Non-trivial solution chemistry between amido-pyridylcalix[4]arenes and some metal salts

Article abstract of DOI:10.1039/c0dt00976h

Mercury ion complexation reactions were carried out between 3 and various mercury(ii) salts. ¹H NMR studies showed that the role of solvent, the anion chosen and the initial reaction conditions were critical and that the formation of a "simple" mercury(ii) complex was non-trivial. The mercury(ii) ion can cause either (i) the formation of an ion-pair system, which have a characteristic doubling of all signals in the ¹H NMR spectrum, (ii) a cleavage reaction to occur resulting in the reformation of the calix[4]arene diester compound 2, but only when the reaction is heated and (iii) "simple" mercury binding to the pyridine rings when the binding studies are carried out using NMR titration techniques. The electrochemistry results, on the same systems, show that the initial reaction involves the removal of the phenoxide protons followed by the resulting catalysis of the mercury species. This proton removal is not observed in the NMR spectra of any of the mercury reactions. It was also found that 3 could bind silver and zinc salts and was not selective for mercury(ii) as was previously described. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c0dt00976h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Non-trivial solution chemistry between amido-pyridylcalix[4]arenes and some
metal salts†
John J. Colleran,a Bernadette S. Creaven,b Denis F. Donlona,b and John McGinley*a
Received 6th August 2010, Accepted 1st September 2010
DOI: 10.1039/c0dt00976h
Mercury ion complexation reactions were carried out between 3 and various mercury(II) salts. 1H NMR
studies showed that the role of solvent, the anion chosen and the initial reaction conditions were critical
and that the formation of a “simple” mercury(II) complex was non-trivial. The mercury(II) ion can
cause either (i) the formation of an ion-pair system, which have a characteristic doubling of all signals
in the 1H NMR spectrum, (ii) a cleavage reaction to occur resulting in the reformation of the
calix[4]arene diester compound 2, but only when the reaction is heated and (iii) “simple” mercury
binding to the pyridine rings when the binding studies are carried out using NMR titration techniques.
The electrochemistry results, on the same systems, show that the initial reaction involves the removal of
the phenoxide protons followed by the resulting catalysis of the mercury species. This proton removal is
not observed in the NMR spectra of any of the mercury reactions. It was also found that 3 could bind
silver and zinc salts and was not selective for mercury(II) as was previously described.
on the lower rim, attached via an amide bond, was interesting as
Introduction
it suggested that the calix[4]arene derivative selectively complexed
The last decade has witnessed a surge of interest in the discovery of
mercury ions. This is contrary to what we have observed and in this
building blocks capable of forming specific molecular arrays under
paper we report the reactions of the same calix[4]arene derivative
certain chemical conditions.1–4 Calix[4]arenes are currently among
with mercury, silver and zinc salts and the subsequent ion pair
the most versatile and useful building blocks in supramolecular
interactions.
chemistry,5–14 as they are intriguing platforms for constructing
selective receptors because of their preorganised basket structure
Results and discussion
assembled from their four phenol rings. They are widely used
for the creation of selective metal cation extractants (including
radionuclides), catalysts and sensor materials and they have been
functionalised by various receptor groups, which can selectively
bind cations, anions, and neutral molecules. Calix[n]arenes (n = 4
or 6) have been used as three-dimensional receptors with multiple
urea, thiourea or amido arms on either the wide or narrow rim to
bind ion pairs in low polarity solvents.15 Calix[4]arene pyridyl
ligands have been reported previously with methanol solvent
molecules contained within the lattice being hydrogen bonded
to the pyridyl nitrogen atoms16 and with the methanol molecule
enclathrated within the calix[4]arene cone-shaped cavity.17 As part
of our continuing research, we have previously reported upper
and lower rim functionalised calix[4]arenes, capable of binding
transition metal salts.18–21 In extending this theme, we targeted
the design of “pre-ligands”, capable of having two or more donor
atoms present, with a view to attaching these to the lower rim of the
calix[4]arene and have reported the copper(II) complexes of two
such ligands.22,23 A recent paper by Rao and co-workers24 on the
formation of calix[4]arene scaffolds containing pyridine ligands
Ligand synthesis and characterisation
The calix[4]arene diester compound 219,25 was heated to reflux for
120 h with 2-(aminomethyl)pyridine in a methanol–toluene sol-
vent mixture (Scheme 1), similar to that previously described.18,26
The reaction yielded the amide 3 as a white solid in good yield.
Compound 3 has recently been reported by Rao using a different
preparative route,24 and the 1H NMR data are comparable in both
1
cases. The H NMR spectrum of 3 consists of two signals being
observed for both the aromatic protons of the calix[4]arene and
the tert-butyl groups, indicative of 1,3-disubstitution at the lower
rim, as well as a singlet for the hydroxyl groups on the lower rim.
Cone conformation of the derivatised calix[4]arene was confirmed
in all cases by the presence of two sharp doublets at approximately
4.4 and 3.3 ppm. The pendant arms attached to the calix[4]arenes
1
gave the expected H NMR data. Four signals in the aromatic
region are observed for the substituted pyridine ring.
