Bis(crown ethers) derived from biphenyl: Extraction and electrochemical properties

Article abstract of DOI:10.1016/j.tet.2004.03.065

Ligands derived from 4,4′-dinitrobiphenyl containing azacrown cavities in the 2,2′ position have been used in extraction and transport experiments. Control experiments with a system containing only one complexing cavity have demonstrated that the capability of forming a sandwich-type complex in the aforementioned ligand not only increases extraction but also the transport across a liquid membrane. Extraction studies have also shown that the complex present in the membrane has a 1:1 stoichiometry with regard to both ligands. Electrochemical studies have also been carried out. The ligand containing two complexing cavities is capable of giving rise to a 2:1 complex under electrochemical conditions.

Full text of DOI:10.1016/j.tet.2004.03.065

Tetrahedron 60 (2004) 4683–4691
Bis(crown ethers) derived from biphenyl: extraction and
electrochemical properties
a
a
a
a
Ana M. Costero,a Joaquin Sanchis, Sergio Peransi, Salvador Gil, Vicente Sanz
,
*
and Antonio Domenechb
a
´
´
Departamento de Quımica Organica, Universidad de Valencia, Edif. Investigacion, Doctor Moliner, 50, 46100 Burjassot, Valencia, Spain
´ ´
Departamento de Quımica Analıtica, Universidad de Valencia, Doctor Moliner, 50, 46100 Burjassot, Valencia, Spain
b
Received 17 November 2003; revised 1 March 2004; accepted 24 March 2004
Abstract—Ligands derived from 4,40-dinitrobiphenyl containing azacrown cavities in the 2,20 position have been used in extraction and
transport experiments. Control experiments with a system containing only one complexing cavity have demonstrated that the capability of
forming a sandwich-type complex in the aforementioned ligand not only increases extraction but also the transport across a liquid membrane.
Extraction studies have also shown that the complex present in the membrane has a 1:1 stoichiometry with regard to both ligands.
Electrochemical studies have also been carried out. The ligand containing two complexing cavities is capable of giving rise to a 2:1 complex
under electrochemical conditions.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
Topology of macrocyclic polyethers strongly contributes to
their cation complexation. The use of bicyclic structures is
an attractive strategy to enhance cation-complexing abilities
and selectivities of monocyclic crown ethers. Macrobicyclic
polyethers carrying two crown ether moieties at the end of a
spacer possess different complexing properties.1 Usually,
these bis(crown ethers) are more likely to form intra-
molecular sandwich-type complexes with certain cations
that are a little larger than the crown ring cavities by means
of a cooperative action of two adjacent crown rings. As a
Chart 1.
result of such a complex formation, the bis(crown ether)
exhibits excellent selectivities towards certain cations as
compared to the corresponding monocycles analogues.2,3
On the other hand, bis(crown ether) compounds have also
been used as carriers in transport experiments, giving rise to
different selectivities and efficiencies. This behaviour can be
related to the strong effect of the stability of the cation-
carrier complex on the transport efficiency.4
2. Results and discussion
2.1. Synthesis
Ligands 1 and 2 have been synthesised as described in
Scheme 1. The reaction of 4,40-dinitrodiphenic acid5 with
thionyl chloride gives rise to the corresponding acid
chloride that reacts with the commercially available
macrocycle 1-aza-18-crown-6 to lead to compound 1. The
asymmetric compound 2 was prepared in a similar way from
the monomethylester of the 4,40-dinitro-diphenic acid (4).
This ester was prepared from the same acid through the
corresponding anhydride (3).
Here we report the use of ligands 1 and 2 (Chart 1) in
extraction and transport experiments, and the influence of
the complex structure in transport efficiency. Additionally,
some electrochemical studies on these ligands have been
carried out.
2.2. Extraction and transport experiments
Keywords: Transport; Extraction; Complexation; Azacrown ligands;
Electrochemistry.
Extraction experiments, between chloroform and aqueous
phases, with ligands 1 and 2 were carried out following the
method described by Cram6 and the log Ka obtained,
*
Corresponding author. Tel.: þ34-963543135; fax: þ34-963543152;
e-mail address: salvador.gil@uv.es
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.03.065

4684
A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
Scheme 1.
