Water-soluble mono- and star-shaped hexanuclear functional organo-iron catalysts for nitrate and nitrite reduction in water: Syntheses and electroanalytical study

Article abstract of DOI:10.1016/S0020-1693(02)00764-8

Ten functional sandwich complexes [Fe^II(η⁵-C ₅H₄CO₂H)(η⁶-(arene)][PF ₆] and [Fe⁺(η⁵-C₅H ₅)(η⁶-(C₆(CH₂CH ₂C₆H₄OH)₆)][PF₆] have been synthesized to provide water-soluble redox catalysts with a range of redox potentials for the Fe^II-Fe^I redox couple. The CpFe ⁺ induced C-H benzylic activation and the carboxylation of hexamethylbenzene, the CpFe⁺ induced one-pot polymethylation of mesitylene and hexamethylbenzene, and the one-pot CpFe⁺ induced hexa-p-methoxybenzylation of hexamethylbenzene have led to the desired functionalized complexes. These compounds catalyze the cathodic reduction of nitrate and nitrite in a basic aqueous medium on a Hg cathode. The rates of the rate-limiting steps of the redox-catalytic processes involving the homogeneous reduction of nitrate and nitrite by the '19-electron' states of the redox catalysts have been measured using enhancement of the intensity of the monoelectronic cathodic wave Fe^II→Fe^I in the presence of the substrate by cyclic voltammetry, polarography and chronoamperometry. For six iron sandwich complexes without excessive bulk around the iron center, a linear Marcus-type correlation between the kinetics and the thermodynamics was found by plotting the logarithm of this rate constant versus the redox potential, which indicates that the homogeneous electron-transfer from the 19-electron catalyst to the substrate is rate-limiting. With four complexes in which chains of the arene ligand introduced more bulk around the iron center, the rates were found to be an order of magnitude lower, showing the crucial effect of the distance on the outer-sphere electron transfer. A small inner-sphere contribution cannot be discarded, however. When the redox catalysts are attached to the termini of the branches of hexa-branch star-shaped cores, the rate constants are close to those of unencumbered catalysts, i.e. no significant loss of activity is observed.

Full text of DOI:10.1016/S0020-1693(02)00764-8

Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
www.elsevier.com/locate/ica
Water-soluble mono- and star-shaped hexanuclear functional organo-
iron catalysts for nitrate and nitrite reduction in water: syntheses and
electroanalytical studyꢀ
Ste´phane Rigaut a, Marie-He´le`ne Delville a, Jose´ Losada b, Didier Astruc a,*
a Laboratoire de Chimie Organique et Organome´tallique, UMR CNRS No. 5802, Universite´ Bordeaux I, 33405 Talence Cedex, France
b Departamento Quimica Inorganica, Universidad Autonoma Madrid, Cantoblanco, 28049 Madrid, Spain
Received 9 October 2001; accepted 16 January 2002
This article is dedicated to Professor Andrew Wojcicki, a distinguished friend and colleague with high sense of ethics, professionalism and humor
who was a pioneer of modern inorganic/organometallic chemistry.
Abstract
Ten functional sandwich complexes [FeII(h5-C5H4CO2H)(h6-(arene)][PF6] and [Feꢀ(h5-C5H5)(h6-(C6(CH2CH2C6H4OH)6)][PF6]
have been synthesized to provide water-soluble redox catalysts with a range of redox potentials for the FeIIꢁFeI redox couple. The
/
CpFeꢀ induced Cꢀ
/
H benzylic activation and the carboxylation of hexamethylbenzene, the CpFeꢀ induced one-pot polymethyla-
tion of mesitylene and hexamethylbenzene, and the one-pot CpFeꢀ induced hexa-p-methoxybenzylation of hexamethylbenzene
have led to the desired functionalized complexes. These compounds catalyze the cathodic reduction of nitrate and nitrite in a basic
aqueous medium on a Hg cathode. The rates of the rate-limiting steps of the redox-catalytic processes involving the homogeneous
reduction of nitrate and nitrite by the ‘19-electron’ states of the redox catalysts have been measured using enhancement of the
intensity of the monoelectronic cathodic wave FeII0
/
FeI in the presence of the substrate by cyclic voltammetry, polarography and
chronoamperometry. For six iron sandwich complexes without excessive bulk around the iron center, a linear Marcus-type
correlation between the kinetics and the thermodynamics was found by plotting the logarithm of this rate constant versus the redox
potential, which indicates that the homogeneous electron-transfer from the 19-electron catalyst to the substrate is rate-limiting. With
four complexes in which chains of the arene ligand introduced more bulk around the iron center, the rates were found to be an order
of magnitude lower, showing the crucial effect of the distance on the outer-sphere electron transfer. A small inner-sphere
contribution cannot be discarded, however. When the redox catalysts are attached to the termini of the branches of hexa-branch
star-shaped cores, the rate constants are close to those of unencumbered catalysts, i.e. no significant loss of activity is observed.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Organo-iron catalysts; Nitrate and nitrite reduction; Electroanalytical study; Redox catalysis; Catalysis in water
1. Introduction
catalysis in water or fluorophases or using dendrimers
[2]. We have been interested in catalysis in water using
dendrimers [2,3], and report here the syntheses and
electroanalytical studies of nitrate and nitrite reduction
by water-soluble mono-and star-shaped hexairon com-
plexes of the [FeCp(arene)] type. Redox processes using
dendritic-type molecules and electrochemistry have
recently been addressed by several research groups
because of their relevance to electron transfer in
The considerable interest for nanoscopic devices has
recently generated an increased amount of electroche-
mical studies of supramolecular systems in order to
improve the venue of sensors and catalysts [1]. New
areas relevant to green chemistry have emerged such as
ꢀ This article is part of the Ph.D. Thesis of S.R.
biological systems [4ꢁ13].
/
* Corresponding author. Tel.: ꢀ
6646.
E-mail address: d.astruc@lcoo.u-bordeaux.fr (D. Astruc).
/
33-556-84 6271; fax: ꢀ33-556-84
/
The elimination of nitrates and nitrites from current
water is an environmentally essential issue [14]. In
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 0 7 6 4 - 8

226
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
nature, nitrates and nitrites are reduced by nitrate
reductase and nitrite reductase enzymes as part of the
nitrogen cycle [14b]. The direct electroreduction of
nitrates and nitrites, however, leads to mixtures of N2,
N2O, NH2OH and NH3 with little selectivity. The
compositions of the reduction products vary with the
nature of the electrode and electrolyte, the applied
potential and the concentrations [14c,14d]. The electro-
reduction also requires overpotentials, high currents,
especially in basic media because of the low adsorption
of substrates on electrodes [14d]. Various electrode
materials have been used for nitrate reduction in acidic,
basic or neutral conditions [15,16]. On mercury cath-
odes, the electroreduction of nitrates is not observable in
basic aqueous media, and indeed, nitrates are used as
supporting electrolytes in polarography. Thus, the redox
catalysis of nitrate reduction on a mercury cathode has
been known for a long time [17]. Recently, it has been
possible to derivatize an electrode with a nitrate
reductase enzyme [18]. Inorganic complexes whose
metal centers are known to bind nitrogen oxides as
ligands have been used as redox catalysts [18b]. Cyclam
complexes of CoIII and NiII have been used as redox
catalysts in basic aqueous media to reduce nitrates [19]
and Ma et al. have deposited these complexes on
electrodes or derivatized electrodes using these catalysts
[20,21]. Tanaka et al. have used FeMoS clusters in such
conditions as Moinet et al. These conditions, in basic
aqueous medium, lead to the formation of ammonia as
described above. Indeed, the 19-electron [FeCp(arene)]
species which are active in the reduction of substrates
[29] can either transfer their extra electron by an outer-
sphere process [30] or undergo a fast pre-equilibrium
with a 17-electron species to which the substrate could
coordinate prior to inner-sphere electron transfer [31ꢁ
/
35]. The activation of aromatics by CpFeꢀ provides rich
possibilities to undergo stereoelectronic transformations
of these arene ligands [36]. In this way, this leads to
modifications of the driving force as well as of the bulk
around the iron center, in order to study the influence of
these essential parameters on the reaction rates. In
addition, it was of interest to envisage synthetic strate-
gies of branching these stable catalysts onto star-shaped
cores or dendrimers and to examine if their catalytic
activity was retained. Since the reaction products were
known from Moinet et al.’s studies, we are focusing in
this paper on the synthesis of various water-soluble iron
complexes and their use for the electro-analytical study
of the relationship between the driving force of the 19-
electron reductants and the kinetics of nitrate and nitrite
reduction in basic aqueous medium.
2. Results
media and showed the implication of Moꢀ
/
Oꢀ
/
N type
coordination of nitrate and nitrite in the course of their
catalytic reduction [22,23].
2.1. Syntheses of catalysts [FeII(h5-C5H4CO2H)(h6-
(arene)][PF6] soluble in basic aqueous medium
In a preliminary communication, Moinet et al. [24]
reported the redox catalysis of cathodic reduction of
nitrate and nitrite by FeCp(arene) complexes in basic
The complexes [FeII(h5-C5H4CO2H)(h6-arene)][PF6]
are accessible since it was found that only ferrocene
derivatives bearing a single electron-withdrawing carbo-
nylated function conjugated with a Cp ring lead to the
complexes [FeII(h5-C5H4COR)(h6-arene)][PF6] by selec-
tive substitution of the free Cp with the arene [25,37,38].
Thus, these complexes can be directly obtained by ligand
substitution in the presence of the arene and aluminum
chloride from ferrocenylcarboxylic acid (Eq. (1)) or
from ferrocene itself in the presence of CO2 (Eq. (2))
[25,28a,39].
aqueous medium on an Hg cathode at ꢂ
SCE. NH3 was produced in 63% chemical yield and 57%
electrical yield using
[FeII(h5-C5H4CO2H)(h6-
/1.8 V versus
C6Me6)][PF6], 1a [24,25]. Under these conditions, the
intensity of the wave due to the iron catalyst remained
unchanged after electroreduction, indicating the re-
markable stability of the catalyst. Nitrite, hydroxyla-
mine and hydrazine were shown to be intermediates in
the nitrate reduction and a minute amount of nitrogen
was also detected. The reduction stopped at the hydro-
xylamine level when it was conducted at neutral pH.
Kinetic studies were carried out by polarography using
1a and two other [FeCp(h6-arene)] catalysts (areneꢃ
benzene, m-xylene) whose neutral FeI form is insoluble
in water [24ꢁ27]. The rate constant of reduction of
nitrate by the 19-electron state [28] was found to be 39
1ꢄ
102 l mol sꢂ1 with these three catalysts, and the
/
ð1Þ
/
/
/
authors concluded that the rate of catalysis was
independent of the catalyst structure in these series
[24]. Since these catalysts are robust and are the first
organometallic complexes which were found to be active
redox catalysts for the reduction of nitrogen oxides, we
were interested in their mechanism in using the same
ð2Þ
In basic aqueous medium (NaOH 0.1 N), these
carboxylic acids give soluble and stable carboxylate
zwitterions. The variation of the number of methyl

