Calixpyrrole chemistry: A study of a new ditopic receptor highlighting some fundamental concepts in assessing thermodynamic selectivity

Article abstract of DOI:10.1039/b915332b

Following the synthesis and characterization of meso-tetramethyl tetrakis (4-phenoxy acetone) calix[4]pyrrole, 1, the solution properties of this receptor in various solvents were investigated. Particular emphasis is placed on the selection of the solvent

Full text of DOI:10.1039/b915332b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Calixpyrrole chemistry: a study of a new ditopic receptor highlighting
some fundamental concepts in assessing thermodynamic selectivity
Angela F. Danil de Namor* and Rasha Khalife
Received 28th July 2009, Accepted 19th October 2009
First published as an Advance Article on the web 27th November 2009
DOI: 10.1039/b915332b
Following the synthesis and characterization of meso-tetramethyl tetrakis (4-phenoxy acetone)
calix[4]pyrrole, 1, the solution properties of this receptor in various solvents were investigated.
Particular emphasis is placed on the selection of the solvent in assessing thermodynamic
selectivity in ion complexation studies involving calixpyrrole receptors. Thus, statements recently
made in the literature are addressed.
The interaction of 1 with anions and cations was assessed through 1H NMR in CD3CN at 298 K.
Among anions, significant chemical shift changes were observed in the pyrrolic protons by the addition
of halide salts to 1 in the deuterated solvent. The sequence observed in the Dd values is that observed in
the transfer Gibbs energies of these anions from a dipolar aprotic (acetonitrile) to a protic medium
(representative of 1) based on the Ph4AsPh4B convention. As far as metal cations in CD3CN are
concerned, the most significant changes are observed (relative to those of the ligand) in the protons of
the phenoxy acetone functionality upon the addition of the mercury(II) salt to 1. Conductance
measurements reveal the formation of 1 : 1 complexes with these ions. Thermodynamic data derived
from titration calorimetry are reported and compared with available data for analogous ligands.
A quantitative evaluation of the thermodynamic stability was carried out. The applications of these
ligands as anchor groups in oligomeric frameworks are discussed. Final conclusions are given.
addressed in view of statements and recommendations made in
the literature concerning calixpyrrole chemistry.6 These issues
1. Introduction
Calixpyrroles are an old class of macrocycles first synthesised
are highlighted in this paper, in which we report
by Baeyer using the condensation of pyrrole and acetone.1 The
(i) the synthesis and characterization (structural and thermo-
remarkable growth of interest in these macrocycles was
dynamic) of meso-tetramethyl tetrakis-(4-phenoxy acetone)
initiated in the 1990s by the extensive work of Floriani and
coworkers2 on the metabolism and synthetic chemistry of
calix[4]pyrrole, 1. The interest in this ligand relies on the
presence of the phenyl moiety in the meso position which
deprotonated calixpyrroles, followed by the contributions
made by Sessler and coworkers.3 The latter authors reported
allows its polymerisation via a treatment with formaldehyde
and formic acid.4,5
that calix[4]pyrroles are not only effective anion-binding
(ii) proton NMR and conductance studies aiming to identify
ligands but also selective ones.3 They showed that these
the process taking place in solution in order to ensure that the
receptors have a marked preference for fluoride relative to
thermodynamic data are referring to a well defined process.
other halide guests.3 A detailed study carried out by Danil de
(iii) a detailed thermodynamic investigation of the anion
Namor and Shehab4,5 involving not only the thermodynamics
and cation complexation in acetonitrile at 298.15 K.
of anion-complexation of calix[4]pyrrole in dipolar aprotic
media, but also the solution thermodynamics of the reactants
and the product participating in the binding process, confirmed
that this receptor interacts selectively with halides ions.
In fact we demonstrated for the first time the solvent effect
on the complexation process and discussed it in terms of the
solvation changes that the reactants and the product undergo
in moving from one medium to another. In sharp contrast to
the excellent progress made on the synthetic and the structural
studies on calix[4]pyrrole and derivatives, the fundamental
thermodynamics of these systems leading to a quantitative
evaluation of selectivity is still a matter which needs to be
Laboratory of Thermochemistry, Chemical Sciences,
Faculty of Health and Medical Sciences, University of Surrey,
Guildford, Surrey, UK GU2 7XH.
E-mail: a.danil-de-namor@surrey.ac.uk; Tel: 01483689581
(iv) a discussion on the effect of the receptor on ion
complexation. Thus, the results obtained from this
ꢀc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 753–760 | 753

View Article Online
investigation are compared with those for calix[4]pyrrole, 2,4
an analogous derivative, 3,7 and a calix[4]arene ketone deri-
vative, 4, previously reported by our group.8
(s, 4H, OH); 6.78 (d, 8H, ArH); 6.66 (d, 8H, ArH); 5.96 (d, 8H,
PyH); 1.81 (s, 12 H, CH3).
2.3 Synthesis of meso-tetramethyl tetrakis-(4-phenoxy
acetone) calix[4]pyrrole, 1
2. Experimental
Chloroacetone (4.3 ml) and 18-crown-6 (18-C-6) (0.6 g) were
dissolved in acetonitrile (80 ml). The solution was placed in a
three-necked round bottom flask equipped with a magnetic
stirrer, a refluxing condenser carrying a guard-tube filled with
anhydrous calcium chloride. Potassium iodide (13.44 g) was
added and the mixture was heated under vigorous stirring at
70 1C for one hour. A yellow-brown suspension was observed.
