Macrocyclic ligand design. Structure-function relationships involving the interaction of pyridinyl-containing, oxygen-nitrogen donor macrocycles with selected transition and post transition metal ions on progressive N-benzylation of their secondary amines

Article abstract of DOI:10.1039/b412437e

Structure-function relationships underlying the interaction of progressively N-benzylated N₄O₂-donor macrocycles with cobalt(II), nickel(II), copper(II), zinc(II), cadmium(II), silver(I) and lead(II) have been probed using a range of techniques that include X-ray diffraction, DFT computations, solvent extraction, potentiometric stability constant determinations and competitive membrane transport experiments. Collectively, the results indicate that N-benzylation of the secondary amine donor groups of the parent macrocyclic ring results in an enhanced tendency towards selectivity for silver(I) relative to the other six metals investigated. The observed behaviour serves as additional exemplification of the previously proposed concept of selective 'detuning' as a mechanism for metal ion discrimination.

Full text of DOI:10.1039/b412437e

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Macrocyclic ligand design. Structure–function relationships involving
the interaction of pyridinyl-containing, oxygen–nitrogen donor
macrocycles with selected transition and post transition metal ions on
progressive N-benzylation of their secondary amines†
Jason R. Price,a Marina Fainerman-Melnikova,a Ronald R. Fenton,a Karsten Gloe,b
Leonard F. Lindoy,*a Torsten Rambusch,a,b Brian W. Skelton,c Peter Turner,a Allan H. Whitec
and Kathrin Wichmanna,b
a
Centre for Heavy Metals Research, School of Chemistry, the University of Sydney,
NSW 2006, Australia
Institut für Anorganische Chemie, TU Dresden, 01062 Dresden, Germany
Chemistry M313, The University of Western Australia, Crawley, WA 6009, Australia
b
c
Received 12th August 2004, Accepted 14th September 2004
First published as an Advance Article on the web 8th October 2004
Structure–function relationships underlying the interaction of progressively N-benzylated N4O2-donor macrocycles
with cobalt(II), nickel(II), copper(II), zinc(II), cadmium(II), silver(I) and lead(II) have been probed using a range
of techniques that include X-ray diffraction, DFT computations, solvent extraction, potentiometric stability
constant determinations and competitive membrane transport experiments. Collectively, the results indicate that
N-benzylation of the secondary amine donor groups of the parent macrocyclic ring results in an enhanced tendency
towards selectivity for silver(I) relative to the other six metals investigated. The observed behaviour serves as
additional exemplification of the previously proposed concept of selective ‘detuning’ as a mechanism for metal ion
discrimination.
parent cyclam. A variety of influences have been proposed to
account for such behaviour, with solvation effects predicted to
be dominant, lower oxidation states are expected to be stabilised
on N-alkylation.
We now report an extension of our prior studies involving
the interaction of the seven metal-ion series mentioned above
with the progressively N-benzylated, O2N4-donor macrocycles
1–6, each incorporating a 20-membered macrocyclic ring.
Introduction
Metal-ion recognition by organic substrates is an area of funda-
mental importance to numerous aspects of both chemistry and
biochemistry. In previous studies our group has investigated
the effect on N-alkylation of secondary nitrogen atoms in
both mixed-donor1–3 and N4-donor4 macrocyclic ligands on
their metal-ion recognition properties towards the following
transition and post-transition metal ions: cobalt(II), nickel(II),
copper(II), zinc(II), cadmium(II), lead(II) and silver(I). The aim
The related 17-membered ring species of type 7 have been the
focus of a prior study by us.2 It was of particular interest in
of these studies has been to probe structure/function effects in
the present investigation to probe the effect of the larger ring
terms of metal-ion discrimination behaviour. Over several such
ligand systems it has been documented that there is a tendency
for N-benzylation to reduce the affinity (often very significantly)
and the presence of a pyridyl nitrogen donor in each of 1–6 on
the relative ‘detuning’ behaviour towards the metal-ion series
mentioned above.
of such ligand systems for the first six of the above metal ions
while the affinity for silver(I) remains much less affected. We
have termed behaviour of this type ‘selective detuning’.5 The
latter represents a potential metal ion discrimination mechanism
that has received little attention by others. Nevertheless, in this
context it needs to be noted that a limited number of examples
of N-alkylation leading to enhanced metal ion selectivity and
related behaviour have been reported previously.6,7 For example,
while alkylation of a secondary amine generally results in a
reduction of its metal ion affinity due to steric factors, it has
been documented that tertiary amines sometimes bind more
strongly to silver(I) compared with secondary amines and
that the relative affinities can be solvent dependant.6 Similarly,
the N-methylation of secondary amines has been shown by
Meyerstein et al.8 to favour stabilisation of the monovalent
state of copper relative to its divalent state as exemplified by
the electrochemical behaviour of copper(I) and copper(II) in
the presence of the tetra-N-methylated cyclam derivative versus
† Electronic supplementary information (ESI) available: Table S1
(coordinated anion and other geometries [M(2)(NO3)2] (M = Co, Zn,
Cd, Ni)) and Table S2 (ligand conformational descriptors); Fig. S1 (cal-
culated geometry of the lowest found minimum structure of [Ag(4)]+)
and Fig. S2 (calculated geometry of the minor disordered component
of [Ag(4)]+. See http://www.rsc.org/suppdata/dt/b4/b412437e/
T h i s j o u r n a l i s
©
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 1 5

