Metal ion recognition. Selective interaction of silver(I) with tri-linked N2S2-donor macrocycles and their single-ring analogues

Article abstract of DOI:10.1039/b107574h

The interaction of four tri-linked N₂S₂-donor macrocyclic ligands and their single ring analogues with silver(I) has been investigated. The 3:1 (metal:ligand) stoichiometries of the silver(I) complexes of the tri-linked species in di

Full text of DOI:10.1039/b107574h

View Article Online / Journal Homepage / Table of Contents for this issue
Metal ion recognition. Selective interaction of silver(I) with tri-
linked N2S2-donor macrocycles and their single-ring analogues
Jy D. Chartres,a Andrew M. Groth,a Leonard F. Lindoy*b and George V. Meehan*a
a School of Pharmacy and Molecular Sciences, James Cook University, Townsville, Qld. 4811,
Australia
b Centre for Heavy Metals Research, School of Chemistry F11, University of Sydney,
N.S.W., 2006, Australia
Received 22nd August 2001, Accepted 30th October 2001
First published as an Advance Article on the web 3rd January 2002
The interaction of four tri-linked N2S2-donor macrocyclic ligands and their single ring analogues with silver() has
been investigated. The 3 : 1 (metal : ligand) stoichiometries of the silver() complexes of the tri-linked species in
dimethyl sulfoxide-d6/CDCl3 were confirmed by means of NMR titrations involving the incremental addition of
silver() nitrate to the respective ligands in the above solvent mixture and following the corresponding induced shifts
in the 1H NMR spectrum. Solid 3 : 1 (metal : ligand) complexes were isolated for each of the parent (unsubstituted)
tri-linked ligands. Competitive solvent extraction experiments (water/chloroform) and related bulk membrane
transport (water/chloroform/water) experiments have been performed in which each of the four tri-linked ligands as
well as their single ring analogues were employed as the extractant/ionophore in the chloroform phase. In both sets of
experiments the respective aqueous source phases (buffered at pH 4.9) contained an equimolar mixture of cobalt(),
nickel(), copper(), zinc(), cadmium(), silver() and lead() nitrates. For membrane transport the aqueous
receiving phase was buffered at pH 3. Under the conditions employed, both the solvent extraction and the bulk
membrane transport experiments resulted in high extraction/transport selectivity for silver() relative to the other six
metal ions present.
Introduction
Arising from their often unique properties, such as their
tendency to yield both very kinetically and thermodynamically
stable metal complexes,1 macrocyclic rings have frequently
proved to be desirable building blocks for incorporation into
new supermolecular and supramolecular systems.2 Initially,
such studies have mainly involved crown or azacrown rings3–
largely due to the ready synthesis of suitably functionalised
derivatives from these macrocyclic categories.4 However, more
recently, porphyrin ring and other amine-containing systems
have also been increasingly employed in this context.5
In a recent study by our group, the application of a pro-
tecting group strategy involving introduction of comple-
mentary protecting groups into the precursors of the parent,
16-membered, S2N2-heteroatom macrocycle has enabled a
clean and efficient synthesis of the new tri-linked macrocycles 1
(R = H or CH2Ph) and 2 (R = H or CH2Ph).6 We report here
the results of an investigation of the interaction of silver(),
with all four of these tri-linked species as well as with their
corresponding single ring analogues 3 (R = H or CH2Ph) and 4
(R = H or CH2Ph). The study includes the use of these systems
as ionophores in competitive extraction and membrane trans-
port experiments involving cobalt(), nickel(), copper(),
zinc(), cadmium(), silver() and lead().
The behaviour of 1 (R = H or CH2Ph) – 4 (R = H or CH2Ph)
towards silver() was of particular interest since these ring
systems incorporate structural features that might be expected
to enhance their relative affinity for this ion. Thus, when con-
sidering the coordination chemistry of silver(), two features
appear amenable for exploitation in ligand design. The first
is silver’s preference for linear diammine coordination, as
exemplified by the structure of [Ag(NH3)2]ϩ, while the second is
its well documented affinity for soft donors7 such as thioether
sulfur. Both these aspects are capable of being accommodated
by the present ring systems of type 1 (R = H or CH2Ph) – 4
(R = H or CH2Ph).