Table 1. Extraction experiments with ligands 1 and 2
Ligand
log Ka
Selectivity
Naþ/Kþ
% Extraction
Liþ
Naþ
Kþ
Csþ
Naþ/Liþ
Naþ/Csþ
Liþ
Naþ
Kþ
Csþ
1
2
5.23
3.42
5.35
4.64
5.09
4.09
5.21
3.16
1.32
16.59
1.82
3.55
1.38
30.20
22.0
0.3
32.1
8.0
27.2
3.5
19.3
0.2
assuming a 1/1 stoichiometry of the crown ether and the
cation as summarised in Table 1, along with the selectivity
ratios Naþ/Liþ, Naþ/Kþ and Naþ/Csþ, as well as the
extraction percentage for each cation.
observed for the bis(crown ether) when compared with the
selectivity observed for the corresponding monocyclic
counterpart (Fig. 2). In addition, the results observed for
Mg2þ are similar to those shown by Liþ, which led us to
propose that the ionic radius is one of the most important
factors in extraction processes. The low extraction observed
for Ca2þ may be due to the large Kd shown by the
corresponding picrate. Thus, the diffusion gave rise to
inaccurate results that have not been included.
The bis(crown ether) 1 showed a cation extracting ability,
which is greater than those observed in the monocycle
compound 2, probably due to the formation of intra-
molecular sandwich-type complexes. In specific terms, the
complexation of Liþ and Csþ by 1 were enhanced by the
log Ka values of 1.81 and 2.05, respectively, as compared to
the corresponding monocycle analogue 2 (Fig. 1). The ionic
radius of these two cations are too small and too large,
respectively, to fit into a single crown cavity, but the
simultaneous complexation with both crown subunits
allows them to show a large enhancement of the extraction
properties.
By contrast, a poor selectivity for alkali cations was
Figure 1. Association constants of ligands 1 and 2. (CHCl3 at 298 K).
Figure 2. Extraction selectivity for ligands 1 and 2.

A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
4685
As Figure 3 shows, the slopes are mostly 1 in the log [M] vs
log [L] representations (where [M] is the concentration of
the cation in the organic phase at the end of the experiment,
and where [L] is the concentration of the ligand in the
membrane) for the alkali metals studied. These results agree
with a 1:1 stoichiometry of the complexes in the organic
phase7 not only with ligand 2 as expected, but also with
ligand 1, as proposed above.
Figure 4. Transport experiments with ligands 1 and 2.
in transport efficiency cannot only be explained by the mere
presence of two cavities in its structure, i.e. ligand 1
transports more than twice ligand 2 does. One explanation
for this behaviour could be due a sandwich-type complex
formation that allows the cation to be completely
encapsulated by the lipophilic part of both coronand
subunits (Fig. 5). As expected, the effect of the sandwich-
type complex formation in transport is more efficient the
larger the cation is (e.g. Csþ or Ba2þ).
Figure 3. Extraction of LiPic (4.81 mM), NaPic (4.76 mM) and CsPic
(2.40 mM) with ligand 1 (top) and 2 (bottom) (concentration of ligands
from 23.73 to 2.73 mM).
On the other hand, transport experiments were carried out in
a U-tube at 293^1 K.8 Chloroform was used as the liquid
membrane and deionised water was used with both the
source and the receiving phase. The concentration of cations
was deduced from that of the corresponding picrate anions
determined by UV spectroscopy. The results of these
experiments are shown in Table 2.
Figure 5. Lipophilic surface in complexes of ligand 1.
Table 2. Transport efficiency [mM (Mþ)/ml (receiving phase) mM (ligand)]
through a CHCl3 bulky membrane, 293^1 K after 48 h
Additionally, experiments have been carried out to know the
influence of cation source-phase concentration in transport.
Thus, two starting conditions (in the first case cation
concentration at the source phase and at the membrane were
four times the concentration of the second experiment) were
used in NaPic extraction. The results observed demonstrated
that there was not a direct correlation between cation
concentration and transport, even though the concentration
is higher the transport is more highly efficient. However, the
effect was different in both ligands. Ligand 2 transported
around 5 times the amount in the first experiment than in the
second one. By contrast, ligand 1 shows more important
increments, this ligand transported around 10 times more
cation in the concentrated conditions than in the more
diluted source phase (see Table 3).
Ligand
Transport efficiency
Naþ
Kþ
Csþ
Ba2þ
1
2
17.80
4.03
10.95
2.66
9.56
0.99
9.46
0.43
These results demonstrate that the stronger the cation
complexation is, the more efficiently it is transported.
Although Naþ is the more efficiently transported cation with
both ligands, the behaviour of these two ligands is
completely different. Thus, ligand 1 is a more efficient
carrier than ligand 2 is (around 4.4 times for Naþ, 4.1 for
Kþ, 9.6 for Csþ and 22 for Ba2þ) (Fig. 4). These increments

4686
A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
Table 3. Transport experiment results with ligands 1 and 2
Picrate salt
1 at 400 nm
Ligand 1
c. r.p. (M)
Ligand 2
c. r.p. (M)
c. of H (M)
c. s.p. (M)
f.e.
c of H (M)
c. s.p. (M)
f.e.