The reaction vessel was removed from the hot oil bath and
allowed to cool down to room temperature. Then meso-
tetramethyl-tetrakis-(4-hydroxyphenyl) calix[4]pyrrole (4 g),
18-C-6 (0.5 g) and potassium carbonate (7.46 g) were added
and the reaction was left to reflux overnight at 60 1C. The
reaction vessel was cooled down to room temperature and the
mixture was evaporated to remove the acetonitrile. The solid
compound was dissolved in dichloromethane and extracted
with an aqueous solution of sodium hydrogen carbonate
(0.2 mol dmꢁ3). The organic layer was collected and dried
with MgSO4. The solvent was removed by evaporation and a
brown oil was obtained. The oil was re-crystallized using
methanol. The compound was collected by filtration and dried
under vacuum at 90 1C. 1H NMR (d6-DMSO, 300 MHz,
298 K, ppm): 7.97 (s; 4H, NH); 6.81 (d, 8H, ArH); 6.75 (d, 8H,
ArH); 6.01 (d, 8H, PyH); 4.66 (s, 8H, CH2); 2.18 (s, 12H,
CH3); 1.83 (s, 12H, CH3).
2.1 Chemicals used
Pyrrole (99%), methane sulfonic acid (99.5%), p-hydroxyaceto-
phenone (99%), tri-ethyl amine (99%), lithium perchlorate,
sodium perchlorate (99%), potassium perchlorate (99%),
calcium perchlorate tetrahydrate (99%), magnesium perchlorate
(80%), cadmium perchlorate, mercury perchlorate mono-
hydrate (99.99%), tetra-n-butylammonium bromide (99%),
iodide (97%), nitrate (97%), 18-crown-6 (99%), chloroacetone
(95%) and acetonitrile (CD3CN) were all purchased from
Aldrich Chemical Company.
Magnesium sulfate (62–70%), anhydrous potassium carbo-
nate, potassium iodide (99%), methanol (HPLC grade),
diethyl ether (laboratory reagent), acetic acid (analytical
reagent), acetone (analytical reagent), acetonitrile (HPLC
grade) and dimethyl sulfoxide (HPLC grade) were purchased
from Fisher UK Scientific International. Acetonitrile and
dimethyl sulfoxide were refluxed in a nitrogen atmosphere
and distilled over calcium hydride.9
Perchlorate salts of rubidium (Alfa Aesar) and strontium
hexahydrate (99.9%, Fluka Chemical Company), tetra-
n-butylammonium fluoride trihydrate (97%, Fluka Chemical
Company) and chloride (97%, Fluka Chemical Company)
were dried over P4O10 under vacuum for several days. The
absence of a water peak in the NMR spectra upon addition of
salts indicated that they were dry.
2.4 Solubility measurements
These measurements were carried out as previously described.4
Analysis of the saturated solution was carried out gravi-
metrically by triplicate.
2.2 Synthesis of meso-tetramethyl-tetrakis-(4-hydroxyphenyl)
calix[4]pyrrole
The preparation of meso-tetramethyl-tetrakis-(4-hydroxy-
phenyl) calix[4]pyrrole was achieved by applying the procedure
reported in the literature10,11 Thus, pyrrole (5 g, 75 mmol)
was placed in a 250 ml round bottom flask equipped with
a magnetic stirrer and filled with methanol (100 ml). Methane-
sulfonic acid (1 ml) was then added to the mixture and stirred
for 30 min. This was followed by the stepwise addition of
a solution of p-hydroxyacetophenone (11 g, 80 mmol) in methanol
(50 ml). The reaction mixture was monitored by thin layer
chromatography (TLC) using a dichloromethane : methanol
mixture (9.5 : 0.5) as the developing solvent and was left to stir
overnight. It was then stopped by pouring the reaction mixture
into distilled water (200 ml). An orange colored precipitate
was obtained. The residue was filtered off, collected and then
dissolved in diethyl ether (400 ml). The solution was filtered
gravitationally to remove the black tar obtained. This was
re-crystallised from acetic acid and left to cool down. Green
crystals containing the aaaa isomer acetic acid complex were
obtained. The acid was removed by further re-crystallisation
with an acetonitrile : acetone mixture (1 : 1). White crystals
of the calixpyrrole derivative (65% yield) were obtained.
2.5 1H NMR measurements
Using a Bruker AC-300E pulsed Fourier transform NMR
spectrometer, 1H NMR measurements were recorded at
298 K. Typical operating conditions for the routine proton
measurements involved a pulse or flip angle of 301, spectral
frequency of 300.135 MHz, delay time of 1.60 s, acquisition
time of 1.819 s, and line broadening of 0.55 Hz.
A solution of 1 was prepared in the appropriate deuterated
solvent and then placed in a 5 mm NMR tube using tetra-
methylsilane as the internal reference.
The complexation behavior of the ligand toward the anions
and cations was studied by adding the salts (8.0 ꢂ 10ꢁ3 to
1.6 ꢂ 10ꢁ2 mol dmꢁ3) into the NMR tube containing 1
dissolved in the same solvent (1.0 ꢂ 10ꢁ3 mol dmꢁ3). An
excess amount of the appropriate salt was added and the
chemical shifts were recorded. Changes in the chemical shifts
upon addition of the salts relative to the free ligand were
calculated.
2.6 Conductometric measurements
1
These were dried in vacuum at 90 1C. H NMR (d6-acetone,
300 MHz 298 K, d in ppm); 8.75 (s broad, 4H, NH); 8.22
Conductance measurements were performed as discussed
elsewhere.7 The experimental conditions used were as
ꢀc
This journal is the Owner Societies 2010
754 | Phys. Chem. Chem. Phys., 2010, 12, 753–760

View Article Online
follows: concentration of the anion or cation salts were in the
5.00 ꢂ 10ꢁ5–1.00 ꢂ 10ꢁ4 mol dmꢁ3 range and the concentration
of the ligand was 1.00 ꢂ 10ꢁ3 mol dmꢁ3 in the same solvent.