View Article Online
The metal concentration in both phases was determined
radiometrically by means of c-counting using a NaI(Tl)
Scintillation Counter (Cobra II, Canberra-Packard). The
following metal isotopes were employed: 110mAg, 65Zn, 60Co and
64Cu. The values from the duplicate analyses were averaged, then
taken as the ‘working’ concentration in each case.
Experimental
All commercial reagents and solvents were of analytical (or
HPLC) grade where available. Tetrakis(acetonitrile)copper(I)
hexafluorophosphate was synthesized by the literature proce-
dure.9 The preparation of 1–6 has been reported previously.10
Physical measurements
Membrane transport
UV-VIS spectra were obtained in methanol on a Cary 5E
The transport experiments employed a ‘concentric cell’ in
which the aqueous source phase (10 mL) and receiving phase
(30 mL) were separated by a chloroform phase (50 mL).
Details of the cell design have been reported elsewhere.13 For
each experiment both the aqueous phases and the organic
phase were stirred separately at 10 rpm; the cell was enclosed
by a water jacket and thermostatted at 25 °C. The aqueous
source phase was buffered (CH3CO2H–CH3CO2Na) at pH 4.9
(±0.1) (6.95 mL of 2 M sodium acetate solution and 3.05 mL
of 2 M acetic acid made up to 100 mL)14 and contained an
equimolar solution of the nitrate salts of cobalt(II), nickel(II),
zinc(II), copper(II), cadmium(II), lead(II) and silver(I), each at a
concentration of 1 × 10−2 M. The chloroform phase contained
the ligand (1 × 10−3 M), as well as palmitic acid at 4 × 10−3 M.
The aqueous receiving phase was buffered (HCO2H–HCO2Na)
at pH 3.0 ± 0.1 (56.6 mL of 1 M formic acid and 10.0 mL of
1 M sodium hydroxide made up to 100 mL).14 All transport
runs were terminated after 24 hours and atomic absorption
spectroscopy was used to determine the amount of metal ion
transported over this period. Transport rates (J values) are in
mol (24 h)−1. The transport results are quoted as the average
values obtained from duplicate runs (maximum error ca. ±10%
of reported value).
spectrophotomer.
A Finnigan LCQ-8 spectrometer was
employed for low resolution, positive ion ESMS. X-Band
EPR spectra were obtained as the first derivatives from frozen
DMF solution at liquid N2 temperature. A Bruker EMX EPR
spectrometer (operating microwave frequency 9.3 GHz)
equipped with a Bruker EMX 035M NMR gaussmeter coupled
with EMX 048T microwave bridge controller and EMX032T
field controller for calibration, a Bruker EMX 081 magnet and
an ERO 41XG microwave bridge were used. Spectral manipu-
lations and calculations were performed using the WINEPR
software package.11 Magnetic susceptibilities were measured
on a Sherwood Scientific magnetic susceptibility balance
(Cambridge Research Laboratories).
Potentiometric titrations
The stability constants were determined by potentiometric titra-
tion. All reagents were analytical grade, with analytical grade
methanol being fractionated and distilled from magnesium/
iodine before use. The apparatus consisted of a water-jacketed
measuring cell containing a Philips glass electrode (GA-110)
and a water-jacketed calomel reference electrode connected
by a salt bridge. A solution of base (tetraethylammonium
hydroxide) was introduced into the measuring cell by means
of a Metrohm dosimat 665 automatic titrator under personal
computer control.
Crystal structure determinations
Full spheres of CCD area-detector diffractometer data were
measured (Bruker AXS instrument, x-scans; monochromatic
Mo-Ka radiation, k = 0.71073 Å; T ca. 150 K) yielding Nt(total)
reflections, merging to N unique (Rint cited) after ‘empirical’/
multiscan absorption correction (proprietary software), No
with F > 4r(F) being considered ‘observed’ and used in the full
matrix least squares refinement, refining anisotropic thermal
parameter forms for the non-hydrogen atoms, (x, y, z, Uiso)H
being constrained at estimates, unless otherwise stated. Conven-
tional residuals R, Rw on |F| are cited at convergence (weights:
(r2(F) + 0.0004F2)−1). Neutral atom complex scattering factors
were used within the Xtal 3.7 program system.15 Pertinent results
are given below and in Tables 3–5 and in Electronic Supplemen-
tary Information† Table S1 (Coordinated anion and other
geometries [M(2)(NO3)2] (M = Co, Zn, Cd, Ni)) and Table S2
(Ligand conformational descriptors) as well as in Figs. 6–13
(these showing 50% displacement amplitude ellipsoids for the
non-hydrogen atoms, hydrogen atoms having arbitrary radii of
0.1 Å).
The protonation constants for the respective ligands and the
corresponding metal stability constants were obtained in 95%
methanol (I = 0.1 M, Et4NClO4), with the data processed using
a local version of MINIQUAD.12 The titration data for the metal
complexation were successfully refined assuming the presence of
only 1:1 metal–ligand species in solution; in virtually all cases
only data corresponding to the lower portions of the titration
curves were employed for the calculations in order to avoid
complications arising from competing hydrolysis/precipitation
at higher pH values. Quoted log K values (see Table 2) are
the mean of two or three separate determinations obtained at
different metal to ligand ratios (in each case measured on dupli-
cate sets of potentiometric titration apparatus). Unless other-
wise specified, all log K values obtained in separate titrations fell
within ±0.1 units (and usually within ±0.05 units) of the mean
value for a given metal/ligand system.
Liquid–liquid extraction experiments
The solvent extraction experiments were carried out at 26 ±
2 °C in polypropylene micro-centrifuge tubes (2 mL) with a
phase ratio V(aq) :V(org) of 1:1 (each phase 0.5 mL). The aqueous
phase contained the metal ion (at 1.0 × 10−4 M), a support-
ing anion, picric acid (5.0 × 10−3 M) or sodium nitrate (5.0 ×
10−3 M), and a selected buffer (used to maintain the chosen pH).
The pH of the aqueous phase was checked before and after each
experiment. The organic phase contained a known concentra-
tion of ligand in chloroform (normally 1.0 × 10−3 M except
where variable concentration experiments were employed).
All experiments involved mechanical shaking of the vials for
30 min, except for extractions involving cobalt(II) where they
were shaken for 3 h. At the end of these periods, the phases were
separated, centrifuged (to ensure full phase separation) and then
duplicate 100 lL samples of the aqueous and organic phases
were removed for analysis. Trial experiments were undertaken
in each case to confirm that equilibrium is attained within the
respective above time periods.
CCDC reference numbers 247455–247466.
See http://www.rsc.org/suppdata/dt/b4/b412437e/ for crystal-
lographic data in CIF or other electronic format.
Crystal/refinement data. [(2)H2](PF6)(PF2O2)·H2O·CH3CN 
C34H43F8N5O5P2, M = 815.7. Monoclinic, space group P21/c
(C52h, No.14), a = 12.0235(6), b = 18.5839(9), c = 17.1673(9) Å,
b = 96.688(1)°, V = 3810 Å3. Dc (Z = 4) = 1.422 g cm−3. lMo
=
2.0 cm−1; specimen: 0.15 × 0.15 × 0.12 mm; ‘T’min/max = 0.87.
2hmax = 75°; Nt = 78600, N = 20042 (Rint = 0.028), No = 13867;
R = 0.055, Rw = 0.065.
Variata. Refinement of (x,y,z,Uiso)H throughout establishes
the credibility of the solvent complement. The anion assign-
ment is based on refinement and geometrical considerations;
the PF6 anion is modelled with the fluorine atoms disordered
over two sets of sites, occupancies 0.875(2) and complement,
isotropic displacement parameter forms refined for the minor
component.
3 7 1 6
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6