Liþþ
Na
9564
9532
16,820
9532
10,626
10,895
19,074
0.0012
0.0012
0.0012
0.0003
0.0003
0.0003
0.0003
2.39£1024
1.58£1024
1.20£1024
8.12£1025
4.73£1025
3.84£1025
3.90£1025
2.00£1022
1.99£1022
2.05£1022
4.94£1023
5.01£1023
5.04£1023
4.14£1023
14.43
88.68
7.26
17.79
10.95
9.55
0.0012
0.0012
0.0012
0.0003
0.0003
0.0003
0.0003
2.63£1025
7.31£1024
2.09£1025
1.88£1025
1.25£1025
4.68£1026
1.99£1026
2.68£1022
1.70£1022
1.87£1022
4.57£1023
4.08£1023
4.21£1023
3.61£1023
1.44
40.16
1.13
4.03
2.66
0.99
0.43
Mg2þ
Naþ
Kþ
Csþ
Ba2þ
9.45
c. of H¼concentration of host in the membrane; c.r.p.¼final concentration of salt at the receiving phase; c.s.p.¼final concentration of salt at the source phase;
f.e.¼factor of efficiency.
2.3. Electrochemical studies
The interaction of 1 with Cd2þ ions is illustrated in Figure 7.
For ‘free’ Cd2þ (see Fig. 7(a)), a prominent reduction peak
appears at 20.73 V (C4) followed by a tall oxidation peak at
Ligands 1 and 2 present a 4,40-dinitrobiphenyl subunit
which could be considered as a possible electrochemistry
switch. Furthermore, electrochemistry has been shown to be
an appropriate technique to study the complexation of
related ligands with transition and post-transition metals.9,10
It was for these reasons that we have decided to carry out
some electrochemical studies in the presence of Cd2þ and
Zn2þ. These cations were selected in relation to the fact of
the most striking contrast that the biological activity of
Cd2þ has in comparison to the activity of Zn2þ. The latter is
crucial in a number of important biological processes,
whereas Cd2þ is one of the most toxic metals.
The cyclic voltammetric (CV) response of the receptors
dissolved in MeCN is illustrated in Figure 6(a), as far as 1 is
concerned. In the initial cathodic scan, two well-defined
reduction peaks appear at 21.02 (C1) and 21.94 V (C2),
with the first one being preceded by a weak shoulder near to
20.74 V (C3). In the subsequent anodic scan, an ill-defined
shoulder appears at 21.55 V (A2), followed by two
overlapped peaks at 20.88 (A1) and 20.61 V (A3). This
last peak disappears in CVs if the potential is reversed at
potentials close to 21.25 V, suggesting that the peak A3
corresponds to the oxidation of any species generated during
the electrode process C2 (Fig. 6(b)). The voltammetric
response of the ligand 2 in MeCN (0.10 M Bu4NPF6) is
almost identical to that previously described for 1.
The observed response can be described on the basis of the
well-known electrochemistry of aromatic and nitroaromatic
compounds in aprotic solvents, consisting of two successive
one-electron transfer processes, yielding an anion radical
and a dianion.11–13 The voltammetric profile, however,
depends strongly on the stability of the intermediate radical
anion. If this is unstable with respect to a disproportionisa-
tion into the dianion and the parent aromatic compound, the
voltammogram looks like a single two-electron wave. Such
is the case of the studied receptors; here the obtained value
of the half-peak width of the process C1, measured in square
wave voltammetrys (SQWVs), 115 mV, is clearly lower
than that expected for a reversible one-electron process
(126 mV).14 The overall electrochemical process can be
represented as:
L þ 2e2 ! L22
where L¼1, 2.
ð1Þ
Figure 6. CVs at the GCE of a 2.0 mM solution of 1 in MeCN (0.10 M
Bu4NPF6). Potential scan rate 200 mV/s. (a) Potential range þ1.25/
22.05 V; (b) potential range þ1.65/21.35 V.

A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
4687
profile becomes similar to that recorded in the Cd2þ plus
receptor solutions, with a prominent peak near to 20.45 V
(C5) and two overlapped peaks at 20.95 and 21.03 V (C1).
As expected, the reduction of complexed Zn2þ to Zn metal
occurs at more negative potentials than those of the free
Zn2þ ions. Also the peak C6 is absent in SQWVs of the Zn2þ
plus receptor solutions.
The interaction of 2 with Cd2þ and Zn2þ gives rise to
analogous results that these described for 1. The most
remarkable feature is the appearance of a well-defined
reduction peak at 20.45 V (C5), which accompanies the
peaks corresponding to the uncomplexed ligand.