Surprisingly, there have been relatively few prior reports of
the interaction of silver() with fully saturated N2S2-donor
macrocycles related to present systems. In one previous investi-
gation, Kaden et al.8 have probed the solution complexation
behaviour of silver() with N-substituted, cis-S2N2 macrocycles
related to the present monomeric (trans-donor) rings. The
isolation and X-ray structure determinations of solid silver()
complexes of the type AgLX (X = NO3, OAc), incorporating a
related octamethylated trans-N2S2-donor macrocycle, have also
been reported.9,10
In keeping with the expectation that ligands of the present
type will show enhanced binding towards ‘softer’ metals, an
DOI: 10.1039/b107574h
J. Chem. Soc., Dalton Trans., 2002, 371–376
This journal is © The Royal Society of Chemistry 2002
371

View Article Online
initial study by our group has recently demonstrated that the
‘unsubstituted’ derivatives, 1 (R = H)–4 (R = H), yield stable
complexes with both palladium() and platinum().11
tained the required macrocyclic ligand (1.00 × 10Ϫ3 mol dmϪ3
for the single ring ligands and 3.33 × 10Ϫ4 mol dmϪ3 for the tri-
linked ligands). The receiving phase was buffered (formic acid/
sodium formate) at pH 3.0
0.1. All transport runs were
terminated after 24 h and were performed in duplicate. A con-
trol experiment in which the chloroform phase did not contain
ionophore, as expected, yielded no evidence of metal ion trans-
port, indicating that entrainment of the aqueous phase in
the organic phase was not a problem. Atomic absorption
spectrometry was used for metal ion analysis of the receiving
phase. Transport fluxes are expressed as J values in mol hϪ1
and represent mean values from duplicate runs measured over
24 h.
Experimental
Where available, all reagents and solvents were A.R. grade.
High resolution mass spectra were obtained on a Kraytos
M25RFA spectrometer or a Bruker APEX 47e spectrometer
(ESI). NMR spectra were recorded on a Bruker AM-300
spectrometer; δH values are relative to Me4Si and δC values are
relative to CDCl3 at 77.0 ppm, J are in Hz.
NMR titrations
Ligand synthesis
NMR titrations were employed to investigate silver complex-
ation by the tri-linked ring derivatives 1 (R = H or CH2Ph) and
2 (R = H or CH2Ph). Titrations were performed by sequential
15 µL additions of silver() nitrate solution (≈0.5 mol dmϪ3) in
CDCl3/DMSO-d6 (1 : 1) to a solution of the ligand (≈0.02 mol
dmϪ3) in the same solvent mixture contained in the NMR tube.
This procedure was carried out in the dark to minimise light
catalysed decomposition of the corresponding silver com-
plex. For each system, the change in the chemical shift of the
methylene protons adjacent to the thioether sulfur atoms in
the macrocyclic ring was monitored throughout the titration.
The induced chemical shift corresponding to each addition of
silver ion was then plotted against the ratio of metal ion con-
centration to ligand concentration present in the NMR tube.
The NMR titration data were analysed using a local version of
the program, EQNMR.12 EQNMR is used to evaluate stability
constants in systems where the species are in rapid equilibrium
and where the NMR chemical shift data reflect the degree
of complex formation. The program fits a titration curve to the
measured data using an iterative process that starts with initial
estimates of the stepwise stability constants.
1,3,5-Tris[(1,9-dithia-5,13-diazacyclohexadec-5-yl)methyl]-
benzene 1 (R = H),6 1,3,5-tris[(2-(1,9-dithia-5,13-diazacyclo-
hexadec-5-yl)ethoxy]benzene 2 (R = H),6 1,3,5-tris[(N-benzyl-
1,9-dithia-5,13-diazacyclohexadec-5-yl)methyl]benzene 1 (R =
CH2PH),6
1,3,5-tris[2-(N-benzyl-1,9-dithia-5,13-diazacyclo-
hexadec-5-yl)ethoxy]benzene 2 (R = CH2PH),6 5-benzyl-1,9-
dithia-5,13-diazacyclohexadecane 3 (R = H),6 2-(1,9-dithia-
5,13-diazacyclohexadec-5-yl)ethoxybenzene 4 (R = H)6 and 5-
benzyl-1,9-dithia-13-chloroacetyl-5,13-diazacyclohexadecane6
were synthesised as described elsewhere. The syntheses of 3
(R = CH2Ph) and 4 (R = CH2Ph), which were performed under
an atmosphere of dry nitrogen, are given below.
5,13-Dibenzyl-1,9-dithia-5,13-diazacyclohexadecane 3 (R ؍
CH2Ph). 5-Benzyl-1,9-dithia-5,13-diazacyclohexadecane
3
(R = H) (1.33 g, 3.7 mmol) was dissolved in dry tetrahydrofuran
(40 cm3). Triethylamine (0.43 g, 4.2 mmol) and then benzoyl
chloride (0.58 g, 4.15 mmol) were added by syringe. The
reaction mixture was stirred at room temperature for 12 h
after which the tetrahydrofuran was removed under reduced
pressure. The residue was partitioned between dichloromethane
(100 cm3) and water (50 cm3). The organic phase was washed
with a further 50 cm3 of water and then dried (sodium sulfate).