View Article Online
Nt = 31495, N = 8234 (Rint = 0.025), No = 6797; R1(F) = 0.029,
wR2(F2) = 0.062 (A = 0.062, B = 0).
[M(2)(NO3)2]·2CH3OH  C34H44N6O10M (M = Co, Zn, Cd).
Monoclinic, space group P21/n (C52h, No.14 (variant)), Z = 4
(x,y,z,Uiso)H refined for the Co, Cd adducts.
(i) [Co(2)(NO3)2]·2CH3OH. M = 755.7. a = 17.586(2), b =
9.1457(8), c = 23.063(2) Å, b = 100.249(2)°, V = 3480 Å3. Dc =
1.442 g cm−3. lMo = 5.6 cm−1; specimen: 0.30 × 0.14 × 0.11 mm;
‘T’min/max = 0.95. 2hmax = 58°; Nt = 64342, N = 9231 (Rint = 0.043),
No = 7094; R = 0.042, Rw = 0.048.
[Ni(3)](NO3)2·H2O  C32H38N6NiO10, M = 709.4. Monoclinic,
space group P21/n, a = 11.013(4), b = 23.351(9), c = 12.836(5) Å,
b = 91.737(12)°, V = 3300 Å3. Dc (Z = 4) = 1.424 g cm−3. lMo
=
6.5 cm−1; specimen: 0.45 × 0.42 × 0.41 mm; ‘T’min/max = 0.90.
2hmax = 56.6°; Nt = 32067, N = 7807 (Rint = 0.018), No = 7274;
R1(F) = 0.074, Rw(F2) = 0.019 (A = 0.005, B = 8).
Variata. Difference map residues were modelled in terms of a
disordered water molecule oxygen, associated hydrogen atoms
not being located. The nitrate moieties also exhibit significant
disorder, the components being modelled as rigid bodies.
[Ni(5)](NO3)2·3CH3OH  C42H54N6NiO11, M = 877.6.
Triclinic, space group P1, a = 9.498(1), b = 11.094(2), c =
21.743(3) Å, a = 92.006(3), b = 95.444(3), c = 111.507(3)°, V =
2116 Å3. Dc (Z = 2) = 1.378 g cm−3. lMo = 5.3 cm−1; specimen:
0.38 × 0.14 × 0.11 mm; Tmin,max (Gaussian correction) = 0.84,
0.95. 2hmax = 56.7°; Nt = 21470, N = 9931 (Rint = 0.034),
No = 8015; R1(F) = 0.035, wR2(F2) = 0.080 (A = 0.03, B = 0).
(ii) [Zn(2)(NO3)2]·2CH3OH. M = 762.2. a = 17.637(4), b =
9.133(2), c = 23.161(4) Å, b = 110.049(6)°, V = 3505 Å3. Dc =
1.444 g cm−3. lMo = 7.7 cm−1; specimen: 0.14 × 0.10 × 0.09 mm;
‘T’min/max = 0.85. 2hmax = 55°; Nt = 12819 (hemisphere), N = 8070
(Rint = 0.067), No = 5492; R = 0.055, Rw = 0.070.
(iii) [Cd(2)(NO3)2]·2CH3OH. M = 809.2. a = 17.693(2),
b = 9.338(1), c = 22.992(3) Å, b = 109.128(3)°, V = 3589 Å3.
Dc = 1.497 g cm−3. lMo = 6.7 cm−1; specimen: 0.28 × 0.18 ×
0.12 mm; ‘T’min/max = 0.87. 2hmax = 75°; Nt = 46795, N = 17963
(Rint = 0.033), No = 13449; R = 0.038, Rw = 0.047.
[Ag(2)]BF4  C32H36AgBF4N4O2, M = 703.3. Monoclinic,
space group P21/c (C52h, No.14), a = 9.427(1), b = 16.268(2),
c = 20.346(3) Å, b = 101.348(5)°, V = 3059 Å3. Dc (Z = 4) =
1.527 g cm−3. lMo = 7.2 cm−1; specimen: 0.35 × 0.18 × 0.11 mm;
‘T’min/max = 0.87. 2hmax = 60°; Nt = 38639, N = 8774 (Rint = 0.048),
No = 6659; R = 0.049, Rw = 0.061.
Computational study
The DFT computations were performed with the Amsterdam
Density Functional 1999.02 (ADF)17 program package on
a Silicon Graphics Origin 2000 computer (56 MIPS R10000
64-Bit CPU with 195 MHz and 17 GB Memory). All struc-
tures were optimised without imposing symmetry constraints.
The molecular orbitals were expanded in an uncontracted set
of Slater-type orbitals (STO’s) containing diffuse functions.
The basis is of triple-zeta quality, augmented with one or two
polarisation functions. The cores were treated by the frozen-
core approximation.18 The numerical integration was performed
using the procedure developed by Baerends et al.19 The Becke–
Perdew (BP86) functional was used for the calculations, this
uses Becke’s20 gradient correction for the local expression of the
exchange energy and Perdew’s21 gradient correction for the local
expression of the correlation energy. The convergence criteria for
the geometry optimisations, which use analytical derivatives,22
were set to 1 × 10−3 hartree for the changes in energy, 1 × 10−3
hartree Å−1 for the energy gradient, 1 × 10−3 Å for the changes
between old and new bond lengths, and 0.3° for changes in bond
and dihedral angles. All stationary points on the hypersurface
were characterised as true minima by frequency analysis.
[Ag(4)]PF6  C39H42AgF6N4O2P, M = 851.6. Monoclinic,
space group P21/c, a = 10.362(1), b = 17.501(3), c = 20.358(3) Å,
b = 90.094(2)°, V = 3692 Å3. Dc (Z = 4) = 1.532 g cm−3. lMo
=
6.6 cm−1; specimen: 0.30 × 0.15 × 0.10 mm; ‘T’min/max = 0.89.
2hmax = 58°; Nt = 43278, N = 9393 (Rint = 0.032), No = 7121;
R = 0.040, Rw = 0.049.
Variata. The anion was modelled as rotationally disordered
about the PF(1,2) line, F(3–6) modelled over two sets of sites,
occupancies refining to 0.816(5) and complement. Disorder is
also observed in the cation, the silver atom being modelled over
two sets of sites, occupancies refining to 0.9081(7) and comple-
ment; no associated components of disorder in the ligand were
resolvable.
[Ag(5)]BF4  C39H42AgBF4N4O2, M = 793.5. Triclinic,
space group P1 (C1i, No. 2), a = 10.807(2), b = 11.020(2),
c = 16.153(2) Å, a = 73.702(2), b = 78.784(2), c = 72.043(2)°,
V = 1744 Å3. Dc (Z = 2) = 1.511 g cm−3. lMo = 6.4 cm−1;
specimen: 0.13 × 0.10 × 0.065 mm; ‘T’min/max = 0.87. 2hmax = 53°;
Nt = 16914, N = 7069 (Rint = 0.032), No = 5778; R = 0.042,
Rw = 0.048.
Syntheses
[Ag(6)]PF6·CH3CN  C46H48AgF6N4O2P·CH3CN, M = 982.8.
Monoclinic, space group P21/c, a = 15.995(2), b = 16.038(2),
c = 17.855(3) Å, b = 95.680(2)°, V = 4558 Å3. Dc (Z = 4) =
1.432 g cm−3. lMo = 5.5 cm−1; specimen: 0.80 × 0.22 × 0.04 mm;
The syntheses of 110,23 have been described previously.
Cobalt(II), nickel(II), zinc(II), and cadmium(II) nitrate com-
plexes of 2. The required hydrated metal(II) nitrate in methanol
(1.31 mL, 0.200 M) was added to a solution of 2 in methanol
(3.00 mL, 0.0872 M). The respective solutions were filtered
and crystallisation of the corresponding metal complexes were
induced by diffusion of diethyl ether vapour into each solution.
In each case, the product was removed by filtration and dried
under vacuum over phosphorus pentoxide. In most instances, in
separate experiments, crystals suitable for X-ray crystallography
were grown by slow diffusion of diethyl ether vapour into a
methanol or acetonitrile solution of the respective complexes.
‘T’min/max = 0.87. 2hmax = 58°; Nt = 53765, N = 11682 (Rint
=
0.039), No = 8704; R = 0.039, Rw = 0.043. (x,y,z,Uiso)H (ligand)
were refined.
[Cu(6)(OH)]PF6·2H2O  C46H49CuF6N4O3P·2H2O, M =
15
2h
950.5. Orthorhombic, space group Pbca (D , No. 61) a =
16.347(4), b = 20.768(5), c = 24.865(6) Å, V = 8442 Å3. Dc
(Z = 8) = 1.496 g cm−3. lMo = 6.4 cm−1; specimen: 0.38 ×
0.32 × 0.10 mm; ‘T’min/max = 0.81. 2hmax = 50°; Nt = 55561, N =
7400 (Rint = 0.075), No = 5000; R = 0.068, Rw = 0.095.
Variata. Difference map residues were modelled in terms of
oxygen atoms, tentatively assigned as hydroxide (coordinated)
or (lattice) water; associated hydrogen atoms were not resolved.
The following determinations were undertaken with the
following general variations in procedure: the refinement was
executed using SHELXL-97,16 Gaussian absorption correction
being applied in two instances, as indicated; reflection weights
[Co(2)(NO3)2]·0.5CH3OH·0.5H2O. Yield 52% [Found
(CoLNO3)+, m/z 629.1 (ES) C32H36CoN5O5 requires 629.2]
(Found:C,54.2;H,5.4;N,11.8.Calc.forC32H36CoN6O8·0.5CH3-
OH·0.5H2O:C, 54.47; H, 5.49; N, 11.73%).
2
were (r2(Fo ) + AP2 + BP)−1, P = (Fo2 + 2Fc2)/3 (A,B cited
[Ni(2)(NO3)2]·CH3OH·3H2O. Yield 62% [Found (NiL −
H)+, m/z 565.3 (ES) C32H35N4NiO2 requires 565.2] (Found: C,
51.2; H, 6.1; N, 10.8. Calc. for C32H36N6NiO8·CH3OH·3H2O: C,
50.98; H, 5.96; N, 10.81%).
below). R1(F), wR2(F2) are quoted.
[Ni(2)(NO3)2]·CH3CN·H2O  C34H41N7NiO9, M = 750.5.
Monoclinic, space group P21/n, a = 18.416(4), b = 8.693(2),
c = 23.139(5) Å, b = 109.485(4)°, V = 3492 Å3. Dc (Z = 4) =
1.427 g cm−3. lMo = 6.2 cm−1; specimen: 0.33 × 0.31 × 0.22 mm;
Tmin,max (Gaussian correction) = 0.80, 0.89. 2hmax = 56.5°;
[Zn(2)(NO3)2]·1.5CH3OH·2.5H2O. Yield 43% [Found
(ZnL − H)+, m/z 571.3 (ES) C32H35N4O2Zn requires 571.2]
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 1 7

