Ca2+ vs. Ba2+ electrochemical detection by two disubstituted ferrocenyl chalcone chemosensors. Study of the ligand-metal interactions in CH3CN

Article abstract of DOI:10.1016/j.jorganchem.2006.10.045

We show here that the disubstituted ferrocenyl chalcones 1 and 2 are good electrochemical sensors for calcium and barium in CH₃CN. However, these two triflate salts are detected in a different way by both ligands. To clarify this point, a thorough and informative NMR study of the ligand-salt interactions is presented. The unusual shapes of the titration curves obtained depend on both the ligand and cation used. For example, they illustrate that ligand 1 mainly interacts with the metal by its CO functions, while ligand 2 also interacts by its azacrown groups. These curves also reflect complex equilibriums in solution involving several ligand-salt adducts detected by mass spectrometry. To evaluate the strength of these interactions, the association constants of all the species formed have been determined by fitting the NMR data. It is noteworthy that changing the diethylamino groups in molecule 1 by the azacrown residue enhances the selectivity for the calcium salt, as pointed out by the value of the association constant of the 2Ca²⁺ species. The synthesis of the protonated counterparts 3 and 4 was useful to clarify the electrochemical behaviour of 1 and 2. Although the two ligand-salt interactions present several common points, the whole results obtained allow us to propose an original explanation for the difference observed between the Ca²⁺ and Ba²⁺ electrochemical sensing.

Full text of DOI:10.1016/j.jorganchem.2006.10.045

Journal of Organometallic Chemistry 692 (2007) 874–886
www.elsevier.com/locate/jorganchem
Ca2+ vs. Ba2+ electrochemical detection by two disubstituted
ferrocenyl chalcone chemosensors. Study of the
ligand–metal interactions in CH3CN
a,
a
b
*
´
´ ˆ
´
Beatrice Delavaux-Nicot , Jerome Maynadie , Dominique Lavabre ,
b
Suzanne Fery-Forgues
a
CNRS: National Center of Scientific Research, Laboratoire de Chimie de Coordination du CNRS, UPR 8241,
205, route de Narbonne, F-31077 Toulouse cedex 04, France
b
Laboratoire des Interactions Mole´culaires et Re´activite´ Chimique et Photochimique, UMR 5623 du CNRS,
Universite´ Paul Sabatier, F-31062 Toulouse cedex 9, France
Received 2 October 2006; accepted 18 October 2006
Available online 27 October 2006
Abstract
We show here that the disubstituted ferrocenyl chalcones 1 and 2 are good electrochemical sensors for calcium and barium in CH3CN.
However, these two triflate salts are detected in a different way by both ligands. To clarify this point, a thorough and informative NMR
study of the ligand–salt interactions is presented. The unusual shapes of the titration curves obtained depend on both the ligand and
cation used. For example, they illustrate that ligand 1 mainly interacts with the metal by its CO functions, while ligand 2 also interacts
by its azacrown groups. These curves also reflect complex equilibriums in solution involving several ligand–salt adducts detected by mass
spectrometry. To evaluate the strength of these interactions, the association constants of all the species formed have been determined by
fitting the NMR data. It is noteworthy that changing the diethylamino groups in molecule 1 by the azacrown residue enhances the selec-
tivity for the calcium salt, as pointed out by the value of the association constant of the 2Ca2+ species. The synthesis of the protonated
counterparts 3 and 4 was useful to clarify the electrochemical behaviour of 1 and 2. Although the two ligand–salt interactions present
several common points, the whole results obtained allow us to propose an original explanation for the difference observed between the
Ca2+ and Ba2+ electrochemical sensing.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Chalcone; Ferrocene; Chemosensors; Ligand–cation interaction; Electrochemical properties
1. Introduction
conjugated system’’ concept [3c]. In contrast with reported
systems [3d,3e,4], in these compounds the ferrocenyl, com-
plexing, and fluorescent moieties are not connected by a
saturated link but belong to the same conjugated electron
system. In particular, we have showed that the monosubsti-
tuted ferrocenyl derivatives containing the basic fragment
(–CO(CH@CH)C6H4-p–R), where R is an amino alkyl or
azacrown group, are good electrochemical [5,6] and optical
(UV–vis absorption) [7] calcium sensors, though they are
not fluorescent. However, the fact that the corresponding
disubstituted compounds [Fe(C5H4CO(CH@CH)nC6H4-p-
R)2], n = 1, R = NEt2 (1) [8] (Scheme 1), and n = 2,
R = NMe2 (A) display remarkable fluorescence properties
Ferrocenyl chalcones have regularly been the subject of
many interesting papers dealing with varied aspects of their
chemistry [1] or potential uses in different fields [2]. As far
as we are concerned, we have recently demonstrated their
ability in behaving as ion chemosensors. Actually, to get
a new generation of original electrochemical and optical
ferrocenyl ion chemosensors [3], we have designed a family
of ferrocenyl compounds according to a ‘‘three-component
*
Corresponding author. Tel.: +33 5 61 33 31 00; fax: +33 5 61 55 30 03.
E-mail address: delavaux@lcc-toulouse.fr (B. Delavaux-Nicot).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.10.045

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
875
p
q
iour in the presence of cations. The interactions that take
place between ligands 1–2 and the Ca2+ and Ba2+ cations
have been thoroughly investigated by NMR spectroscopy.
Thus, we show that each interaction gives rise to a complex
equilibrium between several species in solution, and is rep-
resentative of both the nature of the ligand and that of the
cation involved. Interestingly, we give evidence that for
both receptors 1 and 2, the Ca2+ and the Ba2+ electrochem-
ical detection processes are different in nature.
c
d
O
e
b
f
N
a
Fe
N
O
O
1
r
q
s
p
c
d
O
O
e
b
f
O
t
N
a
O
2. Results and discussion
Fe
O
O
O
O
N
2.1. Synthesis and characterization of compounds 1–4
O
2
Compounds 1 and 2 were obtained by reaction of
[Fe(C5H4COMe)2] with p-dimethylaminobenzaldehyde,
and with crown aldehyde [N-(4-formylphenyl)aza-15-
crown-5] respectively, in basic aqueous solution [3c]. Our
newly published procedure [5] used for the synthesis of
the monosubstituted ferrocenyl chalcones was successfully
applied to the synthesis of the disubstituted compounds
1–2. It allows the facile isolation of compound 2 and a sig-
nificant improvement (from 7% to 60%) on the isolated
Scheme 1. Compounds 1 and 2.
in acetonitrile is of particular interest, and we have recently
reported the capacity of A for behaving as a multirespon-
sive calcium chemosensor [9].
We are interested in the different factors that could influ-
ence the physical properties in this family, and in the pro-
cesses leading to the detection phenomena. Therefore, in
this article, we focus on the capacity of compound 1 for
behaving as a cation electrochemical chemosensor. Some
preliminary results about this compound have already been
published [8]. We compare the behaviour of 1 with that of
its disubstituted counterpart [Fe(C5H4COCH@CHC6H4-
p–aza-15-crown-5)2] (2) (Scheme 1), where the diethyla-
mino groups have been replaced by the azacrown residues.
The electrochemical response of both compounds in the
presence of different cations was investigated. In particular,
we show that, in contrast to the monosubstituted deriva-
tives, compound 1 and 2 are in fact clearly electrochemi-
cally sensitive to the presence of the Ba2+ cation but in a
‘‘non-classical way’’. We describe the synthesis and charac-
terization of the two new protonated compounds
½ð1Þ2H2þꢀ½BF4ꢁꢀ2 (3) and ½ð2Þ2H2þꢀ½BF4ꢁꢀ2 (4) (Scheme 2),
which were necessary to clarify the electrochemical behav-
1
yield of 1 when compared to our original procedure. H
and 13C 2D NMR characterization was undertaken in
CD3CN (see Section 4). It provides a good assignment of
each signal necessary to compare 1 and 2 with their proton-
ated derivatives, and to monitor their interaction with salts
by NMR spectroscopy. In comparison with 1, which has
nearly the same signals in this solvent, the CH2 crown sig-
nals situated in the d = 3.5–3.8 ppm range appear as the 1H
NMR signature of the compound (see Fig. 1). The values
of the Hp–t chemical shifts are close to those reported
for other crown compounds [6,10]. The NMR spectra
reveal the symmetry of the molecule in solution, and
confirm that the structure is close to that of its X-ray char-
acterized monosubstituted counterpart [(C5H5)Fe(C5H4-
COCH@CHC6H4-p–aza15-crown-5)] but with a second
organic arm.
We will see in Section 2.4. that comparing the ligands
with their protonated derivatives proved to be useful to
clarify the electrochemical behaviour. Consequently, the
protonated derivatives of compounds 1 and 2 were pre-
pared here. Treatment of a CH3CN solution of compound
2 with HBF4 Æ Et2O in a 1/2 stoichiometry turned the mix-
ture from orange to pink and afforded the protonated spe-
cies ½ð2Þ2H2þꢀ½BF4ꢁꢀ2 (4, Scheme 2). This new compound
was isolated as a violet powder in a 65% yield. As described
in the Experimental Section, elemental analyses and mass
spectra are in agreement with the proposed formula for
4, and the characterization of this compound was fully
-
2 BF4
H
O
O
O
N
N
N+
N+
i)
Fe
Fe
O
O
H
1
3
-
2 BF4
O
O
O
O
O
O
O
N+ H
O
N
N
O
ii)
Fe
Fe
O
O
O
O
O
O
O
+ H
1
achieved by H and 13C 2D NMR measurements. The sig-
N
O
O
O
nal situated at d = 8.52 ppm is attributed to the protons of
the NH+ groups. The protonation of the nitrogen atom
affects the whole molecule 2 as illustrated by Table 1. In
particular, when compared to 2, the most significant down-
field-shifted chemical shift variations are observed for the
2
4
Scheme 2. Synthetic route for compounds 3 and 4. Reagents and
conditions: (i) Et4NBF4 2 equiv, Ca(CF3SO3)2 2 equiv, CD3CN, 4 h,
20 ꢁC. (ii) HBF4 Æ Et2O 2 equiv, CH3CN, 20 ꢁC.