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
227
substituents on the arene ligand leads to desirable
Fe(II) redox potential insofar
2.1.2. Bulky functional complexes
variations of the Fe(I)ꢁ
/
The most challenging synthetic problem concerned
the route to the complexes [FeCp(arene)][PF6] which
would still be water soluble, yet present a relatively large
bulk around the metal center in order to inhibit the
potential approach of a substrate onto the metal center.
Since ligand exchange reactions between ferrocene and
bulky arenes lead to low yields or to no product at all
[43], we had to address the build up of the bulk after
complexation using suitable overall engineering. The
CpFeꢀ induced polyalkylation and polyfunctionaliza-
tion of polymethyl aromatic ligands would be a viable
possibility [28,44], but it had not been used in the
presence of a functional substituent on the Cp ring. Such
a substituent may indeed have a considerable stereo-
electronic influence on the outcome of these reactions.
Thus, we have attempted the hexamethylation of the
hexamethylbenzene complex 1a and the nonamethyla-
tion of the mesitylene complex 3. By using more drastic
reaction conditions (DME, reflux) than the mild ones
used in the case of the parent Cp complex, we
have obtained the desired carboxylic acid complexes 7
and 8 (Eqs. (5) and (6)) in 25 and 21% yields,
respectively.
as the 1e1 and 2e1 orbitals of the arene ligand contribute
to the antibonding e*1 orbital (19th electron) of the
Fe(I) state [40]. For this purpose, we have selected
hexamethylbenzene, toluene, mesitylene and durene,
which led to the complexes [Fe(C5H4COOH)(are-
ne)][PF6] 1aꢁ
/
4, respectively, in 53ꢁ79% yields.
/
2.1.1. Carboxylic acid function located on the arene
ligand
It is also possible to introduce the carboxylate
function on the arene ligand by deprotonation of
[FeCp(C6Me6)][PF6] followed by fast reaction of CO2
at 20 8C according to the known procedure of Eq. (3)
[41], which gives the carboxylic acid complex [FeCp(C6-
Me5CH2COOH)] [PF6], 5.
ð3Þ
We have now developed a variation of this strategy
consisting in applying the deprotonationꢁcarboxylation
/
sequence to the synthesis of a diacid in which a
carboxylic acid function is located on each ligand.
Starting from [Fe(C5H4CO2H)(C6Me6)][PF6], 1a, it is
necessary to protect the first carboxylic acid function in
order to allow the work up of the deprotonation
reaction, i.e. to solubilize the deprotonated species.
This is achieved by esterification of 1a to 1b using
ð5Þ
ð6Þ
BF3×/Et2O in refluxing methanol as the ester function is
stable at 20 8C in anhydrous basic medium [42]. The
ester is saponified using aqueous NaOH after the second
carboxylation, which gives a di-sodium carboxylate 6 in
61% yield from 1b according to Eq. (4). The infrared
spectrum of complex 6 contains two strong bands at
The hexa-p-methoxybenzylation of [FeCp(C6-
Me6)][PF6], followed by the cleavage of the methyl
groups using BBr3 gives 50% of the known hexaphenol
9 [45] which is soluble in basic aqueous medium (Eq.
(7)).
gCꢁO
boxylate function.
ꢃ
1611 and 1590 cmꢂ1 characteristic of the car-
/
ð7Þ
Application of this reaction to 1 gives the new
carboxylic acid 10 in the same way in 17% overall yield
(Eq. (8)). We believe that the requirement of more
drastic reaction conditions and systematically decreased
yields obtained in the presence of the carboxylic acid
ð4Þ

228
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
function are due to the additional steric hindrance
provided by this Cp substituent. The steric congestion
somewhat destabilizes the successive deprotonated com-
plexes in the course of reactions [46]. Nevertheless, all
these new carboxylic acids obtained subsequent to
optimization of the reaction conditions proved to be
spectroscopically and analytically pure.
ð10Þ
ð8Þ
2.1.3. Star-shaped, functional hexanuclear complexes
heterobifunctional
Two
new
complexes
[Fe(C5H4R)(arene)][PF6] have been synthesized and
branched onto the extremities of star-shaped cores.
The first strategy consisted in branching a sandwich
complex bearing a secondary amino group on the arene
ligand to an hexaiodo core using the acidity of the amine
proton induced by the electron-withdrawing character
of the CpFeꢀ moiety [47]. The desired heterobifunc-
tional mononuclear complex 11 was synthesized in two
steps from ferrocenylcarboxylic acid according to Eq.
(9) in 41% overall yield.
ð11Þ
The second strategy consists in branching an hetero-
bifunctional complex bearing a remote primary amino
group onto an hexabenzaldehyde core. Prior attempts to
realize the opposite strategy, i.e. condensation of 15
bearing an aldehyde substituent on the Cp ring onto the
branches of an hexa-amino core, failed, although
monoamines in excess react without a problem to give
the imino group on the Cp ring. This dichotomy led us
to use 15, whose synthesis is described elsewhere [48], for
the synthesis of 16 bearing a primary amine function
remote from the Cp ring (Eq. (13), 81% overall yield).
The hexabenzaldehyde core 17 is synthesized in 96%
yield from the hexaiodo core 12 (Eq. (11)). Then, the
reaction of a fourfold excess of 16 with 17 followed by
hydrogenation of the hexa-imine provides the desired
hexanuclear complex 18 without any decomplexation in
48% overall yield from 17 (Eq. (12)).
ð9Þ
The reaction (Eq. (10)) of a twofold excess of 11 with
the known hexaiodo core 12 (vide infra Eq. (11)) leads
to a complete reaction of the iodo groups giving the
hexaester complex 13. A small amount (7% detected by
1
the arene H NMR signals) of decomplexation occurs,
even after optimization of the reaction conditions.
Elemental analysis (C, H) of the hexanuclear
complex 14 obtained subsequent to saponification of
13 (51% overall yield from 12, Eq. (10)) is satisfactory,
however.

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
229
In MeCN with n-Bu4NBF4 0.1 M, on an Hg cathode, no
adsorption is found for 18. Its redox potential is ꢂ1.84
/
V versus SCE and Ep is 80 mV with ia/ic also equal to
unity. These very close values of the redox potentials in
basic aqueous medium and MeCN are also noted for 1
[49]. These data confirm that the six redox centers that
are almost independent have values of their six redox
potentials of 11 and 18 which are extremely close [50]
(the differences among the six redox potentials being not
detectable), although they cannot, in theory, be strictly
identical [51,52].
2.3. Thermodynamic and kinetic aspects of nitrate and
nitrite reduction. Requirement of redox catalysis
The thermodynamics of nitrate and nitrite electro-
reduction in water are given by Eqs. (13) and (14):
ð12Þ
NOꢂ3 ꢀH2Oꢀ2eꢂ 0 NO2ꢂ ꢀ2OHꢂ
0:251 V versus SCE
(13)
2.2. Cyclic voltammetry studies of the organoiron
carboxylates: redox potentials
NOꢂ2 ꢀ5H2Oꢀ6eꢂ 0 NH3 ꢀ7OHꢂ
0:076 V versus SCE
(14)
The cyclic voltammograms of all the organoiron
carboxylates in basic aqueous medium (0.1 M LiOH)
on a mercury cathode present an electrochemically and
chemically reversible reduction wave corresponding to
Nitrate and nitrite are not electroactive in basic
aqueous medium, even in the most cathodic region
where the reduction would be highly thermodynamically
favorable, given the high overpotential of cathodic
reduction on Hg. On the other hand, the homogeneous
reduction, in the bulk of the solution, is also thermo-
dynamically favorable, since the redox potential of the
the Fe(II)ꢁ
potentials independent of scan rates, ia/ipꢃ
reduction wave being diffusion controlled as checked
by verifying the equation idꢃ
f(v1/2).
/
Fe(I) redox system (DEpꢃ
/
60 mV, peak
/1), the
/
The scale of redox potentials reflecting the electronic
effect of the various ring substituents of the mono-
nuclear systems are represented in Chart 1. Similarly,
the hexanuclear complexes 11 and 18, which are soluble
redox catalyst is located between ꢂ
/
1.4 and ꢂ1.9 V
/
versus SCE. It is now kinetically facilitated, however, by
the tridimentional space of the solution available for the
mutual interaction between the substrate and the 19-
electron redox catalyst. The overall catalytic scheme is
depicted in Eqs. (15) and (16).
in basic aqueous medium, show redox potentials of ꢂ
1.66 and ꢂ1.79 V, respectively. Complex 18 shows some
adsorption, however, and therefore DEp is only 40 mV.
/
/
Chart 1. Redox potentials of the FeII/FeI system in the organo-iron complexes used as redox catalysts for the cathodic reduction of nitrate and
nitrite.

230
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
For cyclic voltammetry, Save´ant and Vianello [54]
and Nicholson and Shain [55] have solved Fick’s
equation in the case of a redox catalyzed process leading
to a simple expression Eq. (17) applicable to a planar
electrode if Eq. (18) is fulfilled.
pffiffiffiffiffiffiffiffi
I
kCz
k ꢃ
(17)
(18)
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Id 0:446 ×
nFv=RT
pffiffiffiffiffiffiffiffi
Cz
kC
pffiffiffiffizffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ꢀ
COx 0:446 × nFv=RT
Since we are using a spherical dropping mercury
electrode, we need to use low time scales (scan rate v ꢁ
/
0.1 V sꢂ1) so that the diffusion layer will be thin enough
for the approximation of a planar electrode to be valid
[56].
If, in addition, the regime is under kinetic control, no
peak, but only a plateau is observed, which corresponds
to the case where kCzRT/nFv ꢁ1.
/
ð15Þ
ð16Þ
This is easily achieved in the case of 5 since the
kinetics of redox catalysis is very fast. The average value
obtained for k using various scan rates between 100 and
300 mV sꢂ1 allowing the fulfillment of both conditions
is 9.80ꢄ
103 l molꢂ1 sꢂ1 and the agreement between the
different values obtained is excellent.
/
For polarography, Fick’s equations have been solved
by Koutecky [57] and the Ik/Id ratio is given by Eq. (19)
in which t is the lifetime of the mercury drop provided
kCzt ꢁ10 and the condition of Eq. (20) concerning the
concentrations is fulfilled. Then polarograms of Fig. 2
are obtained (Fig. 1).
/
Since, at least with the mononuclear complexes,
electrons are transferred one at a time, the first
reduction step is the one-electron reduction of the
substrate to its radical dianion by the 19-electron state
of the redox catalyst (Eq. (16), step E). Whether this step
is rate-limiting will now be examined from the kinetic
study by electrochemistry.
pffiffiffiffiffiffiffiffiffi
I
k ꢃ0:815 kCzt
(19)
(20)
Id
pffiffiffiffiffiffiffiffiffi
Cz
ꢀ0:815 kCzt
COx
The average value obtained from various experiments
with lifetimes of mercury drops between 1.2 and 1.5 s
2.3.1. Comparison of the different electrochemical
techniques
Cyclic voltammetry, polarography and chronoam-
perometry have been compared for the study of the
redox catalysis of the cathodic reduction of nitrate and
nitrite by 5 which has the most negative redox potential,
ꢂ
/
1.86 V versus SCE, in this series of organoiron redox
catalysts. In basic aqueous medium used here (aq. LiOH
0.1 M), reduction of water is observed at ꢂ2.1 V versus
/
SCE on Hg cathode, and 5 is thus electroactive (at a pH
lower than 11.5, the carboxylic group is reduced to
primary alcohol) [53]. For each of the three electro-
chemical techniques, the rate constant k of the homo-
geneous reduction of the substrate by the 19-electron
state of the catalyst (Eq. (16), step E) is deduced from
Fig.
C6Me5CH2CO2ꢂ)] obtained in absence (a) and in presence (b) of
nitrate: COx 3.07ꢄ 3.07ꢄ
10ꢂ4 mole lꢂ1; CNO 10ꢂ3 mol lꢂ1
CNO /COx 10; CLiOH
0.1 mol lꢂ1; vꢃ3
200 mV sꢂ1
1. Cyclic
voltammograms
of
[FeII(h5-C5H5)(h6-
the ratio Ik/Id of the current intensity of the Fe(II)ꢁFe(I)
/
wave in the presence of substrate over the current
intensity of this wave in the absence of substrate.
ꢃ
/
/
ꢂ
ꢃ
/
/
;
ꢂ
ꢃ/
ꢃ/
/
.
3