Standard Gibbs energies of solution, DsG1, were calculated
from the solubility data using eqn (1) only when the solid
phase was not altered by solvate formation.
DsG1 = ꢁRTln K
(1)
2.7 Calorimetric measurements
Calorimetric measurements were carried out to derive the
thermodynamic parameters of complexation of this ligand
and the relevant ions in acetonitrile.
In eqn (1), K is the equilibrium constant which is equal to the
concentration of the ligand (mol dmꢁ3) in the saturated
solution; given that this is a neutral ligand and, therefore,
given the low concentrations used, it is assumed that the
activity coefficient is equal to 1. The data refer to the standard
The stability constant (expressed as log Ks) and enthalpies of
complexation of the ligand with the appropriate anion or
cation at 298.15 K in MeCN were determined by direct
calorimetric titration using a Tronac 450.12 The calibration
of this machine was carried out by determining the enthalpy of
protonation of an aqueous solution of tris-(hydroxymethyl)
aminomethane (THAM) with an aqueous solution of hydro-
chloric acid (0.1 mol cmꢁ3) at 298.15 K. The value obtained
(ꢁ47 kJ molꢁ1) was in good agreement with that reported by
Wilson and Smith13 at 298.15 K. A solution of the ligand in
the selected solvent (10ꢁ4–10ꢁ3 mol dmꢁ3) was placed in the
vessel (50 cm3) and titrated with the corresponding anion or
cation salt (10ꢁ3–10ꢁ2 mol dmꢁ3). The latter was added from a
2 cm3 burette connected by a silicone tube to the reaction
vessel after reaching thermal equilibrium. These experiments
were repeated three times for each solvent. Blank experiments
were carried out to account for dilution effects resulting from
the addition of the fluoride salt to the solvent placed in the
calorimetric vessel.
state of 1 mol dmꢁ3
.
In solvents where solvate formation was observed, it was
not possible to determine the DsG1 value since the composition
of the solid in equilibrium with the saturated solution is not
the same. The transfer Gibbs free energy, DtG1, was calculated
using acetonitrile as the reference solvent, s1. The standard
Gibbs free energies of transfer of 1 from MeCN to various
solvents presented in Table 1 reflect the difference in solvation
of this neutral ligand in s2 relative to s1. For comparison
purposes, corresponding data for an analogous ligand,7 3 are
also included in Table 1.
Ligand 3 is an ester with the oxygen in conjugation with the
carbonyl functional group. This will tend to keep these atoms
on the same plane with the benzene ring, stabilizing the crystal
lattice (this is also reflected in the melting point of 3, which was
found to be 192 1C, compared to 136 1C for 1). It is probably
more relevant to consider the zwitterionic nature of the ester
functionality. This charge separation would also contribute
to crystal lattice stability relative to the more flexible and
unconjugated ketone in 1, which lacks the charge separation,
thus increasing the solubility of this ligand in non-aqueous
solvents.
For the complexation studies involving 1 and Hg2+ in
MeCN, a solution of 1 (1 ꢂ 10ꢁ3 mol dmꢁ3) was placed in
the vessel (50 cm3) and titrated with the mercury cation salt
(5 ꢂ 10ꢁ2 mol dmꢁ3). The latter was added from a 2 cm3
burette connected as described above.
Analysis of DtG1 values shows that the ability of 1 to
interact with the solvents follows the sequence:
3. Results and discussion
PC 4 DMSO 4 MeCN 4 MeOH 4 EtOH
3.1 Solubilities and derived standard Gibbs energies of
solution. Transfer Gibbs energies from acetonitrile
As far as 3 is concerned, the following trend in solvation is
observed:
Solubility measurements were carried out in order to get
information about the effect of the ligand solvation upon
complexation. Solubility data of 1 in various non-aqueous
solvents (methanol, MeOH; ethanol, EtOH; acetonitrile,
MeCN; dimethylsulfoxide, DMSO; and propylene carbonate,
PC) are reported in Table 1 at 298.15 K.
DMSO 4 PC ꢃ DMF 4 MeCN 4 EtOH 4MeOH
The negative DtG1 (MeCN - s) values for 1 and 3 in PC and
DMSO (also DMF for 3) reflect the high interaction of these
ligands in these solvents relative to acetonitrile.
Table 1 Solubilities and derived standard Gibbs energies of solutions of 1 and 2 in various solvents.a Transfer Gibbs energies from acetonitrile at
298.15 K
Solubility/mol dmꢁ3
DsG1/kJ molꢁ1
DtG1(MeCN-s)/kJ molꢁ1
1
2
1
2
1
2
MeCN
MeOH
EtOH
DMF
DMSO
PC
(4.5 ꢄ 0.5) ꢂ 10ꢁ2
(5.20 ꢄ 0.02) ꢂ 10ꢁ3
(8.6 ꢄ 0.1) ꢂ 10ꢁ4
—
(18.2 ꢄ 0.2) ꢂ 10ꢁ4
(20.5 ꢄ 0.2) ꢂ 10ꢁ4
(19.8 ꢄ 0.4) ꢂ 10ꢁ4
(10.6 ꢄ 0.2) ꢂ 10ꢁ2
(9.6 ꢄ 0.1) ꢂ 10ꢁ2
(1.4 ꢄ 0.1) ꢂ 10ꢁ2
7.68 ꢄ 0.03
13.27 ꢄ 0.09
17.5 ꢄ 0.2
—
18.2 ꢄ 0.2
20.5 ꢄ 0.2
19.8 ꢄ 0.4
10.6 ꢄ 0.2
9.6 ꢄ 0.1
0.00
5.59
9.82
—
0.00
2.23
1.57
ꢁ7.65
ꢁ8.64
ꢁ7.65
(1.15 ꢄ 0.05) ꢂ 10ꢁ1
5.36 ꢄ 0.02
ꢁ2.32
(1.67 ꢄ 0.02) ꢂ 10ꢁ1
4.45 ꢄ 0.03
10.58 ꢄ 0.09
ꢁ3.23
a
Solvent abbreviation: acetonitrile, MeCN, propylene carbonate, PC, dimethyl sulfoxide, DMSO, N,N-dimethylformamide, DMF, methanol,
MeOH, ethanol, EtOH.