876
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
Fig. 1. 1H NMR spectra of compounds 1 (top) and 2 (bottom) (400 MHz, CD3CN, 293 K), in the d = 0.5–8 ppm range.
Hc, Hd, and Hp protons. The HMQC measurements indi-
cated that the Hq and Hr protons of the crown groups are
diastereotopic.
by NMR (see Section 4). As shown by Fig. 2, we note that
the signal attributed to the CH2 protons appears now as a
complex multiplet and that the NH+ signal is situated at
9.14 ppm. Table 1 indicates that, as for 2, the Hc, Hd,
and Hp protons are strongly perturbed by the protonation
reaction. Moreover mass spectrometry measurements
(FAB > 0) performed on this mixture revealed the peak
expected for the [3-CF3SO3]+ species. Unfortunately, it
was not possible to properly isolate the product.
The latter synthetic protonation route did not lead to
clear characterization of compound ½ð1Þ2H2þꢀ½BF4ꢁꢀ2 (3)
starting from 1. Actually, under these conditions, rapid for-
mation of by-products also prevented its successful isola-
tion. However, in the mixture formed, compound 3 was
detected by the presence of its [MꢁBF4]+ = 677 peak in
the MS (FAB) spectrum. In the light of the outcome of pre-
vious study about the reactivity of the monosubstituted
compounds of this family, the synthesis of compound 3
was then performed by subsequently reacting compound
1 with 2 equiv of NEt4BF4 and 2 equiv of Ca(CF3SO3)2
in CD3CN. After 4 h, compound 3 was the only product
obtained in solution and its characterization was achieved
2.2. Electrochemical study
2.2.1. Characterization of ligands 1–2
The electrochemical properties of compounds 1 and 2
have been investigated in CH3CN. For comparison pur-
pose, the solutions were light-protected, as it was also the
Table 1
Selected NMR chemical shift variations (Dd in ppm) induced by protonation of compounds 1 and 2 to give compounds 3 and 4, respectively;
L = 5 · 10ꢁ3 M, CD3CN at 293 K
CHa
CHb
CHc
CHd
CHe
CHf
CHp
CHq
CHr
CO
d(3)–d(1)
1H
13C
0.29
8.42
0.12
ꢁ3.51
0.44
ꢁ0.20
0.92
11.79
0.11
0.48
0.15
1.02
0.23
10.12
–
–
–
–
–
ꢁ0.06
d(4)–d(2)
a
a
1H
13C
a
0.32
7.93
0.14
ꢁ3.23
0.47
ꢁ0.22
1.25
12.34
0.10
0.41
0.14
1.03
0.34
5.90
–
ꢁ5.33
0.36
ꢁ0.04
Diastereoisotopic groups.

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
877
9.5
9.0
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
δ/ppm
Fig. 2. 1H NMR spectrum (250 MHz, CD3CN, 293 K), in the d = 0.5–9.5 ppm range of a mixture of 1 + 2 equiv of Ca(CF3SO3)2 and 2 equiv of Et4NBF4
after 4 h. These conditions lead to the formation of compound 3.
case for their protonated derivatives that are sensitive to
light. A typical cyclic voltammogram (CV) of compound
2 is shown in Fig. 3. It reveals that the electrochemical
characteristics of 2 are similar to those of compound 1.
This electrochemical trend has also been observed for
related compounds, where the alkyl groups have been
replaced by crown groups [6,11a]. In oxidation, the first
wave observed corresponds to an irreversible and complex
oxidation process of compound 2. It may be attributed to
the oxidation of the organic amine moiety [6,7,9,11] whose
first oxidation potential E1/2Org is 0.81 V. The wave
observed at Epa @ 1.12 V is due to the oxidation of the fer-
rocene moiety and corresponds to a quasi-reversible
process (E1/2 = 1.05 V). In reduction, the single wave
observed Epa = ꢁ1.76 V is attributed to a reduction pro-
cess mainly located on the CO function.
The main difference between the CV of 1 and 2 is the
fact that the organic moiety of compound 2 is slightly more
difficult to oxidize than that of 1. Actually, for 2, an anodic
shift of 70 mV of the first oxidation potential of the organic
moiety is observed when compared to that of 1. This is due
to the presence of the electron-withdrawing oxygen atom
which decreases the donor strength of the nitrogen atom
[6].
It is noteworthy that substituting a cyclopentadienyl
proton of the monosubstituted compounds [(C5H5)Fe-
(C5H4COCH@CHC6H4NEt2)] (5) [5] and [(C5H5)Fe(C5H4-
COCH@CHC6H4-p–aza15-crown-5)] (6) [6] by a second
organic aza arm to give the corresponding compounds 1
and 2 induces a remarkable change in the electrochemical
properties. Actually, in both monosubstituted compounds
5 and 6, the ferrocene moiety is clearly responsible for
the first oxidation process and the organic aza moiety for
the following processes. It is the opposite in both disubsti-
tuted compounds 1 and 2: the oxidation of the organic link
precedes the oxidation of the ferrocene moiety. However,
the explanation of this ‘‘inversion of potential’’ is not obvi-
ous since this behaviour is not systematic. For example, for
both the monosubstituted compound [(C5H5)Fe(C5H4-
COCH@CHC6H4NMe2)] (7) [5] and its disubstituted
homologue [Fe(C5H4CO(CH@CH)2C6H4-p–NMe2)2] (A)
[9] the first oxidation potential is that of the organic moi-
ety(ies). A series of theoretical calculations are underway
to understand this intriguing behaviour.
8
6
4
2
0
-2
2 µA
-4
-6
-8
-2.0
-1.0
0.0
1.0
2.0
E (V) / SCE
Fig. 3. Cyclic voltammograms of compounds 1 (dashed line) and 2 (solid
line). Experimental conditions: Pt electrode (1 mm diameter) in 0.1 M
solution of nBu4NBF4 in CH3CN, scan rate 100 mV sꢁ1, ligand concen-
tration 10ꢁ3 M; reference electrode SCE.