Under the studied conditions, peak current data can be used
to determine the stoichiometry and formation constant of
the complexes of Cd2þ and Zn2þ with 1 and 2, using a
generalisation of the molar-ratio method.16–18 By using the
peak currents of C5, which were measured in solution at
constant concentration of M2þ and increasing concen-
trations of each one of the ligands, a 1:1 metal/2
stoichiometry was obtained as expected, where the for-
mation equilibrium constants were (5.8^0.2)103 and
(4.6^0.4)103 (mol dm23), for Cd2þ and Zn2þ, respectively.
Figure 7. CVs at a Pt electrode of (a) 2.16 mM Cd(NO3)2, (b) id. plus
0.86 mM 1 in MeCN (0.10 M Bu4NPF6). Potential scan rate 100 mV/s.
However, a 2:1 metal/receptor stoichiometry was obtained
for 1, where the formation constants were (2.8^0.2)107 and
(4.1^0.4)106 (mol dm23)22, for Cd2þ-1 and Zn2þ-1,
respectively. This stoichiometry must correspond to that
of the reduced complexes formed between M2þ and the
receptor dianion, L22, resulting from the biphenyl-centred
reduction process C5 (Eq. (2)). The reduction of the
biphenyl moiety of 1 gives rise to the formation of a planar
20.26 V (A4) which is then followed by an anodic wave
close to þ0.22 V. The cathodic process corresponds to the
two-electron reduction of Cd2þ ions to form a non-uniform
deposit of Cd metal on the electrode surface, further
oxidised to Cd2þ in solution (stripping peak A4), in
agreement with the literature.15
In the presence of the ligand (see Fig. 7(b)), overlapped
peaks at 20.46 (C5), 20.95 (C1), 21.64 (C6) and 21.95 V
(C2) appear, whereas the anodic region becomes ill-defined.
As the SQWVs show in Figure 8, the non-coordinated 1
displays isolated peaks C1 and C2. On addition of Cd2þ
,
additional peaks appear at 20.49 (C5) and 21.64 V (C6)
whereas the peak C1 is resolved into two overlapped peaks
at 20.92 and 21.03 V.
The electrode process C5 can be described in terms of the
biphenyl-centred reduction of the complex. This can be
represented as:
L þ xM2þ þ 2e2 ! ðM2þÞxðL22
Then, the electrode process C6 can be ascribed to the metal-
centred reduction:
Þ
ð2Þ
ðM2þÞxðL22Þ þ 2xe2 ! xM0 þ L22
This process takes place at considerably more negative
ð3Þ
potentials than those of the reduction of uncomplexed Cd2þ
,
as expected for the electrochemical reduction of metal ions
to metal deposits.
The interaction between the ligands and Zn2þ follows a
scheme similar to that previously described. As shown in the
SQWV depicted in Figure 9(a), for Zn2þ a prominent
reduction peak is recorded at 21.33 V (C7). By adding
increasing amounts of 1 (see Fig. 9(b)), the voltammetric
Figure 8. SQWVs at a Pt electrode of (a) 0.48 mM 1, (b) 2.16 mM Cd2þ
plus 1.69 mM 1 solutions in MeCN (0.10 M Bu4NPF6). Potential step
increment 4 mV, square wave amplitude 25 mV, frequency 15 Hz.

4688
A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
Scheme 2.
dianion with the two crown ethers placed faraway from each
other, as depicted in Scheme 2. Therefore, there is no place
for the formation of a sandwich-type complex, and each of
the crowns is coordinated with one metal ion.
3. Conclusions
The studies carried out with ligands 1 and 2 have
demonstrated that the second cavity present in ligand 1
has a strong influence in transport experiments. The
sandwich-type complex formation clearly modifies the
lipophilic character of this complex, giving rise to an easier
transport across the liquid membrane. However, ligand 1
exhibit no selectivity towards the studied cations. By
contrast, ligand 2 is less effective in transport, but shows a
clear selectivity towards Naþ. In addition, it has been
observed that transport efficiency is dependent on the
concentration at the source phase.
On the other hand, electrochemical studies showed that
ligand 2 is able to complex Cd2þ and Zn2þ with a 1:1
stoichiometry. Electrochemical results obtained with ligand
1 are more interesting because instead of the sandwich
complexes, complexes 2:1 were detected under these
conditions. The presence of this type of complexes seems
to be related to the coplanarity induced in the biphenyl
system by the electrochemical process.
4. Experimental
4.1. General methods
All commercially available reagents were used without
further purification. Benzene was dried over sodium. Water
sensitive reactions were performed under argon. Column
chromatographies were carried out on SDS activated neutral
aluminium oxide (0.05–0.2 mm; activity degree 1). Melting
points were measured with a Cambridge Instrument and are
uncorrected. IR spectra were recorded as KBr pellets on a
Figure 9. SQWVs at a Pt electrode of: (a) 2.26 mM Zn(NO3)2, (b) id. plus
1.01 mM 1 solutions in MeCN (0.10 M Bu4NPF6). Potential step increment
4 mV, square wave amplitude 25 mV, frequency 15 Hz.