The solvent was evaporated and the crude product chromato-
graphed on silica gel (eluting with methanol–dichloromethane,
1 : 99) to give 5-benzoyl-13-benzyl-1,9-dithia-5,13-diazacyclo-
hexadecane as a yellow oil [Found: M ϩ Hϩ, 457.2374
(LSIMS). C27H37N2OS2 requires M ϩ Hϩ, 457.2347]; Rf
(MeOH–dichloromethane, 1 : 19) 0.61; δH (CDCl3, 300 MHz)
≈7.3 (10 H, m, OCPh, CH2Ph), 3.66, 3.38 (4 H, br m,
CH2NCOPh), 3.51 (2 H, s, CH2Ph), 2.59 (4 H, t, J 7.4,
CH2CH2CH2NCH2Ph), 2.48 (4 H, t, J 6.0, PhCH2NCH2),
2.60, 2.36 (4 H, br m, PhCONCH2CH2CH2), 2.06, 1.92
(4 H, br m, PhCONCH2CH2), 1.74 (4 H, br m, CH2CH2-
NCH2Ph); δC (CDCl3, 75 MHz) 171.4, 167.4, 139.1, 136.5,
129.2, 128.9, 128.5, 128.1, 127.7, 126.5, 126.2, 126.0, 121.1,
114.1, 58.7, 52.5, 48.7, 47.0, 46.9, 44.5, 29.5, 29.3, 28.7, 28.5,
27.3, 26.8.
Extraction experiments
The procedure employed involved ‘competitive’ metal extrac-
tions from an aqueous phase into a chloroform phase. The
aqueous phase was buffered at pH 4.9
0.1 with sodium
acetate/acetic acid buffer and contained equal concentrations
(10Ϫ2 mol dmϪ3) of cobalt(), copper(), lead(), silver(),
cadmium(), zinc() and nickel() as their respective nitrate
salts. The extractions were carried out in small sealed flasks in
the absence of light to minimise the possibility of light-induced
silver() decomposition. The flasks were agitated for one
hour on a mechanical shaker after which time an aliquot of
the organic phase was removed and evaporated to dryness.
The residue was then taken up in nitric acid and, after
appropriate dilution, was analysed by atomic absorption
spectrophotometry. Each experiment was performed in quad-
ruplicate and the quoted values are the average from individual
experiments.
5-Benzoyl-13-benzyl-1,9-dithia-5,13-diazacyclohexadecane
(1.09 g, 2.4 mmol) was dissolved in dichloromethane (50 cm3)
at 0 ЊC and a 1.0 mol dmϪ3 solution of diisobutylaluminium
hydride (11.1 cm3, 11.1 mmol) was then added. The reaction
mixture was stirred for 2 h after which the excess diisobutyl-
aluminium hydride was destroyed by careful addition of metha-
nol. The residue was then partitioned between 10% aqueous
sodium hydroxide (100 cm3) and diethyl ether (100 cm3). The
aqueous layer was extracted with ether (100 cm3 × 2) and the
combined organic layers dried (sodium sulfate) and the solvent
evaporated under reduced pressure. The crude product was
purified by column chromatography on silica gel (eluting with
ethyl acetate–hexane, 1 : 9) to give 5,13-dibenzyl-1,9-dithia-5,13-
diazacyclohexadecane 3 (R = CH2Ph) as a yellow oil (0.57 g,
55%) [Found: M ϩ Hϩ, 443.2560 (LSIMS). C26H38N2S2 requires
M ϩ Hϩ, 443.2555); Rf (ethyl acetate–hexane, 1 : 3) 0.26;
Membrane transport
The transport experiments employed a ‘concentric cell’ in
which the aqueous source phase (10 cm3) and receiving phase
(30 cm3) are separated by a chloroform phase (50 cm3). Details
of the cell design have been reported elsewhere13 and the con-
ditions employed were generally similar to those used for a
recent study.14 For each experiment (which was carried out in
subdued light), both aqueous phases and the chloroform phase
were stirred separately at 10 revolutions minϪ1; the cell was
enclosed by a water jacket and thermostatted at 25 ЊC. The
aqueous source phase was buffered (acetic acid/sodium acetate)
at pH 4.9 0.1 and contained 10Ϫ2 mol dmϪ3 of each of the
following nitrate salts: cobalt(), nickel(), copper(), zinc(),
silver(), cadmium() and lead(). The chloroform phase con-
372
J. Chem. Soc., Dalton Trans., 2002, 371–376

View Article Online
δH (CDCl3, 300 MHz) ≈7.3 (10 H, m, Ph), 3.54 (4 H, s, CH2Ph),
2.57 (8 H, t, J 7.7, CH2S), 2.49 (8 H, t, J 6.3, CH2N), 1.77 (8 H,
quin, J 7.0, NCH2CH2CH2S); δC (CDCl3, 75 MHz) 139.7,
128.8, 128.1, 126.8, 59.4, 52.6, 30.0, 27.7.