878
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
2.2.2. Electrochemical cation detection study
To investigate the cations detection properties of com-
pounds 1 and 2, electrochemical tests were performed in
Org
Fe
0.4 µA
acetonitrile in the presence of Na+, Zn2+, Ca2+, Mg2+
,
and Ba2+ triflates. During these new experiments, equiva-
lent of salts were added until a clear reproducible and final
shift was observed in each case. Some common trends
appeared. In particular, we noted that both compounds
were not sensitive to the presence of the Na+ cation, and
hardly sensitive to the presence of Zn2+. However, they
were sensitive to the presence of the Ca2+ and Mg2+ cat-
ions. We were especially interested by these cations for
comparison purpose with other ligands of this chalcone
family. In this case, a cathodic shift of around ꢁ120 mV
of the oxidation potential of the FeII/FeIII couple is
observed upon addition of two [8], and 4 equiv of salt
respectively (see Table 2). In addition, the disappearance
of the wave corresponding to the first oxidation process
of the organic moieties and the marked broadness and evo-
lution of the reduction process were clearly observed, indi-
cating that the whole conjugated organic parts of the
molecule were also strongly affected [8]. In reduction, a
new irreversible process appeared around Epa = ꢁ340 mV
for 1 and ꢁ280 mV for 2, respectively. These latter phe-
nomena are faster and more clearly observed for the
Ca2+ salt.
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
E (V) / SCE
Fig. 4. Cyclic voltammograms of compound 1 in oxidation (bold line) and
1 + 2 equiv of Ba(CF3SO3)2 (solid line). Experimental conditions: Pt
electrode (1 mm diameter) in 0.1 M solution of nBu4NBF4 in CH3CN, scan
rate 100 mV sꢁ1, ligand concentration 10ꢁ3 M; reference electrode SCE.
Interestingly, upon Ba2+ addition (6 and 4 equiv for 1
and 2, respectively), for both compounds the electrochem-
ical detection is different: an anodic shift of the first oxida-
tion potential of the organic moiety, 130 and 30 mV,
respectively, is observed whereas the oxidation potential
of the ferrocene moiety remains nearly the same. Fig. 4 is
provided as a chosen illustration for the (Ba2+, 1) couple
in oxidation. The reduction process is strongly perturbed
here with a broad ill-defined wave situated around
Epa = ꢁ1200 mV. The potential of the latter wave
increases with salt concentration. Fig. 5 also illustrates
the different behaviour of 2 depending on the nature of
the salt used. In the case of compound 1, it is noteworthy
that an anodic shift of the Epa of the organic moiety
(50 mV) and of the ferrocene moiety (20 mV) are observed
for the zinc cation (6 equiv), however the DE1/2 values
remain nearly the same.
Table 2
Selected electrochemical characteristics of compounds 1 and 2 upon
addition of the indicated number of salt equivalents
Fig. 5. Top: from top to bottom: cyclic voltammograms of (a) 2, (b)
2 + 1 equiv Ca(CF3SO3)2, (c) 2 + 2 equiv Ca(CF3SO3)2, and (d) after
10 min. Bottom: cyclic voltammograms of (a) 2, (b) 2 + 1 equiv
Ba(CF3SO3)2, and (c) 2 + 4 equiv Ba(CF3SO3)2. Experimental conditions:
Pt electrode (1 mm diameter) in 0.1 M solution of nBu4NBF4 in CH3CN,
scan rate 100 mV sꢁ1, ligand concentration 10ꢁ3 M; reference electrode
SCE.
Salt
Eq.
DE1/2 Fe
DE1/2 Org
Compound 1
Ba2+
6
2
4
0
120
–
–
Ca2+
ꢁ130
Mg2+
ꢁ110
Compound 2
Ba2+
4
2
4
0
30
–
–
Ca2+
ꢁ120
Generally, cation detection by ferrocene derivatives is
mainly based on the variation of the oxidation potential
of the ferrocene unit [12], and the variations associated
with other oxidation or reduction processes are rarely con-
Mg2+
ꢁ120
DE1/2 (mV), Org = first oxidation process of the organic part of the
molecule; ligand concentration 10ꢁ3 M.

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
879
sidered or explained. In the present cases, the observation
that the barium electrochemical detection did not induce
the same perturbations of the oxidation process as the cal-
cium salt (or magnesium salt) prompted us to get a deeper
insight into this phenomenon. In the light of our knowl-
edge about this family of compounds, the link between cat-
ion interaction and cation electrochemical detection was
investigated for ligands 1 and 2 with both the Ca2+ and
Ba2+ salts. This is the reason why the NMR study concern-
ing the different ligand–cation interactions is presented
below.
compounds, the chemical shift variations on most of the
Ha–f protons confirm that the whole unsaturated core of
the molecule including the Cp rings contributes to the elec-
tronic interaction with calcium. In the case of 2, the aza-
crown ring is also clearly involved as shown by the Hp
protons shift variation. The non-classical shapes of the
curves obtained (especially for 2) are in agreement with
the involvement of multiple equilibriums in solution. A
remarkable fact is the ability of the Ha–c protons of 1,
and Ha–d protons of 2 to undergo different chemical shift
variations (downfield or upfield shift variations) according
to the salt concentration.
2.3. NMR and mass spectrometry investigations of the
ligand–cation interactions
Processing the NMR data acquired can provide the
number and stoichiometry of the calcium adducts in the
L–Ca2+ interactions. The curves obtained by plotting the
chemical shift variation vs. calcium concentration were
processed according to a global curve-fitting method previ-
ously reported [5,9,14], and the analysis was performed
simultaneously for the five and six protons of 1 and 2
2.3.1. Focus on the Ca2+ Cation
The NMR experiments were performed at 293 K, using
first 2 equiv of salts in CD3CN (see Section 4).
As previously reported for compound 1 [8], 2 D NMR
analyses highlighted that the CHb and CO groups are char-
acteristic sites indicating a CO-salt interaction and an influ-
ential role of the double bond. Actually, these groups
display the most important 13C shift variations. This fact
is confirmed by the 1H NMR spectrum (see Table 3) show-
ing that the strongest downfield shift occurs for the Hb
protons. The NEt2 groups are the only groups not to be
affected by this interaction. These trends are reminiscent
of those observed for the monosubstituted homologue 5
under the same conditions. In the case of compound 2,
the Hd protons are strongly deshielded, clearly indicating
an interaction with the azacrown groups. However, the
Hb protons of 2 are also significantly perturbed. The latter
observation reveals that a competitive CO interaction
occurs for 2, which is in line with our previous 13C NMR
study on the monosubstituted analogue 6 [6].
1
respectively. In Fig. 6 top, we display the H NMR exper-
imental data points together with the calculated curves to
show the quality of the fit. Complete agreement between
theory and experiment could only be obtained by taking
into account the existence of four and five LnMm species
of different stoichiometries for compounds 1 and 2 respec-
tively. These LnMm species are formed by interaction of the
ferrocenyl ligand L (1, 2) with the calcium cation M. Their
corresponding association constants (Kn) are given in Table
4.
As shown by Fig. 6 (bottom), the profile of the curve
obtained by calculating the concentrations of the species
formed vs. calcium concentration is quite different for 1
and 2. This is directly related to the nature of the ligand
(N-alkyl or N-crown), and consequently to the nature
and number of the main interacting sites (CO or both
crown and CO). Regarding compound 1, the substoichio-
metric species are the main species at low concentration,
and the concentration of the 1M2 species increases rapidly
until this compound becomes the major species at high
concentration. The 12M species displays the highest associ-
ation constant. For compound 2, the concentration of the
22M and 23M species rapidly decreases. The main charac-
teristic feature is the specific formation of the 2M species
around 1 equiv of calcium, giving rise to a ‘‘bell’’ curve.
Contrary to compound 1, a fifth LM3 (2M3) species can
be formed, probably by interaction of both the CO and
two crowns groups of the same molecule with three differ-
ent calcium cations. The association constant correspond-
ing to the 2M3 species is the lowest of all. This low value
is probably due to repulsive electrostatic effect between
three neighbouring cations.
The 1H NMR behaviour of compounds 1 and 2 was fur-
ther examined in the presence of calcium salt by varying the
calcium concentration in order to get a better insight into
the L–Ca2+ interaction processes. In both cases, the contin-
uous shifts of the sharp peaks observed during the calcium
titration experiments were indicative of the presence of fast
equilibriums on the NMR time scale [13]. By chance, the
chemical shift variations of five protons of 1 and of 6 pro-
tons of 2 could be successfully monitored and were plotted
vs. calcium concentration (Fig. 6, top, points). For both
Table 3
13C NMR CO chemical shift variation and selected 1H NMR shift
variations (Dd, ppm) of the indicated groups for compounds 1–2 and
compounds 5–6 upon addition of 2 equiv, and 1 equiv of calcium salt,
respectivelya
Let us come back to the 2M species. It is striking to see
that this species exhibits the highest association constant in
this family of compounds (2.35 · 105 L molꢁ1). This can be
explained by the fact that the crown ring fits well the cal-
cium ion [15]. Besides, when compared to its monosubsti-
tuted counterpart 6, the presence of the second crown
Compound Ha
Hb
Hc
Hd
Hp
He
Hf
CO
1
5
2
6
ꢁ0.08
0.32 ꢁ0.09 ꢁ0.11 ꢁ0.03 0.16 0.21 3.26
ꢁ0.01
0.17
0.12
0.01 ꢁ0.02 ꢁ0.00 0.11 0.09 2.77
0.06
0.11 ꢁ0.02
0.13
0.15
0.46 ꢁ0.05 0.08 0.14 1.25
0.59 ꢁ0.15 0.06 0.11 1.48
a
[L] = 5 · 10ꢁ3 M; see group labelling in Scheme 2.