A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
4689
Perkin–Elmer 1750 FT-IR and a Bruker Equinox 55 FT-IR.
NMR spectra were recorded with Bruker Avance 300/500
and Varian Unity-300/400 spectrometers. Chemical shifts
are reported in parts per million downfield from TMS.
Spectra were referenced to residual undeuterated solvent.
High-resolution mass spectra were taken with a Fisons
VG-AUTOSPEC. UV spectra were run at 20 8C (thermo-
stated) on a Shimadzu UV-2102 PC.
(d), 131.19 (d), 123.25 (d, s, 2 signals), 71.18 (t), 70.67–
70.19 (t, several signals), 68.73 (t), 49.98 (t, CH2N–), 44.96
(t, CH2-N–); UV (CH3CN): 1¼17398, lmax¼293 nm.
4.1.4. Synthesis of 2. 2-(2-Methyloxycarbonyl-4-nitro-
phenyl)-5-nitrobenzoic acid (0.509 g, 1.47 mmol) was
converted into its acid chloride, as described for the
synthesis of 1. Benzene (30 ml) was added and evaporated
to yield 2-(2-methyloxycarbonyl-4-nitrophenyl)-5-nitro-
benzoic acid chloride as a solid (0.536 g, 1.47 mmol). It
was dissolved in dry CH2Cl2 (25 ml) and used in the
following step.
4.1.1. Synthesis of 4,40-dinitro-2,20-diphenic anhydride
acid (3). 4,40-Dinitro-2,20-diphenic acid (0.197 g,
0.594 mmol) was stirred at 90 8C in acetic anhydride
(40 ml) for 24 h. After removal of the solvent under
vacuum, a clear brown oil was obtained. This oil was
washed with Et2O, allowing separation of the soluble
starting acid (0.078 g) and its anhydride as a brown powder
that was dried under vacuum. (0.119 g, 0.38 mmol, 64%
yield). Mp 230–233 8C. HRMS (EI): found Mþ 314.0181,
C14H6N2O7 requires 314.0175; nmax 1815 (CvO), 1733
(CvO), 1525, 1350 cm21; dH (CDCl3) 8.92 (2H, d,
J¼2.3 Hz, Ar-H), 8.53 (2H, dd, J1¼8.5 Hz, J2¼2.3 Hz,
Ar-H), 7.46 (2H, d, J¼8.5 Hz, Ar-H); dC (CD3COCD3)
166.1 (COOCO), 148.5, 147.0, 131.6, 126.4, 125.3, 124.8
(12 CAr).
To a stirred mixture of 1-aza-18-crown-6 (95%) (0.408 g,
1.47 mmol) and dry triethylamine (99%) (0.150 g,
1.47 mmol) in dry CH2Cl2 (20 ml) at 0 8C, the above
solution was added dropwise under Ar atmosphere. Then,
the mixture was stirred at room temperature and monitored
by TLC. After completion of the reaction (12 h), the
solution was concentrated under reduced pressure and the
crude reaction product was purified by chromatography
through a neutral alumina column using CH2Cl2/EtOAc
(8:2) as eluents to give 2 as a brown oil (0.530 g,
0.887 mmol). (60% yield). HRMS (EI): found Mþ
591.2073. C27H33N3O12 requires 591.2064. nmax 1731
(OCvO), 1634 (NCvO), 1608, 1524 (NvO, asym.),
1349 (NvO, sym.), 1123 cm21; dH (CDCl3) 8.85 (1H, d,
J¼2.4 Hz, Ar-H), 8.39–8.36 (2H, m, Ar-H), 8.27 (1H, dd,
J1¼8.5 Hz, J2¼2.4 Hz, Ar-H), 7.63 (1H, d, J¼8.5 Hz, Ar-
H), 7.40 (1H, d, J¼8.5 Hz, Ar-H), 3.79 (3H, s, OCH3),
3.65–3.54 (20H, m, –OCH2–), 3.42 (2H, m, –NCH2–),
3.36 (2H, m, –NCH2–); dC (CDCl3) 168.20 (CON), 165.23
(–COO), 148.07, 147.66, 145.52, 143.65, 137.35, 133.02,
131.59, 130.78, 126.44, 125.93, 123.50 and 123.11 (12 CAr),
71.49, 70.95, 70.90, 69.50, 69.00 (10 –OCH2–), 53.29
(–OCH3), 49.91 (–NCH2), 45.19 (–NCH2); UV (CH3CN):
1¼1470, lmax¼284 nm.