(M Ϫ 2NO3)2ϩ, 641.6051. Found: C, 35.79; H, 5.92; N 8.45.
C45H84Ag3N9O9S6ؒ5H2O requires C, 36.00; H, 6.31; N, 8.40%].
1,3,5-Tris[2-(1,9-dithia-5,13-diazacyclohexadec-5-yl)ethoxy]-
benzenetrisilver(I) nitrate pentahydrate [Ag3L](NO3)3ؒ5H2O
(L ؍ 2, R ؍ H). Using a similar procedure to that described
for [Ag3L](NO3)3ؒ5H2O (L = 1, R = H), silver nitrate (0.017 g,
0.01 mmol) and 2 (R = H) (0.033 g, 0.03 mmol) yielded
[Ag3L](NO3)3ؒ5H2O (L = 2, R = H), (0.040 g, 75%) as an
2-(N-Benzyl-1,9-dithia-5,13-diazacyclohexadec-5-yl)ethoxy-
benzene 4 (R ؍ CH2Ph). Caesium carbonate (0.923 g, 2.83
mmol) was suspended in a solution of 5-benzyl-1,9-dithia-13-
chloroacetyl-5,13-diazacyclohexadecane6 (1.10 g, 2.6 mmol)
in dry dimethylformamide (25 cm3). Solid phenol (0.267 g,
2.83 mmol) was added and the mixture stirred at 70 ЊC for 48 h.
The dimethylformamide was removed in vacuo and the residue
suspended in dichloromethane (100 cm3). The solid was
removed by filtration through Celite and the solvent evaporated
under reduced pressure. The crude product was purified by
column chromatography on silica gel (eluting with ether–
hexanes, 1 : 1) to give 5-benzyl-13-(2-phenoxyacetyl)-1,9-dithia-
5,13-diazacyclohexadecane (0.80 g, 64%) as a viscous oil
[Found: M ϩ Hϩ, 487.2468 (LSIMS). C27H38N2O2S2 requires M
ϩ Hϩ, 487.2453]. δH (CDCl3, 300 MHz) 7.3–7.0 (10 H, m,
CH2Ph, OPh), 4.70 (2 H, s, PhOCH2), 3.55–3.52 (4 H, over-
lapping t, CH2NCO), 2.54 (8 H, m, CH2S), 2.46 (4 H, t, J 6.0,
CH2NCH2Ph), 1.95 (4 H, overlapping quin, CH2CH2NCO),
1.73 (4 H, quin, J 6.0, CH2CH2NCH2Ph); δC (CDCl3, 75 MHz)
167.9, 157.8, 139.4, 129.6, 128.8, 128.1, 126.9, 121.6, 114.4,
67.7, 59.1, 52.8, 52.7, 47.6, 45.8, 29.9, 29.8, 29.5, 29.2, 28.9,
27.7, 27.4, 27.2.
5-Benzyl-13-(2-phenoxyacetyl)-1,9-dithia-5,13-diazacyclo-
hexadecane (1.59 g, 3.3 mmol) was dissolved in dry tetrahydro-
furan (50 cm3). A 2.0 mol dmϪ3 solution of borane–dimethyl
sulfide complex in tetrahydrofuran (8.2 cm3, 16.4 mmol)
was added slowly and the mixture heated to reflux for 12 h.