880
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
Compound 1
Compound 2
7
6
0.5
0.4
Hd
Hb
Hb
5
0.3
He
Hf
4
0.2
3
0.1
Hf
2
0.0
Ha
Hp
Hc
Ha
1
-0.1
-0.2
-0.3
0
Hc
-1
-2
0.0
0.5
1.0
1.5
2.0
0
1
2
3
4
5
6
7
Ca(CF3SO3)2 Concentration (x 10-2 M)
Ca(CF3SO3)2 Concentration (x 10-2 M)
7
6
5
4
3
2
1
0
4
1M2
3
2
1
0
2M2
2M
2
1
(1)2M
(2)3M
(2)2M
2M3
3
(1)3M
1M
0.5
0.0
1.0
1.5
2.0
0
1
2
4
5
6
7
Ca(CF3SO3)2 Concentration (x10-2 M)
Ca(CF3SO3)2 Concentration (x 10-2 M)
Fig. 6. Top: 1H NMR chemical shift variations of the mentioned protons of 1 (5.0 · 10ꢁ3 M) (left), and 2 (6.5 · 10ꢁ3 M) (right) vs. Ca(CF3SO3)2
concentration in CD3CN at 293 K. Experimental values (dots) and calculated curves (lines) obtained by fitting the data. See atom labelling in Scheme 1.
Bottom: concentration of the formed species with 1 (left) and with 2 (right) vs. the Ca(CF3SO3)2 concentration.
Table 4
Association constants (Kn in L molꢁ1) related to the different species formed with ligand 1 or 2 and calcium triflate in acetonitrile, determined by
processing the NMR data (values given with 15% error)
Compound
Salt
LM
K1
L2M
K2
LM2
K3
L3M
K4
LM3
K5
1
2
1
2
Ca2+
Ca2+
Ba2+
Ba2+
2.22 · 101
2.37 · 105
8.92 · 101
1.94 · 103
1.66 · 105
1.29 · 102
1.01 · 103
2.29 · 101
3.21 · 104
1.27 · 104
2.58 · 102
2.96 · 102
2.93 · 102
2.12 · 103
2.43 · 103
5.38 · 103
–
2.89 · 101
2.60 · 101
–
arm in 2 enhances the value of the association constant of
the 2M species by two orders of magnitude. This 2Ca2+
species is probably formed by a Ca2+ cation sandwiched
between two crown rings of the ligand [16]. In the litera-
ture, although numerous calcium-containing structures
are known, only a few azacrown calcium complex X-ray
structures have been reported for azacrown complexes
[17], and unfortunately, it was not possible to obtain any
X-ray structures for this 2M species.
This NMR analysis has shown that the same interacting
sites are involved for ligands 1 and 5, and for ligands 2 and
6, respectively, as long as their interaction with the Ca2+
salt is considered. However, if comparing the monosubsti-
tuted compounds with the disubstituted counterparts it
appears that a second organic arm in the molecule
increases the number of species involved in the equilibrium
processes and the number of possible associations for a
species formed and a given stoichiometry. Actually, as pre-
viously developed [5,6], several species could be envisaged
for a specific LnMm stoichiometry. Several factors have to
be taken into account, as for example the number of arms
of the molecules or the number of coordinating sites
involved in the formation of the adducts. The coordinating
solvent molecules and the different donor atoms of the
CF3SO3 anion may also interfere [18]. It is noteworthy
that, in these equilibriums, weak intermolecular or electro-
static interactions may also be considered rather than clas-
sical complexation reactions [19].
Thus, for the concentrations used, several species com-
pete in solution. To provide strong support for their forma-
tion in solution, samples of 1 and 2 (5 · 10ꢁ3 M) containing
0.5, 1, and 2 equiv of calcium salt were prepared as for

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
881
Table 5
Assignment of the main peaks obtained by mass spectrometry using the positive FAB and the ESI-MS techniques for compounds 1 and 2
(L = 5 · 10ꢁ3 M) in the presence of calcium
Compound
Salt
[LMX]+
[L2MX]+
[L3M]2+
[LM2X3]+
[LM3X5]+
1
2
1
2
Ca2+
Ca2+
Ba2+
Ba2+
777
1069
875
1366
1951
1463
2048
903
1341
952
1115
1407
1311
1603
–
1745
1746
–
1167
1391
NMR measurements, and their mass spectra were recorded
using the positive FAB technique with an MNBA matrix as
well as the ESI-MS technique. All the peaks expected for
the different compounds were effectively detected (see Table
5).
shift variation contrary to their behaviour with the calcium
salt. However, when comparing with the calcium interac-
tion, the general trends remain the same, and consequently
a CO-centered interaction can also be proposed for Ba2+
.
Regarding compound 2, its NMR behaviour with the
barium salt is quite close to that observed with the calcium
salt. The same trends are observed for the Ha, Hc, Hd, Hp
protons. A difference is noted for the Hb protons, whose
‘‘two-wave’’ behaviour visible in Fig. 6 top disappears,
and whose chemical shift variation magnitude was weaker
when replacing calcium by barium. The crown protons Hp
and Hr undergo a strong variation of their chemical shift.
This confirms that the interaction with barium involves
the whole conjugated electron system and the crown group.
These new data were processed with the curve-fitting
model. For compound 1, in addition to the species already
encountered in the interaction with the calcium, the new
2.3.2. Focus on the Ba2+ Cation
During the titration experiments of compound 1 with
Ba2+, the solution turned from orange to red. The chemical
shift variations of six protons were successfully plotted vs.
barium concentration (Fig. 7 top, left). From a general
point of view, the curves are flatter than those obtained
with the calcium salt and 1. This is in agreement with the
fact that the barium cation has a lower charge density than
the calcium cation [20], thus inducing a lower chemical shift
variation for a given salt concentration. A neat difference is
observed for the Ha protons that depict here a monotonous
Compound 1
Compound 2
0.5
Hd
Hb
0.4
0.2
He
Hf
0.3
0.2
Hr
0.1
Hc
Hb
Ha
0.1
0.0
0.0
Ha
-0.1
Hp
-0.1
Hd
-0.2
Hc
-0.2
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
8
Ba(CF3SO3)2 Concentration (x 10-2 M)
Ba(CF3SO3)2 Concentration (x 10-2 M)
8
7
6
5
4
3
2
1
0
8
7
6
5
4
3
2
1
0
2M2
1M2
1M3
2
1
(1)3M
1M
(1)2M
(
2)3M
2M
5
(2)2M
0
1
2
3
4
5
6
7
8
0
1
2
3
4
6
7
Ba(CF3SO3)2 Concentration (x 10-2 M)
Ba(CF3SO3)2 Concentration (x 10-2 M)
Fig. 7. Top: 1H NMR chemical shift variations of the mentioned protons of 1 (7.8 · 10ꢁ3 M) (left), and 2 (6.8 · 10ꢁ3 M) (right) vs. Ba(CF3SO3)2
concentration in CD3CN at 293 K. Experimental values (dots) and calculated curves (lines) obtained by fitting the data. See atom labelling in Scheme 1.
Bottom: concentration of the species formed with 1 (left), and with 2 (right) vs. Ca(CF3SO3)2 concentration.

