp*CH3), 6.15 (s, 2 H, CpH), 6.95 (s, 1 H, CpH), 7.37–7.42
als: 1-(5-{1-[(2,6-diisopropylphenyl)amino]-1-phenylmethylene}-
cyclopenta-1,3-dien-1-yl)ethanone (2c) (240 mg, 0.81 mmol) and
1
[Cp*RhCl2]2 (200 mg, 0.32 mmol). Yield: 324 mg (78%). H NMR
(300 MHz, CD2Cl2, 298 K): δ = 0.77 [d, 3 H, CH(CH3)2], 1.07 [d,
3 H, CH(CH3)2], 1.21 [dd, 6 H, CH(CH3)2], 2.14 (s, 15 H,
(m, 4 H, o-PhH), 7.51–7.53 (m, 2 H, p-PhH), 7.74–7.77 (m, 4 H, Cp*CH3), 2.39 [s, 3 H, C(O)CH3], 2.59 [sept, 1 H, CH(CH3)2], 3.22
m-PhH) ppm. 13C NMR (300 MHz, CDCl3, 298 K): δ = 10.1
(Cp*CH3), 84.6 (Cp), 86.9 (Cp), 97.1 (Cp*), 128.8–129.4 (br. signal,
Ph), 134.6 (Ph), 136.0 (Ph), 188.4 (C=O) ppm. MS (FAB pos): m/z
[sept, 1 H, CH(CH3)2], 5.88 (br. s, 1 H, CpH), 6.38 (br. s, 1 H,
CpH), 6.82 (br. s, 1 H, CpH), 6.95–7.05 (m, 4 H, PhH), 7.10–7.15
(m, 4 H, PhH) ppm. 13C NMR (300 MHz, CD2Cl2, 298 K): δ =
10.6 (Cp*CH3), 21.9 [CH(CH3)2], 23.2 [CH(CH3)2], 24.2 [CH-
= 600.93 (M+ of cation). IR (ATR): ν = 696 (s), 714 (m), 796 (w),
˜
858 (m), 938 (m), 1023 (m), 1072 (m), 1246 (m), 1335 (m), 1375 (CH3)2], 25.3 [CH(CH3)2], 28.1 [CH(CH3)2], 28.3 [CH(CH3)2], 29.5
1
1
(w), 1446 (m), 1575 (m), 1595 (s), 1650 (s, C=O), 1722 (w), 2910
(w), 2954 (w), 3373 (m) cm–1. C29H28ClIrO2 (636.21): calcd. C
54.75, H 4.44; found C 54.88, H 4.45.
(CH3), 91.0 [d, J(103Rh-13C) = 4.3 Hz, Cp], 91.5 [d, J(103Rh-13C)
= 6.1 Hz, Cp*], 95.2 (unresolved d, Cp), 123.6, 123.8, 124.6, 128.2,
128.7, 130.4, 134.7, 135.1, 137.4, 145.2 (Ph), 161.1 (C=N), 194.1
(C=O) ppm. MS (ESI pos): m/z = 608.18 (M+ of cation). IR
1-(2-{[(2,6-Dimethylphenyl)imino](phenyl)methyl}-1Ј,2Ј,3Ј,4Ј,5Ј-penta-
methylrhodocen-1-yl)ethanone Chloride (6a)/Tetraphenylborate (6b):
Starting materials: 1-(5-{1-[(2,6-dimethylphenyl)amino]-1-
phenylmethylene}cyclopenta-1,3-dien-1-yl)ethanone (2a) (511 mg,
1.62 mmol) and [Cp*RhCl2]2 (500 mg, 0.81 mmol). Yield: 940 mg
(99%) of 6a. M.p. 115 °C. For analytical purposes 6a was converted
into 6b by anion exchange in methanolic solution. Spectral data
for 6b: 1H NMR (300 MHz, CD2Cl2, 298 K): δ = 1.87 (s, 3 H,
CH3), 1.96 (s, 15 H, Cp*CH3), 2.11 (s, 3 H, CH3), 2.36 [s, 3 H,
C(O)CH3], 5.03 (t, 1 H, CpH), 5.24 (d, 1 H, CpH), 5.57 (d, 1 H,
CpH), 6.87–6.91 (m, 6 H, PhH), 6.99–7.06 (m, 11 H, PhH), 7.19
(t, 2 H, PhH), 7.28 (d, 1 H, PhH), 7.35–7.36 (unresolved m, 8 H,
PhH) ppm. 13C NMR (300 MHz, CD2Cl2, 298 K): δ = 10.6
(Cp*CH3), 18.7 [C(O)CH3], 28.9 (CH3), 88.8 [d, 1J(103Rh-13C) =
6.5 Hz, Cp], 89.5 [d, 1J(103Rh-13C) = 6.6 Hz, Cp], 91.2 [d, 1J(103Rh-
(ATR): ν = 696 (s), 762 (s), 902.8 (w), 952 (w), 1026 (m), 1152 (w),
˜
1232 (m), 1343 (m), 1357 (m), 1379 (m), 1456 (s), 1584 (w), 1633
(m, C=N), 1682 (s, C=O), 2861 (w), 2914 (w), 2959 (m), 3057 (w),
3373 (m) cm–1. C36H43ClNORh (644.11): calcd. C 67.13, H 6.73,
N 2.17; found C 67.33, H 6.75, N 2.15.