Crystals of 3 suitable for X-ray crystallographic study were
obtained from methanol, and the structure was confirmed to be
identical to that published by Rao and co-workers.24
aChemistry Department, National University of Ireland Maynooth,
Maynooth, Co. Kildare, Ireland. E-mail: john.mcginley@nuim.ie; Fax: +353
1 708 3815; Tel: +353 1 708 4615
Complexation reactions
bDepartment of Science, Institute of Technology Tallaght Dublin, Dublin
24, Ireland. E-mail: bernie.creaven@it-tallaght.ie; Fax: +353 1 404 2700;
Tel: +353 1 404 2889
The calixarene derivative 3 was reacted with mercury(II), silver(I)
and zinc(II) salts (HgX2, AgX, ZnX2; X = perchlorate, thiocyanate
and chloride) in a 1 : 1 metal to ligand ratio in ethanol at room
temperature for 2 h. Precipitation did not occur in any case,
† Electronic supplementary information (ESI) available: Tables containing
1H NMR complexation data. See DOI: 10.1039/c0dt00976h
10928 | Dalton Trans., 2010, 39, 10928–10936
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Scheme 1 Reagents and conditions: i) ethyl bromoacetate, K2CO3, MeCN, D, 18 h; ii) 2-(aminomethyl)pyridine, MeOH, toluene, 120 h.
and following the removal of the solvent, the remaining white
solids were characterised by both NMR and IR spectroscopies
(see Experimental section). When samples were left for several
days, it was only on evaporation of the entire solution did a solid
appear.
Electrospray (ESI) mass spectrometry of 3 and it’s metal
complexes were obtained in either MeCN or CHCl3 as solvent.
In all cases, only the spectrum of 3 was obtained, with either
sodium or potassium ions attached (see Experimental section). No
evidence of a mercury, silver or zinc complex could be obtained.
amide resulting in a weakening of the hydrogen bonding between
the amide hydrogen and the phenolic oxygen. We also obtained the
1H NMR spectrum in CDCl3. While, all the signals present were
well-defined, the spectrum, shown in part in Fig. 2, was neither
that of the calixarene derivative 3 on its own, nor did it show simple
mercury complexation (as observed in Fig. 1), as there are now
double the number of signals that should be present, for the latter
explanation. This means that either there are either two different
species present in solution or that the complexation of the mercury
ion is giving rise to a single highly asymmetric species in solution,
as indicated below.
Removal of the solvent under reduced pressure and redissolving
the sample in CD3CN or d6-DMSO resulted in spectra similar
to that shown in Fig. 1, with the obvious difference of solvent
peaks. When a sample of the mercury(II) perchlorate complex was
NMR studies
1
The H NMR spectrum of the Hg(ClO4)2 complex of calixarene
3 was run in both d6-DMSO and CD3CN. Both the d6-DMSO
and the CD3CN spectra showed the presence of a single set of
signals and, based on the shifts of the signals in comparison to the
starting calixarene derivative 3, the species present was identified
as the mercury(II) complex (see Fig. 1). In both solvents, there
are significant shifts in the signals related to the pyridyl protons,
indicating binding of the mercury(II) cation at this position.
Furthermore, there are also shifts in the signals for both the
amide protons and the hydroxyl protons on the lower rim of the
calixarene, indicative of the binding of the mercury(II) cation to
the pyridine rings, causing a change in the electron density at the
1
run in CD2Cl2, the resulting H NMR spectrum also contained
a doubling of all signals, similar to that shown in Fig. 2. A
plausible suggestion, based on these spectral results, is that the
doubling of the spectral signals in CDCl3, and also in CD2Cl2, is
due to the lack of solvation of the perchlorate anion, something
which is happening in the donor solvents. We therefore believe
that the mercury(II) cation is binding to the pyridyl arms of
the functionalised calix[4]arene, thereby separating itself from the
perchlorate anion. When this happens in d6-DMSO or CD3CN,
then the anion is solvated by donor solvent molecules, whereas
Fig. 1 300 MHz 1H NMR spectrum of mercury(II) perchlorate complex in CD3CN from 6.25–11 ppm.
The Royal Society of Chemistry 2010 Dalton Trans., 2010, 39, 10928–10936 | 10929
This journal is
©

View Article Online
Fig. 2 500 MHz 1H NMR spectrum of mercury(II) perchlorate complex from 6.25–11 ppm and from 6.25–2.25 ppm in CDCl3.
when the solvent is CDCl3 or CD2Cl2, no solvation occurs due
to the poor donating ability of these solvents and the resulting
close ion-pairing induces a high degree of asymmetry in the
complex.
within coupling distance is now involved in a coupling process.
For example, the methylene group attached to the phenoxide
oxygen and the carbonyl carbon at the upper end of the pendant
arm is a singlet in the 1H NMR spectrum of 3 at 4.56 ppm
but now both protons couple to each other to give a pair of
doublets at 4.17, 3.35, 2.83 and at 2.53 ppm. It is difficult to see
how the mercury(II) ion induces asymmetry into the calix[4]arene
molecule but the complexity of the NMR spectrum confirms
this asymmetry. Several samples of the perchlorate complex, in
both CHCl3 and CDCl3 solution, were left aside in an effort to
Closer inspection of the NMR spectra in ion-pairing solvents
reveals the following: for the methylene protons, there are now
four sets of doublets in the 6.25–2.25 ppm region, as shown in Fig.
2, which would indicate that each proton on the methylene bridge
1
is in a unique environment. Assignment of each signal in the H
NMR spectrum, obtained in CDCl3, has found that every proton
10930 | Dalton Trans., 2010, 39, 10928–10936
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 3 300 MHz 1H NMR spectra measured at 25 ◦C during the titration of 3 with mercury(II) perchlorate in d6-DMSO.
obtain suitable crystals for an X-ray structural determination.
Where crystals were obtained, they analysed as the starting
calix[4]arene 3.
As titration studies in CDCl3 and CD2Cl2 were not successful
it was decide to try to isolate the mercury(II) complexes under
different conditions to see if a number of species existed in solution.
Varying the time of reaction, temperature or solvent did not result
in complexes of different conformation or symmetries but did
give rise to an unusual outcome. The initial reaction between
mercury(II) perchlorate and 3 was also repeated in ethanol but
this time at the reflux temperature of the solvent for 2 h, instead
of being at room temperature for 2 h. On cooling, a white solid
precipitated from the reaction mixture. Analysis of the 1H NMR
spectrum, using CDCl3 as solvent, showed that the calix[4]arene
derivative has lost both pendant pyridyl arms and that the cleavage
has occurred between the amide carbonyl and nitrogen atoms.