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
231
and the agreement with cyclic voltammetry is perfect as
expected. For the kinetic study of all the other redox
catalysts of the series, only cyclic voltammetry is used, as
measurements are easier to carry out without drawback
and this technique is precise in all cases.
2.4. Cyclic voltammetry study of the kinetics of redox
catalysis of the reduction of nitrate and nitrite by the iron
sandwich complexes (Tables 1ꢁ5)
/
2.4.1. Mononuclear complexes
In some cases, the condition kCznFV/RT ꢁ
/1 is not
fulfilled, which means that the regime is not strictly
under kinetic control, and a peak is observed. Such a
case is shown on Fig. 4.
Then, it is necessary to use the working curve reported
by Nicholson and Shain representing the Ik/Id ratio as a
function of (kCznFV/RT)1/2 (Fig. 5).
Fig. 2. Polarograms of [FeII(h5-C5H5)(h6-C6Me5CH2CO2ꢂ)] ob-
tained in absence (a) and in presence (b) of nitrate. COx 1.10ꢄ
10ꢂ4 mol lꢂ1; CNO 10ꢂ3 mol lꢂ1
2.97ꢄ
ꢃ/
/
ꢂ
ꢃ/
/
.
3
The graphical representations of the logarithm of the
rate constant k for nitrate and nitrite reduction versus
the redox potentials of the redox catalysts are shown in
Fig. 5, respectively (for tables of k values including
concentrations and scan rates, see the Section 4). There
are clearly two sets of redox catalysts (Fig. 6).
which allows to fulfill the two above conditions lead to
kꢃ9.40ꢄ
103 l molꢂ1 sꢂ1. It is quite difficult to
measure Ik accurately (see Fig. 2) because of the
proximity of the solvent discharge. In spite of this
problem, the agreement with the k value obtained by
cyclic voltammetry is very good.
For chronoamperometry, the proximity of the solvent
discharge is a major problem, and this technique is
completely dependent on cyclic voltammetry (Fig. 3).
Fick’s equations for this technique have been solved
by Delahay and Stiehl [58], and Popisi [59] and Miller
/
/
The first group does not contain excessively bulky
arene ligands, i.e. arene ligands whose polymethylation
does perturb a potential approach of a substrate near
the iron center in the transition state. These polymethyl-
benzene ligands do not bring too much bulk near the
iron center in the 17-electron Fe(I) state because they
have a planar, rigid carbon framework. These complexes
form a set of redox catalysts for which the rate constants
fall on a straight line, an indication of a Marcus-type
relationship [58] showing that the electron transfer step
is rate limiting. The second group contains very bulky
arene ligands whose substituents clearly inhibit the
potential approach of the substrate near the iron center.
For those catalysts, the points obtained on the graph are
located well below the straight line and the rate constant
values are about an order of magnitude lower than if
they would fit on the line. For instance, 1 and 9 have the
same redox potential and the rate constant for nitrate
reduction obtained in the case of 9 is 14 times lower than
in the case of 1.
[59]. If kCzt ꢁ2, the intensity ratio is given by Eq. (21)
/
provided the condition of Eq. (22) concerning the
concentrations is fulfilled.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I
k ꢃ PkCzt
(21)
Id
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
PkCzt
Cz
ꢀ
(22)
COx
The average k value obtained is 9.77ꢄ
/
103 l mol sꢂ1
2.5. Mechanism of the nitrate and nitrite reduction by the
19-electron [Fe(I)Cp(arene]) catalysts
The linear relationship in this region of the graph
log kꢃf(E8) obtained for the first group (non-bulky) is
/
usually found for outer-sphere processes. It indicates,
according to Marcus theory, that the electron-transfer
mechanism essentially proceeds by outer sphere [58].
Since there is a continuum between inner-sphere and
outer-sphere mechanisms, it is possible that a small
inner-sphere contribution be involved [61]. The rate
Fig. 3. Chronoamperograms of [FeII(h5-C5H5)(h6-C6Me5CH2CO2ꢂ)]
obtained in the absence (a) and presence (b) of nitrate. COx 2.55ꢄ
10ꢂ4 mol lꢂ1 10ꢂ3 mol lꢂ1
CNO 5.53ꢄ CNO /COx 21;
CLiOH 0.80 V; (a) E2ꢃ
0.1 mol lꢂ1. E1ꢃ
2.00 V;3 (b) E2ꢃ
1.93 V; t1ꢃ
0.3 s; sweeping rate: 0.2 s cmꢂ1
ꢃ/
/
;
ꢂ
ꢃ
/
/
;
ꢂ
ꢃ/
3
ꢃ/
/
ꢂ/
/
ꢂ
/
/
ꢂ/
/
.

232
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
Table 1
Rate contants and conditions for the reduction of nitrate by the non-bulky catalysts
EOx (V/SCE) COx (mmol lꢂ1
)
ꢂ
(mmol lꢂ1
)
k (l molꢂ1 sꢂ1 v (mV sꢂ1
) )
8
8
8
CNO
Redox catalyst (Ox)
3
[Feꢀ(h5-C5H5)(h6-(C6Me5CH2CO2ꢂ))] (5)
ꢂ
/
1.86
1.69
1.68
1.62
1.58
1.50
0.25
1.20
0.45
1.25
1.27
1.51
5.5
26.0
9800
255
265
97
150ꢁ
100ꢁ
100ꢁ
100ꢁ
150ꢁ
120ꢁ
/
350
200
300
250
350
350
[FeII(h5-C5H4CO2H)(h6-(C6Me6))]ꢀPF6 (1a)
ꢂ
/
/
ꢂ
[FeII(h5-C5H4CO2Na)(h6-(C6Me5CH2CO2Na))]ꢀPF6 (6)
ꢂ
/
13.3
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6Me4H2))]ꢀPF6
(4)
ꢂ
/
108.0
118.0
110.0
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6Me3H3))]ꢀPF6 (3)
ꢂ
/
60
14
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6MeH5))]ꢀPF6 (2)
ꢂ/
/
ꢂ
ꢂ
ꢂ
Measurement of the rate constant k of the reduction of NO3 and NO2 by the 19-electron complexes: Working conditions and data.
Table 2
Rate constants and conditions for the reduction of nitrite by the non-bulky catalysts
EOx (V/SCE) COx (mmol lꢂ1
)
ꢂ
(mmol lꢂ1
)
k (l molꢂ1 sꢂ1 v (mV sꢂ1
) )
8
8
8
CNO
Redox catalyst (Ox)
2
[Feꢀ(h5-C5H5)(h6-(C6Me5CH2CO2ꢂ))] (5)
ꢂ
/
1.86
1.69
1.68
1.62
1.58
1.50
0.12
0.81
0.45
0.86
0.98
1.24
5.8
17.5
17.9
19.4
34.4
61.0
37 200
1094
1400
394
150ꢁ
100ꢁ
100ꢁ
120ꢁ
100ꢁ
100ꢁ
/
300
250
300
300
200
400
[FeII(h5-C5H4CO2H)(h6-(C6Me6))]ꢀPF6 (1a)
ꢂ
/
/
ꢂ
[FeII(h5-C5H4CO2Na)(h6-(C6Me5CH2CO2Na))]ꢀPF6 (6)
ꢂ
/
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6Me4H2))]ꢀPF6 (4)
ꢂ
/
/
ꢂ
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6Me3H3))]ꢀPF6 (3)
ꢂ
/
208
42
/
[FeII(h5-C5H4CO2H)(h6-(C6MeH5))]ꢀPF6 (2)
ꢂ/
/
ꢂ
Table 3
Rate constants and conditions for the reduction of nitrate by the bulky catalysts
EOx (V/SCE) COx (mmol lꢂ1
)
ꢂ
(mmol lꢂ1
)
k (l molꢂ1 sꢂ1 v (mV sꢂ1
) )
8
8
8
CNO
Redox catalyst (Ox)
3
[FeII(h5-C5H5)(h6-(C6(CH2CH2C6H4OH)6)]ꢀPF6 (9)
ꢂ
/
1.69
1.62
1.60
0.45
0.69
0.81
7.7
24.6
10.7
75
35
38
100ꢁ
100ꢁ
200ꢁ
/
200
200
250
[FeII(h5-C5H4CO2H)(h6-(C6Et6))]ꢀPF6 (7)
ꢂ
/
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6(CH2CH2C6H4OH)6)]ꢀPF6
(10)
ꢂ/
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6H3(CCH3)3))]ꢀPF6 (8)
ꢂ
/
1.55
0.79
54.9
9
100ꢁ
/
250
ꢂ
Table 4
Rate constants and conditions for the reduction of nitrite by the bulky catalysts
EOx (V/SCE) COx (mmol lꢂ1
)
ꢂ
(mmol lꢂ1
)
k (l molꢂ1 sꢂ1 v (mV sꢂ1
) )
8
8
8
CNO
Redox catalyst (Ox)
3
[FeII(h5-C5H5)(h6-(C6(CH2CH2C6H4OH)6)]ꢀPF6 (9)
(7)
ꢂ
/
1.69
1.62
1.60
0.11
0.64
0.72
10.5
15.6
11.6
188
187
120
100ꢁ
100ꢁ
150ꢁ
/
250
250
250
[FeII(h5-C5H4CO2H)(h6-(C6Et6))]ꢀPF6
ꢂ
/
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6(CH2CH2C6H4OH)6)]ꢀPF6
(10)
ꢂ/
/
ꢂ
[FeII(h5-C5H4CO2H)(h6-(C6H3(CCH3)3))]ꢀPF6 (8)
ꢂ
/
1.55
7.69
32.0
49
100ꢁ
/
300
ꢂ
constants are low, however, confirming that this con-
tribution is only small. The low oxidation state involved,
FeI, is indeed not favorable for coordination of an
oxygen atom of the oxo-anions disfavoring a nitrate- or
nitrite-bridged transition state [62,63] that would sig-
nificantly accelerate the electron transfer by such an
inner-sphere contribution. The data obtained for the
second group of catalysts (bulky ones) show that the
heterogeneous electron transfer is disfavored by the
bulk. Recent studies with metal-centered dendrimers
show that outer sphere heterogeneous electron transfer
between the core-centered metal and the electrode is
considerably slowed down as the dendrimer generation
increases. Thus, this rate decrease in the bulky series is
not an indication of any inner-sphere contribution to the
electron-transfer mechanism [60]. Indeed, the lowering
of the rate constant by steric congestion is not as high as
expected for a pure inner-sphere mechanism [59]. Thus,
among the two extreme mechanisms shown in Scheme 1,
the outer-sphere mechanism is highly favored, although
a minor inner-sphere contribution cannot be discarded.
2.6. Hexanuclear star-shaped complexes
The analogous kinetic study was performed by cyclic
voltammetry for the two star-shaped hexanuclear com-

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
233
Table 5
Rate constants and conditions for the reduction of nitrate and nitrite by the hexairon catalysts
COx (mmol lꢂ1
)
ꢂ
(mmol lꢂ1
)
k (l molꢂ1 sꢂ1
)
v (mV sꢂ1
)
8
EOx (V/SCE)
8
8
CNO
Redox catalyst (Ox)
x
Hexaironꢁ
Hexaferꢁ
Hexaferꢁ
Hexaferꢁ
/
toluidine (14): nitrate
ꢂ
/
1.63
1.63
1.79
1.79
0.14
0.22
0.05
0.06
48.4
13.9
110
220
100ꢁ
100ꢁ
100ꢁ
100ꢁ
/
250
250
200
250
/
toluidine (14): nitrite
HMB (18): nitrate
HMB (18): nitrite
ꢂ
/
/
/
ꢂ
/
3.35
3.90
3240
11 000
/
/
ꢂ/
/
Fig. 4. Cyclic voltammograms of [FeII(h5-C5H4COOH)(h6-C6H2Me4)]ꢀPF6ꢂ obtained in absence (a) and presence (b) of nitrite. COx
ꢃ
/
9.01ꢄ
/
10ꢂ4
300
mol lꢂ1; CNO
ꢂ
ꢃ/
1.16ꢄ
/
10ꢂ2 mol lꢂ1; CNO
ꢂ
/COx
ꢃ/12; CLiOH
ꢃ/
0.1 mol lꢂ1; (b) kCznFV/RT ꢁ
/
1 (vꢃ
/
100 mV sꢂ1); (c) kCznFV/RT B
/1 (vꢃ
/
3
3
mV sꢂ1).
plexes 14 and 18, soluble in basic aqueous medium.
Typical cyclic voltammograms of 14 and 18 in the
absence and presence of NO3ꢂ are represented in Figs. 7
and 8.
clearly bring about some additional bulk for the iron
center in 14 and 18 in comparison with the iron center in
the redox catalysts of the non-bulky group. Thus, since
the steric effect is demonstrated by the low k values
obtained for the group of bulky redox catalysts, it is
understandable that a very low decrease of the values of
the rate constants is observed for 14 and 18 compared
with the catalysts fitting in the straight line. This effect is
extremely weak, however, so that it may be stated that
the branching of the redox catalyst onto the termini of
the branches of a star-shaped core does not much
perturb the redox catalytic activity.
ꢂ
For NO3 and NO2ꢂ, respectively, the k values of
110 l molꢂ1 sꢂ1 and 220 l molꢂ1 sꢂ1 for 14 and 3240
and 11 000 l molꢂ1 sꢂ1 for 18 are obtained. These
values are very close to those of the unencumbered
mononuclear redox catalysts having the same driving
force, as can be seen in Fig. 9 where the comparison is
made between the star-shaped hexanuclear redox cata-
lysts and the unencumbered group of redox catalysts.
The points corresponding to the k values for 14 and
18 are close to the straight line. The star branch and core
3. Conclusion
The first study of the relationship between the kinetics
and thermodynamics of the redox catalysis of the
cathodic reduction of nitrate and nitrite in water by
soluble organometallic redox catalysts could be
achieved. The basis for the synthesis of a family of
Fe(C5H4R)(arene) redox catalysts lies in the CpFeꢀ
induced activation of the benzylic protons of the methyl
substituents of the arene ligand and subsequent reac-
tions with electrophiles. In particular, the CpFeꢀ
induced one-pot permethylation and perfunctionaliza-
tion of the mesitylene and hexamethylbenzene ligands
have been shown to be possible when the stereoelec-
tronically perturbing carboxylate substituent is present
on the Cp ring. These reactions have led to the synthesis
Fig. 5. Working curve of degree-5 polynomial regression drawn using
the theoretical Ik /Id values.