4.1.2. Synthesis of 2-(2-methyloxycarbonyl-4-nitro-
phenyl)-5-nitrobenzoic acid (4). 4,40-Dinitro-2,20-diphenic
anhydride acid (0.119 g, 0.38 mmol) was stirred for 5 h in
refluxing MeOH (40 ml). Then, the solvent was evaporated
under vacuum, to give a dark brown solid. This solid was
dissolved in boiling CHCl3 and the hot solution filtered.
After evaporation of the solvent under reduced pressure, a
white powder (0.132 g, 0.38 mmol) was obtained. Mp 176–
180 8C. nmax 2956 (OH), 1728, 1698, 1525, 1350,
1265 cm21; dH (CDCl3) 8.97 (1H, d, J¼2.4 Hz, Ar-H),
8.92 (1H, d, J¼2.4 Hz, Ar-H), 8.46 (1H, dd, J1¼8.5 Hz,
J2¼2.4 Hz, Ar-H), 8.43 (1H, dd, J1¼8.5 Hz, J2¼2.4 Hz, Ar-
H), 7.37 (2H, d, J¼8.5 Hz, Ar-H), 3.75 (3H, s, OCH3); dC
(CD3COCD3) 166.13 (COO), 165.83 (COO), 149.54,
149.42, 148.77, 143.70, 132.81, 132.79, 132.08, 131.76,
127.4, 127.45, 126.14 and 125.82 (12 CAr), 53.25 (OCH3);
MS (EI): Mþ (346, 65%); Mþ21 (345, 100%).
4.2. Transport experiments
Experiments of transport were carried out across an organic
liquid bulky membrane consisting in 15 ml of ethanol-free
CHCl3 confined in a 12 mm of diameter U-shaped tube. A
known amount of carrier was dissolved in this membrane
(see Table 3). Picrate salts were synthesised from picric acid
and the corresponding hydroxide of each metal, and most of
them were purified by recrystallisation in ethanol/water.
Initial solutions of picrate salts in water were prepared: Liþ
(0.02 M), Naþ (0.02 M), Naþ (0.005 M), Kþ (0.005 M),
Csþ (0.005 M), Mg2þ (0.02 M) and Ba2þ (0.005 M).
4.1.3. Synthesis of 1. 4,40-Dinitro-2,20-diphenic acid
(0.063 g, 0.19 mmol) was added to an excess of thionyl
chloride (30 ml) and the stirred suspension was heated under
reflux until disappearance of the precipitate (ca. 2 h). The
resulting solution was concentrated under reduced pressure
until all the SOCl2 had been removed. It was dissolved in
benzene (30 ml) and dropped over a stirred solution of
1-aza-18-crown-6 (0.105 g, 0.38 mmol) and triethylamine
(0.039 g, 0.38 mmol) in benzene (30 ml) under an argon
atmosphere. The resulting solution was stirred overnight at
room temperature, the reaction mixture filtered and the
filtrate then washed with water (3£2 ml), dried over
anhydrous Na2SO4 and concentrated to yield 1 as a brown
oil (0.099 g, 63%). HRMS (EI): found (Mþþ1) 823.3643;
C38H55N4O16 requires 823.3613; nmax: 2867, 1633 (CvO),
Solution of carrier in CHCl3 (15 ml) was transferred to a
U-shaped tube, provided with a magnetic stirring bar. Initial
solutions of the picrate salts (5 ml) were added to one of the
branches of the tube (source phase). 5 ml of pure water was
allocated in the other (receiving phase). In addition, one
blank experiment for each initial solution was carried out, so
as to determine the passive transport, by replacing the 15 ml
of carrier solution with 15 ml of pure ethanol-free CHCl3.
After 48 h stirring (350 rpm) at 293^1 K, aqueous
receiving phase was homogenised and a convenient aliquot
was transferred to a 1 cm pathlength quarz UV cell.
Distilled water was added to the cell till complete 3 ml of
1523 (NvO, asym.), 1349 (NvO, sym.), 1113 (CO) cm21
;
dH (CDCl3) 8.33 (2H, d, J2¼2.7 Hz, Ar-H), 8.29 (2H, dd,
J1¼8.4 Hz, J2¼2.7 Hz, Ar-H), 7.74 (2H, dd, J1¼8.4 Hz, Ar-
H), 3.7–3.5 (48H, m broad band, –CH2– linked to O and
N); dC (CDCl3) 168.24 (s), 147.28 (s), 141.31 (s), 138.17

4690
A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
Table 4. Extraction results with ligands 1 and 2
Picrate salts
1 at 430 nm in CH3CN
Kd£103
log Ke
log Ka
1
2
1
2
Liþþ
Na
8626
1.42 M21
2.38
2.59
2.50
2.95
3.81
3.86
0.57
1.88
1.50
0.89
1.31
2.70
5.23
5.35
5.09
5.21
5.21
3.12
3.42
4.64
4.09
3.16
2.71
1.97
8975.9
9251.5
9586.4
9339.4
10,880
1.74 M21
Kþ
2.55 M21
Csþ
Mg2þ
Ca2þ
5.41 M21
39.8 M22
5.41£10þ3 M22
total volume. Absorbance was registered at 400 nm. Factor
of efficiency was calculated from (4). In order to determine
concentrations, both in the source and in the receiving
phase, Beer–Lambert’s law was used. Extinction coefficient
for each picrate salt in water was determined as in the
section of extraction experiments, but using water as the
solvent instead of CH3CN. Values of 1 were recorded in
Table 4. Experiments of transports were carried out
in triplicate and values of factor of efficiency are recorded
in Table 3.