882
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
species 1M3 is required to have a good fit. This species dis-
plays the lowest association constant with Ba2+, whereas
the substoichiometric 12M and 13M species have the high-
est ones, as for the calcium salt.
trolyte in the absence of calcium salt. It is noteworthy that
the value of the cathodic shift of the oxidation potential of
the FeII/FeIII couple of 1 and 2 upon Ca2+ addition
(around – 120 mV) is the double of that obtained with
the longer disubstituted compound [Fe(C5H4CO-
(CH@CH)2C6H4-p–NMe2)2] (A) (ꢁ60 mV). This is in
agreement with a protonation reaction on the nitrogen
atom. Actually, shortening the distance between the redox
center and the amino sites increases the electronic commu-
nication through the conjugated link. Consequently, upon
protonation, the DE1/2 value of the Fe oxidation process
of ligands 1 and 2 is more important than that of ligand
A, which is in agreement with the literature data [22]. For
all the cited reasons, we propose that the Ca2+ electro-
chemical detection pathway by ligands 1 and 2 involves a
protonation reaction due to the presence of the calcium
salt, the supporting electrolyte and water traces in the elec-
trochemical medium.
For compound 2, the different LnMm species formed
appear in the same order as those formed with the cal-
cium salt. However, the bell curve of the 2M complex is
flattened, and this species finally competes with the 2M2
species at high salt concentration. The 2M3 species does
not exist, probably because of both steric hindrance,
and electrostatic repulsion. It is noteworthy that the bis-
crown compound 2 is better suited for the complexation
of calcium than for that of the bulky barium cation, as
reflected by the respective association constant of the
LM species.
From a general point of view, the association constants
of the species formed with the Ba2+ cation are weaker than
with the Ca2+ cation. For both ligands 1 and 2, the highest
Ba2+ association constant is obtained for the L3M species.
This could be explained by the propensity of the Ba2+ cat-
ion to promote intermolecular interactions [21].
In contrast, the Ba2+ electrochemical detection by
ligands 1 and 2 is based on a phenomenon very different
from that observed in the case of calcium. The variations
induced by the presence of the Ba2+ cation are not com-
mon: the oxidation potential of FeII/FeIII couple remains
roughly the same, while the first oxidation potential of
the organic part increases. For classical electrochemical
ferrocenyl cation sensors, the intensity of the wave corre-
sponding to the latter oxidation potential generally
decreases upon salt addition, or this wave disappears [23].
Most of the time, this oxidation process is also situated
after that of the ferrocene unit. An important fact is that
no additional wave corresponding to the reduction of
NH+ function was detected here with the barium salt even
in excess (50 equiv).
As for calcium, mass spectrometry measurements per-
formed on ligands 1–2 with the barium salt and under
the same experimental conditions support the existence of
the different species in solution (see Table 5).
2.4. Link between cation interaction and cation
electrochemical detection by ligands 1 and 2
The NMR study has shown that, for each disubstituted
ligand, changing the Ca2+ salt by the Ba2+ salt induces
some differences but the interacting sites remain the same.
Therefore, at this point the paramount question is: why is
the Ca2+ electrochemical detection so different from that
Following the same procedure as for the calcium salt, we
decided to prepare CD3CN mixtures incorporating the
of Ba2+
?
n
In previous papers, we have established for the mono-
substituted compounds 5, 6 and for compound A, that
the Ca2+ electrochemical detection is the result of a two-
step process under our electrochemical conditions. Actu-
ally, a complex Ca2+ interaction with the ligand promotes
a subsequent protonation reaction in presence of the
nBu4NBF4 salt and water traces. This protonation reaction
is responsible for the Ca2+ electrochemical detection that
induces a potential shift of the FeII/FeIII couple and the
disappearance of the waves corresponding to the oxidation
processes of the organic arms of the molecules. The new
wave observed in reduction corresponds to the NH+ reduc-
tion process.
These electrochemical changes are close to those
observed with compounds 1 and 2 in Section 2.2.2. It has
also been verified that the CV of ligand 1 and 2 upon addi-
tion of 2 equiv of calcium is quite similar to those of 1 and
2 upon 2 equiv of acid, respectively. Moreover, we have
shown in Section 2.1 that it is possible to synthesize proton-
ated compound 3 by reacting 1 with appropriate amounts
of Ca(CF3SO3)2 salt and Et4NBF4 supporting electrolyte.
But, ligands 1 and 2 do not react with the supporting elec-
ligands (1 or 2), the barium salt, and the Bu4NBF4 sup-
porting electrolyte in different ratio. In these experiments,
we have never detected the formation of the corresponding
protonated compounds. Consequently, we propose that, in
the case of barium, the electrochemical observations corre-
spond to an averaged phenomenon directly related to the
sum of the different metal–ligand interactions. Under our
electrochemical conditions, globally, all these interactions
are not strong enough to induce a perceptible perturbation
of the electron density of the ferrocene unit. The fact that
two possible interacting sites compete in the case of ligand
2 probably contributes to weaken the Ba2+–2 interactions
under the electrochemical conditions, inducing thus a weak
averaged shift of the oxidation potential of the organic
part.
Finally, in the case of zinc and ligand 1, the explanation
of the electrochemical characteristics is not obvious. We
suggest that a complex ligand-zinc interaction occurs (as
for Ca2+ and Ba2+) and competes with a minor proton-
ation reaction. Actually, 1H NMR spectra of CD3CN solu-
tions of compound 1 upon Zn2+ addition are unclear. In
particular, the signals attributed to the phenyl and methyl