234
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
Fig. 6. Representation of log kꢃf (E8) for the reduction of nitrate (a) and nitrite (b).
/
Scheme 1. Compared outer-sphere (top) and inner-sphere (bottom) mechanisms for the redox catalysis of the cathodic reduction of nitrate (or
nitrite).
Fig. 7. Cyclic voltammograms of the hexaironꢁ
/
toluidine complex 14 obtained in absence (a) and presence (b) of nitrate. COx
ꢃ/
2.2ꢄ ;
10ꢂ4 mol lꢂ1
/
CNO 13.9ꢄ
ꢂ
ꢃ/
/
10ꢂ3 mol lꢂ1
.
3

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
235
Fig. 8. Cyclic voltammograms of the hexairon complex 18 recorded in absence (a) and presence (b) of nitrate. COx
3.35ꢄ
10ꢂ3 mol lꢂ1. CLiOH 0.1 mol lꢂ1; vꢃ170 mV s-1.
ꢃ/
5ꢄ
/
10ꢂ5 mol lꢂ1; CNO
ꢂ
ꢃ/
3
/
ꢃ/
/
of a family of ten mononuclear-and two star-shaped
hexanuclear complexes containing the redox active iron-
sandwich unit(s). All these complexes also bear the
carboxylate or the hexaphenol function allowing the
solubilization in basic aqueous medium of both redox
forms of the FeII/FeI redox couple involved in the redox
catalytic process. This family of complexes can be
divided into three groups: (i) a group of six unencum-
bered complexes in which the polymethylbenzene ligand,
with its planar framework, does not perturb the
approach of a substrate onto the FeI center of the
sandwich complexes; (ii) a group of four complexes in
which the bulky substituents of the arene ligand inhibit
the approach of a substrate onto the iron center which
increases the electron-transfer distance and significantly
reduces its rate; (iii) a group of two hexa-arm star-
shaped hexanuclear complexes in which the redox
catalyst is attached to the termini of the branches.
Cyclic voltammetry, polarography and chronoampero-
metry have been compared for the kinetic study of the
redox catalytic process using 1a. The values of the rate
constant k of the homogeneous reduction of nitrate by
the 19-electron form of the redox catalyst are in very
good agreement for the three methods, although cyclic
voltammetry is more precise and convenient. Using the
latter technique, the k values were determined for the 12
redox catalysts. For the first group, the plot of log k
versus the redox potential gives a straight line, pointing
to a Marcus-type correlation which indicates that the
electron transfer is rate limiting. For the second group,
the comparison at identical driving force shows that the
k values are an order of magnitude lower than for the
first group. This shows that bulk around the iron center
lowers the rate of electron transfer, as expected from the
results obtained by several research groups with metal-
lodendrimers. It is concluded that the mechanism is
essentially of the outer-sphere type between FeI and
ꢂ
NO3 or NO2ꢂ. For the third group, the k values
almost fit on a straight line, an indication of relatively
weak steric effect. The attachment of the redox catalyst
onto the termini of the branches of the hexa-arm star-
shaped cores does not really decrease their catalytic
activity.
This study sheds some light on the rate-limiting
primary electron transfer step, but not on the subse-
quent electron-transfer and proton transfer steps that
Fig. 9. Comparison of the catalytic activities of the two hexairon complexes 14 and 18 with those of the non-bulky catalysts for the reduction of
nitrate (left) and nitrite (right).

236
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
are faster. Indeed, we have confirmed in this study that
ꢂ
4.1.1. [FeII(h5-C5H4CO2CH3)(h6-C6Me6)][PF6], 1b
In a flamed Schlenk tube, 1a (1.91 g, 4.05 mmol) was
deaerated under vacuum, then 40 ml MeOH were added
using a cannula, then 1 ml BF3Et2O (48%) was slowly
added using a syringe. The orange solution was refluxed
for 16 h, then concentrated, and nitromethane was
added. The organic phase was washed using a diluted
aqueous HPF6 solution, then using a diluted aqueous
NaHCO3 solution, then dried over sodium sulfate. After
filtration, the solution was concentrated and filtered
over a short alumina column in order to remove traces
of unreacted 1a. The ester [FeII(h5-C5H4CO2CH3)(h6-
C6Me6)][PF6], 1b, was precipitated by addition of ether
and washed using this solvent, giving an orange micro-
crystalline powder (1.73 g, 3.56 mmol, 88% yield).
1H NMR (CD3CN, d ppm, TMS): 4.91 (m, 4H,
the catalyzed reduction proceeds more rapidly for NO2
than for NO3ꢂ. The detection of intermediates in this
system using spectroscopic studies, however, has not yet
been possible.
Subsequent studies with dendritic systems could
provide information on the bulk at the periphery of
the metallodendrimer. In this first approach, the bulk
problem at the dendrimer periphery was avoided by
choosing the metallo-star topography.
4. Experimental
4.1. General data
C5H4), 3.92 (s, 3H, Oꢀ
/
CH3), 2.40 (s, 18H, arene CH3).
13C (CD3CN, d, TMS): 166.96 (CO2H), 100.74 (Cq,
The reactions were carried out under an inert atmo-
sphere using Schlenk techniques. Solvents were freshly
distilled under argon. Dichloromethane and acetonitrile
were dried over calcium hydride and tetrahydrofuranne,
diethylether and dimethoxymethane were dried over
sodium wires using ketyl radical as an indicator, pentane
C6Me6), 81.73, 78.94 (CH, C5H4), 53.49 (Oꢀ
/
CH3), 17.26
(arene CH3). Infrared (Nujol):1725 cmꢂ1 (CO), 840
cmꢂ1 (PF6).
Anal. for C19H25O2FePF6: C 46.93, H 5.18. Found: C
46.94, H 5.27%.
was dried over a NaꢁK alloy and methanol was dried
/
over magnesium. For the syntheses of the 19-electron
FeI complexes, tetrahydrofuranne was dried over
LiAlH4 and degassed under vacuum at low temperature.
Chromatography and filtration were performed using
4.1.2. Disodium salt 6
In a flamed Schlenk tube, [FeII(h5-C5H4CO2CH3)(h6-
C6Me6)][PF6] (1.00g, 2.06 mmol) and t-BuOK (0.700 g,
6.23 mmol) were deaerated under vacuum, and 20 ml
THF were added at 0 8C using a canula. The reaction
medium instantaneously became red, and was stirred at
0 8C until disappearance of the orange solid. THF was
removed under vacuum and the red compound was
extracted using dry pentane into another Schlenk flask.
A stream of dry CO2 was allowed into the solution and a
yellow precipitate immediately appeared while the solu-
tion became colorless. After removing pentane under
vacuum, nitromethane was added and this solution was
washed using a diluted aqueous HPF6 solution. The
organic phase was dried over sodium sulfate, filtered,
and concentrated, and the product was precipitated
using ether and with this solvent, yielding [FeII(h5-
C5H4CO2Me)(h6-C6Me5CH2CO2H)][PF6] (0.666 g, 1.25
mmol). This product was then dissolved in the minimum
alumina (Prolabo 50ꢁ
were carried out using a Paragon 1000 Fourrier
transformed PerkinꢁElmer apparatus, calibrated with
/
160 mm, IIꢁ/III). Infrared spectra
/
polystyrene. Samples were analyzed between two NaCl
plates, in KBr pellets or in solution in 0.1 mm-thick
cells. The 250 MHz 1H and 62.9 13C NMR spectra were
recorded using a Brucker 250 FT spectrometer and
chemical shifts are given in part per million (d, ppm)
versus tetramethylsilane (TMS). The EPR spectra were
recorded using an ESP 300 E Brucker spectrometer
equipped with an Oxford ESR 900 cryostat. The
electrochemical studies were carried out under argon
using a PAR 273 EG & G galvanostat potentiostat for
cyclic voltammetry and chronoamperometry with the
three-electrode format: the working electrode was a
METROHM AG CH 9100 HERISSAU hanging mer-
cury drop, the reference electrode was a saturated
calomel electrode, and the counter electrode was a
platinum wire. For polarography, a Metrohm E 611
three-electrode polarograph equipped with a VA Scan-
ner E 612 was used. All the measurements were effected
in aqueous solution with 0.1 M LiOH as a supporting
electrolyte. Water was purified by inverse osmose and
mercury was tridistilled. Sodium nitrate and nitrite
(Aldrich) were used as received. Elemental analyses
were carried out at the Lyon-Villeurbanne CNRS
center.
amount (10 ml) of H2OꢁEtOH (50/50) together with
/
NaOH (0.100 g, 2.5 mmol), and the mixture was stirred
for 1 h at room temperature (r.t.). After removing water
under vacuum, the disodium salt 6 was obtained as an
orange powder (0.090 g, 1.25 mmol, 61% yield).
1H NMR (D2Oꢁ
4H, C5H4), 3.82 (s, 2H, CH2), 2.36 (s, 3H, CH3), 2.31 (s,
6H, CH3), 2.27 (s, 6H, CH3). 13C NMR (D2Oꢁ
NaOD
/NaOD 1 N, d ppm, TMS): 4.62 (m,
/
0.1 N, d ppm, TMS): 187.0, 172.0 (CO2H), 100.9, 100.7,
101.3, 97.4 (Cq C6Me5), 87.3 (Cq C5H4), 80.8, 79.0 (CH
of C5H4), 40.4 (CH2), 17.3, 16.7 (CH3). Infrared (Nujol,
n cmꢂ1): 1611, 1590 (CO), 841 (PF6).

