Electron-transfer-catalyzed ligand substitution of carboxylato niobocene complex induced by electrochemical oxidation

Article abstract of DOI:10.1016/j.jorganchem.2004.06.043

Electrochemical reduction of niobocene dichloride (η⁵ -C₅H₄SiMe₃)₂NbCl₂ 1 formulated as Cp₂′NbCl₂ in the presence of 3,4-diaminobenzoic acid yields to the complex [Nb(η⁵ -C₅H₄SiMe₃)₂ (κ²-O,O-OOC(C₆H₃)(NH₂) ₂)] 3. When CN(2,6-Me₂C₆H₃) formulated as xylylisonitrile (CNXylyl) is added to a complex 3 solution, a substitution reaction takes place to lead to the complex [Nb(η⁵-C₅H₄SiMe₃) ₂(κ¹-O-OOC(C₆H₃) (NH₂)₂)(CN(2,6-Me₂C₆ H₃)) 4 after 3 h. An alternative way to yield quantitatively and nearly instantaneously 4 consists in a previous oxidation of 3 in the presence of CNXylyl. Hence, we present here a new example of electron-transfer-catalyzed (ETC) ligand substitution of carboxylato niobocene complex induced by electrochemical oxidation. The structure of the complexes, the formation mechanism are described using electrochemical and spectroscopic data. Electrochemical simulation have been done to verify experimental results and to complete them with a kinetic study.

Full text of DOI:10.1016/j.jorganchem.2004.06.043

Journal of Organometallic Chemistry 689 (2004) 3473–3480
www.elsevier.com/locate/jorganchem
Electron-transfer-catalyzed ligand substitution of carboxylato
niobocene complex induced by electrochemical oxidation
a
a
a,
a
b
Laurent Salvi , Alain Vallat , Helene Cattey *, Yves Mugnier , Antonio Antinolo ,
´ `
b
Antonio Otero , Santiago Garcıa-Yuste
b
´
a
` ` ´ ´ ´
Laboratoire de Synthese et Electrosynthese Organometalliques, LSEO UMR 5188 Faculte des Sciences Mirande, 9 Allee Alain Savary,
21000 Dijon, France
´ ´ ´ ´
Departamento de Quımica Inorganica, Organica y Bioquımica, Facultad de Quımica, Universidad de Castilla-la Mancha 13071 Ciudad Real, Spain
b
´
Received 7 May 2004; accepted 28 June 2004
Available online 11 September 2004
Abstract
Electrochemical reduction of niobocene dichloride (g5-C5H4SiMe3)2NbCl2 1 formulated as Cp0 NbCl2 in the presence of 3,4-
2
diaminobenzoic acid yields to the complex [Nb(g5-C5H4SiMe3)2(j2-O,O-OOC(C6H3)(NH2)2)] 3. When CN(2,6-Me2C6H3) formu-
lated as xylylisonitrile (CNXylyl) is added to a complex 3 solution, a substitution reaction takes place to lead to the complex
[Nb(g5-C5H4SiMe3)2(j1-O-OOC(C6H3)(NH2)2)(CN(2,6-Me2C6H3)) 4 after 3 h. An alternative way to yield quantitatively and
nearly instantaneously 4 consists in a previous oxidation of 3 in the presence of CNXylyl. Hence, we present here a new example
of electron-transfer-catalyzed (ETC) ligand substitution of carboxylato niobocene complex induced by electrochemical oxidation.
The structure of the complexes, the formation mechanism are described using electrochemical and spectroscopic data. Electrochem-
ical simulation have been done to verify experimental results and to complete them with a kinetic study.
ꢀ 2004 Elsevier B.V. All rights reserved.
Keywords: Niobocene; Electrochemistry; Carboxylato ligand; Isonitrile; Electron-transfer-catalyzed ligand substitution; Synthesis
1. Introduction
thus fast and complete [5]. This can be performed for
example by oxidative initiation in monomolecular
complexes when the incoming ligand is less electron-
donating than the displaced [3,6]. Few cases are known
with an endergonic cross-ET step [7,8] but in those cases
the catalytic efficiency is low.