1,2-Bis{[(2,6-dimethylphenyl)imino](phenyl)methyl}-1Ј,2Ј,3Ј,4Ј,5Ј-
pentamethylrhodocenium Chloride (7a): Starting materials: 2,6-di-
methyl-N-(1-{5-[1-(2,6-dimethylphenyl)amino-1-phenylmethylene]-
cyclopenta-1,3-dien-1-yl}-1-phenylmethylene)aniline (3a) (778 mg,
1.62 mmol) and [Cp*RhCl2]2 (500 mg, 0.81 mmol). Yield: 535 mg
(44%). 1H NMR (300 MHz, CD2Cl2, 298 K): δ = 1.48 (s, 6 H,
CH3), 2.06 (s, 6 H, CH3), 2.22 (s, 15 H, Cp*CH3), 5.31 (t, 1 H,
CpH), 5.85 (d, 2 H, CpH), 6.75 (d, 4 H, PhH), 6.85–6.88 (m, 2 H,
PhH), 6.95–6.98 (m, 4 H, PhH), 7.12 (t, 4 H, PhH), 7.23 (t, 2 H,
2964
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 2958–2966

Redox-Responsive Rhodocenium N/O-Metalloligands
PhH) ppm. 13C NMR (300 MHz, CD2Cl2, 298 K): δ = 11.3
PhH) ppm. 13C NMR (300 MHz, CD3CN, 298 K): δ = 10.5 (br.,
(Cp*CH3), 17.9 (CH3), 19.2 (CH3), 90.4 [d, 1J(103Rh-13C) = 7.2 Hz, Cp*CH3), 89.6 (br., Cp), 95.9 (br., Cp), 104.6 (br., Cp), 115.8,
1
1
Cp], 91.5 [d, J(103Rh-13C) = 6.8 Hz, Cp], 103.4 [d, J(103Rh-13C)
= 8.1 Hz, Cp*], 105.5 [d, 1J(103Rh-13C) = 6.2 Hz, Cp], 123.5, 124.8,
126.2, 128.0, 128.4, 130.3, 135.7, 148.1 (Ph), 161.7 (C=N) ppm. MS
117.0, 119.3, 122.7, 126.1, 126.5, 129.5, 129.7, 129.9, 130.8, 148.5
(br., Ph) ppm. MS (FAB pos): m/z = 645.13 (M+ of cation). IR
(ATR): ν = 694 (s), 765 (s), 872 (w), 987 (w), 1023 (m), 1129 (s),
˜
(FAB pos): m/z = 717.30 (M+ of cation). IR (ATR): ν = 693 (s),
1165 (s), 1250 (w), 1375 (w), 1446 (m), 1468 (m), 1486 (m), 1592
˜
720 (s), 760 (s), 885 (m), 956 (w), 1019 (m), 1085 (m), 1157 (w),
(s), 1646 (m, C=N), 2852 (w), 2914 (w), 2963 (w), 3373 (w) cm–1.
1219 (w), 1246 (m), 1343 (m), 1379 (w), 1437 (m), 1468 (m), 1588 C31H28ClF6N2Rh (680.93): calcd. C 54.68, H 4.14, N 4.11; found
(m), 1610 (m), 1628 (m, C=N), 2848 (w), 2914 (w), 2963 (w), 3057
(w), 3364 (m) cm–1. C45H46ClN2Rh (753.24): calcd. C 71.76, H
6.16, N 3.72; found C 71.85, H 6.19, N 3.74. Single-crystal struc-
ture analysis: Table 2, Figure 3.
C 54.79, H 4.17, N 4.08.