The solution was allowed to cool to room temperature and the
excess borane destroyed by careful addition of methanol. The
solvent was removed under reduced pressure and the residue
was hydrolysed in refluxing methanol–water–concentrated
hydrochloric acid (40 : 10 : 4; 55 cm3) for 1 h. The methanol was
removed under reduced pressure and the resulting solution was
partitioned between 10% aqueous sodium hydroxide (100 cm3)
and dichloromethane (100 cm3). The aqueous layer was
extracted twice further with dichloromethane (100 cm3) and
the combined organic layers were dried (sodium sulfate)
and evaporated under reduced pressure. Purification by
chromatography on silica gel (eluting with ethyl acetate–
hexanes, 1 : 3) gave 2-(N-benzyl-1,9-dithia-5,13-diazacyclo-
hexadec-5-yl)ethoxybenzene 4 (R = CH2Ph) as a colourless oil
(1.01 g, 66%) [Found: M ϩ Hϩ, 473.2662 (LSIMS). C27H40-
N2OS2 requires M ϩ Hϩ, 473.2660]. δH (CDCl3, 300 MHz)
≈7.3 (5 H, m, CH2Ph), ≈6.9 (5 H, m, OPh), 4.04 (2 H, t, J 6.0,
CH2O), 3.51 (2 H, s, CH2Ph), 2.86 (2 H, t, J 6.0, OCH2CH2N),
2.63 (4 H, t, J 7.2, CH2S), 2.62 (4 H, t, J 6.3, CH2NCH2-
CH2OPh), 2.55 (4 H, t, J 7.4, CH2S), 2.46 (4 H, t, J 6.3,
CH2NCH2Ph), 1.76 (8 H, quin, J 7.4, PhCH2NCH2CH2), 1.74
(8 H, quin, J 7.8, CH2CH2NCH2CH2O); δC (CDCl3, 75 MHz)
158.7,139.6, 129.3, 128.7, 128.0, 126.7, 120.5, 114.3, 66.3, 59.4,
53.5, 53.4, 52.5, 30.0, 29.8, 27.7, 27.5.
off-white microcrystalline product [Found (M Ϫ 2NO3)2ϩ
686.6206 (ESI). C48H90Ag3N9O12S6 requires (M Ϫ 2NO3)2ϩ
,
,
686.6209. Found: C, 36.10; H, 6.09; N 7.87. C48H90Ag3N9O12S6ؒ
5H2O requires C, 36.23; H, 6.33; N, 7.92%].
Results and discussion
Ligand synthesis
The macrocyclic derivatives 1 (R = H, CH2Ph), 2 (R = H,
CH2Ph), 3 (R = H) and 4 (R = H) were synthesised using pro-
cedures reported previously by our group.6 It is noted that the
parent (unsubstituted) 16-membered S2N2 macrocycle has been
reported by Kaden et al.;8 however, the synthetic approach
employed in this previous study was not suitable for use for
the preparation of the present derivatives [with the exception of
the symmetrically substituted derivative 3 (R = CH2Ph)] since
it did not allow easy chemical differentiation between the two
ring nitrogen sites. For the present syntheses, an alternate dis-
connection of the parent macrocycle was employed that placed
the nitrogens in separate synthons, enabling the former to be
differentially protected prior to cyclisation. Namely, the
procedure employed bis-alkylation of the terminal dithiolate
groups of a N-protected bis(thiolopropyl)amine species with a
bis(3-halogenopropyl)amine derivative incorporating different
nitrogen protection.
The new macrocyclic derivative 3 (R = CH2Ph) was obtained
from 3 (R = H)6 by initial acylation with benzoyl chloride in dry
tetrahydrofuran in the presence of base. The resulting amide
derivative was then reduced with diisobutylaluminium hydride
to yield crude 3 (R = CH2Ph) as an oil which was purified
by partitioning between aqueous sodium hydroxide and ether,
followed by chromatography on silica gel.
The related derivative 4 (R = CH2Ph) was obtained using a
similar strategy to that employed for the tri-linked compound 2
(R = CH2Ph)6 in which the required macrocyclic chloroamide
derivative was used to alkylate phenol (rather than phloro-
glucinol (1,3,5–trihydroxybenzene)). The resulting substituted
macrocyclic amide was then reduced using borane/dimethyl
sulfide to give 4 (R = CH2Ph).
Metal complex isolation
Initial attempts to isolate solid silver() nitrate complexes of
the tri-linked ligands of type 1 and 2 resulted in coloured
contaminated products. However, in the case of 1 (R = H) and 2
(R = H), when the syntheses were performed in the absence of
light in dry dichloromethane/ethanol, analytically pure (off-
white) products of type [Ag3L](NO3)3ؒ5H2O were obtained,
confirming the ability of these ligands to bind three silver ions.
Unfortunately, we were not able to obtain suitable crystals for
X-ray structure determinations in either case.