Recently, we have prepared several paramagnetic di-
carboxylato Nb(IV) complexes from electroreduction
of niobocene dichloride (g5-C5H4SiMe3)2NbCl2 or
Cp02NbCl2 1 in the presence of suitable aliphatic acid
such malonic acid, oxalic acid and succinic acid [9]. Con-
tinuing our studies in connection with the synthesis of
new carboxylato niobocene complexes, we are currently
interested in studying the reactivity of different niobo-
cene moities towards aromatic acid and we report
here a new example of ETC ligand substitution from
Electrocatalysis or electron-transfer chain (ETC) ca-
talysis [1] first disclosed in organic chemistry by Kornb-
lum [2] and Russell [3] was applied to ligand exchange
reactions on a transition metal complex by Feldberg
[4] in 1972. Almost all the examples with the ETC mech-
anism have an exergonic cross-ET propagation step and,
Abbreviations: Cp0; (g5-C5H4SiMe3); ETC; Electron-transfer-cata-
lyzed; RDE; Rotating disk electrode; SCE; Saturated calomel elec-
trode; ESR; Electron spin resonance; NMR; Nuclear magnetic
resonance; THF; Tetrahydrofuran; NaBPh4; Sodium tetraphenylbo-
rane; CNXylyl; 2,6-Dimethylphenylisocyanide.
*
Corresponding author. Tel.: +33 03 80 396065; fax: +33 03 80
396091.
E-mail address: hcattey@u-bourgogne.fr (H. Cattey).
0022-328X/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.06.043

3474
L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
carboxylato niobocene complexes in the presence of
xylylisonitrile.
product corresponds to the carboxylato niobocene com-
plex 3 which has been characterized by NMR spectros-
copy (see Section 4). The formation of Nb(III) 3 (18
electrons configuration) can be rationalized in accord-
ance with the global reaction (1).
2. Results and discussion
NH2
2.1. Electrochemical behaviour of 1 in the presence of 3,4-
diaminobenzoic acid
NH2
Cp'2NbCl2 + HOOC
1
+
2 e-
Upon rotating disk electrode (RDE) voltammetry, in
THF in the presence of 0.2 M NaBPh4 as supporting
electrolyte Cp02NbCl2 1 exhibits a reduction wave A
and an oxidation wave E0 (Fig. 1(a)).
NH2
NH2
O
+
2 Cl-
+
H2
1
2
Cp'2Nb
C
O
3
The wave A corresponds to the one electron reduc-
tion of 1 to give Cp02NbCl 2 as previously mentioned
[10]. In the presence of 1 equivalent of 3,4-diaminoben-
zoic acid, no modification of this RDE voltammogram
was observed; however when the electrolysis is per-
formed on a carbon gauze electrode at ꢀ1.2 V (versus
SCE electrode) corresponding to the plateau of wave
A, a quantity of electricity close to 2 equivalents of elec-
trons per mol of 1 was consumed. No signal was detected
by ESR spectroscopy. The RDE voltammogram of the
resulting solution showed one well defined oxidation
wave F0 at ꢀ0.655 V (Fig. 1(b)). This electrogenerated
ð1Þ
When 1 equivalent of 3,4-diaminobenzoic acid was
added to a solution containing 2 (obtained from one elec-
tron reduction of Cp20 NbCl2 1 in THF-0.2 M NaBPh4) a
fast reaction occurs giving 3 and regeneration of 50% of 1
(identified by its reduction wave A and by ESR spectros-
copy). This result explained the consumption of two elec-
trons from electrolysis of 1 in the presence of 1 equivalent
of 3,4-diaminobenzoic acid. Theses electrochemical re-
sults can be rationalized with Scheme 1.
The complex 2 (Nb(III) – 16 electrons configuration)
reacts with aromatic acid to give a paramagnetic nio-
bium(IV) complex 20 and elimination of dihydrogen (re-
action (3)). This intermediate 20 is then reduced by 2
yielding 3 and regeneration of 1 (yield 50%, reaction
(4)). The results necessarily imply that reaction (4)
should be strongly faster than reaction (3).