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
237
4.1.3. [FeII(h5-C5H4CO2H)(h6-
C6(CH2CH3)6)][PF6], 7
C5H4CO2H)(h6-C6H3(C(CH3)3)3][PF6], 8, as an orange
microcrystalline powder (0.108 g, 0.19 mmol, 21%
yield).
In a three-necked flask equipped with a refrigerant
and a dropping funnel, 1a (0.59 g, 1.25 mmol and t-
BuOK (5.6 g, 50 mmol) were deaerated for 1 h at 40 8C
under vacuum, then 50 ml DME were added under
Argon. A red color immediately appeared. Then, CH3I
(3.2 ml, 50 mmol) was added dropwise to the suspension
by the dropping funnel. The reaction mixture was
refluxed for 3 days while the color progressively turned
beige. The solvent was removed under vacuum and the
solid residue was dissolved in 30 ml of CH2Cl2. The
organic phase was washed with water, then with a
diluted aqueous HPF6 solution before drying over
sodium sulfate. After filtering the solution and removing
the solvent under vacuum, the oily residue was washed
1H NMR: (CD3COCD3, d ppm, TMS): 6.51 (s, 3H,
C6H3), 5.52 (m, 4H, C5H4), 1.57 (s, 27H, CH3). 13C:
(CD3COCD3, d ppm, TMS): 167.07 (CO2H), 118.41
(Cq, C6H3), 95.12 (CH, C6H3), 81.16, 80.19 (CH, C5H4,
36.10 (Cq, C(CH3)3), 30.67 (CH3). Cyclic voltammetry:
(H2O, LiOH 0.1 M, 293 K, Hg, vꢃ
/
0.1 V sꢂ1, E8 (V vs.
1). IR (Nujol, n
SCE)): ꢂ
/
1.55 (DE8ꢃ
/
60 mV, Ia/Icꢃ
/
cmꢂ1): 1725 (nCꢁO), 840 (nPF ).
6
4.1.5. [FeII(h5-C5H4CO2H)(h6-
C6(CH2CH2C6H4OMe)6)][PF6], 19
In a flamed Schlenk tube, 1a (0.400 g, 1.00 mmol and
t-BuOK (1.9 g, 16.1 mmol) were deaerated for 1 h at
40 8C under vacuum. After the Schlenk flask had
cooled down to r.t., 10 ml DME were added under
argon. A red color immediately appeared. In another
Schlenk tube, p-MeOC6H4CH2Br (3.6 g, 16.4 mmol)
was dissolved in 50 ml DME, then added by canula to
the suspension contained in the first Schlenk flask,
which darkened the reaction medium. Stirring was
continued for 1 day at 40 8C while the color progres-
sively turned from orange to beige. The solvent was
removed under vacuum and the solid residue was
dissolved in 40 ml of CH2Cl2. The organic phase was
washed with water, then with a diluted aqueous HPF6
solution before drying over sodium sulfate. After
filtering the solution and removing the solvent under
vacuum, the product was washed with pentane in order
to remove excess p-MeOC6H4CH2Br, which gave the
with ether (3ꢄ20 ml) which gave an orange powder.
/
The product was solublilized in the minimum amount of
THF and KOH (0.140 g, 2.50 mmol) in the minimum
amount of water. The yellowꢁ
/
orange mixture was
stirred for 2 h at r.t., and the first part of the extraction
above
was
repeated,
which
gave
[FeII(h5-
C5H4CO2H)(h6-C6(CH2CH3)6)] [PF6], 7, as an orange
microcrystalline powder (0.012 mg, 0.21 mmol, 25%
yield).
1H NMR: (CD3COCD3, d ppm, TMS): 5.16 (m, 4H,
C5H4), 3.05 (m, 12H, CH2), 1.40 (t, 18H, CH3). 13C:
(CD3COCD3, d ppm, TMS): 167.12 (CO2H), 106.13
(Cq, C6R6), 80.97, 79.13 (CH, C5H4), 16.07 (CH2), 14.45
(CH3). IR (Nujol, n cmꢂ1): 1720 (nCꢁO), 840 (nPF ).
6
Cyclic voltammetry: (H2O, LiOH 0.1 M, 293 K, Hg; vꢃ
0.1 V sꢂ1, E8 (V vs. SCE)): ꢂ
1.62 (DE8ꢃ60 mV, Ia/
Icꢃ0.94).
/
/
/
/
hexamethoxybenzylation
product
[FeII(h5-
Anal. for C24H35O2FePF6: Calc.: C, 51.81, H 6.34.
Found: C, 51.92, H, 6.30%.
C5H4CO2H)(h6-C6(CH2CH2C6H4OMe)6)][PF6] 19 as
an orange microcrystalline powder (0.200 g, 0.20
mmol, 20% yield).
4.1.4. [FeII(h5-C5H4CO2H)(h6-
1H NMR: (CD3COCD3, d ppm, TMS): 7.10 (m, 24H,
C6H3(C(CH3)3)3][PF6], 8
C6H4ꢀ
3.75 (m, 12H, CH2ꢀ
CH2ꢀC6H4ꢀ
OMe). 13C: (CD3COCD3, d ppm, TMS):
166.70 (CO2H), 159.66 (CqꢀO, C6H4), 135.36 (Cqꢀ
/
OMe), 5.59 (m, 4H, C5H4), 3.84 (s, 18H, OCH3),
In a flamed Schlenk tube, [FeII(C5H4CO2H)(C6-
Me3H3)][PF6] (0.400 g, 0.93 mmol) and t-BuOK (5.400
g, 48 mmol) were deaerated under vacuum, and 50 ml
THF were added at 0 8C using a canula. The reaction
medium instantaneously became red. Then, CH3I (3.0
ml, 48 mmol) was added dropwise to the suspension
using a syringe. The reaction mixture was stirred for 3
days while the color progressively turned beige. Then t-
BuOK (5.400 g, 48 mmol) and CH3I (3.0 ml, 48 mmol)
were added under argon. The reaction medium imme-
diately turned beige and was stirred 3 more days. The
solvent was removed under vacuum and the solid
residue was dissolved in 40 ml of CH2Cl2. The organic
phase was washed with water, then with a diluted
aqueous HPF6 solution before drying over sodium
sulfate. After filtration and concentration of the solu-
tion, the product was precipitated by adding ether, then
/
C6H4ꢀOMe), 3.24 (m, 12H, CH2ꢀ
/
/
/
/
/
/
CH2, C6H4), 130.08 (CH meta-(O), C6H4), 115.09 (CH
ortho-(O), C6H4), 104,. 5 (Cq, C6R6), 81.66, 79.64 (CH,
C5H4), 80.72 (Cq, C5H4), 55.67 (OCH3), 36.95 (CH2ꢀ
CH2ꢀC6H4ꢀOMe), 33.55 (CH2ꢀCH2ꢀC6H4ꢀOMe). IR
(Nujol, n cmꢂ1): 1700 (nCꢁO), 840 (nPF ).
/
/
/
/
/
/
6
Anal. for C66H71O8FePF6: Calc.: C, 66.42; H, 6.00.
Found: C, 66.41; H, 6.01%.
4.1.6. [FeII(h5-C5H4CO2H)(h6-
C6(CH2CH2C6H4OH)6)][PF6], 10
In a three-necked flask equipped with a dropping
funnel, 19, whose synthesis is described above (0.320 g,
0, 27 mmol), was deaerated under vacuum. Argon was
introduced, 20 ml of freshly distilled CH2Cl2 were added
using a syringe, and a molar solution of BBr3 (3.5 ml,
washed with ether (3ꢄ
20 ml), which gave [FeII(h5-
/

238
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
3.37 mmol) in CH2Cl2 was added dropwise through the
dropping funnel to the reaction medium at ꢂ30 8C,
which provoked a darkening of the orange solution. The
reaction medium is allowed to warm up to r.t. overnight
and is stirred for 1 day. BBr3 was then slowly hydrolyzed
using 30 ml deaerated ice water. The organic phase was
mixture gave [FeII(h5-C5H4CO2H)(h6-p-CH3C6H4Cl)]-
[PF6], 20, as a crystalline orange solid (4.25 g, 9.70
mmol, 43.5% yield).
/
1H NMR: (CD3 CN, d ppm, TMS): 6.39 (m, 4H,
C6H4), 5.33 (m, 4H, C5H4), 2.36 (s, 3H, CH3). 13C:
(CD3CN, d ppm, TMS): 167.53 (CO2H), 107,70 (Cqꢀ
/
filtered over celite, which was then washed with 2ꢄ
/
20
Cl, C6H4), 105.23 (Cqꢀ/CH3, C6H4), 90.44, 90.07 (CH,
ml CH3NO2. The mixed organic fractions were washed
with a diluted aqueous HPF6 solution, then with water,
before being dried over sodium sulfate, filtered and
concentrated. The product 10 was precipitated by
C6H4), 81.25 (Cq, C5H4), 84.24, 80.49 (CH, C5H4), 19.63
(CH3), IR: (nujol, n cmꢂ1): 1725 (nCꢁO), 840 (nPF ).
6
4.1.8. [FeII(h5-C5H4CO2CH3)(h6-p-CH3C6H4Cl)]
[PF6], 21
In a flamed Schlenk tube, 20 (0.400 g, 0.91 mmol) was
deaerated under vacuum, then 30 ml dry MeOH were
adding ether and washed with 2ꢄ50 ml ether, which
/
gave [FeII(h5-C5H4CO2H)(h6-C6(CH2CH2C6H4OH)6)]-
[PF6] 10 as an orange microcristalline powder (0.261 g,
0.23 mmol, 84% yield).
added by canula and 1.2 ml of 48% BF3×
/
Et2O (4.56
1H NMR: (CD3COCD3, d ppm, TMS): 6.95 (m, 24H,
mmol) were added by syringe. The orange solution was
refluxed for 14 h, then concentrated and 30 ml
nitromethane was added. The organic phase was washed
with a diluted aqueous HPF6 solution, then with a
diluted aqueous NaHCO3 solution, and dried over
sodium sulfate. After filtration, the solution was con-
centrated and filtered over alumina in order to remove
the traces of unreacted carboxylic acid. The product was
then precipitated and washed with ether. [FeII(h5-
C5H4CO2CH3)(h6-p-CH3C6H4Cl)][PF6], 21, was ob-
tained as an orange microcrystalline powder (0.40 g,
0.88 mmol, 90% yield).
ꢀ
/
C6H4ꢀ
C6H4ꢀOH), 3.24 (m, 12H, CH2ꢀ
(CD3COCD3, d ppm, TMS): 167,20 (CO2H), 157.23
CH2, C6H4), 130.06 (CH
/
OH), 5.52 (m, 4H, C5H4), 3.75 (m, 12H, CH2ꢀ
/
/
/
CH2ꢀC6H4ꢀ
/
/
OH). 13C:
(Cqꢀ
/
O, C6H4), 13.15 (Cqꢀ
/
meta-(O), C6H4), 116.48 (CH ortho-(O), C6H4), 104.88
(Cq, C6R6), 81.60, 79.86 (CH, C5H4), 81.22 (Cq, C5H4),
37.00 (CH2ꢀ
C6H4ꢀ
OH). IR (Nujol, n cmꢂ1): 3300 (nOꢀH), 1700
(nCꢁO), 840 (nPF ). Cyclic voltammetry: (H2O, LiOH 0.1
M, 293 K, Hg, 6vꢃ
0.1 V sꢂ1, E8 (V vs. SCE)): ꢂ
1.60
(DE8ꢃ60 mV, Ia/Icꢃ0.90).
/
CH2ꢀ
/
C6H4ꢀ
/
OH), 33.49 (CH2ꢀ
/
CH2ꢀ
/
/
/
/
/
/
1H NMR: (CD3CN, d ppm, TMS): 6.40 (m, 4H,
4.1.7. [FeII(h5-C5H4CO2H)(h6-p-CH3C6H4Cl)]
[PF6], 20
C6H4), 5.36 (m, 4H, C5H4), 3.91 (s, 3H, Oꢀ
(s, 3H, CH3 aromatic). 13C: (CD3CN, d ppm, TMS):
166.20 (CO2H), 108.30 (CqꢀCl, C6H4)), 106.50 (Cqꢀ
CH3, C6H4)), 90.36, 90.04 (CH, C6H4), 82.08, 80.01
/CH3), 2.35
In a 250 ml three-necked flask equipped with a
refrigerant, the mixture of the solids ferrocenylcar-
boxylic acid (5 g, 21.7 mmol) and AlCl3 (15 g, 110
mmol) was deaerated under vacuum, then argon was
introduced and 100 ml p-chlorotoluene were added by
canula. The reaction medium was slowly heated and
stirred for 2 h at 60 8C, then deaerated water (0.4 ml,
21.7 mmol was added by a syringe and the reaction
medium was vigorously stirred at 90 8C for 1 day. After
cooling down to 40 8C, AlCl3 was slowly hydrolyzed
with 100 ml deaerated ice water. After decantation, the
aqueous phase was washed with ether until the ether
phase was colorless. The Al3ꢀ ions were removed as
precipitated alumina by adding an ammonia solution
until pH 9 was reached. The pH was then brought to 13
by addition of 100 ml of a 20% aqueous NaOH solution.
Addition of 1 ml aqueous HPF6 provoked the precipita-
tion of the side product [FeII(h5-C5H5)(h6-p-
CH3C6H4Cl)]ꢀPF6. After recrystallization from an
/
/
(CH, C5H4), 53.91 (Oꢀ
/
CH3), 19.75 (CH3). IR: (Nujol, n
cmꢂ1): 1730 (nCꢁO), 1100 (nCꢀOꢀC), 840 (nPF ).
Anal. for: Calc.: C14H14O2FePF6Cl: C, 37.32; H, 3.13.
Found: C, 37.32; H, 3.21%.
6
4.1.9. [FeII(h5-C5H4CO2CH3)(h6-p-
CH3C6H4NHCH3)] [PF6], 11
In a flamed Schlenk tube, 21 (0.606 g, 1.34 mmol),
NH2Me×/HCl (0.361 g, 5.36 mmol) and K2CO3 (1.111 g,
8.04 mmol) were deaerated under vacuum, 10 ml DMF
were introduced under argon by canula, and the orange
suspension was stirred for 1 day at r.t. Water (200 ml)
was added and the aqueous phase was extracted with
nitromethane. The organic phase was washed over
sodium sulfate and then concentrated and filtered over
alumina using acetonitrile as eluent. The product was
precipitated and washed with ether; [FeII(h5-
C5H4CO2CH3)(h6-p-CH3C6H4NHCH3)][PF6], 11, was
obtained as a yellow microcrystalline powder (0.508 g,
1.14 mmol, 85% yield).
acetoneꢁethanol mixture, 0.650 g (9.70 mmol, 7% yield)
/
of this complex resulting from the cleavage of the
carboxylated ring of ferrocenyl carboxylic acid were
obtained. The aqueous phase obtained above was
acidified to pH 1 by addition of concentrated HCl.
Upon addition of HPF6, the carboxylic acid 20 was
1H NMR: (CD3CN, d ppm, TMS): 6.20 (s large, 1H,
NH), 5.88 (m, 4H, C6H4), 5.26 (m, 4H, C5H4), 3.92 (s,
3H, Oꢀ
/
CH3), 2.94 (s, 3H, Nꢀ
/
CH3), 2.37 (s, 3H, CH3
benzene). 13C: (CD3CN, d ppm, TMS): 166.77 (CO2H),
precipitated. Recrystallization from an acetoneꢁethanol
/