Metal complex synthesis
1,3,5-Tris[(1,9-dithia-5,13-diazacyclohexadec-5-yl)methyl]-
benzenetrisilver(I) nitrate pentahydrate [Ag3L](NO3)3ؒ5H2O
(L ؍ 1, R ؍ H). Silver() nitrate (0.017 g, 0.01 mmol) in abso-
lute ethanol (2 cm3) was added dropwise to a solution of 1 (R =
H) (0.030 g, 0.03 mmol) in dry 4 : 1 ethanol/dichloromethane
(1 cm3) in the dark. The reaction was stirred for 30 min during
which a white solid precipitated from solution which was
filtered off, washed with dry ethanol and ether, then dried over
P2O5 under vacuum to give [Ag3L](NO3)3ؒ5H2O (L = 1, R = H)
(0.044 g, 88%) as an off-white microcrystalline product [Found
(M Ϫ 2NO3)2ϩ, 641.6080 (ESI). C45H84Ag3N9O9S6 requires
NMR titrations
The interaction of the tri-linked S2N2-donor ligands 1 (R = H
or CH2Ph) and 2 (R = H or CH2Ph) with silver() was investi-
gated by means of H NMR titration experiments in CDCl3/
DMSO-d6. Titration curves for incremental addition of silver
nitrate to both 1 (R = H) and 2 (R = H) yielded sharp 3 : 1
(metal : ligand) end points (Fig. 1), indicating that binding of
silver by these ligands is relatively strong. The thermodynamic
formation constants for these two systems are clearly too high
1
J. Chem. Soc., Dalton Trans., 2002, 371–376
373

View Article Online
Fig. 1 NMR titration curves for the addition of silver nitrate to the
tri-linked macrocyclic species in DMSO-d6/CDCl3 (1 : 1) (a) 1 (R = H),
(b) 2 (R = H).
to be determined under the conditions employed. However,
the binding of silver() by the tribenzylated derivatives 1 (R =
CH2Ph) and 2 (R = CH2Ph) under similar conditions is sig-
nificantly weaker (Fig. 2), as indicated by the absence of a
sharp end point in the respective titration curves. Analysis of
these latter curves using a local version of EQNMR12 enabled
calculation of the corresponding log β values for Ag() com-
plexation; the values obtained for 1 (R = CH2Ph) are log
β1 = 3.54, log β2 = 6.52 and log β3 = 8.61 and, for 2 (R = CH2Ph),
log β1 = 3.66, log β2 = 6.94 and log β3 = 9.21. No satisfactory fit
of the experimental data was obtained when other simple
models were substituted for the 3 : 1 (metal : ligand) model used
for the calculations.
As expected from statistical considerations, the stepwise con-
stants decrease for successive complexation steps. A possible
further contribution to the diminishing step-wise stability
constants could involve charge repulsion between the positive
metal centres. However, this does not appear to make a major
contribution in the present cases; the respective log βn values for
the complexes of 1 (R = CH2Ph) and 2 (R = CH2Ph) given
above differ only slightly even though a different number of
linking atoms separate the macrocyclic rings in each system.
Fig. 2 NMR titration curves for the addition of silver nitrate to the
tri-linked macrocyclic species (a) 1 (R = CH2Ph) and (b) 2 (R = CH2Ph)
in DMSO-d6/CDCl3 (1 : 1), showing fitted EQNMR plots.
each of cobalt(), copper(), lead(), silver(), cadmium(),
zinc() and nickel() as their nitrate salts. In individual experi-
ments, the tri-linked species, 1 (R = H or CH2Ph), 2 (R = H
or CH2Ph), and corresponding single ring species, 3 (R = H
or CH2Ph) and 4 (R = H or CH2Ph), were employed as the
ionophore in the respective chloroform phases. To enable direct
comparison of results, individual experiments were performed
under ‘equivalent concentrations’ with respect to the number of
macrocyclic rings present in each ionophore. Hence ionophore
concentrations of 3.33 × 10Ϫ4 mol dmϪ3 for the tri-linked
systems and 1.00 × 10Ϫ3 mol dmϪ3 for the single ring systems
were employed. The results are presented in Fig. 3; for ready
comparison of the ‘efficiencies’ of the single and three-ring
species, the degree of extraction is presented as the percentage
of ligand sites occupied in each experiment. Clearly, under the
conditions employed, each single ring macrocycle is a more
effective extracting agent on a ‘per hole’ basis than its tri-linked
analogue. In all experiments, the respective systems showed sole
selectivity for silver()–in confirmation of the expected affinity
of this ion for a N2S2-donor set. Hence, while the macrocyclic
ring substitution (and linkage) pattern clearly influences the
efficiency of the extraction, silver ion selectivity is maintained
in each case.