The calix[4]arene derivative that is obtained as a result of these
cleavage reactions is the calix[4]arene diester 2. It is not clear why
the cleavage should occur under such mild conditions as heating
to reflux temperature in ethanol. One possible mechanism which
would allow the cleavage reactions to occur is outlined in Scheme
2. The ethyl ester which is obtained results from the choice of
ethanol as solvent.
Titration studies between 3 and mercury(II) perchlorate in
various deuterated solvents, particularly in CDCl3, might be
informative about the number of species present in solution. So,
these studies were undertaken using various concentrations of
metal salt to one equivalent of 3. The titration study between 3 and
mercury(II) perchlorate in CDCl3 could not be carried out because
of the poor solubility of mercury(II) perchlorate in the solvent, even
at the low concentrations required for the titration. The titration
studies between mercury(II) perchlorate and 3 in CD3CN and d6-
DMSO at variable temperatures and concentrations were carried
out and showed, in all cases, that metal binding was observed.
In either solvent, the 1H NMR spectra showed no change in
the cone conformation. It was observed that the chemical shift
separation of the proton signals of the methylene bridge in 3
decreased, suggesting that the ligand adopts a more symmetrical
cone conformation upon complexation. However, considerable
shifts in the amide, hydroxyl and pyridyl proton signals were
observed in the spectra. Fig. 3 shows the spectra obtained when 1
equivalent of 3 is titrated with various concentrations of Hg(ClO4)2
up to 2.5 equivalents of metal salt at 25 ◦C. Both the pyridyl
and amide protons showed the largest downfield shift, indicating
that the Hg(II) cation mainly interacts with both the pyridyl
and amide parts of the calix[4]arene derivative. Such downfield
shifts have been previously reported for other mercury(II)-bound
complexes.24,27–29
Based on our room temperature reactions and binding studies,
we believe that the reaction between 3 and the mercury(II)
perchlorate salt results initially in the formation of the cationic
mercury(II)-calix[4]arene complex with perchlorate anions present
in solution. For the purposes of clarity, the mercury ion is only
bound to one pyridine ring in the Scheme. The mercury ion is
also involved in an interaction with the amide carbonyl oxygen
atom. As a result of this interaction, an ethanol molecule can
Scheme 2 Possible mechanism for cleavage reaction. The mercury ion is not shown as being charged in the Scheme.
The Royal Society of Chemistry 2010 Dalton Trans., 2010, 39, 10928–10936 | 10931
This journal is
©

View Article Online
attack the carbonyl carbon atom, resulting in the bonding of the
alcohol to the carbonyl carbon atom. Loss of a proton followed
by a rearrangement of the electrons gives the amide carbonyl
group back, which is still bonded to the mercury ion, and also
cleaves the C–N amide bond to give the ester and, on protonation,
the 2-aminomethylpyridine compound. The ethanol molecule
is only shown attacking one pendant arm of the calixarene
derivative in Scheme 2, but attack does occur at both amide
carbonyls. If this mechanism is feasible, then replacing ethanol
by methanol as solvent should result in the formation of the
methyl ester calix[4]arene derivative. This reaction was carried
out using methanol as solvent and heating the mixture to the
reflux temperature of methanol for two hours. The resulting white
Fluorescence studies
The binding of zinc(II) perchlorate and silver(I) perchlorate to
3 contradicts the findings of Rao and co-workers who published
their work on the binding of metal(II) salts to various amide linked
derivatives of calix[4]arene.24 In the course of their work, they
studied the interaction of eleven different divalent metal ions, in-
cluding both Hg(II) and Zn(II), with 3 in 50% aqueous acetonitrile
by fluorescent spectroscopy. They reported that only in the case
of the Hg(II) cation were any significant changes in fluorescence
observed. We decided to use fluorescence spectroscopy to see if we
could obtain similar results to Rao and co-workers, as there may be
sensitivity issues between 1H NMR spectroscopy and fluorescence
spectroscopy. We dissolved 3 in 50% aqueous acetonitrile to obtain
the same concentration (10 mmolar) as used by Rao and co-
workers and then ran it’s fluorescence spectrum. We also made up
a 10 mmolar 1 : 1 solution of 3 and Hg(ClO4)2 and also obtained
it’s spectrum. Both spectra showed weak fluorescence but it was
not possible to observe a difference in fluorescence between the
two spectra, as observed by Rao and co-workers.24 We therefore
decided to repeat the fluorescence studies using a tenfold increase
in concentration. Using the 100 mmolar solution of 3, we observed
the expected increase in absorption due to the concentration
effect. The spectrum of the metal complex showed an obvious
decrease in fluorescence (see Fig. 4), as expected when using a
mercury(II) salt. When a second equivalent of Hg(ClO4)2 was
added, the fluorescence showed a further decrease. This confirmed
the observations of Rao, although it was at a higher concentration.
We have also found a similar decrease in fluorescence when we
carried out titration studies of 3 with zinc(II) perchlorate.
1
solid was analysed by H NMR spectroscopy and revealed that
the methyl ester derivative was indeed formed, which suggested
that our proposed mechanism was feasible. Furthermore, when
we repeated the reaction in either ethanol or methanol but this
time reduced the reaction time to 1 h, the solid obtained had lost
one pendant arm only.
The reaction was also repeated using perchloric acid instead of
mercury(II) perchlorate in order to see was the presence of the
mercury(II) ion essential for the cleavage reaction to occur. The
reaction was carried out for two hours at both room temperature
and at the reflux temperature of ethanol. Once the excess solvent
had been removed under vacuum, the residue in both cases was
analysed using 1H NMR spectroscopy using CDCl3 as solvent. In
both cases, significant shifts in the pyridyl signals were observed,
which suggested that the corresponding pyridinium cations were
being formed. The differences in the chemical shift values on
going from the pyridine derivative 3 to the pyridinium cation were
similar to those observed and reported by Beer and co-workers.30
Therefore, the presence of the mercury(II) ion is essential for the
cleavage reaction to occur.