20 µA
b
F’
E’
In cyclic voltammetry, 3 exhibits a system F0/F
(Fig. 2); the current ratio ip,a/ip,c is equal to unity (or very
close to it) for sweep rates v between 20 and 200 mV sꢀ1
and the peak current increases linearly with v1/2. The
half-wave potential is independent of the potential scan
-1.4
-1.1
-0.8
-0.5
-0.2
0.1
0.4
a
c
F
A
rate and the peak shape is characterized by jEp,a
ꢀ
Ep,cj ꢁ 60 mV in agreement with a one-electron transfer
Fig. 1. RDE voltammogram of (a, ꢂ) Cp0 NbCl2 1 in THF containing
2
controlled by diffusion [11,12].
This was verified by controlled potential electrolysis at
ꢀ0.2 V (plateau of wave F0) which gave 1.0 0.1 eꢀ. The
0.2 M NaBPh4; (b, j) after adding 1 equivalent of 3,4-diaminobenzoic
acid and a 2 eꢀ reduction at ꢀ1.2 V on carbon electrode; (c, N) after
1 eꢀ oxidation at 0 V (scan rate: 20 mV sꢀ1).
Cl-
Cp'2NbCl2
1
+
1 e-
Cp'2NbCl
2
+
(2)
NH2
NH2
NH2
NH2
Cl
1
2
Cp'2NbCl
2
+
HOOC
Cp'2Nb
+
H2
C
O
O
(3)
(4)
2’
NH2
NH2
NH2
NH2
Cl
O
Cp'2Nb
Cp'2NbCl
2
+
Cp'2NbCl2
1
+
Cp'2Nb
C
O
C
O
O
2’
3
Scheme 1.

L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
3475
The complex 3 was isolated as a green crystalline
solid and characterized spectroscopically. The pseudo-
1
C2v symmetric of 3 gave in the H NMR spectra two
10 µA
F’
signals multiplets each assigned to two protons of the
substituted cyclopentadienyl moiety together with three
signals of the phenyl ring. The 13C NMR spectra for 3
showed the expected signals for the different ligand sys-
tems present (see Section 4).
E (V/SCE)
-0.2
-1.2
-1
-0.8
-0.6
-0.4
F
NH2
Fig. 2. Cyclic voltammogram of 3 on carbon electrode in THF
containing 0.2 mol Lꢀ1 NaBPh4 (scan rate: 50 mV sꢀ1; starting
potential: ꢀ1 V).
O
Cp'2Nb
C
NH2
+
CNXylyl
O
3
RDE voltammogram of the resulting solution exhibited
one reduction wave F at ꢀ0.655 V (Fig. 1(c)). The elec-
trogenerated paramagnetic cationic Nb(IV) complex 3+
has been characterized by ESR spectroscopy (Fig. 3)
and shows the characteristic 10-line spectrum (93Nb nu-
NH2
NH2
O
C
O
Cp'2Nb
CNXylyl
4
cleus, 100% natural abundance, I = 9/2, THF: giso
=
(6)
1.970, aiso,Nb = 48.49 G). The low value of the hyperfine
splitting constant aiso,Nb denotes appreciable delocaliza-
tion of the unpaired spin onto the carboxylato ligand
while conventional Cp2Nb(IV) with two r-donor ligands
show aiso,Nb above 100 G [13,14]. This explanation may
be related to that previously observed for the analogous
dicarboxylato niobocene complexes [9]. Moreover, these
results are in agreement with the reaction (5).
2.3. Electrochemical behaviour of 3 in the presence of
isonitrile ligand
At room temperature, the addition of 1 equivalent of
CNXylyl to a solution of 3 led to 4. This electrogenerated
product has been characterized on the spectroscopic
features and corresponds to the diamagnetic Nb(III)
complex 4 (18 electrons configuration). The involved re-
action (6), followed by NMR analysis, is relatively slow
and the quantitative transformation of 3 to 4 was
achieved in 3 h (see Section 4).
NH2
O
NH2
Cp'2Nb
C
O
3
In order to speed up the reaction, an electrochemical
approach was proposed. This ligand substitution
should be improved by a previous oxidation (initiation
step) of the complex 3. To realize this type of studies, to
analyze the electrochemical results and to specify all ki-
netic and thermodynamic constants, the initial experi-
mental conditions must be chosen as no chemical
evolution of the complex 3 are observed, at least during
a time lower than the time scale used for electrochem-
ical technique.