Extraction experiments
Metal ion solvent extraction experiments (water/chloroform)
have been carried out using each of the tri-linked macrocycles
and their single ring macrocyclic analogues as extractants. The
procedure employed involved ‘competitive’ metal extractions
from an aqueous phase at pH 4.9 (sodium acetate/acetic acid
buffer) containing equal concentrations (≈10Ϫ2 mol dmϪ3) of
374
J. Chem. Soc., Dalton Trans., 2002, 371–376

View Article Online
Both the lipophilicity of a ligand and its binding strength
with a particular cation are factors that are well documented to
influence overall extraction efficiencies.15 In the present study
the N-benzylated single-ring and tri-linked ligands were found
to be more efficient extractors of silver() relative to their
corresponding non-benzylated analogues (Fig. 3). This was so
Fig. 4 Comparative transport flux rates (a) for the tri-linked ligands
ligands 1 (R = H or CH2Ph) and 2 (R = H or CH2Ph) and (b) their
respective single ring analogues 3 (R = H or CH2Ph) and 4 (R = H or
CH2Ph). Flux values J are in mol hϪ1 (all plotted J values are by 10Ϫ9).
[Source phase contained seven metal nitrates, each at ≈10Ϫ2 mol dmϪ3
buffered at pH 4.9; receiving phase buffered at pH 3.0; for individual
experiments the chloroform phase contained a tri-linked ligand at
≈3.33 × 10Ϫ4 mol dmϪ3 or a single ring ligand at 1.00 × 10Ϫ3 mol dmϪ3
runs terminated after 24 h; 25 ЊC.]
;
Fig. 3 Mixed metal ion extraction data (a) for the tri-linked ligands
ligands 1 (R = H or CH2Ph) and 2 (R = H or CH2Ph) and (b) their
respective single ring analogues 3 (R = H or CH2Ph) and 4 (R = H
or CH2Ph). [Source phase contained seven metal nitrates, each at
≈10Ϫ2 mol dmϪ3 buffered at pH 4.9; for individual experiments the
chloroform phase contained a tri-linked ligand at 3.33 × 10Ϫ4 mol dmϪ3
or a single ring ligand at 1.00 × 10Ϫ3 mol dmϪ3; experiments terminated
after 1 h; 25 ЊC.]
transport fluxes than their respective tri-linked analogues – a
result that again directly parallels the corresponding solvent
extraction results. Also, as before, the benzylated ligands (both
tri-linked and mono-ring) are each more efficient ionophores
than their NH-containing parents, with the single ring species
being better transporters than their tri-linked analogues when
compared on a ‘per hole’ basis.
Clearly, the transport and extraction results closely parallel
each other for the present ligand series, a situation also found
to occur in our recent studies14 in which a series of oxygen–
nitrogen donor macrocycles were employed as extractants/
ionophores in comparative experiments of the above type.
even though the NMR titration results (in CDCl3/DMSO-d6)
indicated (see above) that higher complex stabilities occur for
the silver complexes of the non-benzylated tri-linked ligands 1
(R = H) and 2 (R = H) under the conditions employed for these
latter experiments. Hence, it appears that it is the greater
lipophilicity of the benzylated derivatives 1 (R = CH2Ph), 2 (R =
CH2Ph), 3 (R = CH2Ph) and 4 (R = CH2Ph) that favours the
observed enhanced extraction efficiencies of these ring systems
relative to those of the corresponding systems with R = H.
Concluding remarks
Transport experiments
The study demonstrates the high selectivity of all the present
ligands for silver() and confirms that all three macrocyclic
sites in the linked ligands readily bind this cation. Overall, the
results provide a foundation for a study of the metal binding
behaviour of related dendritic systems incorporating the
present tri-linked frameworks to which additional N2S2-donor
macrocyclic units have been appended. An investigation of
this type is currently underway and the results will be presented
in due course.
The study involved metal ion transport from an aqueous
source phase containing the nitrate salts of cobalt(), nickel(),
copper(), zinc(), cadmium(), silver() and lead(), each
at ≈10Ϫ2 mol dmϪ3, across a bulk chloroform membrane
incorporating an ionophore chosen from 1 (R = H or CH2Ph),
2 (R = H or CH2Ph), 3 (R = H or CH2Ph) or 4 (R = H or
CH2Ph). Transport was performed against a back gradient
of protons, maintained by buffering the source and receiving
phases at pH 4.9 and 3.0, respectively.
Even though theory does not decree that it need be the case,16
it was nevertheless of interest to see whether the clear silver()
selectivity observed in the solvent extraction experiments for
each of the present ligands would be maintained in the trans-
port studies. This indeed was found to be the case, with evi-
dence for a minor amount of zinc() transport also occurring
in individual cases. The ‘per macrocyclic hole’ results from the
above experiments are presented in Fig. 4. Under the conditions
employed, the single ring ligands resulted in relatively higher
Acknowledgements
We thank the Australian Research Council for support and
Drs A. J. Leong and I. M. Atkinson for experimental assistance.
References
1 L. F. Lindoy, The Chemistry of Macrocyclic Ligand Complexes,
Cambridge University Press, 1989.