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
883
groups are very broad, suggesting an interaction with this
part of the molecule. Small signals situated around
8.8 ppm could be attributed to NH+ groups. 13C NMR
spectra are not really informative; however, the presence
of two small chemical shifts characteristic of the ethyl
group of its protonated counterpart 3 was detected.
Ba(CF3SO3)2 (97%) (Fluka), Zn(CF3SO3)2 (98%) (Aldrich),
Mg(CF3SO3)2 (98%) (Fluka). All these salts were dried
under vacuum, weighted and added to solution under an
argon atmosphere. [CHOC6H4-p–aza-15-crown-5] was pre-
pared according to a published procedure [24].
4.2. General instrumentation and procedures
3. Concluding remarks
All syntheses were performed under a nitrogen atmo-
sphere using standard Schlenk tube techniques. The solu-
tions of compounds 1 and 2 were light-protected before
each measurement. IR spectra were recorded on a Perkin
Elmer GX FT-IR spectrophotometer. Samples were run
as KBr pellets. Elemental analyses were carried out on a
Perkin–Elmer 2400 B analyser at the L. C. C. Microanalyt-
ical Laboratory in Toulouse. Mass spectra were obtained
In this work, ligands 1 and 2 appear to be good calcium
electrochemical sensors. The electrochemical detection
pathway of this salt remains the same as that already
encountered for their monosubstituted counterparts. These
ligands are also ‘‘non-classical’’ barium electrochemical
sensors. We show that the pathway of the electrochemical
detection of barium is different from that of the calcium.
There is no protonation reaction under our electrochemical
conditions with barium, and the electrochemical detection
is directly related to the ligand–cation interaction. The
ligand–cation interactions of these ligands with both salts
have been thoroughly studied by NMR and mass spec-
trometry. These studies highlight the existence of complex
equilibriums between different LnMm species, all of them
being characteristic of the ligand and cation used. In partic-
ular the NMR studies show that changing the amino alkyl
groups by azacrown rings in the initial framework of 1
enhances the stability of the LM calcium complex. More-
over, this work underlines the fact that the study of the
reactivity of electrochemical sensors, as for example the
synthesis of compounds 3 and 4, and of their different
ligand–metal interactions, can be useful to clarify the nat-
ure of the electrochemical detection phenomena which
occur in solution. To our knowledge, this approach is
rarely considered in the research field of ion sensing by ferr-
ocenyl ligands.
´
at the Service Commun de Spectrometrie de Masse de
´
l’Universite Paul Sabatier et du CNRS de Toulouse. Fast
atom bombardment (FAB > 0) spectra were performed
on a Nermag R 10-10H spectrometer. A 9 kV xenon atom
beam was used to desorb samples from the 3-nitrobenzyl
alcohol matrix. Other spectra were performed on a triple
quadrupole mass spectrometer (Perkin-Elmer Sciex API
365) using electrospray as the ionization mode. The infu-
1
sion rate was 5 lL/min. H and 13C NMR spectra were
performed on Bruker, AM 250, DPX 300 and AMX 400
1
spectrometers. H and 13C NMR spectra are referenced
to external tetramethylsilane. For 2 D NMR experiments,
the observation frequencies were in the range of
1
400.13 MHz for H and 100.62 MHz for 13C in CD3CN
at 293 K.
4.3. Electrochemical studies
Voltammetric measurements were carried out with a
home-made potentiostat [25] using the interrupt method
to minimize the uncompensated resistance (iR drop).
Experiments were performed at room temperature in an
airtight three-electrode cell connected to a vacuum/argon
line. The reference electrode consisted of a saturated cal-
omel electrode (SCE) separated from the solution by a
bridge compartment filled with the same solvent and sup-
porting electrolyte solution. The counter electrode was a
platinum wire of ca. 1 cm2 apparent surface. The working
electrode was a Pt electrode (1 mm diameter). The sup-
Finally, with the barium salt, we give evidence for a new
and original example of cation electrochemical sensing by
ferrocenyl chalcones. Actually, the significant shift of the
first oxidation potential of the organic moiety that is
observed here in the presence of barium had never been
reported beforehand for this class of compounds. As it will
soon be highlighted by optical studies, ligands 1 and 2, and
some of their derivatives constitute intriguing examples of
multi-channel ion chemosensors.
n
4. Experimental
porting electrolyte Bu4NBF4 (99%) (Fluka electrochemi-
cal grade) was melted and dried under vacuum for one
hour. All solutions measured were 1.0 · 10ꢁ3 M in orga-
nometallic complex and 0.1 M in supporting electrolyte.
The solutions were degassed by bubbling argon before
experiments. With the above reference, E1/2 = 0.45 V vs.
SCE was obtained for 1 mM ferrocene (estimated experi-
mental uncertainty of 10 mV). Cyclic voltammetry was
performed in the potential range ꢁ2 to 2 V vs. SCE scan-
ning from 0 to 2 V/SCE for oxidation studies (and from 0
to ꢁ2 V/SCE for reduction studies) at 0.1 V sꢁ1, at room
temperature. Before each measurement, the electrode was
4.1. Materials
Toluene, THF and ether were distilled over sodium/
benzophenone; pentane, dichloromethane and CH3CN
(pure SDS) were distilled over CaH2 and stored under
argon. EtOH analytical grade (purex SDS) was simply
degassed. Fe(C5H4COMe)2 (95%, Aldrich), CHOC6H4NEt2
(99%, Fluka), [C6H5-p–aza-15-crown-5] (98%, Acros), and
HBF4 (54% in Et2O; Aldrich), Ca(CF3SO3)2 (96%, Strem)
were used as received. Other salts: NaCF3SO3 (97%)

884
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
polished with Emery paper (Norton A621). To calculate
the half wave potential (E1/2), a quasi-steady state behav-
iour (at Pt working electrode: 1 mm of diameter) is
0.022 g (0.55 · 10ꢁ3 mol) in ethanol (15 mL); eluent: 2:1.5
toluene/ethyl acetate. The product is obtained as an orange
1
powder in 62% yield (m = 0.299 g). NMR H: d 3.57-3.73
obtained by the use of linear voltammetry at 5 mV sꢁ1
.
(m, 8H, Hs), 3.59 (m, 8H, Hp), 3.62 (m, 8H, Hr), 3.62
(m, 8H, Ht), 3.73 (m, 8H, Hq), 4.59 (m, 4H,
3JHeHf = 1.8 Hz, Hf), 4.91 (m, 4H, He), 6.70 (d, 4H,
For cation detection experiments, concentrated acetoni-
trile solutions of calcium triflate (0.3–10 equiv) were syr-
inged into the ferrocenyl solution under argon
atmosphere, keeping the total volume of the electrochem-
ical mixture constant. The solution was immediately
degassed and examined.
3
3JHcHd = 8.9 Hz, Hd), 7.02 (d, 2H, JHaHb = 15.4 Hz,
3
Ha), 7.57 (d, 4H, JHcHd = 8.9 Hz, Hc), 7.60 (d, 2H,
3JHaHb = 15.4 Hz, Hb). 13C {1H} NMR: d 52.87 (CHp),
68.48 (CHq), 69.76 (CHt), 70.30 (CHs), 70.96 (CHr),
71.35 (CHe), 73.89 (CHf), 83.38 (Cipso-C5H4), 111.97
(CHd), 118.30 (CHa), 122.78 (Cipso-C), 130.87 (CHc),
141.91 (CHb), 150.08 (Cipso-N), 191.71 (CO). IR
(CH3CN): 1521, 1550, 1577, 1611, 1648 (mCO), 2873–3001
(mCH –O–CH ). MS (FAB): Anal. Calcd for 2,
4.4. Proton NMR titration studies
Proton NMR titrations were typically performed as fol-
lowed. A solution (500 lL) of receptors 1–2 in a deuterated
solvent was added (using a microsyringe) into NMR tubes
containing the appropriate quantities of solid Ca(CF3SO3)2
salt under inert atmosphere. The NMR spectrum of the
receptor alone and of each mixture was monitored. The
samples of solid calcium were prepared by evaporating
the corresponding calculated volumes of a calcium guest
solution (10ꢁ2 M) in acetonitrile. Stability constants were
evaluated from titration data using the method indicated
in the main text.
2
2
C48H60N2O10Fe: C, 65.45; H, 6.87; N, 3.18. Found: C,
65.28; H, 6.93; N, 3.12%.
4.7. [Fe(C5H4COCH@CHC6H4NHEt2)2][BF4]2 (3)
See main text. NMR 1H:
d
1.16 (t, 12H,
3JHpHq = 7.2 Hz, Hq), 3.68 (m, 8H, Hp), 4.74 (m, 4H,
3JHeHf = 2,0 Hz Hf), 5.02 (m, 4H, JHeHf = 2,0 Hz He),
3
3
7.30 (d, 2H, JHaHb = 15.7 Hz, Ha), 7.62 (d, 4H,
3JHcHd = 8.7 Hz, Hd), 7.72 (d, 2H, JHaHb = 15.7 Hz,
3
3
4.5. [Fe(C5H4COCH@CHC6H4NEt2)2] (1)
Hb), 8.01 (d, 4H, JHdHc = 8.7 Hz, Hc), 9.14 (br s, 2H,
NH+). 13C {1H} NMR d 10.15 (CHq), 54.64 (CHp),
71.76 (CHe), 74.85 (CHf), 82.93 (Cipso-C5H4), 123.39
(CHd), 126.21 (CHa), 130.86 (CHc), 138.00 (Cipso-N),
138.05 (Cipso-C), 138.55 (CHb), 191.55 (CO). MS
(FAB): [MꢁCF3SO3]+ = 739.
A light-protected mixture of Fe(C5H4COMe)2 0.250 g
(0.93 · 10ꢁ3 mol), CHOC6H4NEt2 0.328 g (1.85 · 10ꢁ3
mol), and 1 equiv of NaOH was dissolved in ethanol
(20 mL) and stirred for 4 h at room temperature. The
mixture was evaporated to dryness. The residue was puri-
fied by column chromatography on dried silica (eluent:
2.5:0.4 toluene/ethyl acetate) and the red phase obtained
was extracted with THF as eluent (thrice). After evapora-
tion of the solvent, the product was washed with pentane
(30 mL · 2) and dried to afford the desired product as a
deep orange powder in 60% yield (m = 0.327 g). NMR
4.8. [Fe(C5H4COCH@CHC6H4-p–aza-15-crown-
5H)2][BF4]2 (4)
HBF4 Æ Et2O (2 equiv) was slowly syringed into a stirred
solution of 2 0.050 g (0.06 · 10ꢁ3 M) in acetonitrile (8 mL).
The light-protected mixture solution was stirred for 1 h.
After solvent evaporation, the product was washed with
ether (30 mL) and pentane (50 mL) and dried under vac-
uum. A violet powder was obtained in 65% yield
(m = 0.039 g). NMR 1H: d 3.10, 3.81 (m, 8H, Hq),
3.37, 3.72 (m, 8H, Hr), 3.66 (m, 8H, Hs), 3.77 (m, 8H,
3
1H: d 1.19 (t, 12H, JHpHq = 7.1 Hz, Hq), 3.45 (q, 8H,
3
3JHpHq = 7.1 Hz, Hp), 4.59 (m, 4H, JHeHf = 1.9 Hz,
3
Hf), 4.91 (m, 4H, JHeHf = 1.9 Hz, He), 6.70 (d, 4H,
3
3JHcHd = 9 Hz, Hd), 7.01 (d, 2H, JHbHa = 15.4 Hz, Ha),
3
7.57 (d, 4H, JHcHd = 9.0 Hz, Hc), 7.60 (d, 2H,
3JHaHb = 15.4 Hz, Hb). 13C {1H} NMR: d 12.29 (CHq),
44.52 (CHp), 71.28 (CHe), 73.83 (CHf), 83.41 (Cipso-
C5H4), 111.60 (CHd), 117.79 (CHa), 122.04 (Cipso-C),
131.06 (CHc), 142.06 (CHb), 150.05 (Cipso-N), 191.61
(CO). IR (CH3CN): 1522, 1552, 1575, 1611, 1647 (mCO),
2873–2975 (mCH). Anal. Calc. for 3, C36H40N2O2Fe: C,
73.47; H, 6.85; N, 4.76. Found: C, 73.38; H, 6.91; N,
4.83%.
Ht), 3.93 (m, 8H, Hp), 4.73 (m, 4H, JHeHf = 2.0 Hz, Hf),
3
3
3
5.01 (m, 4H, JHeHf = 2.0 Hz, He), 7.34 (d, 2H, JHaHb
=
3
15.7 Hz, Ha), 7.74 (d, 2H, JHaHb = 15.4 Hz, Hb), 7.95
(d, 4H, JHcHd = 8.8 Hz, Hd), 8.04 (d, 4H, JHcHd
3
3
=
8.8 Hz, Hc), 8.52 (br s, 2H, NH+). 13C {1H} NMR: d
58.77 (CHp), 63.15 (CHq), 68.90 (CHs), 69.00 (CHt),
70.76 (CHr), 71.76 (CHe), 74.92 (CHf), 82.40 (Cipso-
C5H4), 124.31 (CHd), 126.23 (CHa), 130.65 (CHc),
136.35 (Cipso-N), 138.39 (Cipso-C), 138.68 (CHb),
191.67 (CO). IR (KBr): 1515, 1594, 1605, 1654 (mCO),
2876–2921 (mCH –O–CH ), 3050–3152 (mNH+). ES-MS:
4.6. [Fe(C5H4COCH@CHC6H4-p–aza-15-crown-5)2] (2)
2
2
Same procedure as for 1 with a mixture of [Fe(C5H4-
COMe)2] 0.148 g (0.55 · 10ꢁ3 mol), N-(4-formylphe-
nyl)aza-15-crown-5] 0.370 g (1.10 · 10ꢁ3 mol), and NaOH
[MꢁBF4]+ = 969.7. Anal. Calc. for 2, C48H62N2O10-
FeB2F8: C, 54.57; H, 5.91; N, 2.65. Found: C, 54.38; H,
5.93; N, 2.52%.