ꢀc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 753–760 | 755

View Article Online
As far as DMSO and DMF are concerned, these are
protophilic dipolar aprotic solvents and, as such, their basic
oxygens are likely to interact with the NH array of 1 through
hydrogen bond formation. These solvents are known to interact
with other calix[4]pyrrole derivatives.14
Therefore the data do not refer to a well defined process. As
such, these data are not suitable to assess the thermodynamic
selectivity, S, of the receptor, either as a function of the solvent
(S given by the ratio of stability constants for a given system in
two solvents at a given temperature) or the ion (S given by the
ratio of Ks values for two ions in a given solvent and
temperature). Therefore, there is no need to carry out experi-
mental work with different counter-ions if in the assessment of
selectivity (one of the main features of supramolecular
chemistry), the selection of the solvent is based on existing
information regarding the behaviour of 1 : 1 electrolytes in
nonaqueous solvents. This information shows that in low
permittivity media, such as dichloromethane, chloroform,
etc. (where a great deal of data involving calixpyrroles have
been reported28), ion pair formation is likely to occur. The
extent of ion pair formation is concentration dependent.
In a high permittivity medium, for the salts containing
multicharged ions, it is important to check the concentration
range at which electrolytes are predominantly ionic species in
solution. Within this context, conductance measurements have
proved to be useful in establishing this. These measurements
are also recommended in low dielectric media, where processes
such as
Gibbs energies of transfer of 1 to various solvents provide
relevant information regarding the differences in solvation of
the solute in the two solvents. This parameter is required in
assessing the different solvation factors which contribute to
complex stability in one solvent relative to another, as shown
in previous papers4,14,15 Thus, the optimum medium for the
formation of highly stable complexes is that which offers a
poor solvation medium for the reactants and a good one for
the product (metal–ion complex).16,17
Having determined the solubility of these ligands in various
solvents, in the following section we discuss the selection of the
complexation medium and address some of the statements
made in the literature.6
3.2 Selection of the complexation medium. Addressing
statements recently made in the literature
The selection of the solvent for investigating the binding
properties of 1 with ionic species is an important aspect to
consider. At this point, it is timely to address some of the
statements recently made in the literature regarding the
medium and the counter-ion effects on complexation processes
involving calix[4]pyrroles. Thus, Sessler and co-workers6
recommended that ‘‘anion binding studies involving new receptors
be carried out in several different solvents and with different
counter-ion before a detailed understanding of the anion binding
properties of the receptor-based selectivities is claimed.’’ Besides
the fact that the authors have overlooked previous work in
the general area of macrocyclic chemistry, where the medium
(solvation) and counter-ion effects have been extensively
discussed,18–21 this recommendation may lead to unnecessary
experimental work for the following reason.
MX(s) + L(s) - MLX(s)
(5)
have been suggested without verification. The use of eqn (5)
can only be justified if no conductance is observed during the
course of a titration, a clear indication that ions are not
present in solution.
As far as 1 is concerned, in selecting the solvent to proceed
with complexation studies, we recall the DtG1 values in Table 1
which indicate that this ligand is better solvated in dipolar
aprotic media than in protic media. However, protic solvents
are good solvators for cations and anions through hydrogen
bond formation. On the other hand, dipolar aprotic solvents
are poor solvators for the anions. Based on these statements
and the facts that
It does not establish a distinction between (i) the nature of
the solvents, mainly the properties of 1 : 1 electrolytes in
nonaqueous media to which so many efforts have been
devoted,22–26 (ii) complexation (eqn (2)) and ion pair formation
(eqn (3) and (4)) by implying that the counter-ion may have an
effect on the binding properties of the receptor (L) in
solution.27 A representative example is given in eqn (2). This
equation refers to a 1 : 1 complexation process involving
univalent anions.
(i) there is a considerable amount of complexation data
involving calix[4]pyrrole and anions in acetonitrile
(ii) 1 is least solvated in this solvent than in any of the
dipolar aprotic solvents considered in this work
(iii) acetonitrile is a poor cation solvator (except for Ag+
and Cu2+ 22,23
we selected, for these investigations, this
)
solvent as the complexation medium.
3.3 1H NMR studies on the interactions of 1 with ions
Ks
XꢁðsÞ þ LðsÞ ꢁ! LXꢁðsÞ
ð2Þ
Table 2 reports the chemical shift changes found upon the
addition of halide anions and mercury cation salts to 1 at
298 K, relative to the free ligand in CD3CN. For the halides,
the most significant changes are observed for the pyrrole
protons, indicating that these provide the sites of interaction
between the ligand and the halides in acetonitrile. Thus, the
following sequence is observed.
In eqn (2), Ks is the thermodynamic stability constant. The
validity of this equation requires that the single (Xꢁ) and the
complex (LXꢁ) anions are predominantly ionic species in
solution, the ligand is present as a monomer and the counter-
ion has no participation in the binding process. If it does, ion
pair formation between the single (eqn (3)) or the complex
(eqn (4)) anion and the counter-ion (M+) or both takes place.