+
NH2
NH2
O
+
1 e-
Cp'2Nb
C
O
3+
(5)
2.2. Chemical synthesis of complex 3
At 0 ꢁC, when 1 equivalent of CNXylyl was added to
a complex 3 solution, no modification of the wave F was
observed by RDE voltammetry even after several hours;
however by cyclic voltammetry, an additional system
G0/G with low intensity was detected at less cathodic
potential than the system F0/F (Fig. 4). The height of
peak G0 to that of peak F0 decreases when the sweep rate
increases.
The reaction of the [Nb(g5-C5H4SiMe3)2H3] 5 with
the 3,4-diaminobenzoic acid in THF at 60 ꢁC, gave the
corresponding carboxylate–niobocene complex [Nb(g5-
C5H4SiMe3)2(j2-O,O-OOC(C6H3)(NH2)2)] 3.
NH2
When electrolysis of 3 in the presence of CNXylyl
was performed at the potential of peak F0, the current
dropped to zero after the consumption of a low amount
of electricity (0.1 electron) and the RDE voltammogram
of this resulting solution exhibited the oxidation wave
G0. By cyclic voltammetry, oxidation peak G0 and
reduction peaks G and F were observed (Fig. 5).
NH2
HOOC
+
Cp'2NbH3
5
3
NH2
NH2
O
O
Cp'2Nb
C
+ 2H2

3476
L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
Fig. 3. ESR spectrum of 3+ in THF at 293 K.
The ligand exchange reaction between 3 and 4 in-
duced by electrochemical oxidation can be summarized
as in Scheme 2.
10 µA
F’
From electrochemical oxidation of 3, the correspond-
ing cationic Nb(IV) complex 3+ is obtained (initiation
step). 3+ reacts with CNXylyl to give the cationic deriv-
ative 4+ (17 electrons configuration). This reaction is not
in fact a classical ligand substitution since the bidentate
carboxylate ligand remains on the complex but as a
monodentate ligand (from j2 to j1). Reactivity of 17
electrons organometallic complex towards ligand substi-
tution reaction is well-documented [15] and the associa-
tive and dissociative mechanisms for such exchanges are
usually considered [16,17]. This cationic derivative 4+
oxidized 3 with regeneration of 3+ and formation of
4 (propagation step). The electron transfer between 4+
and 3 is thermodynamically favorable (DE = ꢀ0.655
ꢀ (ꢀ0.41) = ꢀ0.245 V) hence driving the catalytic cycle.
It is interesting to note that by injecting a small amount
of electricity (0.02 F molꢀ1 for 1 equivalent of CNXylyl
for instance and stopping the electrolysis before the cur-
rent reaches zero), we observe that the ligand substitu-
tion goes to completion giving 4 in quantitative yield.
As a result, the turnover number for this specific exam-
ple becomes equal to 49 (i.e., (1 ꢀ 0.02)/0.02) in compar-
ison with 9 (i.e., (1 ꢀ 0.1)/0.1) for the more rapid
exhaustive nonstop electrolysis. So the turnover number
is a function of the initial amount of injected electricity.
At room temperature, the reaction of the complex 3
with 1 equivalent of CNXylyl gave the complex 4. The
substitution reaction is nearly instantaneous (few min-
utes) and quantitative if the induction is performed by ox-
idation of 3 at the plateau of wave F. Without induction,
the substitution reaction is much more slower: to be
G’
a
b
c
d
-1.00
-0.75
-0.50
-0.25
0.00
G
E (V/SCE)
F
Fig. 4. Cyclic voltammogram of 3 (conc. = 4.02 · 10ꢀ3 mol Lꢀ1) in
presence of a slight excess of CNXylyl (conc. = 4.46 · 10ꢀ3 mol Lꢀ1) on
a carbon electrode in THF containing 0.2 mol Lꢀ1 NaBPh4, starting
potential: ꢀ1 V; (—) experimental and (s, j, n, r) simulated curves;
scan rate: (a) 200 mV sꢀ1; (b) 100 mV sꢀ1; (c) 50 mV sꢀ1; (d) 20 mV sꢀ1
.