Studies with other metal ions
Although Rao and co-workers24 had reported that 3 was selective
for mercury(II), we were interested in studying the complexation
reactions with other metal ions to see if the unusual behaviour
displayed by the mercury(II) complexes of 3 was evident in other
complexes. The reaction of 3 with either silver(I) perchlorate
or zinc(II) perchlorate was carried out for two hours at both
room temperature and at the reflux temperature of ethanol. The
resulting white solids, obtained from each reaction after removal
1
of solvent in vacuo, were analysed using H NMR spectroscopy
Fig. 4 Fluorescence titration spectra measured during the titration of 3
with mercury(II) perchlorate complex in 50% aqueous acetonitrile solution;
dark blue line = 3, green line = 3 + Hg(ClO4)2 (1 : 1), red line = 3 + Hg(ClO4)2
(1 : 2), pale blue line = 3 + Hg(ClO4)2 (1 : 5).
using CDCl3 as solvent. In all cases, binding of either the silver(I)
ion or the zinc(II) ion to the pyridine ring was observed, but
without the doubling of peaks indicative of the presence of an
ion-pair either, as shown in Fig. 1 for the mercury(II) case, or the
cleavage reaction occurring, as observed when the reaction with
mercury(II) perchlorate was carried outathigh temperature for two
Electrochemistry studies of mercury(II) and silver(I) complexes
1
hours. H NMR titration studies were subsequently undertaken
between 3 and silver(I) perchlorate in d6-DMSO respectively.
Binding of the silver(I) ion to the pyridine ring was observed,
which is similar to the binding pattern observed when using
mercury(II) perchlorate, suggesting that the mercury(II) and sil-
ver(I) cations interact with the calix[4]arene derivative in the same
manner.
The electrochemical behaviour of 3, and metal binding to 3, was
investigated in acetonitrile, chosen due to its inertness within
a large potential window,31 with 0.1M TBAPF6 as supporting
electrolyte. Cyclic voltammetry (CV) was performed on a So-
lartron potentiostat Model 1285, while a PAR Model 636 was
employed to perform RDE Voltammetry (RDV). The working
10932 | Dalton Trans., 2010, 39, 10928–10936
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 1 Summary of parameters obtained for the different analytes
proposed ECE mechanism. Our values differ from those obtained
for diamide-calix[4]arene,32b which differs from 3 only by the two
non-phenolic pendants. This would indicate, unexpectedly, that
the two pyridyl pendants impose a bearing on the oxidation
process at the phenolic groups. The origin of the cathodic peak
was determined by cycling 3 from +0.8 to -1.0 V, that is to
potentials that did not oxidise 3. The reduction peak at -0.48 V was
absent indicating that the reduction peak is an electrochemically
generated daughter peak of the oxidation process (not shown). All
voltammograms exhibit a reduction peak at +0.35 V, which has
previously been reported as the reduction of protons at Pt.34 To
investigate whether the redox behaviour of 3 was indeed due to
the hydroxyl groups, potential cycling of a 1 ¥ 10-3 M solution
of the parent tetramer (only hydroxyl groups present) was carried
out (Fig. 6, red trace). Both the pre-peak and the larger oxidation
peak are observed at around the same potentials. The slightly
lower oxidation potentials and smaller oxidation peak currents are
observed due to the poor solubility, and thus lower concentration,
of the tetramer in MeCN.
Cyclic voltammetry
Analyte (1 ¥ 10-3 M)
E
p,ox/V
Potential shift/V
3
1.490
1.749
1.620
—
+0.259
+0.130
3-Hg(ClO4)2
3-AgClO4
electrodes were platinum disks (CV, A = 0.0314 cm2; RDV, A =
0.1257 cm2) and platinum wire counter electrodes. Solutions were
deaerated by purging with N2 for 15 mins prior to experiments. All
potentials are quoted versus a Ag/Ag+ reference electrode (E1/2
=
+0.075 V vs. FeCp2), which comprised of a silver wire sealed in
glass, containing 0.01 M AgNO3 and 0.1 M TBAPF6 in MeCN.
Concentrations of 3, and all the metal salts were analysed in 1 ¥
10-3 M concentrations. The voltammograms were swept anodically
(unless otherwise stated) in the potential window -1.0 to +1.9 V at
a scan rate of 100 mV s-1. Ep values and other derived parameters
are summarised in Table 1.
Electrochemistry of 3
The hydroxyl groups on the calix[4]arene derivative are the
prime candidates to infer 3 with electroactivity. Typical cyclic
voltammograms for a 1 ¥ 10-3 M solution of 3, shown in Fig. 5
(black trace), display a small oxidation peak at +1.12 V, a second
very well defined oxidation peak at ca. +1.49 V, and a reduction
peak at ca. -0.48 V. Both anodic peaks represent the oxidation
and accompanying deprotonation of the hydroxyl group on 3, as
has been observed in other studies performed on disubstituted
calix[4]arenes.32 An ECE mechanism has been suggested for this
process and a plot of Ip/n1/2 versus n supports this.33 RDV was used
to analyse the main anodic peak. The slope of the plots of E versus
ln(i/id-i) can be used to determine the oxidation stoichiometry.
A linear plot was observed with a slope of 92 mV dec-1. This
lies midway between the theoretical Nernstian predicted slope
value for one and two electron processes. Our value would suggest
that the first electron transfer is not solely rate determining, and
that deprotonation of the hydroxide groups also contributes to
the kinetic process. This observation would seem to support the
Fig. 6 Typical cyclic voltammograms detailing the redox behaviour of
Hg(ClO4)2 (1 ¥ 10-3 M) in MeCN at a Pt working electrode. Supporting
electrolyte was 0.1 M TBAPF6 and scan rate was 100 mV s-1.