View Article Online
Table 1 Physical properties of the complexes of 2–6
Complex
Colour
Ka
75
86
83
lb/lB
Electronic spectra c/nm (e/M−1 cm−1)
[Co(2)(NO3)2]·0.5CH3OH·0.5H2O
[Ni(2)(NO3)2]·CH3OH·3H2O
[Zn(2)(NO3)2]·1.5CH3OH·2.5H2O
[Ag(2)]BF4
[Cd(2) (NO3)2]·CH3OH·0.5H2O
[Ni(3)](NO3)2·2.5H2O
[Ag(4)]PF6
[Ni(5)](NO3)2·H2O
[Ag(5)]BF4
[Cu(6)OH]PF6·0.5H2O
[Ag(6)]PF6
Pink
4.59
3.37
470 (sh), 500 (e = 34), 545 (sh)
380 (e = 35), 610 (e = 22), 995 (e = 24)
Pale blue
Pale blue
Colourless
Colourless
Pale blue
Colourless
Pale blue
Colourless
Violet
91
110
119
96
111
93
89
89
3.17
3.36
560 (e = 35), 790 (e = 27), 825 (e = 32), 1200 (e = 11)
575 (e = 15), 780 (e = 26), 830 (e = 27), 1290 (e = 12)
515 (e = 194), 645 (e = 193)
Colourless
a Conductance (S cm2 mol−1) at 25 °C in methanol; expected range for a 1:1 electrolyte in methanol is 80–115 S cm2 mol−1 while that for a 2:1 electro-
lyte is 160–220 S cm2 mol−1 see ref. 24. b Magnetic moment at 25 °C. c In methanol.
[Ag(4)]PF6. Yield 64% [Found (AgL)+, m/z 705.3 (ES)
(Found: C, 50.7 H, 5.8; N, 10.7. Calc. for C32H36N6O8Zn·1.5-
C39H42AgN4O2 requires 705.2] (Found: C, 54.7; H, 5.4; N, 7.9.
Calc. for C32H36AgBF4N4O2: C, 54.65; H, 5.16; N, 7.97%).
CH3OH·2.5H2O: C, 50.86 H, 5.99; N, 10.62%).
[Cd(2)(NO3)2]·CH3OH·0.5H2O. Yield 47% [Found (CdL −
H)+, m/z 621.2 (ES) C32H35CdN4O2 requires 621.2] (Found: C,
50.3; H, 5.0; N, 10.5. Calc. for C32H36CdN6O8·CH3OH·0.5H2O:
C, 50.42; H, 5.26; N, 10.69%).
[Ag(6)]PF6. Yield 54% [Found (AgL)+, m/z 795.4 (ES)
C46H48AgN4O2 requires 795.3] (Found: C, 58.7; H, 5.1; N, 6.1.
Calc. for C46H48AgBF4N4O2: C, 58.67; H, 5.14; N, 5.95T%).
[Cu(6)OH]PF6·0.5H2O. Copper(I) hexafluorophosphate
in acetonitrile (2.96 mL, 0.05 M) was added to a solution of
6 in acetonitrile (10.0 mL, 0.015 M). The solution was filtered
and crystallisation of the metal complex was induced by slow
evaporation of the solution. Recrystallisation from aqueous
ethanol yielded the pure product as fine violet needles. The
product was removed by filtration and dried under vacuum
over phosphorus pentoxide. Yield 0.027 g, 14% [Found (CuL)2+,
m/z 345.3 (ES) (C46H48CuN4O2)/2 requires 345.2] (Found: C,
59.7; H, 5.3; N, 6.1. Calc. for C46H48CuF6N4O2P·5H2O: C,
59.83; H, 5.46; N, 6.07%). Crystals suitable for X-ray crystallo-
graphy were grown by slow evaporation of a solution of the
complex in acetonitrile.
An attempt to synthesise the silver(I) hexafluorophosphate
complex of 2 yielded an X-ray quality crystal whose X-ray
structure indicates that hydrolysis of hexafluorophosphate
has occurred; with the product being of type [(2)H2](PF6)-
(PF2O2)·H2O·CH3CN incorporating a doubly protonated ligand
together with PF2O2 and PF6 anions.
Silver(I) tetrafluoroborate complexes of 2 and 5. Silver(I)
tetrafluoroborate (51 mg, 0.26 mmol) in methanol (2 mL)
was added to a solution of the respective ligand in methanol
(3.00 mL, 0.26 mmol). The solution was filtered and crystal-
lisation of the metal complex was induced by diffusion of
diethyl ether vapour into the reaction solution. The product was
removed by filtration and dried under vacuum over phosphorus
pentoxide. Crystals suitable for X-ray crystallography were
grown by slow diffusion of diethyl ether into a methanol
solution of the respective complexes.
[Ag(2)]BF4. Yield 41% [Found (AgL)+, m/z 615.2 (ES)
C32H36AgN4O2 requires 615.2] (Found: C, 54.7; H, 5.4; N, 7.9.
Calc. for C32H36AgBF4N4O2: C, 54.65; H, 5.16; N, 7.97%).
[Ag(5)]BF4. Yield 47% [Found (AgL)+, m/z 705.2 (ES)
C39H42AgN4O2 requires 705.2] (Found: C, 59.0; H, 5.4; N, 7.1.
Calc. for C39H42AgBF4N4O2: C, 59.04; H, 5.38; N, 7.12%).
Nickel(II) nitrate complexes of 3 and 5. These compounds
were obtained by an analogous procedure to that described
above for the corresponding complex of 2. Crystals suitable for
X-ray crystallography were grown by slow diffusion of diethyl
ether into a methanol solution of the respective complexes.
Results and discussion
Metal complexes
In an attempt to obtain crystalline products suitable for X-ray
diffraction studies, solid metal complexes of 2–6 with cobalt(II),
nickel(II), copper(II), zinc(II), cadmium(II) and silver(I) was
isolated and characterised (Table 1). The conductance values
in methanol (Table 1) for all the complexes approximate those
expected for a 1:1 electrolyte thus indicating that the silver(I)
complexes are essentially completely ionised under the condi-
tions employed, while the cobalt(II), nickel(II), copper(II),
zinc(II) and cadmium(II) complexes retain one bound anion in
this solvent.24
The magnetic moments (Table 1) of the pink cobalt complex
of 2 and the blue nickel complexes of 2, 3 and 5 confirm their
high spin nature. The visible spectra in methanol are consistent
with these complexes having distorted octahedral coordination
geometries.25
[Ni(3)](NO3)2·2.5H2O. Yield 47% [Found (NiL − H)+, m/z
565.3 (ES) C32H35N4NiO2 requires 565.2] (Found: C, 52.4; H,
5.4; N, 11.5. Calc. for C32H36NiN6O8·2.5H2O: C, 52.19; H, 5.61;
N, 11.41%).
[Ni(5)](NO3)2·H2O. Yield 52% [Found (NiLNO3)+, m/z
718.1 (ES) C39H42N5NiO5 requires 718.3] (Found: C, 58.3; H,
5.5; N, 10.4. Calc. for C39H42N6NiO8·H2O: C, 58.59; H, 5.55; N,
10.51%).
Silver(I) hexafluorophosphate complexes of 4 and 6. Silver(I)
hexafluorophosphate (25 mg, 0.10 mmol) in acetonitrile (2 mL)
was added to 4 in methanol or 6 in 1:1 acetonitrile:chloroform
(4 mL, 0.10 mmol). The solution was filtered and crystallisation
of the metal complex was induced by diffusion of diethyl ether
vapour into the filtrate. The product was removed by filtration
and dried under vacuum over phosphorus pentoxide. Crystals
suitable for X-ray crystallography were grown by slow diffusion
of diethyl ether into an acetonitrile solution of the respective
complexes.
The visible spectrum of the copper complex [Cu(6)OH]-
PF6·2H2O showed
a broad envelope of bands in the
450–750 nm region with maxima occurring at 516 and 644 nm
respectively. The shape of the observed EPR signal at liquid
nitrogen temperature (DMF glass) suggests the presence of
two copper species under the conditions employed, each with
the metal in an axially symmetrical environment elongated
3 7 1 8
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6