1,3-Bis{1-[(2,6-dichlorophenyl)imino]-2,2,2-trifluoroethyl}-1Ј,2Ј,3Ј,
4Ј,5Ј-pentamethylrhodocenium Chloride (8b): Starting materials: N-
[2,2,2-trifluoro-1-(5-{2,2,2-trifluoro-1-[(2,6-dichlorophenyl)amino]-
ethylene}cyclopenta-1,3-dien-2-yl)ethylene]-2,6-dichloroaniline (4b)
(464 mg, 0.85 mmol) and [Cp*RhCl2]2 (250 mg, 0.40 mmol). Yield:
507 mg (77%). 1H NMR [300 MHz, CD3S(O)CD3, 313 K]: δ = 2.0
(br. s, 15 H, Cp*CH3), 5.30 (s, 1 H, CpH), 5.72 (s, 1 H, CpH), 6.29
(s, 1 H, CpH), 6.99 (t, 2 H, PhH), 7.35 (d, 4 H, PhH), 7.61 (t, 2
1,2-Bis{[(2,6-dichlorophenyl)imino](phenyl)methyl}-1Ј,2Ј,3Ј,4Ј,5Ј-
pentamethylrhodocenium Chloride (7b): Starting materials: 2,6-
dichloro-N-(1-{5-[1-(2,6-dichlorphenyl)amino-1-phenylmethylene]-
cyclopenta-1,3-dien-1-yl}-1-phenylmethylene)aniline (3b) (287 mg,
0.51 mmol) and [Cp*RhCl2]2 (150 mg, 0.24 mmol). Yield: 272 mg
(67%). 1H NMR (300 MHz, CD2Cl2, 298 K): δ = 2.19 (s, 15 H, H, PhH) ppm. 13C NMR (300 MHz, CDCl3, 298 K): δ = 10.0–11.2
Cp*CH3), 5.31 (t, 1 H, CpH), 5.94 (d, 2 H, CpH), 6.65, 6.81, 6.83, (br., Cp*CH3), 87.5 (br., Cp), 90.9 (br., Cp), 97.6 (br., Cp), 103.6
6.86, 7.05, 7.08, 7.11–7.13, 7.16, 7.19 (16 H, PhH) ppm. 13C NMR (br., Cp), 115.9, 116.0, 121.8, 122.9, 125.2, 127.4, 128.1, 129.1,
1
(300 MHz, CD2Cl2, 298 K): δ = 11.2 (Cp*CH3), 90.8 [d, J(103Rh-
129.4, 131.2, 135.2, 142.3, 147.0 (Ph), 169.1 (C=N) ppm. MS (FAB
1
13C) = 6.4 Hz, Cp], 91.9 [d, J(103Rh-13C) = 6.5 Hz, Cp], 103.8 [d,
pos): m/z = 782.75 (M+ of cation). IR (ATR): ν = 717 (m), 739 (s),
˜
1
1J(103Rh-13C) = 8.0 Hz, Cp*], 104.1 [d, J(103Rh-13C) = 6.5 Hz,
771 (s), 787 (s), 872 (m), 987 (m), 1027 (m), 1063 (m), 1085 (m),
Cp], 124.1, 124.7, 125.2, 127.7, 128.0, 128.3, 128.9, 130.7, 131.1, 1133 (s), 1170 (m), 1232 (w), 1259 (w), 1294 (w), 1357 (w), 1375
135.2 (Ph), 166.4 (C=N) ppm. MS (FAB pos): m/z = 798.87 (M+
(w), 1432 (s), 1481 (m), 1584 (m), 1619 (m, C=N), 2914 (w), 2954
of cation). IR (ATR): ν = 694 (s), 725 (s), 745 (s), 768 (s), 789 (s),
(w) cm–1. C31H24Cl5F6N2Rh (818.71): calcd. C 45.48, H 2.95, N
˜
885 (m), 956 (w), 1023 (s), 1067 (m), 1228 (w), 1259 (s), 1343 (w), 3.42; found C 45.40, H 2.98, N 3.39.
1375 (w), 1427 (s), 1459 (w), 1548 (m), 1579 (w), 1615 (m, C=N),
Electrochemistry: The electrochemical work was performed with a
1722 (w), 1735 (w), 2848 (w), 2914 (w), 2959 (w), 3061 (w) cm–1.
C45H42Cl5N2Rh (891.02): calcd. C 60.66, H 4.75, N 3.14; found C
60.89, H 4.78, N 3.11.