The daughter reduction peak was not evident in the voltammet-
ric profile of the tetramer, which may indicate that the pyridyl
pendant groups promote the generation of this species. The
oxidation peak of compound 3 has a diffusion-like character and
a scan rate analysis was performed to investigate. The resultant
Randles–Sevcik plot, shown as inset in Fig. 5, was linear and
had a slight non-zero intercept. The non-zero intercept was not
unexpected due to the fact that only the second oxidation peak was
considered in the analysis. This is typical of R–S behaviour and
thus the oxidation of 3 was deemed to be under diffusion control.
Redox behaviour of Hg(ClO4)2 in MeCN
Mercury(II) is well known to undergo a reversible two electron
reduction to form the insoluble mercury(0). This transition leads
to the formation of thin films on electrode surfaces and is described
by Reaction 1:
Hg(II) + 2e- ↔ Hg0
(1)
Fig. 5 Typical cyclic voltammograms detailing the redox behaviour 1 ¥
10-3 M of 3 (black trace), tetramer (red trace) and MeCN (blue trace) at a
Pt working electrode. Supporting electrolyte was 0.1 M TBAPF6 and scan
rate was 100 mV s-1. Inset: Randles–Sevcik plot.
The corresponding oxidation is evident in cyclic voltammo-
grams as a sharp stripping peak which causes the mercury
dissolution. We chose to study the electrochemical properties of
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10928–10936 | 10933
©

View Article Online
Hg(ClO4)2 and its interaction with 3 because Hg(ClO4)2 dissociates
readily into Hg2+ cations and ClO4- anions in MeCN. Therefore the
solution should be rich in Hg2+ ions, as is evident in the first cyclic
voltammetric cycle outlined in Fig. 6 (dotted trace). We observe a
sharp oxidation peak at +0.59 V, corresponding to the stripping
of Hg0 from the Pt surface. The coupled reduction peak, Hg2+ to
Hg0, lies at +0.15 V but progressive scans document a shift to
more cathodic values, which settle out at 0.07 V. The second peak
associated with the stripping process,35,36 similar in height to the
original peak, appears in cycle 2 at +0.51 V. Both stripping peaks
decrease in magnitude with cycle number, while the reduction of
Hg2+ to the insoluble Hg0 remains relatively stable.
of electrons and protons from the hydroxide moieties, at the Pt
electrode surface, then becomes kinetically more sluggish.
Additional energy (~290 mV - anodic shift) is then required to
oxidise the newly formed complex. Bound mercury ions have not
affected diffusion of this complex to the electrode surface enough
to observe a decline of the oxidation peak currents of 3.
Other perchlorate salts
The effect of AgClO4 on the oxidation of 3 was also investigated
and little bearing on the electrochemistry of 3 was observed. Unlike
with Hg(ClO4)2, in the case of AgClO4, a Ag(0)–Ag(I) stripping
peak was present in all cyclic voltammetric potential sweeps (Cycle
1, black trace & Cycle 10, blue trace, Fig. 8). The electrogenerated
daughter peak appears after the first cycle and is observed in the
blue and red traces. In the presence of AgClO4, the oxidation
potential of 3 is shifted anodically by ~130 mV, again indicative
of a binding process. However, the magnitude of the Ag0 stripping
peak remained constant over successive potential cycles in the
presence of 3. This would indicate that if the binding of Ag+ occurs,
it is reversible i.e. an equilibrium process, and the generation of
Ag0 is not kinetically affected by complexation. We observed an
opposite effect with Hg(ClO4)2 where a lower concentration of
Hg2+ was free in solution to form the deposited Hg0 species upon
reduction. An equilibrium binding process may account for these
observations. The results presented here indicate that although
metallic ions do interact with 3 when highly dissociated in MeCN,
only mercuric species exhibit a dramatic electrochemical effect on
the oxidation of 3. The rigidity inferred on the complex through
binding should vary for the Hg2+ and Ag+ ions. Hg is over twice the
mass of Ag which may affect the magnitude of interaction strength
between the different metal ions and the pyridyl pendants. Varying
degrees of complex rigidity, inferred through the different species
of metal binding, should then be reflected in the different potentials
required to oxidise 3.
Electrochemistry of Hg(ClO4)2 - 3 complex
Due to the ready dissociation of Hg(ClO4)2 in MeCN, mercury will
exist as Hg2+ ions, which are known to favour interaction with soft
Lewis bases. The nitrogen on the pendant pyridine should be an
ideal binding site candidate, as proposed by NMR analysis. In the
presence of Hg(ClO4)2, we observe a relatively large anodic shift in
the oxidation potential of 3, which is indicative of mercury-ligand
binding (Fig. 7). The Ep,ox for 3 is present at +1.46 V on cycle
1, but shifts to +1.75 V thereafter. The same peaks that appear
in the parent Hg(ClO4)2 and ligand analyte solutions are again
evident in this solution mixture. But after cycle 1, the familiar
Hg0 stripping peak at ca. +0.46 V, indicative of mercury film
formation, is no longer evident, which further supports a mercury-
ligand binding process. The peak at +0.07 V, however, actually
increases in magnitude. Another interesting feature is the increase
in peak intensity at ca. +1.17 V with cycle number. One possibility
that could account for this observation is that the electrochemical
process, observed as the pre-peak, now becomes more pronounced
and facile in the presence of Hg(ClO4)2. But since we observe a
general increase in Hg2+ reduction, Hg0 is being oxidised, so we
tentatively assign the peak at +1.17 V to the combined oxidation
processes of both mercury (Hg0 to Hg2+) and ligand species. So,
in conclusion, the presence of Hg(ClO4)2 actually impedes the
oxidation of 3, possibly at the phenoxide sites, and may result
because of a binding process. These observations imply that if
binding occurs at the pyridyl pendant nitrogen, re-oxidation of
metallic mercury may be hindered. Binding of Hg2+ to both pyridyl
pendants may infer rigidity onto the mercury-calixarene complex
and free rotation of the pendant arms is no longer facile. Loss
Fig. 8 Cyclic voltammograms describing the electrochemical behaviour
of 1 : 1 solution mixtures of AgClO4 - 3, cycle 1 (solid black trace) and
AgClO4 - 3, cycle 10 (solid blue trace). The solid red trace is 1 ¥ 10-3 M 3
only. All data recorded in 0.1 M TBAPF6 in MeCN at a scan rate of 100
mV s-1.