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
239
128.59 (Cqꢀ
(CH m o-(N), C6H4), 79.64 (Cq, C5H4), 78.96, 77.44
(CH, C5H4), 67.98 (CH ortho-(N), C6H4), 53.28 (Oꢀ
/
N, C6H4)), 96.93 (Cqꢀ
/
CH3, C6H4)), 88.28
complex 14 was precipitated and washed with ether and
obtained as an orange microcrystalline powder (0.099 g,
0.034 mmol, 86% yield).
/
CH3), 30.00 (Nꢀ
/CH3), 19.26 (CH3 benzene). IR: (Nujol,
1H NMR: (CD3CN, d ppm, TMS): 7.00 (m, 4H, C6H4
decomplexed), 5.65 (m, 24H, C6H4), 5.06 (m, 24H,
n cmꢂ1): 3510 (nNꢀH), 1720 (nCꢁO), 1570 (dNꢀH), 1280
(nCꢀN), 1100 (nCꢀOꢀC), 840 (nPF ).
C5H4), 3.42 (m, 12H, Nꢀ
/
CH2), 3.00 (s, 18H, Nꢀ
/
CH3),
Anal. for C15H18NO2FePF6: 6Calc.: C, 40.48; H, 4.08;
N, 3.14. Found: C, 40.52; H, 4.07; N, 2.95%.
2.67 (m, 12H, CH2ꢀC6R5), 2.27 (s, 18H, CH3 benzene),
/
1.61 (m, 24H, CH2ꢀ
/
CH2ꢀ
/
C6R5ꢀ
/
CH2ꢀ
/
CH2ꢀ
/
CH2ꢀ
/
C6R5).13C: (CD3CN, d ppm, TMS): 168.81 (CO2H),
137.88 (Cq, C6R5), 132.17 (CH meta-(N), C6H4 decom-
4.1.10. 13 (Hexaironꢁtoluidine)
/
In a flamed Schlenk tube, 11 (0.400 g, 0.90 mmol)
were deaerated, then 20 ml of freshly distilled THF were
added under argon. The suspension was frozen with
liquid nitrogen and t-BuOK (0.092 g, 0.90 mmol) was
added under argon. The reaction mixture was warmed
up in 5 min and slowly became red. When the starting
cationic complex 11 completely disappeared, the tem-
perature was set to 0 8C and 12 (0.080 g, 0.070 mmol) in
4 ml THF was added dropwise using a dropping funnel.
The reaction mixture was then allowed to warm up to
r.t. in 1 h and stirred for 16 h. The red color had then
completely disappeared and a yellow precipitate had
formed in the orange solution. THF was filtered and the
solid was washed with THF until the solvent remained
colorless in order to remove excess cation. CH2Cl2 was
added, then the organic phase was washed with water,
then with a diluted aqueous HPF6 solution, and finally
with an aqueous NaHCO3 solution. After drying over
sodium sulfate, filtering and concentrating, the product
plexed), 129.79 (Cqꢀ
C6H4 decomplexed), 96.70 (Cqꢀ
(CH meta-(N), C6H4), 79.44, 78.20 (CH, C5H4), 66.92
(CH ortho-(N), C6H4), 53.73 (NꢀCH2), 40.13 (NꢀCH3),
/
N, C6H4)), 122.81 (CH ortho-(N),
/CH3, C6H4)), 88.62
/
/
30.63, 29.90, 28.42 (CH2), 19.77 (CH3). IR: (Nujol, n
cmꢂ1): 1730 (nCꢁO), 840 (nPF ). Cyclic voltammetry:
(H2O, LiOH 0.1 M, 293 K, Hg,6vꢃ
/
0.1 V sꢂ1, E8 (V vs.
0.95).
114H138O12Fe6N6P6F36 (for six iron
SCE): ꢂ
Anal. for
/
1.63 (DE8ꢃ
/
80 mV, Ia/Icꢃ
/
C
atoms): Calc.: C, 45.78; H, 4.65; N, 2.81. Found: C,
46.00; H, 4.93; N, 2.91%.
4.1.11. 17 (Hexaaldehyde)
In a flamed Schlenk tube, 12 (0.220 g, 0.19 mmol), p-
HOC6H4CHO (1.37 g, 11.3 mmol) and K2CO3 (3.12 g,
22.6 mmol) were deaerated, then 25 ml of dry DMF
were added by canula under argon, and the reaction
mixture was refluxed for 1 day at 45 8C. The solution
was then added to 100 ml water in order to precipitate
the product, and the suspension was filtered over celite.
was precipitated with 20 ml ether and washed with 2ꢄ
/
20 ml ether. The hexaironꢁtoluidine complex 13 was
/
Celite was washed with 5ꢄ
/
50 ml water, then with 2ꢄ
/
obtained as an orange microcrystalline powder (0.120 g,
0.04 mmol, the yield is estimated to 60% given the 7%
15 ml ether. The precipitate was dissolved in 20 ml of
CH2Cl2. The organic phase was washed with 20 ml of an
aqueous NaCl solution, then with 10 ml water, and
finally dried over sodium sulfate. After fitration,
CH2Cl2 was removed under vacuum, and the solid
product was washed with pentane, which gave the
hexaaldehyde 17 as a pale yellow powder (0.205 g,
0.18 mmol, 96% yield).
1
decomplexation determined by H NMR).
1H NMR: (CD3CN, d ppm, TMS): 6,70 (m, 4H, C6H4
decomplexed), 5.67 (m, 24H, C6H4), 5.04 (m, 24H,
C5H4), 3.83 (s, 18H, Oꢀ
2.98 (s, 18H, NꢀCH3), 2.62 (m, 12H, CH2ꢀ
(s, 18H, CH3 benzene), 1.77 (m, 12H, CH2ꢀCH2ꢀ
1.49 (m, 12H, CH2ꢀCH2ꢀCH2ꢀ
C6R5). 13C: (CD3CN, d
ppm, TMS): 168.35 (CO), 137.79 (Cq, C6R6), 130.66
(CH meta-(N), C6H4 decomplexed), 129.35 (CqꢀN,
C6H4)), 113.66 (CH ortho-(N), C6H4 decomplexed),
96.84 (CqꢀCH3, C6H4)), 87.97 (CH meta-(N), C6H4),
78.59, 77.16 (CH, C5H4), 79.13 (Cq, C5H4), 66.20 (CH
ortho-(N), C6H4), 53.53 (NꢀCH2), 52.81 (OꢀCH3),
CH3), 30.40, 29.84, 28.19 (CH2), 19.06
/
CH3), 3.38 (m, 12H, Nꢀ
C6R5), 2.27
C6R5),
/
CH2),
/
/
/
/
/
/
/
1H NMR: (CDCl3, d ppm, TMS): 9.84 (s, 6H, CHO),
7.37 (m, 24H, C6H4), 4.02 (t, 12H, Oꢀ
12H, CH2ꢀC6), 1.83 (m, 12H, CH2ꢀCH2ꢀ
1.68 (m, 12H, CH2ꢀCH2ꢀ
C6). 13C: (CDCl3, d ppm,
TMS): 190.68 (CO), 163.98 (CqꢀO, C6H4), 136.51 (Cq,
C6), 132.06 (CH meta-(O), C6H4), 129.99 (CqꢀCHO,
O),
/CH2), 2.56 (m,
/
/
/
/
CH2ꢀC6),
/
/
/
/
/
/
/
/
C6H4), 114.70 (CH ortho-(O), C6H4), 67.71 (CH2ꢀ
/
38.79 (Nꢀ
/
30.63, 29.29, 27.97 (CH2). IR (Nujol, n cmꢂ1): 2746
(nHCꢁO), 1685 (nCꢁO), 1265 (nCꢀOꢀC).
(CH3). IR: (Nujol, n cmꢂ1): 1720 (nCꢁO), 840 (nPF ).
14 NaOH (0.024 g, 0.04 mmol) and 13 (0.120 g,6 0.04
mmol) were dissolved in a minimum amount 20 ml of a
Anal. For C72H78O12: Calc.: C, 76.20; H, 6.87. Found:
C, 76.13; H, 6.93%.
H2O*EtOH (50/50) mixture and left 2 h at r.t. THF
/
was removed under vacuum, then nitromethane was
added. The organic phase was washed with water, then
with a diluted aqueous HPF6 solution and with an
aqueous NaHCO3 solution, so that the pH of the
4.1.12. [FeII(h5-C5H4CH2NHCH2CH2NH2)(h6-
C6(CH3)5CH2COOH)][PF6], 16
In a flamed Schlenk tube, 15 ((0.210 mg, 0.42 mmol)
(0.220 g, 0.19 mmol), were deaerated, 5 ml of a
aqueous solution is around 5. The hexaironꢁ
/toluidine
CH3CNꢁCH2Cl2 mixture (1:3) were introduced by
/