BAS CV50 potentiostat in a custom-made vacuum-tight one-com-
partment cell that used a 1.6 mm diameter Pt-disk working elec-
trode from BAS, a platinum spiral as the counter electrode, and a
silver spiral as a pseudoreference electrode. Each of the spiral-
shaped electrodes was welded to Vycon wire and sealed into a glass
tube. The counter and reference electrodes were introduced into the
cell by appropriate fittings in the side-wall and were sealed by using
a Quickfit screw. The potential calibration was performed by add-
ing an appropriate amount of ferrocene or decamethylferrocene
1Ј,2Ј,3Ј,4Ј,5Ј-Pentamethyl-1,3-bis[2,2,2-trifluoro-1-(phenylimino)eth-
yl]rhodocenium Chloride (8a): Starting materials: N-(2,2,2-trifluoro-
1-{5-[2,2,2-trifluoro-1-(phenylamino)ethylene]cyclopenta-1,3-dien-
2-yl}ethylene)aniline (4a) (212 mg, 0.52 mmol) and [Cp*RhCl2]2
(160 mg, 0.26 mmol). Yield: 143 mg (40%). 1H NMR (300 MHz,
CD2Cl2, 298 K): δ = 2.01 (br. s, 15 H, Cp*CH3), 6.71–6.76 (br. m,
2 H, CpH), 6.83–6.87 (br. m, 1 H, CpH), 7.16–7.51 (br. m, 10 H, (E1/2 = –0.55 V vs. ferrocene) to the analyte solution and then re-
Table 2. Crystallographic data for 5a, 6b, and 7a.
Compound
5a
6b
7a
Empirical formula
M
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C29H28ClO2Rh·4H2O
618.94
monoclinic
P21/c
16.3853(2)
11.8272(2)
14.9842(3)
90
99.093(1)
90
C56H55BNORh·CH2Cl2
C45H46ClN2Rh·1.5CH2Cl2
956.66
orthorhombic
Pbcn
41.9474(8)
12.1687(2)
19.4506(4)
90
90
90
8
9928.5(3)
233(2)
2.1–25.0
34200
880.59
triclinic
¯
P1
11.2946(3)
13.9477(3)
28.8910(5)
90.857(1)
95.315(1)
100.546(1)
4
4452.82(17)
233(2)
1.5–24.0
21714
γ [°]
Z
4
V [Å3]
T [K]
2867.33(8)
233(2)
2.1–27.5
22472
θ range [°]
Data
Unique data
Rint
R1
6555
0.0203
0.0245
0.0595
371
1.040
0.327/–0.414
8280
13465
0.0324
0.0768
0.2119
1051
1.032
1.044/–0.763
0.0309
0.0377
0.0881
594
1.027
0.340/–0.398
wR2
Parameters
GooF
Peak/hole [e/Å3]
Eur. J. Inorg. Chem. 2011, 2958–2966
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2965

R. F. Winter, B. Bildstein et al.
FULL PAPER
Laszlo, R. Hoffmann, Angew. Chem. 2000, 112, 127; Angew.
Chem. Int. Ed. 2000, 39, 123.
peating all of the necessary scans that were previously performed
on the pure analyte in the presence of the standard. The cooling
was performed by immersing the cell into a Dewar vessel that had
been cooled with a 2-propanol/dry ice slush bath and allowing ther-
mal equilibration to take place for at least 30 min. The electrodes
were polished with diamond paste, 1 μm and 0.2 μm, on the appro-
priate polishing cloths from Wirtz & Bühler.
[5] SciFinder® literature search based on substructure Cp2M (M
= FeII) and Cp2M+ (M = CoIII, RhIII, IrIII).
[6] a) B. Enk, H. Kopacka, K. Wurst, T. Müller, B. Bildstein, Or-
ganometallics 2009, 28, 5575; b) B. Enk, D. Eisenstecken, H.
Kopacka, K. Wurst, T. Müller, F. Pevny, R. F. Winter, B. Bild-
stein, Organometallics 2010, 29, 3169.
[7] a) J. E. Sheats, W. C. Spink, R. A. Nabinger, D. Nicol, G.
Hlatky, J. Organomet. Chem. 1983, 251, 93; b) A. Philippo-
poulos, N. Hadjiliadis, C. E. Hart, B. Donnadieu, P.
Mc Gowan, R. Poilblanc, Inorg. Chem. 1997, 36, 1842.
[8] a) W. J. Linn, W. H. Sharkey, J. Am. Chem. Soc. 1957, 79, 4970;
b) D. Lloyd, N. W. Preston, J. Chem. Soc. C 1969, 2464; c) B.
Enk, B. Bildstein, K. Wurst, Z. Kristallogr., New Cryst. Struct.
2009, 224, 544.