10 µA
G’
a
b
c
d
-1.00
-0.85
-0.70
-0.55
-0.40
-0.25
-0.10
E (V/ECS)
G
F
Fig. 5. Cyclic voltammogram of 4 (conc. = 4.02 · 10ꢀ3 mol Lꢀ1) on a
carbon electrode in THF containing 0.2 mol Lꢀ1 NaBPh4, starting
potential: ꢀ1 V; (—) experimental and (s, j, n, r) simulated curves;
scan rate: (a) 200 mV sꢀ1; (b) 100 mV sꢀ1; (c) 50 mV sꢀ1; (d) 20 mV sꢀ1
.

L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
3477
NH2
NH2
O
O
Cp'2Nb
C
3
- 1 e-
NH2
NH2
O
Cp'2Nb
C
O
3+
NH2
NH2
O
C
CNXylyl
O
Cp'2Nb
CNXylyl
4
NH2
NH2
O
Cp'2Nb
C
O
3
NH2
O
C
O
NH2
Cp'2Nb
CNXylyl
+
4
Scheme 2.
NH2
NH2
NH2
E˚2
O
O
Cp'2Nb
Cp'2Nb
C
NH2
C
+
e-
O
O
3+
3
+ CNXylyl K2
K1 + CNXylyl
NH2
NH2
NH2
NH2
O
O
C
E˚1
O
C
+
e-
O
Cp'2Nb
Cp'2Nb
CNXylyl
4
CNXylyl
4+
NH2
NH2
O
NH2
NH2
NH2
NH2
O
C
NH2
NH2
O
O
C
K
O
+
Cp'2Nb
C
O
Cp'2Nb
O
Cp'2Nb
+
C
Cp'2Nb
O
CNXylyl
4
CNXylyl
4+
3+
3
[4+]
b1 [3 ].[CNXylyl]
[4].[3+]
[3].[4+]
[4]
k
k
f 2
k
k
f
k
k
f 1
with
K2 =
=
K1 =
=
K=
=
b2 [3].[CNXylyl]
b
Scheme 3.
1
complete, the transformation reaction required 3 h. The
final product 4 was isolated as a green crystalline solid
and characterized spectroscopically. The Cs symmetric 4
gave in the H NMR spectra, four signals assigned to
the four protons of the substituted cyclopentadienyl moi-
ety, one singlet signal at 2.36 ppm due to the two methyl

3478
L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
Table 1
Diffusion coefficients obtained from waves height in RDE voltammetry
Complex 3
Complex 3+
Complex 4
Complex 4+
2.1 · 10ꢀ6 cm2 sꢀ1
0.9 · 10ꢀ6 cm2 sꢀ1
1.1 · 10ꢀ6 cm2 sꢀ1
0.6 · 10ꢀ6 cm2 sꢀ1
At 0 ꢁC, kinematic viscosity of THF is equal to m = 0.0070 cm2 sꢀ1
.
Table 2
Thermodynamic electrochemical and kinetic parameters at 0 ꢁC for the reaction between 3 and CNXylyl (khi, heterogeneous charge transfer scan
rate; ai, charge transfer coefficient; Ki, kfi, chemical reaction constants)
E0i vs. SCE
khi (cm sꢀ1
)
ai
Ki
kfi
E01 ¼ ꢀ0:41
kh1 = 0.01
kh2 = 0.01
a1 = 0.5
a2 = 0.5
K1 = 20 molꢀ1 l
kf1 = 40 molꢀ1 l sꢀ1
kf = 3000 molꢀ1 l sꢀ1
E02 ¼ ꢀ0:655
K = 3.33 · 104
K2 = 6.66 · 105 molꢀ1 l
group of the isocyanide ligand and several signals as-
signed to the phenyl rings. The 13C NMR spectra for 4
showed the expected signals for the different ligand sys-
tems present (see Section 4).
the transfer coefficients ai were then obtained by simula-
tion of the RDE voltammograms. The best fitting is ob-
tained with ai = 0.5 and khi = 0.01 cm sꢀ1 and confirms
that the charge transfers are not very fast (Table 2).
Afterwards, the simulation results of CV voltammo-
grams (Figs. 4 and 5) provide by fitting both the equilib-
rium constant K1 and the rate constants kf1 and k. K was
in relation with E10 and E20, and K2 was calculated since
K = K2/K1. K1 is also equal to kf1/kb1, then kb1 is able
to be obtained.