B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
885
(p) J. Boichard, J.P. Monin, J. Tirouflet, Bull. Soc. Chim. Fr. 4
(1963) 851–856;
4.9. Interaction of compound 1 with 2 equiv of Ca(CF3SO3)2
(q) T.A. Mashburn Jr., C.E. Cain, C.R. Hauser, J. Org. Chem. 25
(1960) 1982–1986.
[2] (a) X. Wu, E.R.T. Tieking, I. Kostetski, N. Kocherginsky, A.L.C.
Tan, S.B. Khoo, P. Vilairat, M.-L. Go, Eur. J. Pharm. Sci. 27 (2006)
175–187;
1
3
NMR H: d 1.15 (t, 12H, Jp–q = 7,1 Hz, Hq), 3.42 (q,
3
8H, Jq–p = 7.1 Hz, Hp), 4.80 (br s, 4H, Hf), 5,07 (br s,
3
4H, He), 6.59 (d, 4H, Jc–d = 8.4 Hz, Hd), 6.93 (d, 2H,
3Jb–a = 15.2 Hz, Ha), 7.48 (d, 4H, Jd–c = 8,4 Hz, Hc),
3
7,92 (d, 2H, Ja–b = 15.2 Hz, Hb). NMR 13C {1H}: d
3
(b) Q.-B. Song, X.-N. Li, T.-H. Shen, S.-D. Yang, G.-R. Qiang, X.-
L. Wu, Y.-X. Ma, Synth. Commun. 33 (2003) 3935–3941;
(c) X. Wu, P. Wilairat, M.-L. Go, Bioorg. Med. Chem. Lett. 12
(2002) 2299–2302;
(d) A. Ferle-Vidovic, M. Poljak-Blazi, V. Rapic, D. Skare, Cancer
Biother. Radiopharm. 15 (2000) 617–624.
11.67 (CHq), 44.67 (CHp), 72.93 (CHe), 75.11 (CHf),
82.76 (Cipso-C5H4), 111.61 (CHd), 115.47 (CHa),
121.37 (Cipso-C), 132.48 (CHc), 148.35 (CHb), 151.09
1
(Cipso-N), 194.87 (CO), 121.04 (q, JC–F = 320 Hz,
[3] (a) F. Sancenon, A. Benito, F.J. Hernandez, J.M. Lloris, R.
Martinez-Manez, T. Pardo, R. Soto, Eur. J. Inorg. Chem. (2002)
866–875;
CF3SO3).
4.10. Interaction of compound 2 with 2 equiv of
Ca(CF3SO3)2
(b) S. Fery-Forgues, B. Delavaux-Nicot, J. Photochem. Photobiol. A
Chem. 132 (2000) 137–159;
(c) B. Delavaux-Nicot, S. Fery-Forgues, Eur. J. Inorg. Chem. (1999)
1821–1825;
1
3
NMR H: d 3.54 (t, 8H, JHqHp = 5.6 Hz, Hp), 3.52–
`
(d) P.D. Beer, F. Szemes, V. Balzani, C.M. Sala, M.G.B. Drew, S.W.
3
3.91 (m, 24H, Hr, Hs, Ht), 3.88 (t, 8H, JHqHp = 5.6 Hz,
Dent, M. Maestri, J. Am. Chem. Soc. 119 (1997) 11864–11875;
(e) P.D. Beer, A.R. Graydon, L.R. Sutton, Polyhedron 15 (1996)
2457–2461.
3
Hq), 4.73 (m, 4H, JHeHf = 2.0 Hz, Hf), 4.99 (m, 4H,
3JHeHf = 2.0 Hz, He), 7.08 (d, 2H, JHaHb = 15.6 Hz, Ha),
3
3
[4] (a) F. Sancenon, A. Benito, F.J. Hernandez, J.M. Lloris, R.
Martinez-Manez, T. Pardo, J. Soto, Eur. J. Inorg. Chem. (2002)
866–875;
7.16 (br d, 4H, JHcHd = 8.4 Hz, Hd), 7.70 (d, 4H,
3JHcHd = 8.4 Hz, Hc), 7.72 (d, 2H, JHaHb = 15.6 Hz,
3
Hb). 13C {1H} NMR: d 53.03 (CHp), 68.93 (CHq), 68.98,
69.90, 70.11 (CHr, CHs, CHt), 72.11 (CHe), 74.77 (CHf),
82.76 (Cipso-C5H4), 118.81 (CHd), 121.03 (CHa), 128.11
(Cipso-C), 130.65 (CHc), 142.81 (CHb), 151.57 (Cipso-
N), 192.96 (CO).
(b) F. Oton, A. Tarraga, M.D. Velasco, A. Espinosa, P. Molina,
Chem. Commun. (2004) 1658–1659;
(c) L.-J. Kuo, J.-H. Liao, C.-T. Chen, C.-H. Huang, C.-H. Chen, J.-
M. Fang, Org. Lett. 5 (2003) 1821–1824.
´
[5] J. Maynadie, B. Delavaux-Nicot, D. Lavabre, B. Donnadieu, J.C.
Daran, A. Sournia- Saquet, Inorg. Chem. 43 (2004) 2064–2077.
[6] B. Delavaux-Nicot, J. Maynadie´, D. Lavabre, C. Lepetit, B.
Donnadieu, Eur. J. Inorg. Chem. (2005) 2493–2505.
Acknowledgement
´
[7] J. Maynadie, B. Delavaux-Nicot, D. Lavabre, S. Fery-Forgues, J.
Organomet. Chem. 691 (2006) 1101–1109.
This work was supported by the CNRS and the French
Ministry of Research (doctoral fellowship to J. M.).
´
[8] J. Maynadie, B. Delavaux-Nicot, S. Fery-Forgues, D. Lavabre, R.
Mathieu, Inorg. Chem. 41 (2002) 5002–5004.
´
[9] B. Delavaux-Nicot, J. Maynadie, D. Lavabre, S. Fery-Forgues,
References
Inorg. Chem. 45 (2006) 5691–5702.
[10] K. Rurack, J.L. Bricks, G. Reck, R. Radeglia, U. Resh-Genger, J.
Phys. Chem. A. 104 (2000) 3087–3109.
[11] (a) P.D. Beer, C. Blackburn, J.F. McAleer, H. Sikaniaka, Inorg.
Chem. 29 (1990) 378–381;
[1] (a) For examples: W.-J. Zhou, S.-J. Ji, Z.-L. Shen, J. Organomet.
Chem. 691 (2006) 1356–1360;
(b) Z.-L. Shen, S.-J. Ji, W.-J. Zhou, Synth. Commun. 35 (2005) 1903–
1909;
(c) M. Sun, Q. Shi, G. Huang, Y. Liang, Y. Ma, Synthesis 15 (2005)
2482–2486;
(d) Z.-L. Shen, D.-G. Gu, S.Y. Wang, J. Organomet. Chem. 689
(2004) 1843–1848;
(e) M. Prokesova, E. Solcaniova, S. Toma, K.W. Muir, A.A. Torabi,
G. Knox, J. Org. Chem. 61 (1996) 3392–3397;
(f) D. Villemin, B. Martin, M. Puciova, S. Toma, J. Organomet.
Chem. 484 (1994) 27–31;
(g) A.G. Nagy, P. Sohar, J. Marton, J. Organomet. Chem. 410 (1991)
357–364;
(h) A.G. Nagy, J. Organomet. Chem. 291 (1985) 335–340;
(i) A.G. Nagy, S. Toma, J. Organomet. Chem. 266 (1984) 257–268;
(j) A.M. El-Khawaga, K.M. Hassan, A.A. Khalaf, Z. Naturforsch.
36b (1981) 119–122;
(b) M. Spescha, N.W. Duffy, B.H. Robinson, J. Simpson, Organo-
metallics 13 (1994) 4895–4904;
(c) N.W. Duffy, J. Harper, P. Ramani, R. Ranatunge-Bandarage,
B.H. Robinson, J. Simpson, J. Organomet. Chem. 564 (1998) 125–131.
[12] (a) P.D. Beer, J. Cadman, Coord. Chem. Rev. 205 (2000) 131–155;
(b) P.D. Beer, M.G.B. Drew, J. Pilgrim, Inorg. Chim. Acta 225
(1994) 137–144.
[13] (a) M.J. Hynes, J. Chem. Soc. Dalton Trans. (1993) 311–312;
(b) L. Fielding, Tetrahedron 56 (2000) 6151–6170.
[14] (a) N. Marcotte, S. Fery-Forgues, D. Lavabre, S. Marguet, V.G.
Pivovarenko, J. Phys. Chem. A 103 (1999) 3163–3170;
(b) S. Fery-Forgues, D. Lavabre, D. Rochal, New J. Chem. 22 (1998)
1531–1538.
[15] (a) C.J. Pedersen, H.K. Frendsdorff, Angew. Chem. Int., Ed. Engl. 11
(1972) 16–25;
(k) A.N. Nesmeyanov, G.B. Shulp’in, L.V. Rybin, N.T. Gubenko,
M.I. Rybinskaya, P.V. Petrovkii, V.I. Robas, J. Gen. Chem. USSR
44 (1974) 1994–2001;
(l) S. Stankovianski, A. Beno, S. Toma, E. Gono, Chem. Zvesti 24
(1970) 19–27;
(b) B.J. Dietrich, J. Chem. Educ. 62 (1985) 954–964.
[16] (a) P.D. Beer, A.D. Keefe, H. Sikanyika, C. Blackburn, J.M.
McAleer, J. Chem. Soc. Dalton. Trans. (1990) 3289–3294;
(b) P.D. Beer, H. Sikanyika, A.M.Z. Slavin, D.J. Williams, Polyhe-
dron 8 (1989) 879–886;
(m) S. Toma, A. Perjessy, Chem. Zvesti 23 (1969) 343–351;
(n) J.P.C.G. Dubosc, U.S. Patent 3, 1967, 335,008;
(o) M. Furdik, S. Toma, Chem. Zvesti 20 (1966) 326–335;
(c) T. Akutagawa, N. Takamatsu, K. Shitagami, T. Hasegawa, T.
Nakamura, T. Inabe, W. Fujita, K. Awaga, J. Mater. Chem. 11
(2001) 2118–2124;