X-ray Crystallography: The single-crystal X-ray measurements and
structure determinations for 5a, 6b, and 7a (Table 2, Figures 1, 2,
and 3) were performed as described below. The data collection was
performed with a Nonius Kappa CCD diffractometer that was
equipped with graphite-monochromatized Mo-Kα-radiation (λ =
0.71073 Å) and a nominal crystal-to-area detector distance of
36 mm. The intensities were integrated by using DENZO and were
scaled with SCALEPACK.[19] Several scans were made in the φ-
and ω-directions in order to increase the number of redundant re-
flections, which were averaged in the refinement cycles. This pro-
cedure replaced an empirical absorption correction. The structures
were solved with direct methods (SHELXS-86) and were refined
against F2 (SHELX-97).[20] The hydrogen atoms at the carbon
atoms were added geometrically and were refined by using a riding
model. All of the non-hydrogen atoms were refined with aniso-
tropic displacement parameters. 7a crystallizes with two molecules
in the asymmetric unit, and the high R value was probably caused
by position disordering of the solvent molecules and the two chlo-
rine anions, which are split over four positions. The high aniso-
tropic displacement factors for all of the atoms showed that the
major molecules are also slightly disordered. CCDC-791688 (for
5a), -791689 (for 6b), and -791690 (for 7a) contain the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[9] U. Kölle, B. Fuss, Chem. Ber. 1984, 117, 743.
[10] C. E. Wallace, J. P. Selegue, A. Carillo, Organometallics 1998,
17, 3390.
[11] A. Solchinger, K. Wurst, H. Kopacka, B. Bildstein, Cryst.
Growth Des. 2007, 7, 2380 and references cited therein.
[12] Reviews: a) J. M. Ernsting, S. Gaemers, C. J. Elsevier, Magn.
Reson. Chem. 2004, 42, 721; b) W. von Philipsborn, Chem. Soc.
Rev. 1999, 28, 95; c) B. E. Mann, “Rhodium-103” in Studies
in Inorganic Chemistry 13: Transition Metal Nuclear Magnetic
Resonance (Ed.: P. S. Pregosin), Elsevier Science Publishers
B. V., Amsterdam, 1991, p. 177; d) R. Benn, A. Rufinska, An-
gew. Chem. 1986, 98, 851; Angew. Chem. Int. Ed. Engl. 1986,
25, 861.
[13] E. Maurer, S. Rieker, M. Schollbach, A. Schwenk, T. Egolf, W.
von Philipsborn, Helv. Chim. Acta 1982, 65, 26.
[14] The chemical shift value of Cp2Rh+Cl– versus external
Rh(acac)3 was calculated by comparing the published values
[12c]
for [Rh(acac)cod] against Rh(acac)3
and against the com-
monly used 103Rh reference frequency Ξ = 3.16 MHz:[12]
[δ{Rh(acac)cod}
=
–7014 ppm[12c]
–
(+1294.2) ppm[13]
=
=
–8308 ppm]; [δ(Cp2Rh+Cl–) = –8308 ppm – 1838.7 ppm[13]
Supporting Information (see footnote on the first page of this
article): Electrochemical Figures S1–S7.
–10147 ppm].
[15] E. O. Fischer, H. Wawersik, J. Organomet. Chem. 1966, 5, 559.
[16] N. El Murr, J. E. Sheats, W. E. Geiger, J. D. L. Holloway, Inorg.
Chem. 1979, 18, 1443.
[17] O. V. Gusev, L. I. Denisovich, M. G. Peterleitner, A. Z. Rub-
ezhov, N. A. Ustynyuk, P. M. Maitlis, J. Organomet. Chem.
1993, 452, 219.
[18] N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96, 877.
[19] Z. Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307.
[20] G. M. Sheldrick, SHELXTL V.5.1, Bruker Analytical X-ray In-
struments Inc, Madison, USA, 1997.
[1] a) G. Wilkinson, J. Am. Chem. Soc. 1952, 74, 6148; b) E. O.
Fischer, W. Pfab, Z. Naturforsch. B 1952, 7, 377; c) E. O. Fi-
scher, R. Jira, Z. Naturforsch. B 1953, 8, 1.
[2] a) F. A. Cotton, R. O. Whipple, G. Wilkinson, J. Am. Chem.
Soc. 1953, 75, 3586; b) R. J. Angelici, E. O. Fischer, J. Am.
Chem. Soc. 1963, 85, 3733.
[3] E. O. Fischer, U. Zahn, Chem. Ber. 1959, 92, 1624.
[4] a) T. J. Kealy, P. L. Pauson, Nature 1951, 168, 1039; b) S. A.
Miller, J. A. Tebboth, J. F. Tremaine, J. Chem. Soc. 1952, 632;
c) Review on the history of the discovery of ferrocene: P.
Received: March 4, 2011
Published Online: May 25, 2011
2966
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 2958–2966