Fꢁ 4 Clꢁ 4 Brꢁ
M+(s) + Xꢁ(s) - M+Xꢁ(s)
(3)
(4)
At this stage, it is relevant to mention some of the work
published in the literature regarding the possibility of a
deprotonation of the NH functionality in the presence of
M+(s) + LXꢁ(s) - M+LXꢁ(s)
ꢀc
This journal is the Owner Societies 2010
756 | Phys. Chem. Chem. Phys., 2010, 12, 753–760

View Article Online
Table 2 1H NMR chemical shift changes upon the addition of halide
anion and mercury cation saltsa to 1 relative to the free ligand in
CD3CN at 298 K
Dd (ppm)
Fig. 1 Relationship between Dd values for the pyrrole proton (NH) of 1
and 3 and DtG1 values of the halides moving from acetonitrile to water.
1 2
7.91 6.80
3
6.74
4
6.00
5
4.65
6
2.11
7
1.82
Ligand
Fꢁ
4.42 0.12
3.40 ꢁ0.01
2.01 0.01
0.02 0.03
0.01 0.03
0.12
0.11
0.12
0.03
ꢁ0.06
ꢁ0.05
0.01
0.03
0.11
Clꢁ
Brꢁ
Iꢁ
ꢁ0.08 ꢁ0.09
ꢁ0.07 0.04
From the results obtained, it is quite clear that 1 shows
selectivity for the halide anions while 3 only interacts with the
fluoride ion in this solvent and it discriminates against chloride
and bromide. It becomes evident that the absence of the
methylene moiety between the ethereal and the carbonyl
oxygens has implications for the ability of the NH array of 3
to enter complexation with the largest anions.
0.01
0.02
0.06
0.01
0.02
0.07
0.01
0.02
0.23
0.03
0.01
0.18
Hg2+
a
The counter ion is tetrabutyl ammonium for the anions and
perchlorate for the cation.
As far as metal cations in CD3CN are concerned, the only
significant chemical shift change was observed in the ligand
protons upon the addition of the mercury(II) salt (Table 2).
Thus, the addition of this salt to 1 in CD3CN leads to downfield
changes in H-5, H-6 and H-7. These findings may suggest that
both the carbonyl and the ethereal oxygen donor atoms provide
the sites of interaction of this ligand with this cation in this
solvent.
fluoride due to the basicity of this anion. Boiocchi et al.29
reported that 1,3-bis(4-nitrophenyl) urea interacts with the
fluoride anion through hydrogen bond formation to give a 1 : 1
complex. However, further addition of the fluoride salt leads
to urea deprotonation and the production of HF2ꢁ. A similar
result was reported by Descalzo et al.30 in their investigation
on the interaction of 1,3-indane-based chromogenic calixpyrroles
with the fluoride anion in DMSO-d6. Nishiyabu and
Anzenbacher31 observed deprotonation of the NH functionality
in the calixpyrrole derivative only when the fluoride anion salt
was present in a very large excess. In this work, we tested the
possibility of deprotonation of the NH group of the pyrrole
moiety within the fluoride salt concentration range used. The
1H NMR spectrum of 1 in the presence of the fluoride anion
does not show any signal at 16 ppm. Therefore, the chemical
shift changes observed are attributed to the complexation
reaction, through hydrogen bond formation, between the
fluoride anion and the pyrrolic NH proton.
As far as 1 and 3 are concerned, the chemical shifts observed
for 3 upon the addition of the mercury(II) salt in CD3CN
shows a greater downfield shift than that for 1, suggesting that
the stability of the complex formed with 3 is higher than that
of 1 and this cation in this solvent.
As indicated in the introduction, the pendant arms of the
methylene bridge of 1 are the same as those found in the
tetramethyl ketone p-tert-butyl calix[4]arene, 4, for which a
detailed study on its complexation with metal cations was
previously reported by us.33 However, the latter was shown to
complex with a variety of bivalent cations (Mg2+, Ca2+, Sr2+
,
Chemical shift changes for 1 are lower than those for
calix[4]pyrrole and these anions in CD3CN, which may reflect
that the complex stability of 1 with the halides in acetonitrile is
lower than that involving calix[4]pyrrole, 2 and the same
anions in this solvent. However, in both cases, the sequence
observed in the Dd values for the pyrrolic protons is that found
for the transfer Gibbs energies of these anions when moving
from a dipolar aprotic (acetonitrile) to a protic medium (water
as representative of 1). This is corroborated in Fig. 1, where Dd
values for the pyrrolic protons on addition of halides in
CD3CN are plotted against the DtG1 values of the halides
moving from acetonitrile to water (data based on the
Ph4AsPh4B convention23,32). This finding is in agreement with
a previous statement4 in that the strongest hydrogen bond
formation takes place between the calixpyrrole or its derivative
and the anion with the highest charge density. Thus, neither 1
nor the calix[4]pyrrole, 2, are able to interact with a large
anion (high polarizability), such as iodide.
Ba2+, Pb2+, Cd2+ and Hg2+) in acetonitrile. Undoubtedly,
the calixarene derivative is better pre-organised to interact
with cations than 1.33 However, the fact that 1 interacts only
with Hg2+ and Fꢁ leads to the suggestion that this receptor
may offer potential for the extraction of these polluting ions
from contaminated sources. We will strongly argue that
although the stability constant of 1 and Hg2+ in acetonitrile
is not high, the presence of 1 in a low dielectric medium, such
as dichloromethane, will enhance the extraction of this cation
from the aqueous phase. This statement is made on the basis
that, as previously discussed,18,19 it is not only the stability
constant that contributes to the extraction process but also the
ion-pair formation, which is bound to be present in a low
permittivity medium.