Experimental
1H and 13C NMR (d ppm; J Hz) spectra were recorded on
either a JEOL JNM-LA300 FT-NMR spectrometer or a Bruker
Bruker Avance III 500 MHz spectrometer using saturated CDCl3
solutions with Me4Si reference, unless indicated otherwise, with
Fig. 7 The effect of Hg(ClO4)2 on the oxidation of 3 in MeCN and 0.1
M TBAPF6. The scan rate employed was 100 mV s-1.
10934 | Dalton Trans., 2010, 39, 10928–10936
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
resolutions of 0.18 Hz and 0.01 ppm for the 300 MHz ma-
chine and 0.005 Hz for the 500 MHz machine, respectively.
Infrared spectra (cm-1) were recorded as KBr discs or liquid
films between KBr plates using a Perkin Elmer System 2000 FT-
IR spectrometer. Melting point analysis was carried out using
a Stewart Scientific SMP 1 melting point apparatus and are
uncorrected. Microanalysis was carried out at the Microanalytical
Laboratory of either University College, Dublin, the National
University of Ireland Cork or the National University of Ireland
Maynooth. Electrospray (ESI) mass spectra were collected on an
Agilent Technologies 6410 Time of Flight LC/MS. Complexes
were dissolved in acetonitrile–water (1 : 1) solutions containing
0.1% formic acid, unless otherwise stated. The interpretation of
mass spectra was made with the help of the program “Agilent
Masshunter Workstation Software”. Standard Schlenk techniques
were used throughout. Starting materials were commercially
obtained and used without further purification. Caution! Although
not encountered in our experiments, perchlorate salts of metal ions
are potentially explosive and should be manipulated with care and
used only in small quantities. The synthesis of compound 2 has
been described in the literature previously.19,25
t, py-H), 9.04 (1H, t, py-H), 9.35 (1H, t, NHCH2), 9.61 (1H, d,
py–H), 10.12 (1H, d, py–H), 10.35 (1H, t, NHCH2); ESI-MS, m/z
(%): 945.54 [3 + H]+, 968.53 [3 + Na]+.
A mixture of 3 (0.05 g, 0.53 mmol) and mercury(II) perchlorate
(0.21 g, 1.06 mmol) in ethanol (25 ml) was heated to reflux while
stirring for 2 h. After cooling, the volatiles were removed under
pressure to give a crude solid which was then examined by NMR
spectroscopy. dH(300 MHz; CDCl3; Me4Si) 0.99 (9H, s, t-Bu), 1.02
(9H, s, t-Bu), 1.16 (6H, t, J = 7.1 Hz, OCH2CH3), 1.28 (18H, s, t-
Bu), 3.34 (2H, d, J = 13.1 Hz, ArCH2Ar), 3.40 (2H, d, J = 13.1 Hz,
ArCH2Ar), 3.99 (4H, q, J = 7.1 Hz, OCH2CH3), 4.23 (2H, d, J =
13.1 Hz, ArCH2Ar), 4.25 (2H, d, J = 13.1 Hz, ArCH2Ar), 4.62
(4H, s, OCH2CO), 4.92 (4H, br d, NHCH2), 6.86 (2H, s, Ar-H),
6.90 (2H, s, Ar-H), 7.07 (6H, m, Ar-H & calix-OH), 7.60 (3H, m,
py-H), 8.52 (1H, d, py-H), 9.24 (1H, br t, NHCH2).
NMR Reactions
A mixture of 3 (0.01 g, 0.01 mmol) and either a mercury salt
(i.e. mercury(II) perchlorate, mercury(II) chloride, mercury(II)
thiocyanate), a zinc salt (zinc(II) perchlorate) or a silver salt
(silver(I) perchlorate) (0.01 mmol) was dissolved in a deuterated
solvent (i.e. CDCl3, d6-DMSO, CD3CN or CD2Cl2). After shaking,
the mixture was placed into an NMR tube and analysed using
NMR spectroscopy. The resulting NMR data is presented in the
supplementary information.†
Synthesis of 3
A solution of 2 (0.5 g, 0.6 mmol) and 2-aminomethylpyridine
(0.58 g, 2.4 mmol) in a mixture of methanol (15 ml) and toluene
(15 ml) were allowed to stir at reflux temperature for 120 h.
After completion, the reaction solution was concentrated under
reduced pressure. The resulting solid residue was crystallised from
methanol, filtered and oven dried to yield the title compound.
NMR data for 3: dH(300 MHz; CDCl3; Me4Si) 0.99 (18H, s, t-Bu),
1.28 (18H, s, t-Bu), 3.35 (4H, d, J = 13.5 Hz, ArCH2Ar), 3.65
(3H, s, MeOH), 4.25 (4H, d, J = 13.5 Hz, ArCH2Ar), 4.56 (4H, s,
OCH2CO), 4.78 (4H, d, J = 6 Hz, NHCH2), 6.85 (4H, s, Ar-H),
7.06 (4H, s, Ar-H), 7.15 (2H, m, py-H), 7.43 (2H, s, calix-OH),
7.82 (2H, br t, py-H), 8.42 (2H, br d, py-H), 8.70 (2H, br d, py-H),
9.29 (2H, br t, NHCH2); ESI-MS, m/z (%): 945.55 [3 + H]+, 968.54
[3 + Na]+, 983.51 [3 + K]+.