View Article Online
Table 2 Ligand protonation constants and metal stability constants for 1, 2, 5 and 6 a
b
Log KH
Log KML
Ligand
K1
K2
K3
Co2+
Ni2+
Zn2+
Cd2+
Cu2+
Ag+
Pb2+
1
2
5
6
9.86
10.26
9.64
8.58
7.55
6.14
5.97
—
—
—
—
9.2
10
—
9.4
7.8
-
9.5
8.4
6.1
>14
>14
12.5
>13
>14
10.2
>8
9.8
7.5
5.5
c
-
-
c
c
4.7
c
c
8.95
<4
-
<4
<4
-
<4
a In 95% methanol; I = 0.1 M, (C2H5)4NClO4, 25 °C; unless otherwise noted ligand protonation constants (log K1 and log K2) are ±0.05 while log K
values for complex formation are ±0.1; ligand 4 proved to be of too low solubility for study. b Each value is the mean of at least two (and up to five)
c
determinations at different metal:ligand ratios rounded to the first decimal place; data for 1 taken from ref. 23. Precipitation, slow approach to
equilibrium or competing hydrolysis inhibited log K determination in this case.
along the z-axis (as g|| > g⊥). Two g|| values could be assigned
(2.204 and 2.154); however, only one g⊥ value could be dis-
cerned, 2.082, reflecting the complex nature of the spectrum.
No superhyperfine splitting due to coupling to 14N nuclei was
observed. While the two species perhaps arise from the presence
of ‘frozen’ conformers, in the absence of further evidence it
appears inappropriate to speculate further concerning this.
Potentiometric titrations
The protonation constants for the respective macrocycles were
obtained by potentiometric (pH) titration in 95% methanol
(I = 0.1 M, N(C2H5)4ClO4) and are listed in Table 2. As in pre-
vious studies, this solvent mixture was employed to overcome
the generally restricted solubility of ligands of the present type
Fig. 1 Silver(I) extraction (water/chloroform) by ligands 1–6 with
(and/or their metal complexes) in water. Where possible, log K
picrate or nitrate as the anion at pH 6.2 (MES/NaOH buffer) and
values corresponding to the formation of 1:1 (metal:ligand)
species were obtained in each case by analysis of the initial por-
tions of the respective titration curves. The log K data for 2, 5
and 6 are summarised in Table 2 and, for comparison, stability
data for the parent mixed-donor system 123 is also included. The
general trend of a decrease in the pKa values on progressive
N-benzylation follows previously observed trends.5
Unfortunately, owing to precipitation and/or competing
hydrolysis, it was only possible to obtain log K values for a
selection of the 1:1 complexes of the seven metals investigated
with 2, 5 and 6 (see Table 2). Nevertheless, a comparison of
the values that were obtained with those for the unsubstituted
parent ring, 1 indicates that, in general, selectivity for copper(II)
and silver(I) appears to be maintained as the degree of
N-benzylation is increased along 2–6. While the situation is not
clear cut, overall the results are not inconsistent with ‘selective
detuning’ taking place,5 especially when the relative log K values
for the respective ligand complexes of the remaining five metals
are compared.
26 ± 2 °C; [L] = 1 × 10−3 M.
order of Hofmeister.27 This unexpected effect of the nitrate
counter ion on the silver extraction could reflect either coordi-
nation of the nitrate direct to the silver ion or its binding to a
(secondary) amino hydrogen site on the ligand. The first situa-
tion was observed for the nitrate complexes of 2 with cobalt(II),
nickel(II), zinc(II) and cadmium(II) (see Figs. 7 and 9) in the
solid state, while the second corresponds to the situation in the
solid copper(II) nitrate complex of 1.23 Similar possible binding
patterns for the extraction system leading to increased silver
extraction has also been discussed for amino cryptands28 and
the silver complex of an azathia-macrocycle incorporating an
urea side arm.29 In the case of ligand 1, the extraction of silver(I)
in the presence of nitrate is clearly lower than when picrate anion
is present. This most likely reflects the lower lipophilicity of
this unsubstituted ligand leading to protonated ligand and/or
complex bleeding into the aqueous phase. For the interpreta-
tion of the extraction results it is interesting to note that for the
lipophilic tribenzylated ligand, 6, extraction of silver(I) in the
presence of nitrate is also lower than when picrate is present. The
latter presumably reflects the lack of secondary amines in the
structure of 6 in comparison with 2–5 leading to the inhibition
of (hydrogen bond) association of this ligand with the nitrate
anion in the organic phase — thus favouring the observed high
silver extraction with ligands 2–5 in the presence of nitrate.
In an attempt to determine the stoichiometry of the extracted
species, extraction experiments were performed in which the
concentration of the ligand was varied while the metal ion con-
centration was maintained at 1 × 10−4 M and the anion concen-
tration was effectively constant. The log DM (= [Mn+]org/[Mn+]aq)
was then plotted against the log of the macrocycle concentra-
tion and, provided a ‘simple’ equilibrium is involved, the slope
of this plot gives the stoichiometry of the extracted species
directly. Thus:
Liquid–liquid extraction studies
Liquid–liquid extraction experiments (water/chloroform) using
the radiotracer technique employing the isotopes 110mAg, 65Zn,
60Co and 64Cu and 1–6 have been carried out. The chloroform
phase consisted of a solution of the macrocycle (1.0 × 10−3 M)
while the aqueous phase contained the metal (1.0 × 10−4 M)
and picric acid (5.0 × 10−3 M). This phase was buffered at pH
6.2 with MES/NaOH buffer. The initial experiments involved
the extraction of silver(I) under these conditions and extraction
values of between 85 and 95% were obtained for all systems
(Fig. 1). That is, 1–6 are all excellent extractants for this ion
under the conditions employed.
For comparison, extractions in the absence of picrate ion but
in the presence of a corresponding concentration of sodium
nitrate (5.0 × 10−3 M), employing the same buffer system, were
carried out (Fig. 1). Surprisingly, the extraction efficiency using
the partial benzylated systems 2–5 is comparable in the presence
of either nitrate or picrate. This result is in contrast to most
metal extraction systems with organic ligands26 showing the
anion influence on metal extraction follows the lipophilicity
Mn+(aq) + nA−(aq) + sL(org)  MLsAn(org)
(1)
[MLsAn ](org)
KEx
=
(2)
[Mn+
]
[L](sorg) [A− ](naq)
(aq)
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 1 9

View Article Online
log DM = log KEx + n log[A−](aq) + s log[L](org)
(3)
For ligand 4 (and silver nitrate) variable ligand concentra-
tion experiments were carried out at two different pH values
(Fig. 2). The gradient in both cases approximated one, reflect-
ing the expected 1:1 (ligand to metal) complex stoichiometry
for this system in solution. Ligand 1 was also used for a range
of similar extraction experiments with cobalt(II), zinc(II),
copper(II), and silver(I) in the presence of nitrate and picrate,
but unfortunately, the results indicated that bleeding of the
respective complexes into the aqueous layer occurred for this
less lipophilic ligand system.
Fig. 4 Percent copper(II) and silver(I) extraction by 1 and 6 with
picrate as the anion at pH 6.2 (MES/NaOH buffer) and 26 ± 2 °C.
Membrane transport
In addition to the single ion extraction studies competitive
mixed metal transport experiments (water/chloroform/water)
have been undertaken for which the chloroform membrane
phase contained the ionophore, chosen from 1–6 respectively,
at 1 × 10−3 M. The aqueous source phase contained equimolar
concentrations of the nitrate salts of cobalt(II), nickel(II),
copper(II), zinc(II), cadmium(II), silver(I) and lead(II), with indi-
vidual metal ion concentrations being 1.0 × 10−2 M. Transport
was performed against a back gradient of protons with palmitic
acid (4 × 10−3 M) also present in the organic phase. A main role
of the latter is to aid the transport process by providing a lipo-
philic counterion in the organic phase (after proton loss to the
aqueous source phase) for charge neutralisation of the metal
cation being transported.
Transport flux values for both copper(II) and silver(I) for
each of 1–6 are illustrated in Fig. 5. The values for cobalt(II),
nickel(II), zinc(II), cadmium(II) and lead(II) were all less than
20 × 10−7 mol (24 h)−1 indicating negligible transport under
the present conditions (values are within experimental error of
zero).30 Clearly, the parent ligand 1 is a poor ionophore for all
seven metals (although marginal transport of copper(II) was
detected), a fact that is again probably related to its lower lipo-
philicity arising from the absence of N-benzyl substituents. In
accord with this, the singly benzylated species 2 and 3 yielded
enhanced transport of copper(II) but now silver(I) transport is
observed as well. Moving to the di-N-benzylated derivatives 4
and 5 these both show enhanced selectivity and efficiency for
silver(I) transport. Within the limits of accuracy of the experi-
ments, sole selectivity for silver was attained in the case of 4,
with 5 yielding similar behaviour towards silver(I) but with a
minor amount of copper(II) also being transported in this case.
The final ionophore, 6, in which all three ring nitrogens are
benzylated, also yielded sole transport selectivity for silver(I),
although the efficiency of transport was somewhat less in this
case relative to that observed for 4 and 5. Under the present
conditions, the di- and tri-benzylated ligands 4–6 each show
marked selectivity for silver over the remaining six metal ions
present in the aqueous phase. As before, this result is in general
accord with these ligand systems inducing discrimination for
silver via a ‘selective detuning’ mechanism.
Fig. 2 Variable ligand concentration experiments for 4 with silver(I)
and nitrate as the anion at pH 4.2 (citric acid/NaOH buffer) and pH.
6.2 ± 0.1 (MES/NaOH buffer), 26 ± 2 °C.
A comparison of the percent extraction for each of the
ligands 1–6 for the three metals cobalt(II), zinc(II), and silver(I),
with picrate as the anion (MES/NaOH buffer, pH 6.2 ± 0.1,) is
given in Fig. 3. High extraction efficiencies for silver(I) remain
effectively little changed on progressive N-benzylation along
this ligand series, whereas for cobalt(II) and zinc(II), with the
possible exception of the system incorporating 1, the percent
extraction was minimal (or zero) in all cases for 1.
Fig. 3 Percent cobalt(II), zinc(II), and silver(I) extraction by ligands
1–6 with picrate as the anion at pH. 6.2 (MES/NaOH buffer) and
26 ± 2 °C.
A comparison of the extraction behaviour of copper(II) by 1
and 6 against that for silver(I) is shown in Fig. 4. While extrac-
tion of silver(I) remains high for the tri-N-benzyl derivative 6,
the degree of extraction decreases significantly for copper(II)
for this ligand. That is, overall, the latter system yields consider-
ably enhanced extraction discrimination for silver over copper
relative to its parent (unsubstituted) ring. Combining this result
with the results for cobalt and zinc (Fig. 3) indicates that, over-
all, 6 gives rise to excellent discrimination for silver over each
of the other metal ions investigated.
X-Ray structures
Single crystal ‘low’-temperature X-ray structure determi-
nations have been carried out for a salt of the ligand 2,
three of its divalent metal nitrate complexes and a silver(I)
tetrafluoroborate complex. In all cases, one formula unit devoid
of crystallographic symmetry comprises the asymmetric unit
of the structure, the stoichiometries and connectivities being
consistent with the formulations given above; the three divalent
metal complexes, [M(2)(NO3)2]·methanol solvate (M = Co,
3 7 2 0
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6