Having established the active sites of interaction of these
ligands with anions and metal cations in acetonitrile at 298 K,
conductance studies were performed in order to determine
the composition of the ligand–ion complexes and to gain
ꢀc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 753–760 | 757

View Article Online
semi-quantitative information regarding the strength of com-
plexation of these ligands with ionic species in a given medium.
The data are now discussed.
3.4 Conductometric titrations
Plots of molar conductance, Lm (S cm2 molꢁ1), against the
ligand/ion concentration ratio [1]/[Xꢁ] (Xꢁ = Fꢁ, Clꢁ, Brꢁ) in
MeCN are shown in Fig. 2a–c.
The complexation of Fꢁ with 1 in acetonitrile showed a
decrease in the molar conductance of the complex due to the
conversion of the free anion, Fꢁ, into the 1-Fꢁ complex, where
the size of the latter is larger than that of the former
and therefore the mobility decreases. Fig. 2a shows that the
conductometric curve is the result of a combination of two
linear segments with a well-defined change in the curvature at
the stoichiometry of the reaction (1 : 1). The slope of the
curvature changes in the titration of chloride (Fig. 2b) with
1 in acetonitrile at 298.15 K, which suggests that the chloride
complex is weaker than the fluoride complex. The conducto-
metric curve of bromide with 1 (Fig. 2c) shows a weak change
in the curvature, suggesting that the interaction between the
ligand and the anion is weak. In accordance with the chemical
shift changes observed in the 1H NMR spectra by the addition
of the halides to 1 in CD3CN, conductance measurements
indicate that the ligand is able to discriminate between the
halide anions, following the same sequence as that observed
in 1H NMR measurements. In all cases, the Lm values at
Fig. 3 Conductometric titration curve of mercury (as perchlorate)
with 1 in MeCN at 298.15 K.
To obtain quantitative information about the strength of
the ligand–ion interaction, the thermodynamic characterization
of the complexation process is now discussed.
3.5 Thermodynamics of complexation
Having established the stoichiometry of the complexes in
acetonitrile at 298.15 K, titration calorimetry was the technique
used for the determination of the thermodynamic parameters.
An equation representative of the binding process can be
formulated (eqn (6)) for the interaction of 1 with the halides
and the mercury metal cation in acetonitrile. This equation is
based on the fact that (i) the working range of concentration is
that at which the electrolytes are predominately in their ionic
forms in solution33,34 and (ii) the composition of the complex
was established through conductance measurements and
verified in the calorimetric runs.
[1]/[Xꢁ] = 0 are close to the L0 values of the halide salts in
m
this solvent (164.96 S cm2 molꢁ1 for fluoride, 160.1 S cm2 molꢁ1
for chloride and 162.1 S cm2 molꢁ1 for bromide).22
As far the mercury is concerned, Fig. 3 shows the plot of
molar conductance against the [1]/[Hg2+] molar ratio in
acetonitrile at 298.15 K. Inspection of the plot reveals a
decrease of the molar conductance upon the addition of the
ligand, indicating that an interaction is taking place between
the ligand and mercury in acetonitrile. No changes in the
molar conductance are observed after the break at the 1 : 1
([1]/[Mn+]) stoichiometry, suggesting that one cation is hosted
per unit of 1. The molar conductance value at infinite dilution
for mercury perchlorate investigated in acetonitrile was found
to be within the range expected, as compared with the L0
1(s) + Xꢁ(s) - 1Xꢁ(s) + M+(s)
1(s) + M2+(s) + 2Xꢁ(s) - M2+1(s) + 2Xꢁ(s)
(6)
Stability constants (expressed as log Ks) and derived standard
Gibbs energies, DcG1 (referring to the standard state of
1 mol dmꢁ3 for the reactants and the products), standard
enthalpies, DcH1, and entropies, DcS1, for this ligand with the
halides in acetonitrile are listed in Table 3. At this stage, it is
relevant to mention that the results obtained from titration
calorimetry (Tronac 450) were corroborated by titration micro-
calorimetry in the case of Hg(II) and excellent agreement
was found.
value reported in the literature,8 which was 342.17 S cm2 molꢁ1m
.
As pointed out previously, the conductance data reported in
Fig. 2 and 3 are good enough to determine the stoichiometry
of the process and to give semiquantitative information about
the complex stability, but the equipment used is not sensitive
enough to derive very accurate stability constant data. Therefore,
we selected the calorimetric technique for the determination of
accurate stability constant data.
Inspection of the data shows that the enthalpy and entropy
contribute favorably to the complex stability of 1 with halides
and this ligand is able to selectively recognize these anions in
the following sequence:
Fꢁ 4 Clꢁ 4 Brꢁ
Fig. 2 Conductometric titration curves of fluoride (a), chloride (b),
bromide (c) anions (as tetra-n-butyl ammonium counter ion) with 1 in
MeCN at 298.15 K.