NMR Titrations
Titrations are only described for d6-DMSO solutions, but titra-
tions in other solvents were carried out in a similar manner.
Standard solutions of 3 and the mercury(II) salts (mercury(II)
perchlorate, mercury(II) chloride and mercury(II) thiocyanate)
were prepared in d6-DMSO. These solutions were then diluted
to a concentration of 1 ¥ 10-4 M respectively. 0.5 ml of the
calixarene were then placed in an NMR tube and examined
1
using H NMR spectroscopy. 50 mm (0.2 Eq) of the appropriate
mercury(II) salt solution was then added to the NMR tube and
the resulting complex was examined as before. These titrations
continued until the complex formation had been examined with
0.2, 0.4, 0.6, 0.8, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 equivalents of
Hg salt present. The resulting NMR data is presented in the
supplementary information.†
Reaction of 3 with Hg(ClO4)2
A mixture of 3 (0.05 g, 0.53 mmol) and mercury(II) perchlorate
(0.21 g, 0.53 mmol) in ethanol (25 ml) was stirred at r.t. for 2 h.
Upon completion of the reaction the volatiles were removed under
pressure to give a crude solid which was then examined by NMR
spectroscopy. dH(500 MHz; CDCl3; Me4Si) 0.79 (9H, s, t-Bu), 0.91
(9H, s, t-Bu), 1.23 (9H, s, t-Bu), 1.28 (9H, s, t-Bu), 2.53 (1H, d,
J = 13.2 Hz, OCH2CO), 2.83 (1H, d, J = 13.0 Hz, OCH2CO),
3.35 (1H, d, J = 12.9 Hz, OCH2CO), 3.43 (1H, d, J = 14.6 Hz,
ArCH2Ar), 3.46 (1H, d, J = 14.6 Hz, ArCH2Ar), 4.02 (1H, d, J =
13.9 Hz, ArCH2Ar), 4.17 (1H, d, J = 12.8 Hz, OCH2CO), 4.36
(1H, d, J = 13.2 Hz, ArCH2Ar), 4.38 (1H, d, NHCH2), 4.73 (d,
J = 8.4 Hz), 5.18 (1H, dd, NHCH2), 5.28 (1H, d, J = 17.5 Hz,
ArCH2Ar), 5.85 (1H, d, J = 15.7 Hz, ArCH2Ar), 6.08 (1H, dd,
NHCH2), 6.49 (1H, s, Ar-H), 6.56 (1H, s, Ar-H), 6.71 (1H, s, Ar-
H), 6.73 (1H, s, Ar-H), 6.84 (1H, s, Ar-H), 7.00 (1H, s, Ar-H),
7.03 (1H, s, Ar-H), 7.08 (1H, br t, py-H), 7.12 (1H, s, Ar-H), 7.24
(1H, s, Ar-H), 7.24 (1H, s, calix-OH), 7.36 (1H, s, calix-OH), 7.38
(1H, t, py-H), 7.49 (1H, d, py-H), 7.95 (1H, d, py-H), 8.10 (1H,
Conclusions
The mercury ion complexation reactions were carried out between
the calix[4]arene derivative and various mercury(II) salts. 1H NMR
studies showed that the role of solvent, the anion chosen and the
initial reaction conditions were critical and that the formation
of a “simple” mercury(II) complex was non-trivial. The mercury
ion can cause either (i) the formation of an ion-pair system,
which have a characteristic downfield shift of all signals in the
1H NMR spectrum, (ii) a cleavage reaction to occur resulting in
the reformation of the calix[4]arene diester compound 2, but only
when the reaction is heated and (iii) “simple” mercury binding to
the pyridine rings when the binding studies are carried out using
NMR titration techniques. The electrochemistry data show a large
shift in ligand oxidation potential in the presence of Hg(ClO4)2
which is characteristic of complexation. The magnitude of the
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10928–10936 | 10935
©

View Article Online
shift is ca. 260 mV and, thus, we can infer that the Hg2+ species
is strongly complexed. This is also reflected in the diminished
Hg0–Hg2+ stripping peak observed after the first cycle. AgClO4
also causes a shift, albeit smaller, in the ligand oxidation potential
but the Ag0–Ag+ transition stripping peak remains constant over
the range of voltammetric cycles. This indicates that while Ag+
may bind to the ligand, the binding is not tight and may be
a binding/non-binding equilibrium process. However, activation
energy or power is being applied in both the electrochemical
reactions and also in the case where the reaction is heated. In
all of these cases, simple metal binding is not observed but more
complicated reaction pathways occur. Further investigations into
this system are under way.
15 For recent papers on ion-pair binding with calixarene receptors, see,
for example: (a) L. Pescatori, A. Arduini, A. Pochini, A. Secchi, C.
Massera and F. Ugozzoli, Org. Biomol. Chem., 2009, 7, 3698–3708;
(b) M. Hamon, M. Me´nand, S. Le Gac, M. Luhmer, V. Dalla and I.
Jabin, J. Org. Chem., 2008, 73, 7067–7071; (c) M. D. Lankshear, I. M.
Dudley, K.-M. Chan, A. R. Cowley, S. M. Santos, V. Felix and P. D.
Beer, Chem.–Eur. J., 2008, 14, 2248–2263; (d) S. Le Gac and I. Jabin,
Chem.–Eur. J., 2008, 14, 548–557; (e) M. D. Lankshear, N. H. Evans,
S. R. Bayly and P. D. Beer, Chem.–Eur. J., 2007, 13, 3861–3870; (f) T.
Nabeshima, T. Saiki, J. Iwabuchi and S. Akine, J. Am. Chem. Soc.,
2005, 127, 5507–5511.