The square scheme is not in equilibrium, the rates
of the reaction 3 ꢀ 4 are very low at 0 ꢁC. The chem-
ical reaction constants kf2 and kb2 are not able to be
deduct from voltammetric results. The simulated
curves do not change anymore as soon as kf2 is infe-
rior to 1 molꢀ1 L sꢀ1. Experimentally at 0 ꢁC, when
1 equivalent of CNXylyl was added to complex 3 solu-
tion, no modification of the wave F was observed,
even after several hours. This result means that kf2 is
widly inferior to 1 molꢀ1 L sꢀ1. The reaction SET is
fast but the value of k cannot be determined with great
accuracy because the simulation gives an approximate
value.
2.4. Simulations
The CV and RDE voltammograms of 3 and 4 have
been simulated in relation with the four-species square
scheme [18–20] and with the Solution Electron-Transfer
Red-Ox reaction (SET) between 4+ and 3, as shown in
Scheme 3. When the equilibrium is not established, the
situation can be designated as catalytic. This is the case
if 3 is only present initially and the reaction 3 ! 4 is
very slow (metastable equilibrium). In this case, at the
potential where 3 is oxidized, the 4+ formed from 3+
can be reduced to 4, either at the electrode or via
SET reaction.
The computations were performed using the commer-
cially available program called Digisim (Bioanalytical
Systems) with disabled option of the equilibrium condi-
tions. The uncompensated resistance is Ru = 500 X and
the double-layer capacitance value was estimated as Cdl/
A = 3.04 lF cmꢀ2 using a blank solution prior to vol-
tammetric measurements. The experimental curves were
corrected for residual current, also determined using
voltammetric measurement of blank solutions.
3. Conclusion
The first necessary input parameters for these simu-
lations are the diffusion coefficients for the species 3,
3+, 4, and 4+. Using standard methodologies using
RDE voltammograms at 0 ꢁC, the values were listed
in Table 1.
The knowledge of E10 and E02 is required and were
identified using the half-wave potentials extracted from
the RDE voltammograms for 4 (obtained after electrol-
ysis at F0 peak potential and adduct of a slight CNXylyl
excess and with knowing that the equilibrium 3+/4+ is
much in favor of 4+) and 3 (reaction (5)). The standard
heterogeneous electron transfer rate constants khi and
We have presented here a new electron-transfer-cata-
lyzed ligand substitution of carboxylato niobocene com-
plex induced by electrochemical oxidation. This reaction
permits, via a previous oxidation of 3 to generate 4
nearly instantaneously and quantitatively. 3 was already
obtained by reduction of 1, in presence of 1 equivalent
of 3,4-diaminobenzoic acid.
These two complexes have been obtained by two dif-
ferent ways: chemically or electrochemically. Their
structure and their formation mechanism have been de-
scribed using electrochemical and spectroscopic data.
Electrochemical simulation have been done to verify

L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
3479
experimental results and have been completed by a ki-
netic study.
4.3. Synthesis of [Nb(g5-C5H4SiMe3)2(j2-O,O-
OOC(C6H3)(NH2)2)] 3
A mixture of [Nb(g5-C5H4SiMe3)2H3] 5 (0.28 g; 0.75
mmol) and the 3,4-diaminobenzoic acid [3,4-(NH2)2–
C6H3(COOH)], (0.12 g; 0.75 mmol) was stirred with 30
mL of dry THF, at 60 ꢁC for 4 h. After that time, the
solution become dark green colour, the solvent was
evaporated under vacuum to dryness. The dark green
oily residue was extracted with 10 mL of hexane. The re-
sulting solution was filtered and evaporated to dryness.
4. Experimental
4.1. Chemical experiments
All reactions were carried out by using Schlenk tech-
niques. Oxygen and water were excluded by the use of
vacuum lines supplied with purified N2. Toluene was
distilled from sodium. Pentane was distilled from sodium/
potassium alloy. Diethyl ether and tetrahydrofuran
(THF) were distilled from sodium benzophenone. All
solvents were deoxygenated prior to use. The complex
[Nb(g5-C5H4SiMe3)2(H)3] 5 was prepared as described
in the literature [21]. Deuterated solvents were dried
1
A dark green solid was isolated, yielding 90% of 3: H
NMR (toluene-d8): d 0.22 (s, 18H, SiMe3), 4.30, 5.80
(4H each a complex signal, C5H4SiMe3), 6.15 (d, 5JH–H
=
8.05 Hz, 1H, C6H3), 6.70 (s, 1H, C6H3). 7.68 (d,
5JH–H = 8.05 Hz, 1H, C6H3), 13C{1H} NMR (C6D6): d
0.4 (SiMe3), 93.1 (C1, C5H4SiMe3), 104.1, 107.1 (C2–5
,
˚
over 4 A molecular sieves and degassed prior to use.