240
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
canula under argon. After stirring a few minutes, this
solution was added into a second Schlenk tube contain-
ing ethylene diamine (2 ml, 20 mmol) by syringe. After
stirring for 1 h, the orange solution was concentrated,
and the orange product was precipitated by addition of
dry ether under argon, the solvent was removed by
cipitated and washed with ether, giving 18 as a yellow
solid (0.185 g, 0.030 mmol, 81% yield).
1H NMR: (CD3CN, d ppm, TMS): 7.37 (m, 24H,
C6H4), 4.59 (m, 4H, C5H4), 4.00 (m, 24H, C5H4ꢀ
CH2COOH), 2.60 (m, 12H, CH2ꢀC6R5ꢀNꢀ
CH2ꢀNꢀNꢀCH2ꢀCH2ꢀN), 2.40 (m, 15H, CH3 benze-
nic), 1.78 (m, 12H, CH2ꢀCH2ꢀCH2ꢀC6R5), 1.61 (m,
12H, CH2ꢀCH2ꢀ
C6R5). 13C: (CD3CN, d ppm, TMS):
172.27 (CO), 163.54 (CqꢀO, C6H4), 139.96 (CqꢀCHO,
/
CH2ꢀ
/
/
/
/
CH2ꢀ
/
/
/
/
/
/
canula, 20 mg of 10% Pdꢁ/C and 20 mL of dry methanol
/
/
/
were introduced. A pressure of H2 slightly above 1 atm
was introduced, and the solution was vigorously stirred
for 12 h, then filtered over celite. After removing the
solvent under vacuum, the residue was dissolved in 10
ml of nitromethane. The organic phase was washed with
a diluted aqueous HPF6 solution, then with an aqueous
NaHCO3 solution, so that the pH of the aqueous phase
is around 5. The organic phase was then dried over
sodium sulfate, filtered and concentrated. The amino-
acid complex [FeII(h5-C5H4CH2NHCH2CH2NH2)(h6-
C6(CH3)5CH2COOH)][PF6] 16 was precipitated and
washed with ether, which gave 16 as a yellow solid
/
/
/
/
C6H4), 136.31 (Cq, C6R6), 132.62 (CH meta-(O), C6H4),
116.38 (CH ortho-(O), C6H4), 100.40, 99.90, 99.61, 99.40
(Cq, C6Me5CH2), 68.70 (CH2ꢀ
/
O), 45.12 (C5H4ꢀ
/
CH2),
44.95 (CH2ꢀNH2), 39.88 (Nꢀ
/
/
CH2ꢀCH2), 38.58
/
(CH2COOH), 30.91, 29.71, 2858 (CH2 core), 17.69,
17.64 (CH3 benzene). IR (Nujol, n cmꢂ1): 3445 (nNꢀH),
1731 (nCꢁO), 1254 (nCꢀOꢀC), 844 (nPF ). Cyclic voltam-
metry: (H2O, LiOH 0.1 M, 293 K, Hg,6 vꢃ
/
0.1 V sꢂ1, E8
40 mV), (CH3CN, n-
Bu4NBF4 0.1 M, 293 K, Hg, vꢃ
0.1 V sꢂ1, E8 (V vs.
SCE)): ꢂ1.80 (DE8ꢃ80 mV, Ia/Icꢃ0.96).
Anal. for 198H256N12O18Fe6P6F36K6: Calc.: C,
52.48; H, 5.89. Found (100%): C, 36.92 (52.33); H,
4.26 (6.03)% (Tables 1ꢁ5).
(V vs. SCE)): ꢂ
/
1.79 (DE8ꢃ
/
/
1
(0.185 mg, 0.30 mmol, 81% yield). H NMR: (CD3CN,
/
/
/
d ppm, TMS): 4.52 (m, 4H, C5H4), 3.93 (s, 2H,
C
CH2COOH), 3.64 (s, 2H, C5H4ꢀ
/
CH2), 2.96 (m, 2H,
NꢀCH2ꢀCH2), 2.71 (m, 2H, CH2ꢀ
/
/
/NH2), 2.46 (s, 3H,
/
CH3 benzene), 2.41 (s, 6H, CH3 benzene), 2.38 (s, 6H,
CH3 benzene). 13C: (CD3CN, d ppm, TMS): 173.25
(CO), 101.92, 100.85, 100.23, 99.89 (Cq, C6Me5), 96.66
Acknowledgements
(Cq, C5H4), 80.85, 79.67 (CH, C5H4), 45.00 (C5H4ꢀ
/
CH2), 44.90 (CH2ꢀ
/
NH2), 40.14 (Nꢀ
/
CH2ꢀ
/
CH2), 38.30
(CH2COOH), 17.71, 17.26 (CH3 benzene). IR: (Nujol, n
We are grateful to the Institut Universitaire de
France, the University Bordeaux I, the CNRS, the
Re´gion Aquitaine, the Ministe`re de la Recherche et de
cmꢂ1): 3446 (nNꢀH),1710 (nCꢁO), 844 (nPF ).
6
Anal. for C21H31N2O2FePF6×
/
H2O: Calc.: C, 44.80; H,
la Technologie (Thesis Grant to S.R.), and the Frenchꢁ
/
5.86; N, 4.97. Found: C, 44.80: H, 5.97; N, 5.11%.
Spanish Picasso program No 97054 for financial sup-
port.
4.2. Complex 18
In a flamed Schlenk tube, 12 (0.220 g, 0.19 mmol) was
deaerated, then 5 ml of a CH2Cl2ꢁDMF mixture (3:1)
/
References
were added by canula under argon. After stirring a few
minutes, the solution was added using a syringe into a
second Schlenk tube containing a solution of ethylene
diamine (20 mmol) in 2 ml CH2Cl2, and the reaction
mixture was stirred for 1 h. The orange solution was
concentrated, the orange product was precipitated with
dry ether under argon, and the solvent was removed by
[1] (a) A.E. Kaifer, M. Gomez-Kaifer, Supramolecular Electrochem-
istry, Wiley-VCH, Weinheim, 1999;
(b) D. Astruc, Electron Transfer and Radical Processes in
Transition-Metal Chemistry, VCH, New York, 1995.
[2] (a) G.E. Oosterom, J.N.H. Reek, P.C.J. Kamer, P.W.N.M. van
Leeuwen, Angew. Chem., Int. Ed. Engl. 40 (2001) 1828;
(b) D. Astruc, F. Chardac, Chem. Rev. 101 (2001) 2991.
canula. Twenty milligram of 20% PdꢁC were added,
/
[3] (a) G.R. Newkome, C.N. Moorefield, F. Vogtle, Dendrimers and
¨
Dendrons: Concepts, Syntheses, Application, VCH, New York,
2001;
then 20 ml dry methanol were introduced by canula.
After removing argon under vacuum, a slight over-
pressure of H2 was introduced with a balloon. The
reaction mixture was vigorously stirred, then filtrated
over celite, the solvent was removed under vacuum and
the product was dissolved in 20 ml of MeNO2. The
organic phase was washed with a diluted aqueous HPF6
solution, then with an aqueous NaHCO3 solution, so
that the pH of an aqueous phase is around 5. The
organic phase was dried over sodium sulfate, filtered
and concentrated. The hexairon complex 18 was pre-
(b) R.M. Crooks, M. Zhao, L. Sun, V. Chechik, L.K. Yeung, Acc.
Chem. Res. 34 (2001) 181.
[4] (a) P.J. Dandliker, F. Diederich, M. Gross, C.B. Knobler, A.
Louati, E.M. Stanford, Angew. Chem., Int. Ed. Engl. 106 (1994)
1821;
(b) P.J. Dandliker, F. Diederich, J.-P. Giesselbrecht, A. Louati,
M. Gross, Angew. Chem., Int. Ed. Engl. 107 (1995) 2906;
(c) P.J. Dandliker, F. Diederich, A. Zingg, J.-P. Giesselbrecht, M.
Gross, A. Louati, E.M. Stanford, Helv. Chim. Acta 80 (1997)
1773;
(d) D.K. Smith, F. Diederich, Chem. Eur. J. 4 (1998) 1353;

S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ
/
242
241
(e) P. Weyermann, J.-P. Giesselbrecht, C. Boudon, F. Diederich,
M. Gross, Angew. Chem., Int. Ed. Engl. 38 (1999) 3215.
[5] (a) J. Isberner, F. Vogtle, L. De Cola, V. Balzani, Chem. Eur. J. 3
¨
(e) G. Horanyi, E.M. Rizmayer, J. Electroanal. Chem. 188 (1985)
265.
[15] (a) S.H. Cheng, Y.O. Su, Inorg. Chem. 33 (1994) 5847;
(b) S. Cattarin, J. Appl. Electrochem. 22 (1992) 1077;
(c) H.-L. Li, J.Q. Chambers, D.T. Hobbs, J. Appl. Electrochem.
18 (1988) 454;
(1997) 706;
(b) V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M.
Venturi, Acc. Chem. Res. 31 (1998) 26;
(c) F. Vogtle, M. Plevoetz, M. Nieger, G.C. Azzellini, A. Credi, L.
¨
(d) J.D. Genders, D. Hartsough, D.T. Hobbs, J. Appl. Electro-
chem. 26 (1996) 1;
De Cola, V. De Marchis, M. Venturi, V. Balzani, J. Am. Chem.
Soc. 121 (1999) 6290;
(d) P. Ceroni, V. Vicinelli, M. Maestri, V. Balzani, W.M. Muller,
(e) H.-L. Li, D.H. Robertson, J.Q. Chambers, D.T. Hobbs, J.
Electrochem. Soc. 135 (1988) 1154;
¨
U. Muller, U. Han, F. Osswald, F. Vogtle, New J. Chem. 25
¨
¨
(2001) 989.
(f) E.E. Kalu, R.E. White, D.T. Hobbs, J. Electrochem. Soc. 143
(1996) 3094;
[6] (a) R. Sadamoto, N. Tomioka, T. Aida, J. Am. Chem. Soc. 118
(1996) 3978;
(g) N. Chebotareva, T. Nyokong, J. Appl. Electrochem. 27 (1997)
975;
(b) H.F. Chow, I.Y.-K. Chan, R.W.M. Kwok, Chem. Eur. J. 2
(1996) 1085.
(h) J. O’M Bockris, J. Kim, J. Electrochem. Soc. 143 (1996) 3801;
(i) P.F. Vicento, A.R. Tamarit, J.A. Alberola Enguidanos, Span,
ES 2,065,807 16 Feb 1995, Chem. Abstr. 122 (1995) 31807;
(j) N.D. Jensen, 19,517,652 Ger. Offen. DE (Cl.CO2F1/461), 21
Nov 1996, Appl. 19,517,652, 17 May 1995, Chem. Abstr. 126
(1997) 65150;
[7] (a) M. Kimura, Y. Sugihara, T. Muto, K. Hanabusa, H. Shirai,
N. Kobayashi, Chem. Eur. J. 5 (1999) 3495;
(b) J.J. Apperlo, R.A. Janssen, P.R.L. Malenfant, L. Groenendal,
J.M.J. Fre´chet, J. Am. Chem. Soc. 122 (2000) 7042.
[8] (a) C.B. Gorman, B.L. Parkhurst, W.Y. Su, K.-Y. Chen, J. Am.
Chem. Soc. 119 (1997) 1141;
(k) E.A. Stadlbauer, H. Loehr, J. Eberheim, B. Weber, Ger.
Offen. DE 19,512,955 (Cl.Co2F1/70), 17 Oct 1996, Appl.
19,512,955, 10 Apr 1995, Chem. Abstr. 126 (1997) 65044.
[16] For a recent review on the reduction of nitrate in aqueous
solution, see: J.C. Fanning, Coord. Chem. Rev. 199 (2000) 159.
[17] (a) M. Tokuoka, Collect. Czech. Chem. Commun. 4 (1932) 444;
(b) M. Tokuoka, Collect. Czech. Chem. Commun. 6 (1934) 339;
(c) I.M. Kolthoff, W.E. Haris, G. Matsuyama, J. Am. Chem. Soc.
66 (1944) 1782;
(b) C.B. Gorman, Adv. Mater. 9 (1997) 1117;
(c) C.B. Gorman, Adv. Mater. 10 (1998) 295;
(d) C.B. Gorman, J.C. Smith, M.W. Hager, B.L. Parkhurst, H.
Sierzputowska-Gracz, C.A. Haney, J. Am. Chem. Soc. 121 (1999)
9958;
(e) C.B. Gorman, J.C. Smith, Acc. Chem. Res. 34 (2001) 60.
[9] (a) G.R. Newkome, F. Cardullo, C.N. Moorefield, A.M.W.C.
Thompson, Chem. Commun. (1993) 925ꢁ926;
/
(d) W. Haris, I.M. Kolthoff, J. Am. Chem. Soc. 67 (1945) 1484.
[18] (a) M. Sharf, C. More´no, C. Costa, C. Van Djik, W.J. Payne, J.
LeGall, I. Moura, J.G. Moura, Biochem. Biophys. Res. Commun.
209 (1995) 1018;
(b) G.R. Newkome, G. Guther, C.N. Moorefield, F. Cardullo, L.
¨
Echegoyen, E. Pe´rez-Cordero, H. Luftmann, Angew. Chem., Int.
Ed. Engl. 34 (1995) 2023;
(c) G.R. Newkome, R.V.V. Narayanan, L.A. Godinez, J. Org.
Chem. 65 (2000) 1643.
(b) see Ref. [1b], Chapter 7.
[19] (a) I. Taniguchi, N. Nakashima, K. Yasukouchi, J. Chem. Soc.,
Chem. Commun. (1986) 1814;
[10] C.M. Cardona, A.E. Kaifer, J. Am. Chem. Soc. 120 (1998) 4023.
[11] S. Nlate, J. Ruiz, V. Sartor, R. Navarro, J.-C. Blais, D. Astruc,
Chem. Eur. J. 6 (2000) 2544.
(b) H. Li, W.C. Anderson, J.Q. Chambers, D.T. Hobs, Inorg.
Chem. 28 (1989) 863.
[12] Reviews on metallodendrimers: (a) G.R. Newkome, E. He, C.N.
Moorefield, Chem. Rev. 99 (1999) 1689;
(b) I. Cuadrado, M. Mora´n, C.M. Casado, B. Alonso, J. Losada,
[20] L. Ma, B.-Y. Zhang, H. Li, J. Electroanal. Chem. 362 (1993) 201.
[21] L. Ma, H. Li, Electroanalysis 8 (1995) 7.
[22] S. Kuwabata, S. Uezumi, K. Tanaka, T. Tanaka, Inorg. Chem. 25
(1986) 3018.
Coord. Chem. Rev. (1999) 193ꢁ
(c) M.A. Hearshaw, J.R. Moss, Chem. Commun. (1999) 1ꢁ
(d) M. Venturi, S. Serroni, A. Juris, S. Campagna, V. Balzani, in:
/195, 395;
/8;
[23] K. Tanaka, N. Komeda, T. Tanaka, Inorg. Chem. 30 (1991) 3282.
[24] A. Buet, A. Darchen, C. Moinet, J. Chem. Soc., Chem. Commun.
(1979) 447.
F. Vogtle (Ed.), Dendrimers II; Architecture, Nanostructure and
¨
Supramolecular Chemistry, Topics in Current Chemistry 210,
Springer, Berlin, 2000, pp. 193;
[25] E. Roma´n, R. Dabard, C. Moinet, D. Astruc, Tetrahedron Lett.
16 (1979) 1433.
(e) A. Juris, in: V. Balzani (Ed.), Electron Transfer in Chemistry,
Wiley-VCH, New York, 2001, vol. 3, Biological and Artificial
Supramolecular Systems, H.B. Gray, J. Winkler, V. Balzani, (Ass.
Eds.), section 2, Chapter 9, pp. 655.
[26] A. Darchen, C. Moinet, D. Astruc, Colloque D.G.R.S.T Report,
Paris, December 1978.
[27] C. Moinet, E. Roma´n, D. Astruc, J. Electroanal. Interf. Chem.
121 (1981) 241.
[13] Reviews on ferrocenyl dendrimers: (a) C.M. Casado, B. Gonzale`s,
I. Cuadrado, B. Alonso, M. Mora´n, J. Losada, Angew. Chem.
Int. Ed. 39 (2000) 2135;
[28] (a) D. Astruc, J.-R. Hamon, G. Althoff, E. Roma´n, P. Batail, P.
Michaud, J.-P. Mariot, F. Varret, D. Cozak, D. Astruc, J. Am.
Chem. Soc. 101 (1979) 5545;
(b) I. Cuadrado, M. Mora´n, J. Losada, C.M. Casado, C. Pascual,
B. Alonso, F. Lobete, in: G.R. Newkome (Ed.), Advances in
Dendritic Macromolecules, vol. 3, JAI Press, Greenwich, CT,
1996, p. 151;
(b) J.-R. Hamon, D. Astruc, P. Michaud, J. Am. Chem. Soc. 103
(1981) 758;
(c) V. Gerchais, E. Roma´n, D. Astruc, Organometallics 5 (1986)
2505.
[29] Reference [1b], Chapter 4.
(c) C.M. Casado, I. Cuadrado, M. Mora´n, B. Alonso, B. Garcia,
B. Gonzales, J. Losada, Coord. Chem. Rev. 185ꢁ
[14] (a) M. Schelef, Chem. Rev. 95 (1995) 209;
/
186 (1999). 53.
[30] (a) H. Taube, H. Myers, J. Am. Chem. Soc. 76 (1954) 2103;
(b) H. Taube, J. Chem. Educ. 45 (1968) 452;
(c) H. Taube, Angew. Chem., Int. Ed. Engl. 23 (1994) 329;
(d) H. Taube, Electron Transfer Reactions of Complex Ions in
Solution, Academic Press, New York, 1970;
(b) S.E. Manahan, Fundamentals of Environmental Chemistry,
Lewis Publishers, Boca Raton, FL, 1993, p. 401;
(c) H. Li, D.H. Roberson, J.Q. Chambers, D.T. Hobbs, J. Appl.
Electrochem. 18 (1988) 454;
(e) S.J. Lippard (Ed.),Progress in Inorganic Chemistry, An
Appreciation of Henry Taube, vol. 30, Wiley, New York, 1983.
[31] A. Poe, Trans. Met. Chem. 7 (1982) 65.
(d) R.R. Gadde, S. Bruckenstein, Electroanal. Chem. Interf.
Electrochem. 50 (1974) 163;
¨