16 S. Pappalardo, L. Giunta, M. Foti, G. Ferguson, J. F. Gallagher and B.
Kaitner, J. Org. Chem., 1992, 57, 2611–2624.
17 G. Ferguson, J. F. Gallagher and S. Pappalardo, J. Inclusion Phenom.
Mol. Recognit. Chem., 1992, 14, 349–356.
18 A. D. Bond, B. S. Creaven, D. F. Donlon, T. L. Gernon, J. McGinley
and H. Toftlund, Eur. J. Inorg. Chem., 2007, 749–756.
19 B. S. Creaven, T. L. Gernon, J. McGinley, A.-M. Moore and H.
Toftlund, Tetrahedron, 2006, 62, 9066–9071.
20 B. S. Creaven, T. L. Gernon, T. McCormac, J. McGinley, A.-M. Moore
and H Toftlund, Inorg. Chim. Acta, 2005, 358, 2661–2670.
21 B. S. Creaven, M. Deasy, J. F. Gallagher, J. McGinley and B. A. Murray,
Tetrahedron, 2001, 57, 8883–8887.
Acknowledgements
We thank John F. Gallagher, John C. Stephens and Carmel B.
Breslin for useful discussions, Orla Fenelon for mass spectrometry
data and Andrew D. Bond for confirming the X-ray crystal
structure of 3. DFD acknowledges financial assistance from
the Postgraduate Research and Development Skills Programme
(Technological Sector Research, Strand I) and the Institute of
Technology Tallaght Dublin PhD continuation fund.
22 B. S. Creaven, M. F. Mahon, J. McGinley and A.-M. Moore, Inorg.
Chem. Commun., 2006, 9, 231–234.
23 J. McGinley, V. McKee, H. Toftlund and J. M. D. Walsh, Dalton Trans.,
2009, 8406–8412.
24 R. Joseph, B. Ramanujam, A. Acharya, A. Khutia and C. P. Rao, J.
Org. Chem., 2008, 73, 5745–5758.
25 E. M. Collins, M. A. McKervey and S. J. Harris, J. Chem. Soc., Perkin
Trans. 1, 1989, 372–374.
26 B. Bala´zs, G. To´th, G. Horva´th, A. Gru¨n, V. Csokai, L. To¨ke and I.
Bitter, Eur. J. Org. Chem., 2001, 61–71.
27 A. F. D. De Namor, S. Chahine, E. E. Castellano and O. E. Piro, J.
Phys. Chem. A, 2005, 109, 6743–6751.
Notes and references
1 C. J. Elsevier, J. Reedijk, P. H. Walton and M. D. Ward, J. Chem. Soc.,
Dalton Trans., 2003, 1869–1880.
2 D. L. Caulder and K. N. Raymond, Acc. Chem. Res., 1999, 32, 975–982.
3 D. W. Bruce and D. O’Hare, Inorganic Materials, Wiley, Chicester,
1996.
28 A. F. D. De Namor, A. Aguilar-Cornejo, R. Soualhi, M. Shehab, K.
B. Nolan, N. Ouazzani and L. Mandi, J. Phys. Chem. B, 2005, 109,
14735–14741.
4 F. Vo¨gtle, Supramolecular Chemistry, Wiley, Chicester 1991.
5 A. Ikeda and S. Shinkai, Chem. Rev., 1997, 97, 1713–1734.
6 C. Wieser, C. B. Dieleman and D. Matt, Coord. Chem. Rev., 1997, 165,
93–161.
29 A. F. D. De Namor, S. Chahine, D. Kowalska, E. E. Castellano and O.
E. Piro, J. Am. Chem. Soc., 2002, 124, 12824–12836.
30 P. D. Beer, M. G. B. Drew and K. Gradwell, J. Chem. Soc., Perkin Trans.
2, 2000, 511–519.
7 C. D. Gutsche, Calixarenes Revisited: Monographs in Supramolecular
Chemistry, Royal Society of Chemistry, London, 1998.
8 L. Mandolini and R. Ungaro, Calixarenes in Action, Imperial College,
London, 2000.
9 Z. Asfari, V. Bo¨hmer, J. Harrowfield and J. Vicens, Calixarenes 2001,
Kluwer Academic, Dordrecht, 2001.
31 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fundamentals
and Applications, 2nd ed., Wiley, New York, 2001.
32 (a) A. Pailleret, G. Herzog and D. W. M. Arrigan, Electrochem.
Commun., 2003, 5, 68–72; (b) R. Vataj, A. Louati, C. Jeunesse and
D. Matt, Electrochem. Commun., 2000, 2, 769–775.
33 (a) R. S. Nicholson and I. Shain, Anal. Chem., 1964, 36, 706–723; (b) R.
S. Nicholson and I. Shain, Anal. Chem., 1965, 37, 178–190; (c) R. S.
Nicholson and I. Shain, Anal. Chem., 1965, 37, 190–195.
34 R. C. Nelson and R. N. Adams, J. Electroanal. Chem., 1968, 16, 439–
442.
35 L. N. Balyatinskaya, Russ. Chem. Rev., 1979, 48, 418–429.
36 S. B. Khoo and S. X. Guo, Electroanalysis, 2002, 14, 813–
822.
10 P. D. Beer and E. J. Hayes, Coord. Chem. Rev., 2003, 240, 167–189.
11 C. Redshaw, Coord. Chem. Rev., 2003, 244, 45–70.
´
12 W. Sliwa, J. Inclusion Phenom. Macrocyclic Chem., 2005, 52, 13–37.
13 J. Vicens and J. Harrowfield, Calixarenes in the Nanoworld, Springer,
Netherlands, 2007.
14 B. S. Creaven, D. F. Donlon and J. McGinley, Coord. Chem. Rev., 2009,
253, 893–962.
10936 | Dalton Trans., 2010, 39, 10928–10936
This journal is
The Royal Society of Chemistry 2010
©