3,4-diaminobenzoic acid was used as purchased from
Aldrich. NMR spectra were recorded on a Varian Unity
exact assignment not possible, C5H4SiMe3), 140.0,
133.1, 132.9, 132.1, 122.1 (C6H3), 191.2 (COOꢀ). Anal.
Calcd. (Found) for C23H33NbN2O2Si2: C, 56.22
(56.11); H, 6.72 (6.59); N, 5.70 (5.60)%.
300 (300 MHz for H, 75 MHz for 13C) spectrometer.
1
Chemical shifts were measured related to partially deu-
terated solvent peaks and reported relative to TMS IR
spectra were recorded on a Perkin–Elmer 883 spectrom-
eter in Nujol mulls over CsI windows.
4.4. Synthesis of [Nb(g5-C5H4SiMe3)2(j1-O-OOC-
(C6H3)(NH2)2)(CN(2,6-Me2C6H3))] 4
A
mixture
of
[Nb(g5-C5H4SiMe3)2(j2-O,O-
4.2. Electrochemical experiments
OOC(C6H3)(NH2)2)] 3 (0.13 g; 0.75 mmol), and the
2,6-dimethylphenylisocyanide [CN(2,6-Me2C6H3)] (0.06
g; 0.75 mmol) was stirred with 30 mL of dry THF, at
room temperature for 3 h. After that time, the solution
took a green colour, the solvent was evaporated under
vacuum to dryness. The green oily residue was extracted
with 10 mL of hexane. The resulting solution was fil-
tered and evaporated to dryness. A green solid was iso-
lated, yielding 80% of 4: 1H NMR (toluene-d8): d 0.06 (s,
18H, SiMe3), 2.36 (s, 6H, CN(2,6-Me2C6H3)), 4.95, 5.16,
5.42, 5.72 (2H each a complex signal, C5H4SiMe3), 6.64
All manipulations were performed using Schlenk
techniques in an atmosphere of dry oxygen-free argon
gas and using dry solvents. The supporting electrolyte
was degassed under vacuum before use and then solubi-
lized at a concentration of 0.2 mol Lꢀ1. For cyclic vol-
tammetry experiments, the concentration of the
analyte was nearly 4 · 10ꢀ3 mol Lꢀ1, CNXylyl was in-
troduced in a slight excess. Voltammetric analyses were
carried out in a standard three-electrode cell with a
Princeton Applied Research, Model 263A. The reference
electrode was a saturated calomel electrode (SCE) sepa-
rated from the solution by a sintered glass disk. The aux-
5
(s, 3H, CN(2,6-Me2C6H3)), 6.30 (t, JH–H = 8.0 Hz, 1H,
5
C6H3), 7.13 (s, 1H, C6H3) 6.62 (d, JH–H = 8.0 Hz, 1H,
C6H3). 13C{1H} NMR (toluene-d8): d 0.2 (SiMe3), 19.1
(CN(2,6-Me2C6H3)), 92.9 (C1 C5H4), 96.7, 101.5,
104.5, 110.2 (C2–5, exact assignment not possible,
C5H4SiMe3), 126.5, 129.8, 130.3 and 130.7 (CN(2,6-
Me2C6H3)), 140.0, 133.1, 132.9, 132.1, 122.1 (C6H3),
174.6 (C OOꢀ), 212.2 (CN(2,6-Me2C6H3)). Anal. Calcd.