View Article Online
Fig. 5 The copper(II) and silver(I) transport fluxes for the six ligands
1–6 in the chloroform membrane transport experiment.
Zn, Cd), are isomorphous. The protonated ligand, [(2)H2]2+,
adopts a conformation that is close to m symmetry (Fig. 6); in
the metal complexes this is followed closely except in the bonds
to either side of the central nitrogen which differ, breaking that
symmetry.
Fig. 7 Projection of the [Cd(2)(NO3)2] molecule (as broadly represen-
tative of the cobalt(II) and zinc(II) species also).
In the cobalt(II), zinc(II) and cadmium(II) complexes of
2, the metal supplants the anion oxygen atom in respect of
its interactions with the deprotonated chelating amines, the
local symmetry about N(0) distorting so that the pair of five-
membered metallacycle rings are quasi-2, rather than quasi-m,
in their relationship. The coordination sphere about the metal
comprises this pair of nitrogen atoms and N(0) between them,
as a tridentate, with a pair of chelating nitrate groups. The
latter vary in their interactions with the metal, nitrate 2 essen-
tially symmetrically chelating, while nitrate 1 is unsymmetrically
chelating, the more so as the metal size diminishes (Table 3).
The differences in the nature of the associated oxygens are
reflected in the distances and angles about the central nitrogen
atoms. In the copper(II) complex of 6 (Fig. 8), wherein the
effective coordination number and geometry are driven toward
planar four-coordinate by the electron configuration, the ligand
(indeed the cation) conformation becomes quasi-m once more,
but with a different disposition consequent upon the N3-donors
of the pair of metallacycles becoming mer rather than fac in
the coordination sphere. Four-coordinate copper complexes
that partner a hydroxo ligand with three coordinating nitrogen
atoms are rare, with only two being listed on the current release
of the Cambridge Structural Database.31
Fig. 6 Projection of the [(2)H2(H2O)(PF2O2)]+ aggregate.
The formulation of the ligand salt is adventitious, complex
and interesting (Fig. 6). Hydrogen atom refinement establishes
the cation to be doubly protonated, [(2)H2]2+. Charge balance is
achieved viaonehexafluorophosphateanion, slightlydisordered,
and a second counterion, modelled plausibly in the refinement
−
as F2O2 , a commonly occurring hexafluorophosphate hydro-
lysis product. The latter, together with a water molecule
make up a cationic cluster (Fig. 6), [(2)H2(H2O)(PF2O2)]+,
the final component an acetonitrile molecule, devoid of close
interactions. The structure of the cluster is interesting, the
oxygen of the water molecule ‘chelated’ by a pair of protonated
N(3,3) amine hydrogen atoms O(01)H(3a,3b) 2.19(2),
2.11(2) Å, while unprotonated N(61) interacts with a water
hydrogen, N(61)H(01a) 2.02(3) Å, a motif that also occurs in
the solvating methanol/water interactions of the set of divalent
metal complexes incorporating 2 (see Fig. 7). The other pair of
protonated amine hydrogen atoms chelates one of the anionic
oxygen atoms: O(11)H(3b,3a) 1.91(2), 1.97(2) Å. The other
water molecule hydrogen interacts with the other (symmetry
related) anionic oxygen O(12)H(O1b) (x, ꢀ − y, ꢀ + z)
1.99(2) Å, linking the clusters into strings.
Fig. 8 Projection of the [Cu(6)(OH)]+ cation, projected normal to the
coordination plane.
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 2 1

View Article Online
Table 3 Metal atom environments, [M(2)(NO3)2]; r is the metal–donor atom distance (Å), the other entries being the angles (degrees) subtended by
the relevant atoms at the head of the row and column. The four values in each entry are for M = Co, Zn, Cd, Ni respectively
Atom
N(0)
r
N(3)
N(3)
O(11)
O(12)
O(21)
O(22)
2.215(2)
2.221(3)
2.412(2)
2.122(1)
2.106(2)
2.086(3)
2.302(1)
2.079(1)
2.117(2)
2.103(3)
2.308(1)
2.068(1)
2.863(2)
2.905(4)
2.526(2)
3.289(1)
2.057(2)
2.072(3)
2.343(2)
2.077(1)
2.179(2)
2.232(3)
2.478(2)
2.118(1)
2.263(2)
2.367(3)
2.400(1)
2.129(1)
81.43(7)
83.0(1)
77.17(5)
84.69(5)
81.70(6)
82.8(1)
141.32(6)
143.7(1)
146.09(5)
152.13(4)
74.49(6)
74.5(1)
80.20(5)
118.94(4)
76.52(6)
77.6(1)
166.88(7)
166.3(1)
154.01(5)
166.84(4)
111.62(7)
110.3(1)
128.64(6)
86.87(4)
95.46(6)
95.7(1)
90.18(5)
88.08(5)
48.61(6)
47.8(1)
88.88(7)
89.7(1)
97.45(6)
94.7(1)
99.13(5)
98.83(4)
90.81(6)
90.3(1)
87.23(5)
96.35(4)
162.84(6)
161.0(1)
163.63(5)
157.16(4)
112.49(6)
113.3(1)
104.67(5)
66.02(3)
81.47(6)
82.3(1)
77.16(5)
84.65(5)
105.93(6)
108.0(1)
107.15(5)
106.46(4)
88.59(5)
92.22(4)
145.62(6)
144.7(1)
134.70(5)
156.59(4)
105.15(6)
105.3(1)
111.19(5)
96.32(4)
127.42(6)
124.8(1)
125.11(5)
60.13(4)
79.46(7)
77.5(1)
N(3)
N(3)
O(11)
O(12)
O(21)
O(22)
85.89(5)
100.84(4)
52.59(5)
40.47(3)
74.80(6)
99.49(4)
86.62(5)
92.12(4)
57.70(5)
55.8(1)
52.51(5)
61.12(3)
The structure of the nickel(II) counterpart of 2 has also been
determined (Fig. 9). With a different solvate complement, this
complex is not isomorphous with its cobalt(II), zinc(II) and
cadmium(II) analogues, presumably as a consequence of the
effective nickel atom radius being smaller than that of the other
metal atoms (Table 3). Although the relative dispositions of the
components of the molecule are qualitatively similar to those
of the other arrays (Fig. 7), the differences are significant and
substantial, chiefly arising from changes in the disposition of
nitrate 1 which is now clearly unidentate with Ni–O(11) being
very long at 3.289(1) Å. (In the atom numbering O(12) defines
the shorter of the pair of Ni–O bonds from this nitrate; however,
a more appropriate perception is that, in terms of the Co, Zn and
Cd numbering, the longer and shorter of the two bonds have
interchanged their roles, rather than the nitrate group having
rotated about the N–O(13) (uncoordinated) bond by ca. 180°.)
With the silver(I) complexes (Figs. 10–13), a wide diversity
of forms is found; in each case, one formula unit comprises the
asymmetric unit, the [AgL]+ (L = 2, 4–6) cations devoid of sym-
metry and showing close interactions with other species. Given
the predominance of nitrogen donors, some tendency toward
linear N–Ag–N arrays would not be surprising. Although all
silver complexes have coordination numbers greater than two,
such a tendency towards linear coordination is nonetheless
Fig. 9 Projection of the [Ni(2)(NO3)2] molecule showing the modifi-
cation of the coordination of nitrate 1 to become unidentate, and the
replacement of the hydrogen bonded methanol solvent molecule by
water.
Fig. 10 Proection of the [Ag(2)]+ cation.
3 7 2 2
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6