Extinction coefficients (1) in HPLC-quality CH3CN on the
dynamic range of 1025 to 106 M were determined using a
set of standard different-concentration solutions of each salt
and the values were recorded in Table 4.
Values of Ka for monovalent cations were calculated using
the model described by Cram.19,20 Following a similar
procedure, Ka for divalent cations were calculated from (6).
Values of the constant of distribution (Kd) for alkaline
cations were reported by Cram et al.6 In the case of alkaline-
earth cations, Kd were determined following the same
experimental procedure as Cram et al., but using (5) instead,
which regards the change on the stoichiometry. Values of Kd
thereby obtained are recorded in Table 4.
mmol transported cation
ml receiving phase
E ¼
ð4Þ
mmol carrier
Ke
2
H2O 2CHCl3
2
L
þ M2þ þ 2A
Y M2þLA
CHCl3
H2O
4.3. Extraction experiments
½M2þLA22 ꢀCHCl
½M2þLA22 ꢀCHCl
3
3
Ke ¼
¼
2þ
The method described by Cram et al.6 was followed in order
to determine values of extractions constants (Ke) and
association constants (Ka) for ligands 1 and 2 against
several cations in picrates. Calculations were based on
absorbance at l¼430 nm of the picric salts in the UV
spectra. All the measurements were carried out at 293^1 K
(thermostated) on a double beam Shimadzu UV-2102 PC
spectrophotometer using a 1 cm pathlength quartz UV- cell.
Deionised water and ethanol-free CDCl3 constituted the
aqueous and organic phase respectively. Initial 0.015 M
solutions in water of monovalent (Liþ, Naþ, Kþ) and
divalent (Mg2þ, Ca2þ) cations as picrate salts were
prepared, except for Csþ, which was a 0.0076 M one.
Solution of the host was 0.076 M in CHCl3.
3
½LꢀCHCl ½M
ꢀ
½A2ꢀ2
½LꢀCHCl 4½M2þꢀ3H
H2O
H2O
3
2O
R
¼
ð1 2 RÞ4Q3
being
½M2þLA22 ꢀCHCl
i CHCl3
3
R ¼
½L ꢀ
and
!
V
CHCl3
2þ
Q ¼ ½M
ꢀ
¼ ½M2þA22iꢀH O 2 R
½LiꢀCHCl
H2O
2
3
V
H2O
½M2þA22 ꢀCHCl
H2O
3
Kd ¼
ð5Þ
ð6Þ
Into a 20 ml centrifuge tube was introduced 0.2 ml of the
host solution. A volume of 0.5 ml of water was added to one
of the tubes. Picrate solution (0.5 ml) was added to each of
the others. The tubes were stoppered and shacked
vigorously by hand for 3 min. Phases were separated after
that by centrifugation.
2þ 3
4½M
ꢀ
Ka
Y M2þLA
2
2CHCl3
2
2CHCl3
L
þ M2þ
A
CHCl3
½M2þLA22 ꢀCHCl
Ke
Kd
3
Ka ¼
¼
½Lꢀ
½M2þA2ꢀ
CHCl3
CHCl3
2
A 10 ml aliquot of the organic layer was transferred by
automatic pipette into a 3.5 ml UV cell. 2.99 ml of CH3CN
were added to complete a total volume of 3 ml. The UV
absorption was measured against a cell containing 10 ml of
the organic layer at the blank experiment and 2.99 ml of
CH3CN. Absorbance of both cells when they were filled
with CH3CN was balanced by placing the blank one across
the reference beam, the sample one across the sample beam
and running an automatic baseline correction. Three
repetitions were run for a given cation and concentrations
were found out using Beer–Lambert’s law relationship.
Where M refers to the metal cation, A to the picrate anion, L
to the ligand behaving as a host with only one complexing
site, VCHCl and VH O to the total volume in the organic and
3
2
the aqueous phases.