(Found) for C31H42NbN3O2Si2: C, 56.96 (56.85); H,
6.43 (6.39); N, 6.43 (5.35)%.
iliary electrode was
a
platinum wire. For all
voltammetric measurements, the working electrode
was a vitreous carbon electrode (/ = 3 mm). A
CTV101 Speed Control unit was used to adjust the rota-
tion speed (- = 500 rpm) of the EDI101 electrode (Ra-
diometer). In these conditions, when operating in
THF, the formal potential for the ferrocene+/ꢀ couple
is found to be +0.56 V versus SCE. The controlled po-
tential electrolysis was performed with an Amel 552 po-
tentiostat coupled with an Amel 721 electronic
integrator. High scale electrolyses were performed in a
cell with three compartments separated with fritted
glasses of medium porosity. A carbon gauze was used
as the working electrode, a platinum plate as the coun-
ter-electrode and a saturated calomel electrode as the
reference electrode.
References
[1] (a) M. Chanon, M.L. Tobe, Angew. Chem., Int. Ed. Engl. 21
(1982) 1;
(b) M. Chanon, Acc. Chem. Res. 20 (1987) 214.
[2] N. Kornblum, R.E. Michel, R.C. Kerber, J. Am. Chem. Soc. 88
(1966) 5662.

3480
L. Salvi et al. / Journal of Organometallic Chemistry 689 (2004) 3473–3480
[3] G.A. Russell, in: Essays in Free Radical Chemistry, The Chemical
Society, Special Publication No. 24, 1970, p. 27.
[12] R.S. Nicholson, Anal. Chem. 35 (1965) 1351.
[13] P.B. Hichcock, M.F. Lappert, C.R.C. Milne, J. Chem. Soc.,
Dalton Trans. (1981) 180.
[4] S.W. Feldberg, L. Jeftic, J. Phys. Chem. 76 (1972) 2439.
[5] D. Astruc, Angew. Chem., Int. Ed. Engl. 27 (1988) 643.
[6] J.W. Hershberger, J.C. Kochi, J. Chem. Soc., Chem. Commun.
212 (1982).
[14] (a) L.E. Manser, Inorg. Chem. 16 (1977) 525;
(b) R. Broussier, H. Normant, B. Gautheron, J. Organomet.
Chem. 155 (1978) 346;
[7] C. Amatore, J. Pinsot, J.M. Saveant, A. Thiebault, J. Electroanal.
Chem. Interfacial Electrochem. 107 (1980) 59.
(c) A. Al-Mowali, W.A.A. Kuder, J. Organomet. Chem. 155
(1978) 346.
[15] M.C. Baird, Chem. Rev. 88 (1988) 1217.
[8] (a) Selected: J.M. Saveant, Acc. Chem. Res. 13 (1980) 323;
(b) C. Amatore, J.N. Verpeaux, A. Madonik, M.H. Desbois, D.
Astruc, J. Chem. Soc., Chem. Commun. (1988) 200.
[9] D. Lucas, Z. Modarres Tehrani, Y. Mugnier, A. Antinolo, A.
Otero, M. Fayardo, J. Organomet. Chem. 629 (2001) 54.
[10] (a) H. Naboui, A. Fakhr, Y. Mugnier, A. Antinolo, M. Fajardo,
A. Otero, P. Royo, J. Organomet. Chem. C17 (1988) 338;
(b) D. Lucas, Y. Mugnier, A. Antinolo, A. Otero, M. Fajardo, J.
Organomet. Chem. C 3 (1992) 435.
[16] R.M. Kowaleski, F. Basolo, W.C. Trogler, R.W. Gedridge, T.D.
Newbound, R.D. Ernst, J. Am. Chem. Soc. 109 (1987) 4860.
[17] Y. Huang, C.C. Neto, K.A. Pevear, M.M. Banaszak Holl, D.A.
Sweigart, Y.K. Chung, Inorg. Chim. Acta 226 (1994) 53.
[18] E. Laviron, J. Electroanal. Chem. 169 (1984) 29.
[19] E. Laviron, L. Roullier, J. Electroanal. Chem. 186 (1985) 1.
[20] D.H. Evans, Chem. Rev. 90 (1990) 739.
[21] A. Antinolo, B. Chaudret, G. Commenges, M. Fajardo, F. Jalon,
R.H. Morris, A. Otero, C.T. Schweitzer, J. Chem. Soc., Chem.
Commun. (1988) 1210.
[11] R.W. Murray, C.N. Reitley, Electroanalytical Principles, Inter-
science, New York, 1963.