242
S. Rigaut et al. / Inorganica Chimica Acta 334 (2002) 225ꢁ242
/
[32] Q.-Z. Shi, T.G. Richmond, W.C. Trogler, F. Basolo, J. Am.
Chem. Soc. 104 (1982) 4032.
Astruc, Section 2, f Inorganic and Organometallic Coumpounds,
Chapter 4, pp. 714ꢁ803.
/
[50] The redox potentials of the identical redox sites around the
molecule are not strictly identical, but are statistically distributed
around a mean value: see (a) F. Ammar, J.-M. Save´ant, J.
Electroanal. Chem. 47 (1973) 215;
[33] J.K. Kochi, J. Organomet. Chem. 300 (1986) 139.
[34] D. Tyler, in: S.J. Lippard (Ed.), Progress in Inorganic Chemistry,
vol. 36, Wiley, New York, 1988, p. 125.
[35] J. Ruiz, M. Lacoste, D. Astruc, J. Am. Chem. Soc. 12 (1990) 5471.
[36] (a) D. Astruc, Top. Curr. Chem. 160 (1991) 49;
(b) D. Astruc, New. J. Chem. 16 (1992) 305.
(b) J.B. Flanagan, S. Margel, A.J. Bard, F.C. Anson, J. Am.
Chem. Soc. 100 (1978) 4268;
(c) In the present case, the voltammogram looks as those of single-
electron waves, however, because the redox centers are far away
from one another. Under these conditions, the electrostatic factor
is very weak.
[37] D. Astruc, R. Dabard, Tetrahedron 32 (1976) 245.
[38] (a) Earlier reports according which 1,1?-diacylferrocenes gave
Fe(C5H4COR)(arene)ꢀ upon reactions with arenas [38b] were
erroneous. Only Fe(arene)2ꢀ complexes are obtained under any
conditions [37];
[51] For previous examples, see reference 52 and F. Moulines, L.
Djakovitch, R. Boese, B. Gloaguen, W. Thiel, J.-L. Fillaut, M.-H.
Delville, D. Astruc, Angew. Chem. Int., Ed. Engl. 32 (1993) 1075.
[52] C. Vale´rio, J.-L. Fillaut, J. Ruiz, J. Guittard, J.-C. Blais, D.
Astruc, J. Am. Chem. Soc. 119 (1997) 2588.
(b) A.N. Nesmeyanov, N.A. Vol’kenau, I.N. Bolesova, Dokl.
Akad. Nauk SSSR 149 (1963) 615;
(c) A.N. Nesmeyanov, Adv. Organomet. Chem. 10 (1972) 1.
[39] E. Alonso, C. Vale´rio, J. Ruiz, D. Astruc, New J. Chem. 21 (1997)
1139.
[53] E. Roma´n, D. Astruc, A. Darchen, J. Chem. Soc., Chem.
Commun. (1976) 512.
[40] The ratio between the variation of the redox potential of the
complex and the variation of the redox potential of the arene
ligand upon variation of the nature of the arene gives a good
estimation of the covalency: see M. Lacoste, H. Rabaa, D. Astruc,
A. Le Beuze, J.-Y. Saillard, G. Pre´cigoux, C. Courseille, N.
Ardoin, W. Bowyer, Organometallics 8 (1989) 2233.
[41] D. Astruc, J.-R. Hamon, E. Roma´n, P. Michaud, J. Am. Chem.
Soc. 103 (1981) 7502.
[54] J.-M. Save´ant, E. Vianello, in: P. Zuman (Ed.), Advances in
Polarography, vol. II (Chapter 17), Interscience, New York, 1962.
[55] R.S. Nicholson, I. Shain, Anal. Chem. 36 (1964) 706.
[56] A.J. Bard, L.R. Faulkner, Electrochemichal Methods, Wiley,
New York, 1980, p. 220.
[57] J. Koutecky, Chem. Listy 47 (1953) 9.
[58] (a) R.A. Marcus, N. Sutin, Biochim. Biophys. Acta 811 (1985)
265;
[42] T.W. Greene, in: T.W. Greene, P.G.M. Wuts (Eds.), Protective
Groups in Organic Synthesis, Wiley, New York, 1999, pp. 152
and 315.
(b) R.A. Marcus, Discuss. Faraday Soc. 29 (1960) 21;
(c) R.A. Marcus, Angew. Chem., Int. Ed. Engl. 32 (1993) 1111.
[59] (a) J.-M. Save´ant, Adv. Phys. Org. Chem. 26 (1990) 1;
(b) C.P. Andrieux, P. Hapiot, J.-M. Save´ant, Chem. Rev. 90
(1990) 723;
[43] (a) D. Astruc, Tetrahedron 39 (1983) 4027;
(b) D. Astruc, in: F.R. Hartley, S. Patai (Eds.), The Chemistry of
the MetalꢁCarbon Bond, vol. 4, Wiley, New York, 1987, p. 625.
/
(c) C.P. Andrieux, C. Blocman, J.-M. Dumat-Bouchiat, J.-M.
Save´ant, J. Am. Chem. Soc. 101 (1979) 3431.
[44] J.R. Hamon, J.-Y. Saillard, A. Le Beuze, M. McGlinchey, D.
Astruc, J. Am. Chem. Soc. 104 (1982) 7549.
[60] F. Nordmeyer, H. Taube, J. Am. Chem. Soc. 82 (1968) 1162.
[61] (a) J.K. Kochi, Angew. Chem., Int. Ed. Engl. 27 (1988) 1227;
(b) L. Eberson, Electron Transfer in Organic Chemistry, Springer,
Berlin, 1987;
[45] (a) J.-L. Fillaut, R. Boese, D. Astruc, Synlett (1992) 55;
(b) J.-L. Fillaut, D. Astruc, New J. Chem. 20 (1996) 945.
[46] D. Buchholz, B. Gloaguen, J.-L. Fillaut, M. Cotrait, D. Astruc,
Chem. Eur. J. 1 (1995) 374.
(c) H. Lund, K. Daasbjerg, T. Lund, S.U. Pedersen, Acc. Chem.
Res. 28 (1995) 313;
(d) A. Pross, Acc. Chem. Res. 18 (1985) 212;
[47] C.C. Lee, U.S. Gill, R.G. Sutherland, J. Organomet. Chem. 206
(1981) 89.
[48] S. Rigaut, M.-H. Delville, D. Astruc, J. Am. Chem. Soc. 119
(1997) 11132.
(e) B. Speiser, Angew. Chem., Int. Ed. Engl. 35 (1996) 2471.
[62] J.P. Collman, L.S. Hegedus, J.R. Norton, R.G. Finke, Principles
[49] (a) J. Ruiz, D. Astruc, C.R. Acad. Sci. Paris, t.1, se´rie Iic (1998)
21;
and Applications of Organotransition*Metal Chemistry, Uni-
/
versity Science Books, Mills Valley, CA, 1987, p. 59.
[63] R.H. Crabtree, The Organometallic Chemistry of the Transition
Metals, 3rd ed., 3rd ed., Wiley, New York, 2000, p. 16.
(b) D. Astruc, in: V. Balzani (Ed.), Electron Transfer in
Chemistry, vol. 2, VCH-Wiley, New York, 2001, P. Mattay, D.