View Article Online
Fig. 12 Projection of the [Ag(5)]+ cation.
Fig. 13 Projection of the [Ag(6)]2+ cation.
Fig. 11 (a, b) Projections of the major and minor components of the
[Ag(4)]+ cation.
familiar Ag/aromatic ring interactions, are observed [Fig. 11(b)];
while such behaviour for the silver complexes is not widespread,
contacts of this type could, of course, be of greater significance
in solution.
Similar nickel environments occur in the nickel(II) complexes
of 3 (Fig. 14) and 5 (Fig. 15), with the ligands in both cases
behaving as sexadentates with similar conformations, presum-
ably in response to a more stringent quasi-octahedral coordina-
tion imperative for this metal ion.
Clear-cut strong interactions in the crystal packings of the
metal complexes are relatively limited. Well-defined contacts
from the coordinated anions to the hydroxylic solvent mole-
cules are observed in the cobalt(II), zinc(II), cadmium(II) and
nickel(II) complexes of 2 (Table 5), NH(3,3) also ‘chelating’.
In the copper complex of 6, close approaches are found to the
putative coordinated hydroxide by the putative lattice water
molecules (O(01)O(02,03) 2.68(1), 2.61(1) Å. No associated
hydrogen atoms were located and there are diverse contacts to
the ordered PF6− anion; the crystal packing has incipient higher
symmetry (Fig. 16). Among the silver complexes, only one is
suggested by the observation that in [Ag(2)]+ (Fig. 10), the
N(0)–Ag–N(61) angle is 176.5(1)° and in the [Ag(6)]+ the angle
is 160.53(8)° (Fig. 13); the shortest Ag–N distance among these
is 2.353(2) Å. With coordination numbers effectively greater
than two, the N(3)N(0)N(3) triad might be expected to form
an effective tridentate. In [Ag(2)]+ this might be considered to be
the case (Fig. 10), with the N(3)–Ag–N(3) angle being the larg-
est among the four [AgL]+ cations, but still only 147.0(1)° (see
Table 4). In [Ag(4)]+, N(3) is well removed from the silver atom
and the above motif is effectively non-existent (Fig. 11). The
effects of the pendant groups on the macrocycle conformation
also are diverse, and, one suspects, predominantly indirect by
way of crystal packing effects. Parallel approaches of aromatic
planes are seen to be widespread. Of particular interest is the
disorder observed in [Ag(4)]+ where an alternative Ag+ site at
10% occupancy is identified, well removed from the main com-
ponent. In this component the pendant ring C(3n) is located
in such a manner that Agring contacts, suggestive of the
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 2 3

View Article Online
Table 4 Silver environments, [AgL]+, (L = 2, 4, 5 and 6) and nickel environments [NiL]2+, L = 3 and 5. Presentation as in Table 1. The six values in
each entry are for [Ag(2)]+, [Ag(4)]+,a [Ag(5)]+, [Ag(6)]+, [Ni(3)]2+, [Ni(5)]2+ respectively
Atoms
N(0)
r
N(3)
N(3)
O(5)
O(5)
N(61)
2.455(3)
73.03(11)
(–)
74.86(10)
76.09(7)
83.3(1)
83.80(5)
74.60(10)
76.49(8)
73.44(9)
70.41(7)
82.2(1)
83.10(5)
147.00(10)
(–)
103.35(10)
131.45(6)
164.8(1)
164.75(6)
118.38(8)
104.42(7)
90.99(8)
134.09(6)
105.4(1)
105.63(5)
69.95(8)
(–)
71.26(8)
72.12(6)
88.3(1)
87.60(4)
133.40(8)
142.34(6)
164.42(8)
108.51(6)
91.0(1)
120.03(8)
106.10(7)
138.05(10)
102.04(6)
99.6(1)
102.07(5)
121.46(8)
(–)
133.42(8)
156.52(6)
98.7(1)
100.81(4)
70.90(9)
78.70(6)
70.14(8)
66.77(6)
88.3(1)
89.55(5)
121.11(7)
133.92(5)
124.43(7)
120.07(5)
154.7(1)
151.76(5)
176.48(11)
139.85(8)
133.70(8)
160.53(8)
177.5(1)
177.95(6)
103.50(11)
(–)
122.47(11)
119.35(7)
98.1(1)
98.06(5)
108.78(10)
137.32(7)
129.98(10)
102.40(7)
96.5(1)
95.18(5)
60.17(8)
63.79(6)
61.82(8)
65.07(6)
76.8(1)
75.37(5)
61.07(8)
70.58(6)
64.25(8)
58.93(6)
78.2(1)
2.232(2)
2.523(3)
2.425(2)
2.046(3)
2.129(2)
2.520(3)
(3.737(2))
2.493(3)
2.513(2)
2.176(3)
2.154(1)
2.510(3)
2.508(2)
2.483(3)
2.700(2)
2.107(3)
2.073(1)
2.965(3)
3.061(2)
2.938(3)
2.687(2)
2.130(3)
2.197(1)
2.957(3)
2.645(2)
2.887(2)
2.891(2)
2.130(2)
2.151(1)
2.353(3)
2.240(2)
2.414(3)
2.373(2)
2.019(3)
2.055(2)
N(3)
N(3)
O(5)
88.44(4)
O(5)
N(61)
76.77(5)
a In [Ag(4)]+, Ag(1) distances are 2.266(3), 2.565(3), 3.414(3), 2.912(3), 4.985(3), 4.152(3) Å respectively. Ag(1)Ag(1) is 2.402(2) Å; Ag(1) also
has contacts to C(32,32,33,46) at 2.751(4), 2.472(4), 2.771(4), 3.044(3) Å.
Fig. 14 Projection of the [Ni(3)]2+ dication.
solvated [Ag(6)]PF6·CH3CN, the solvent showing high displace-
ment parameters and seemingly having little role in determining
the structure, a comment generally true of the quasi-spherical
(XFn)− anions which, albeit, are in most cases well-ordered, but
with high displacement parameters. The NH(3,3) hydrogens
are no longer appropriately oriented for chelation of anion or
Fig. 15 Projection of the [Ni(5)]2+ dication.
solvent in either these or the nickel(II) complexes of 3 and 5.
carried out to evaluate the relative energies of the two forms
as well as to probe the nature of the interaction between the
DFT calculations
As mentioned above, the X-ray structure of [Ag(4)]PF6 shows
disorder of the silver over two sites. DFT calculations were
silver ion and the donor atoms and/or the p-faces of particular
aromatic rings within the structures. The calculated geometry of
3 7 2 4
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6