This is in accordance with the chemical shift changes observed
in the 1H NMR spectra for the pyrrolic proton. This statement
ꢀc
This journal is the Owner Societies 2010
758 | Phys. Chem. Chem. Phys., 2010, 12, 753–760

View Article Online
Table 3 Thermodynamic parameters of complexation of halide anions and mercury with various ligands in acetonitrile at 298.15 K
Ion
Fꢁ
Receptor
log Ks
DcG1/kJ molꢁ1
DcH1/kJ molꢁ1
DcS1/J molꢁ1 Kꢁ1
1
2
1
1
1
3
4
5.0 ꢄ 0.4
3.3a
ꢁ28.7 ꢄ 0.2
ꢁ19.2a
ꢁ20.2 ꢄ 0.5
ꢁ81.1a
29
ꢁ205a
Clꢁ
3.3 ꢄ 0.7
2.2 ꢄ 0.1
3.3 ꢄ 0.1
5.64a
ꢁ18.7 ꢄ 0.4
ꢁ12.7 ꢄ 0.1
ꢁ18.9 ꢄ 0.2
ꢁ32.2a
ꢁ3.1 ꢄ 0.2
ꢁ1.8 ꢄ 0.2
ꢁ32.8 ꢄ 0.1
ꢁ118.5a
52
36
Brꢁ
Hg2+
ꢁ48
ꢁ288a
5b
5.90b
ꢁ33.7b
ꢁ32.3b
a
b
Data from ref. 4. Data from ref. 8.
is corroborated by the linear relationship (r2 = 0.96) found
between the stability constants of 1 and halide anions in
acetonitrile at 298.15 K and the chemical shift changes for
the pyrrolic proton as shown in Fig. 4.
A plot of DcG1 values for 1 and anions (Fꢁ, Clꢁ and Brꢁ) in
acetonitrile against the DtG123 of these anions moving from
acetonitrile to water is shown in Fig. 5. These data are in
accordance with the chemical shift changes observed in the
1H NMR studies and reinforce the statements made in that
section regarding the correlation between the electronegativity
of the ions and the strength in hydrogen bond formation
between 1 and the halide anions.
different from those of 3 and this anion in this solvent. While 3
follows the pattern expected from the combination of two
reactants to give one product (loss of entropy), this is not the
case for 1. In fact, the considerable loss of enthalpic stability
and the gain in entropy which accompany the formation of the
1-Fꢁ relative to the 3-Fꢁ complex is typical of processes in
which a strong desolvation occurs upon complexation. The higher
stability of 1-Fꢁ relative to 3-Fꢁ results from the more favourable
entropy contribution of the former relative to the latter.
The ability of 1 to discriminate among anions can be assessed
from the selectivity factor, SFXꢁ , defined as the ratio between the
ꢁ
thermodynamic stability constant of 1 and the fluoride anion
Thus, the interaction of the halides with 1 in acetonitrile is
enthalpically and entropically favoured. However, for the
fluoride anion and this ligand in this solvent, the process is
enthalpically controlled. For chloride and bromide this is
entropically controlled. For comparison purposes, data for
the fluoride anion and 3 in acetonitrile are also included in
Table 3. Enthalpy and entropy contributions to the stability of
the complex between the fluoride anion and 1 are dramatically
relative to chloride and bromide in acetonitrile (eqn (7)).
ꢁ
KsðFꢁÞ
SFXꢁ
¼
ð7Þ
KsðXꢁÞ
These calculations indicate that 1 is more selective for fluoride
relative to chloride and bromide in acetonitrile by factors of
50 and 630, respectively.
As far as Hg2+ is concerned, thermodynamic data for the
complexation of this cation and 1 in acetonitrile at 298.15 K
are also listed in Table 3. The process is enthalpically controlled
but entropically disfavoured. Data for 3 and this cation are
included in this table, together with those for 4 and mercury(II)
in acetonitrile.
Comparison of these data with those for analogous ligands
(3 and 4) and this cation in this solvent shows that receptors 3
and 4 are able to form stronger complexes with Hg2+ than 1.
In fact, stability constant values for Hg2+-3 and Hg2+-4
are very similar. This is the result of an enthalpy-entropy
compensation effect.34 Thus, the higher enthalpic stability and
the greater entropy loss for 3 relative to 4 in their complexation
with the Hg2+ cation in acetonitrile may be attributed to the
reduced flexibility of the carbonyl oxygen in 3 relative to 4.
While the enthalpic stability upon complexation of Hg2+ with
1 is very close to that for 4 and this cation in acetonitrile, it is
the entropic parameter that is the decisive factor in the
different stabilities observed for this cation and these ligands
in this solvent.
Fig. 4 Linear relationship between log Ks and Dd values for the
pyrrolic proton of 1 at 298 K.
Conclusions
From the above discussion, the following conclusions can
be drawn:
(i) In assessing ion-selectivity in calixpyrrole chemistry, as
well as in the general area of macrocyclic chemistry, a great
deal of experimental work can be avoided by an appropriate
Fig. 5 Standard Gibbs energies of complexation of 1 and anions in
acetonitrile against the transfer Gibbs energy of ions moving from
acetonitrile to water (data based on the Ph4AsPh4B convention).
ꢀc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 753–760 | 759

View Article Online
selection of the solvent. In doing so, the experimental design,
as well as the conclusions, should make use of previous work
in this area, particularly the extensive research on the solution
properties of electrolytes and the thermodynamics of ion
complexation involving other macrocycles. Calixpyrroles do
not stand on their own. The effect of the solvent is expected,
since it is mainly the result of the solvation changes that the
species participating in the binding process moves from one
solvent to another.
7 A. F. Danil de Namor and R. Khalife, J. Phys. Chem. B, 2008,
112(49), 15766.
8 A. F. Danil de Namor, A. Aguilar-Cornejo, R. Soualhi,
M. Shehab, K. B. Nolan, N. Ouazzani and L. Mandi, J. Phys.
Chem. B, 2005, 109, 14735.
9 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, Pergamon Press, Ltd, Oxford, UK, 2nd edn, 1988.
10 L. Bonomo, E. Solari, G. Toraman, R. Scopelliti, M. Laatronico
and C. Floriani, Chem. Commun., 1999, 2413.
11 P. Anzenbacher, Jr, K. Jursikova, V. M. Lynch, P. A. Gale and
J. L. Sessler, J. Am. Chem. Soc., 1999, 121, 11020.
12 J. J. Christensen, R. M. Izatt and L. D. Hansen, Rev. Sci. Instrum.,
1965, 36, 779.