J. Chem. Soc., Dalton Trans., 2002, 371–376
375

View Article Online
2 Self-assembly in Supramolecular Chemistry, eds. L. F. Lindoy and
I. M. Atkinson, Royal Society of Chemistry, Cambridge, 2000.
3 C. Seel and F. Vogtle, Crown Compounds: Towards Future
Applications, ed. S. R. Cooper, VCR, New York, 1992, ch. 11,
pp. 191–206; F. Vogtle and E. Weber, J. Inclusion Phenom., 1992,
12, 75.
13 P. S. K. Chia, L. F. Lindoy, G. W. Walker and G. W. Everett, Pure
Appl. Chem., 1993, 65, 521.
14 J.-H. Kim, A. J. Leong, L. F. Lindoy, J. Kim, J. Nachbaur,
Z. Nezhadali, G. Rounaghi and G. Wei, J. Chem. Soc., Dalton
Trans., 2000, 3453.
15 B. G. Cox and H. Schneider, Coordination and Transport Properties
of Macrocyclic Compounds in Solution, Elsevier Science Publishers
B. V., Amsterdam, 1992.
4 K. E. Krakowiak, J. S. Bradshaw and D. J. Zamecka-Krakowiak,
Chem. Rev., 1989, 89, 929.
5 L. F. Lindoy, Adv. Inorg. Chem., 1997, 45, 75; J. D. Chartres,
L. F. Lindoy and G. V. Meehan, Coord. Chem. Rev., 2001, 216–217,
249.
6 A. M. Groth, L. F. Lindoy and George V. Meehan, J. Chem. Soc.,
Perkin Trans. 1, 1996, 1553.
7 S. Ahrland, J. Chatt and N. R. Davies, Quart. Rev., 1958, 12, 265;
R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 3533; R. G. Pearson,
Coord. Chem. Rev., 1990, 100, 403.
8 P. C. Reisen and T. A. Kaden, Helv. Chim. Acta, 1995, 1325.
9 G. Ferguson, R. McCrindle and M. Parvez, Acta Crystallogr., Sect.
C, 1984, 40, 354.
10 R. McCrindle, G. Ferguson, A. J. McAlees, M. Parvez, B. L. Ruhl,
D. K. Stephenson and T. Wieckowski, J. Chem. Soc., Dalton Trans.,
1986, 2351.
11 I. M. Atkinson, J. D. Chartres, A. M. Groth, L. F. Lindoy,
M. P. Lowe, G. V. Meehan, B. W. Skelton and A. H. White, J. Chem.
Soc., Dalton Trans., 2001, 2388.
16 J.-M Lehn, A. Moradpour and J.-P. Behr, J. Am. Chem. Soc., 1975,
97, 2532; J. D. Lamb, J. J. Christensen, S. R. Izatt, K. Bedke,
M. S. Astin and R. M. Izatt, J. Am. Chem. Soc., 1980, 102, 467;
J. D. Goddard, J. Phys. Chem., 1985, 89, 1825; T. M. Fyles,
J. Membr. Sci., 1985, 24, 229; S. Yoshida and S. Hayano, J. Am.
Chem. Soc., 1986, 108, 3903; T. B. Stolwijk, E. J. R. Sudhölter and
D. N. Reinhoudt, J. Am. Chem. Soc., 1987, 109, 7042; T. M. Fyles,
Can. J. Chem., 1987, 65, 884; T. M. Fyles and S. P. Hansen, Can.
J. Chem., 1988, 66, 1445; S. Yoshida and T. Watanabe, J. Coord.
Chem., 1988, 18, 63; P. J. Dutton, T. M. Fyles and S. P. Hansen,
J. Inclusion Phenom., 1989, 7, 173; R. M. Izatt, R. L. Bruening,
M. L. Bruening, G. C. Lindh and J. J. Christensen, Anal. Chem.,
1989, 61, 1140; W. F. Nijenhuis, J. J. B. Walhof, E. J. R. Sudhölter
and D. N. Reinhoudt, Recl. Trav. Chim. Pays-Bas, 1991, 110, 265;
J. C. Hernandez, J. E. Trafton and G. W. Gokel, Tetrahedron. Lett.,
1991, 32, 6269; A. Dindi, R. D. Noble and C. A. Koval, J. Membr.
Sci., 1992, 65, 39; Y. Nakatsuji and M. Okahara, Pure Appl. Chem.,
1993, 65, 557.
12 M. J. Hynes, J. Chem. Soc., Dalton Trans., 1993, 311.
376
J. Chem. Soc., Dalton Trans., 2002, 371–376