4.4. Electrochemical studies
Electrochemical measurements have been performed at
298^1 K in a conventional three-electrode cell under argon
atmosphere. Nominal ca. 2.0 mM concentrations of the

A. M. Costero et al. / Tetrahedron 60 (2004) 4683–4691
4691
Asay, B. W.; Izatt, R. M. J. Am. Chem. Soc. 1980, 102,
6820–6824.
ligand were used in dry MeCN using tetrabuthylammonium
hexafluorophosphate (0.10 M) as a supporting electrolyte.
5. Costero, A. M.; Andreu, C.; Pitarch, M. Tetrahedron 1996, 52,
3683.
Experiments were performed using a BAS CV 50 W
equipment using a BAS MF2012 glassy carbon electrode
(GCE) (geometrical area 0.071 cm2), and a BAS MF2014
platinum electrode (geometrical area 0.018 cm2) as a
working electrode. A platinum wire auxiliary electrode
and a AgCl (3 M NaCl)/Ag reference electrode separated
from the test solution by a salt bridge only containing
supporting electrolyte completed the electrode arrangement.
The potential of such reference electrode was 235 mV vs
the saturated calomel reference electrode (SCE).
6. Moore, S. S.; Tarnowski, T. L.; Newcomb, M.; Cram, D. J.
J. Am. Chem. Soc 1977, 99, 6398–6405.
7. Gloe, K.; Graubaum, H.; Wu¨st, M.; Rambusch, T.; Seichter,
W. Coord. Chem. Rev. 2001, 222, 103–126.
¨
8. Stolwijk, T. B.; Sudholter, E. J. R.; Reinhoudt, D. N.; Van
Eerden, J.; Harkema, S. J. Org. Chem. 1989, 54, 1000–1005.
9. Costero, A. M.; Monrabal, E.; Sanjuan, F.; Martinez-Man˜ez,
R.; Padilla-Tosta, M.; Pardo, T.; Soto, J. Tetrahedron 1999,
55, 15141–15150.
10. Costero, A. M.; Aurell, M. J.; Ban˜uls, M. J.; Ochando, L. E.;
Tamarit, F. J.; Domenech, A. J. Inclusion Phenom. Mol.
Recognit. Chem. 2003, 45, 241–249.
Acknowledgements
11. Perrin, C. L. Mechanisms of organic polarography. In Organic
polarography; Zuman, P., Perrin, C. L., Eds.; Interscience:
New York, 1969; pp 210–214.
´
We thank the Ministerio de Ciencia y Tecnologıa PPQ2002-
00986 and Generalitat Valenciana Grupos03/206 for sup-
port. We also thank the SCSIE of the Universitat de
12. Faulkner, L. R.; Bard, J. Am. Chem. Soc. 1968, 90,
6284–6290.
`
Valencia for technical support. J. Sanchis is grateful to the
Universitat de Valencia for a predoctoral fellowship.
13. Marcoux, L. S.; Fritsch, J. M.; Adams, R. N. J. Am. Chem. Soc.
1959, 89, 5766–5769.
`
14. Aten, A. C.; Buthker, C.; Hoijtink, J. Chem. Soc. Faraday
Trans. 1959, 55, 324–330.
References and notes
15. Allendoerfer, R. D.; Rieger, P. H. J. Am. Chem. Soc. 1965, 87,
2336–2344.
1. Sakamoto, H.; Kimura, K.; Koseki, Y.; Matsuo, M.; Shono, T.
J. Org. Chem. 1986, 51, 4974–4979, and references cited
therein.
16. Lovric, M.; Komorsky-Lovric, S.; Bond, A. M. J. Electroanal.
Chem. 1991, 319, 1–7.
´ ´
17. Bianchi, A.; Domenech, A.; Garcıa-Espan˜a, E.; Luis, S. V.
2. Sinkai, S.; Ogawa, T.; Kusano, Y.; Manabe, O.; Kikukawa, K.;
Goto, T.; Matsuda, T. J. Am. Chem. Soc. 1982, 104,
1960–1967.
Anal. Chem. 1993, 65, 3137–3142.
´ ´
18. Domenech, A.; Garcıa-Espan˜a, E.; Navarro, P.; Reviriego, F.
Talanta 2000, 51, 625–636.
19. Newcomb, M.; Timko, J. M.; Walba, D. M.; Cram, D. J. J. Am.
Chem. Soc 1977, 99, 6392–6398.
3. Cho, Y. H.; Rha, S. G.; Chang, S. K.; Chung, T. D.; Cho, K. C.;
Kim, H. J. Inclusion Phenom. Mol. Recognit. Chem. 1998, 31,
119–129.
20. Timko, J. M.; Moore, S. S.; Walba, D. M.; Hiberty, P.; Cram,
D. J. J. Am. Chem. Soc 1977, 99, 4207–4219.
4. Lamb, J. D.; Christensen, J. J.; Oscarson, J. L.; Nielsen, B. L.;