(ii) Once again, the selective behaviour of calix[4]pyrrole
and its derivatives for anions and the factors contributing to it
have been further corroborated.
(ii) Receptors 1 and 3 interact with Hg2+, while other
13 E. Wilson and D. F. Smith, Anal. Chem., 1969, 41, 1903.
14 A. F. Danil de Namor and M. Shehab, J. Phys. Chem. B, 2005,
109, 17440.
15 A. F. Danil de Namor and I. Abbas, J. Phys. Chem. B, 2007, 111,
5803.
cations tested are discriminated against.
(iii) Receptor 1 can be classified as a ditopic receptor given
its interaction with an anion (Fꢁ) and a cation (Hg2+).
Attempts to establish a cooperative effect of cation and anion
by carrying out experimental work with HgF2 is being
considered. This is by no means a trivial issue. There are
experimental limitations. An obvious one is that high concen-
trations of HgF2 are required, given that in this particular
case, the first ion (if dissociation of HgF2 takes place at all in
acetonitrile) to complex will attract its counter-ion (ion–ion
interactions are stronger than ion–dipole or hydrogen bond
formation) given that ion pair formation will occur. In addition,
the use of high concentrations of HgF2 may lead to deproto-
nation of the calixpyrrole and the formation of HFꢁ as
discussed above. Other approaches are now being considered.
16 A. F. Danil de Namor, M. Shehab, R. Khalife and I. Abbas,
J. Phys. Chem. B, 2007, 111, 12177.
17 A. F. Danil de Namor, I. Abbas and H. Hammud, J. Phys. Chem.
B, 2007, 111, 3098.
18 B. G. Cox and H. Schneider, Coordination and Transport Properties of
Macrocyclic Compounds in Solution, Elsevier, New York, 1992.
19 A. F. Danil de Namor, R. M. Cleverley and M. L. Zapata-
Ormachea, Chem. Rev., 1998, 98, 2495.
20 A. F. Danil de Namor, in Calixarenes 2001, ed. Z. Asfari,
V. Bohmer, J. Harrowfield and J. Vicens, Kluwer Academic
¨
Publishers, Dordrecht, Netherlands, 2001, ch. 19.
21 A. F. Danil de Namor and I. Abbas, Calixpyrrole-Fluoride Inter-
actions. From Fundamental Research to the Applications in the
Environmental Field, in Advances in Fluorine Science, ed.
A. Tressaud, Elsevier Science, New York, 2006, vol. II, ch. 3.
22 A. K. Covington and T. Dickinson, Physical Chemistry of Organic
Solvent Systems, Plenum Press, New York, 1973.
23 Y. Marcus, Ion Solvation, Wiley, Chichester, 1985.
24 O. Popovych and R. Tomkins, Nonaqueous Solution Chemistry,
John Wiley & Sons, New York, 1981.
25 A. F. Danil de Namor, R. Traboulssi, F. Fernandez Salazar,
V. Dianderas de Acosta, I. Fernandez de ViVizcardo and
J. Munoz Portugal, J. Chem. Soc., Faraday Trans. 1, 1989, 85,
2705.
26 Y. Marcus and G. Hefter, Chem. Rev., 2006, 106, 4585.
27 F. P. Schmidtchen, Org. Lett., 2002, 4, 421.
Acknowledgements
The authors thank the European Commission for the financial
support provided to carry out this work under Contract
INCO-CT-2004-509159.
28 W. Dehaen, P. A. Gale, S. E. Garcia-Garrido, M. Kostermans and
M. E. Light, New J. Chem., 2007, 31, 691.
References
29 M. Biocchi, L. Boca, D. E. Go
E. Monzani, J. Am. Chem. Soc., 2004, 126, 16507.
30 A. B. Descalzo, K. Rurack, H. Weisshoff, R. Martinez-Ma
M. D. Marcos, P. Amoros, K. Hoffmann and J. Soto, J. Am.
Chem. Soc., 2005, 127, 184.
´
mez, L. Fabbrizzi, M. Licchelli and
1 A. Baeyer, Ber. Dtsch. Chem. Ges., 1886, 19, 2184.
2 C. Stern, F. Franceschi, E. Solari, C. Floriani, E. Nazzareno and
R. Scopelliti, J. Organomet. Chem., 2000, 593–594, 86.
3 P. A. Gale, J. L. Sessler, V. Kral and V. Lynch, J. Am. Chem. Soc.,
1996, 118, 5140.
4 A. F. Danil de Namor and M. Shehab, J. Phys. Chem. B, 2003,
107, 6462.
5 A. F. Danil de Namor and M. Shehab, J. Phys. Chem. A, 2004,
108, 7324.
´
nez,
´
31 R. Nishiyabu and P. Anzenbacher, Jr, Org. Lett., 2006, 8(3),
359.
32 B. G. Cox, G. R. Hedwig, A. J. Parker and D. W. Watts, Aust. J.
Chem., 1974, 27, 477.
33 A. F. Danil de Namor, S. Chahine, D. Kowalska, E. E. Castellano
and O. E. Piro, J. Am. Chem. Soc., 2002, 124, 12824.
34 A. F. Danil de Namor, S. Chahine, E. E. Castellano and
O. E. Piro, J. Phys. Chem. B, 2004, 108, 11384.
6 J. L. Sessler, D. E. Gross, W.-S. Cho, V. M. Lynch,
F. P. Schmidtchen, G. W. Bates, M. E. Light and P. A. Gale,
J. Am. Chem. Soc., 2006, 128, 12281.
ꢀc
This journal is the Owner Societies 2010
760 | Phys. Chem. Chem. Phys., 2010, 12, 753–760