View Article Online
Table 5 Hydrogen-bonding interactions in [M(2)(NO3)2]. Non-hydrogen distances are given for M = Co, Zn, Cd, the second entries (in parentheses)
being the refined hydrogen distancea
Atoms
Distances/Å
N(3)(H(3))O(10)
N(3)(H(3))O(10)
O(10)(H(10))N(61)
O(20)(H(20))O(22)
3.066(2) (2.27(2))
3.012(3) (2.32(2))
2.824(3) (2.06(4))
2.878(3) (1.99(5))
3.056(4)
3.011(4)
2.827(5)
2.885(5)
3.115(2) (2.34(2))
3.053(4) (2.35(2))
2.802(3) (2.01(3))
2.839(2) (1.97(4))
a In the nickel complex, N(3,3)O(10) are 3.102(2), 2.976(2) and O(10)N(61), O(11) are 2.860(2), 2.848(2) Å.
for silver(I) relative to the other six metals investigated. Such
behaviour is thus in accord with the use of selective ‘detuning’
as a mechanism for achieving metal ion discrimination.5 Based
on the discussion on the effects of N-alkylation presented
in the Introduction, it appears likely that the origins of the
observed behaviour involve a number of influences. Since, N-
benzylation is involved in the present systems, an additional
factor may contribute in such cases. That is, there is a prospect
that the relative stabilities of the silver complexes are enhanced
by the occurrence of aryl–silver p-interactions in solution;
as mentioned already, such interactions have now been well
demonstrated for silver(I) in the complexes of a wide variety of
aryl-containing ligands.32–34 However, even though the present
X-ray and computational studies suggest that such interactions
may occur in the structures of particular complexes in both the
solid state and gas phase, in the absence of further evidence it
appears inappropriate to comment further on the likely effect (if
any) of such p-interactions on the selective detuning behaviour
documented in the present report.
Fig. 16 Unit cell contents of [Cu(6)(OH)]PF6⋅2H2O projected
down a.
the lowest found minimum structure (which corresponds to the
‘major’ component in the crystal structure) is given in the elec-
tronic supplementary information as Fig. S1.† As observed in
the crystal structure, this calculated minimum structure shows a
distorted four-coordinate array about the silver ion that involves
the secondary nitrogen atom, one tertiary nitrogen atom, one
pyridine nitrogen atom and one oxygen atom. While two
C–Haryl p interactions are evident (2.65 and 2.81 Å) within
the folded ligand structure, the calculated charge distributions
indicate that charge transfer from each p region of the aromatic
ring to the corresponding C–H group is quite small.
Acknowledgements
We thank the Australian Research Council, and the Deutsche
Forschungsgemeinschaft for funding to support the collabo-
ration between the research groups involved in this work. We
acknowledge Ms U. Stöckgen of TU Dresden for assistance
with the radiotracer studies.
References
1 T. W. Hambley, L. F. Lindoy, J. R. Reimers, P. Turner, G. Wei and
A. N. Widmer-Cooper, J. Chem. Soc., Dalton Trans., 2001, 614;
A. N. Widmer-Cooper, L. F. Lindoy and J. R. Reimers, J. Phys.
Chem. A, 2001, 105, 6567.
2 J. Kim, T.-H. Ahn, M. Lee, A. J. Leong, L. F. Lindoy, B. R. Rumbel,
B. W. Skelton, T. Strixner, G. Wei and A. H. White, J. Chem. Soc.,
Dalton Trans., 2002, 3993.
3 M. Fainerman-Melnikova, A. Nezhadali, G. Rounaghi, J. C.
McMurtrie, J. Kim, K. Gloe, M. Langer, S. S. Lee, L. F. Lindoy,
T. Nishimura, K.-M. Park and J. Seo, Dalton Trans., 2004, 122.
4 Y. Dong, S. Farquhar, K. Gloe, L. F. Lindoy, B. R. Rumbel,
P. Turner and K. Wichmann, Dalton Trans., 2003, 1558.
The calculated geometry of the second (minimum energy)
structure corresponding to the minor position within the dis-
ordered crystal structure is given as Fig. S2.† Once again, the
generated structure strongly resembles the X-ray structure.
A coordination arrangement in which the silver is bound to
the same macrocyclic ligand donors as those observed in the
first structure is present; however, in this case one aryl ring also
makes a close contact with the metal ion, with an Ag–C dis-
tance of 2.45 Å. Previously it has been noted that experimental
Ag–C distances in the range 2.45–2.49 Å are typical of g1 or g2
coordination in silver–arene complexes,31 although a g1 distance
5 L. F. Lindoy, Pure Appl. Chem., 1997, 69, 2179.
6 A. S. Craig, R. Kataky, R. C. Matthews, D. Parker, G. Ferguson,
A. Lough, H. Adams, N. Bailey and H. Schneider, J. Chem. Soc.,
Perkin Trans. 2, 1990, 1523; H.-J. Buschmann, E. Schollmeyer,
R. Trueltzsch and J. Beger, Thermochim. Acta, 1993, 213, 11.
7 V. P. Solov’ev, N. N. Strakhova, V. P. Kazachenko, A. F. Solotnov,
V. E. Baulin, O. A. Raevsky, V. Rüdiger, F. Eblinger and H.-J.
Schneider, Eur. J. Org. Chem., 1998, 1379; R. M. Izatt, K. Pawlak,
J. S. Bradshaw and R. L. Bruening, Chem. Rev., 1991, 91, 1712;
R. D. Hancock, Coord. Chem. Rev., 1994, 133, 39.
as short as 2.40 Å has been reported.32 The total energy (Etotal
)
obtained for this species is 51.5 kJ mol−1 larger than for the
above lowest energy structure for the major component. Once
again, two C–Haryl p interactions are maintained within
the bound ligand superstructure at almost identical distances
(2.59 Å, 2.77 Å) to those observed in the structure of the
major component. Both C–Haryl p interactions in each case
appear to help define the molecular geometries of the respective
[Ag(4)]+ cations observed for [Ag(4)]PF6.
8 D. Meyerstein, Coord. Chem. Rev., 1999, 185–186, 141; G. Golub,
H. Cohen, P. Paoletti, A. Bencini, L. Messori, I. Bertini and
D. Meyerstein, J. Am. Chem. Soc., 1995, 117, 8353; G. Golub,
H. Cohen, P. Paoletti, A. Bencini and D. Meyerstein, J. Chem. Soc.,
Dalton Trans., 1996, 2055; G. Golub, I. Zilberman, H. Cohen and
D. Meyerstein, Supramol. Chem., 1996, 6, 275; N. Navon, G. Golub,
H. Cohen, P. Paoletti, B. Valtancoli, A. Bencini and D. Meyerstein,
Inorg. Chem., 1999, 38, 3484; T. Clark, M. Hennemann, R. van Eldik
and D. Meyerstein, Inorg. Chem., 2002, 41, 2927.
Concluding remarks
In the present study the metal complexes of macrocycles 2–6
were investigated in the solid state and solution. Collectively,
the potentiometric, liquid–liquid extraction and competitive
membrane transport studies indicate that the effect of
N-benzylation of the secondary amine donor groups of 1 to
produce 2–6 yields an enhanced tendency towards selectivity
9 G. J. Kubas, Inorg. Synth., 1979, 19, 90.
10 R. R. Fenton, L. F. Lindoy, J. R. Price, B. W. Skelton and A. H.
White, Aust. J. Chem., 2003, 56, 1141.
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6
3 7 2 5

View Article Online
11 Bruker WINEPR System 2.11 (shareware), Bruker-Franzen
Analytic GmbH, Bremen, Germany, 1996.
23 R. R. Fenton, R. Gauci, P. C. Junk, L. F. Lindoy, R. C. Luckay,
G. V. Meehan, J. R. Price, P. Turner and G. Wei, J. Chem. Soc.,
Dalton Trans., 2002, 2185.
12 P. Gans, A. Sabatini and A. Vacca, Inorg. Chim. Acta, 1976, 18,
237.
24 W. J. Geary, Coord. Chem. Rev., 1971, 7, 81.
13 P. S. K. Chia, L. F. Lindoy, G. W. Walker and G. W. Everett, Pure
Appl. Chem., 1993, 65, 521.
14 D. D. Perrin and B. Dempsey, Buffers for pH and metal ion control,
Chapman and Hall, London, 1979.
15 S. R. Hall, D. J. du Boulay and R. Olthof-Hazekamp (Editors), The
Xtal 3.7 System, University of Western Australia, 2001.
16 G. M. Sheldrick, SHELX97, Programs for Crystal Structure
Analysis, University of Göttingen, Institüt für Anorganische
Chemie der Universität, Göttingen, Germany, 1998.
17 ADF version 1999.02, Chemistry Department, Vrije Universiteit,
De Boelelaan, Amsterdam, 1999; G. T. Velde and E. J. Baerends,
J. Comput. Phys., 1992, 99, 84; C. F. Guerra, J. G. Snijders,
G. T. Velde and E. J. Baerends, Theor. Chem. Acc., 1998,
99, 391.
18 E. J. Baerends, D. E. Ellis and P. Ros, Chem. Phys., 1973, 2, 41.
19 P. Boerrigter, G. T. Velde and E. J. Baerends, Int. J. Quantum Chem.,
1988, 33, 87; G. T. Velde and E. J. Baerends, Int. J. Quantum
Chem., 1992, 37, 84.
20 A. D. Becke, Phys. Rev. A, 1988, 38, 3089.
21 J. P. Perdew, Phys. Rev. B, 1986, 33, 8822.
22 L. Versluis and T. Zeigler, J. Chem. Phys., 1988, 88, 322.
25 F. A. Cotton, G. Wilkinson, C. A. Murillo and M. Bochmann,
Advanced Inorganic Chemistry, 6th ed, Wiley, New York, 1999.
26 K. Gloe, P. Muhl and J. Beger, Z. Chem., 1988, 28, 1 and refs. therein.
27 F. Hofmeister, Arch. Exp. Pathol. Pharmakol., 1888, 24, 247.
28 C. Chatroux, K. Wichmann, G. Goretzki, T. Rambusch, K. Gloe,
U. Muller, W. Muller and F. Vögtle, Ind. Eng. Chem. Res., 2000, 10,
3616.
29 M. W. Glenny, A. J. Blake, C. Wilson and M. Schröder, Dalton
Trans., 2003, 1941.
30 J. Kim, T.-H. Ahn, M. Lee, A. J. Leong, L. F. Lindoy, B. R. Rumbel,
B. W. Skelton, T. Strixner, G. Wei and A. H. White, J. Chem. Soc.,
Dalton Trans., 2002, 3993.
31 F. H. Allen, Acta Crystallogr., Sect. B, 2002, 58, 380.
32 E. A. H. Griffith and E. L. Amma, J. Am. Chem. Soc., 1974, 96,
743 and refs. therein.
33 K. Shelly, D. C. Finster, Y. J. Lee, W. R. Scheidt and C. A. Reed,
J. Am. Chem. Soc., 1985, 107, 5955.
34 M. Mascal, J.-L. Kerdelhue, A. J. Blake, P. A. Cooke, R. J. Mortimer
and S. J. Teat, Eur. J. Inorg. Chem., 2000, 485; A. N. Khlobystov,
A. J. Blake, N. R. Champness, D. A. Lemenovskii, A. G. Majouga,
N. V. Zyk and M. Schröder, Coord. Chem. Rev., 2001, 222, 155.
3 7 2 6
D a l t o n T r a n s . , 2 0 0 4 , 3 7 1 5 – 3 7 2 6