886
B. Delavaux-Nicot et al. / Journal of Organometallic Chemistry 692 (2007) 874–886
(d) J. Otsuki, T. Yamagata, K. Ohmuro, K. Arabi, T. Takido, M.
Seno, Bull. Chem. Soc. Jpn. 74 (2001) 333.
(b) E.S. Meadous, S.L. De Wall, L.J. Barbour, G.W. Gokel, J. Am.
Chem. Soc. 123 (2001) 3092–3107;
[17] (a) S. Itoh, H. Kumey, S. Nagatomo, T. Kitagawa, S. Fukuzumi, J.
Am. Chem. Soc. 123 (2001) 2165–2175;
(c) J.C. Ma, D.A. Dougherty, Chem. Rev. (1997) 1303–1324;
(d) T. Futterer, A. Merz, J. Lex, Angew. Chem., Int. Ed. Engl. 36
(1997) 611–613.
(b) D. Hall, A. Leineweber, J.H.R. Tucker, D. Williams, J. Organo-
met. Chem. 523 (1996) 13–22;
(c) J. Hu, L.J. Barbour, R. Ferdani, G.W. Gokel, J. Chem. Commun.
(2002) 1806–1807.
[20] R.C. Weast (Ed.), Handbook of Chemistry and Physics, 65th ed.,
CRC Press, Boca Raton, FL, 1985.
[21] F. Vo¨gtle, Supramolecular Chemistry, Wiley, New York, 1991.
[22] H. Plenio, J. Yang, R. Diodone, J. Heinze, Inorg. Chem. 33 (1994)
4098–4104.
[23] M.P. Andrews, C. Blackburn, J.F. McAleer, V.D. Patel, J. Chem.
Soc., Chem. Commun. (1987) 1122–1124.
[24] J.P. Dix, F. Vo¨gtle, Chem. Ber. 113 (1980) 457–470.
[25] P. Cassoux, R. Dartiguepeyron, D. de Montauzon, J.B. Tommasino,
P.L. Fabre, Actual. Chim. 1 (1994) 49–55.
[18] (a) G.A. Lawrance, Chem. Rev. 86 (1986) 17–33;
(b) A.D. Frankland, P.B. Hitchcok, M.F. Lappert, G.A. Lawless, J.
Chem. Soc., Chem. Commun. (1994) 2435–2436;
(c) A. Onoda, Y. Yamada, M. Doi, T. Okamura, N. Ueyama, Inorg.
Chem. 40 (2001) 516–521.
[19] (a) G.W. Gokel, S.L. De Wall, E.S. Meadous, Eur. J. Org. Chem.
(2000) 2967–2978;