Three-ring, branched cyclam derivatives and their interaction with nickel(II), copper(II), zinc(II) and cadmium(II)

Article abstract of DOI:10.1039/b401254m

The interaction of two symmetrically branched tris-cyclam derivatives based on 1,3,5-trimethylenebenzene and phloroglucinol cores with nickel(II), copper(II), zinc(II) and cadmium(II) is reported. All four metal ions yield solid complexes in which the metal : ligand ratio is 3 : 1. For both ligand types, spectrophotometric titrations confirm the formation of nickel(II) and copper(II) complexes of similar 3 : 1 stoichiometry in dimethyl sulfoxide. Visible spectral, electrochemical, magnetic moment, ESR and NMR studies have been performed to probe the nature of the respective complexes. Where appropriate, the results from the above metal-ion studies are compared with those from parallel investigations in which the corresponding (substituted) mono-cyclam analogues were employed as the ligands. A structural determination employing a poorly diffracting crystal of the trinuclear nickel(II) complex of the tris-cyclam ligand incorporating a 1,3,5-trimethylenebenzene core was successfully carried out with the aid of a synchrotron radiation source. A nickel ion occupies each cyclam ring in a square-planar coordination arrangement, with each cyclam ring adopting the stable trans-III configuration.

Full text of DOI:10.1039/b401254m

View Online / Journal Homepage / Table of Contents for this issue
Three-ring, branched cyclam derivatives and their interaction with
nickel(II), copper(II), zinc(II) and cadmium(II)†
Ying Dong, Leonard F. Lindoy,* Peter Turner and Gang Wei
Centre for Heavy Metals Research, School of Chemistry, University of Sydney, NSW 2006,
Australia
Received 26th January 2004, Accepted 4th March 2004
First published as an Advance Article on the web 18th March 2004
The interaction of two symmetrically branched tris-cyclam derivatives based on 1,3,5-trimethylenebenzene and
phloroglucinol cores with nickel(), copper(), zinc() and cadmium() is reported. All four metal ions yield solid
complexes in which the metal : ligand ratio is 3 : 1. For both ligand types, spectrophotometric titrations confirm
the formation of nickel() and copper() complexes of similar 3 : 1 stoichiometry in dimethyl sulfoxide. Visible
spectral, electrochemical, magnetic moment, ESR and NMR studies have been performed to probe the nature of the
respective complexes. Where appropriate, the results from the above metal-ion studies are compared with those
from parallel investigations in which the corresponding (substituted) mono-cyclam analogues were employed as the
ligands. A structural determination employing a poorly diffracting crystal of the trinuclear nickel() complex of
the tris-cyclam ligand incorporating a 1,3,5-trimethylenebenzene core was successfully carried out with the aid of a
synchrotron radiation source. A nickel ion occupies each cyclam ring in a square-planar coordination arrangement,
with each cyclam ring adopting the stable trans-III configuration.
Introduction
Recently there has been increasing interest in the design and
synthesis of larger molecular units, incorporating multi-metal
ion binding sites, that are frequently of nanometer dimensions.
Because of their tendency to yield both kinetically and thermo-
dynamically stable complexes,1 macrocyclic ligands have proved
desirable building blocks for incorporation into such systems.
In this context the use of the cyclam ring is particularly attrac-
tive because of its well documented rich transition and post-
transition metal chemistry and its tendency to yield complexes
showing unusual properties such as the stabilisation of less-
common metal oxidation states.2 The metal coordination
chemistry of linked cyclam rings,3,4 including co-facially linked
systems,5 capable of binding two metal ions has now received
considerable attention.
Previously we have investigated the interaction of 1 with a
range of transition and post-transition ions that include
nickel(), copper(), zinc() and cadmium().6,7 We now present
the results of a further investigation in which a comparative
study of the complexation behaviour of the single ring cyclam
derivative 2 and the related tris(cyclam) derivatives 3 and 4
towards the above metals is reported.
Relative to that of linked bis(cyclam) derivatives, the metal-
reported.10 The X-ray structure of the resulting tri-copper()
complex [Cu3(5)](ClO4)6 has been determined; each copper()
ion is five coordinate being bound by four secondary amines
from a macrocyclic subunit and an O-donor in an axial
position from either a water or a perchlorate anion. While
corresponding bond lengths and angles in each subunit are
similar, one of the macrocyclic subunits is positioned on the
opposite side of the plane of the aromatic core to the other
ion chemistry of corresponding tris-derivatives of this ring has
received little attention. We have recently reported8 nickel(),
copper(), zinc() and cadmium() complexation studies
involving the ligand in which three cyclam rings are linked in a
linear fashion by two ‘p-xylyl’ (CH2C6H4CH2) bridges between
nitrogens on adjacent rings.
A communication outlining the synthesis of the tris(cyclam)
macrocycle 3 appeared in 1998,9 with the mercury() com-
plex of this species being reported to act as a substrate for the
recognition of tris(histidine).
two.
In other work, the synthesis of the related tris-macrocycle 5
by means of a copper() template reaction has also been
† Electronic supplementary information (ESI) available: Fig. 1S:
Spectrophotometric titration curve for the addition of nickel() nitrate
hexahydrate to 3. Fig. 2S: EPR spectra for the copper() perchlorate
complexes of (a) 1, (b) 2 and (c) 4. Fig. 3S: EPR spectrum for the
copper() perchlorate complex of 3. Figs 4S–6S: Cyclic voltammo-
grams for [Ni(2)](ClO4)2, [Ni3(3)](ClO4)6 and [Ni3(4)](ClO4)6. See
http://www.rsc.org/suppdata/dt/b4/b401254m/
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
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
1264

View Online
Table 1 Metal coordination distances (Å) for trinuclear [Ni3(3)]-
(ClO4)6
Finally, it is noted that recent studies by our group have
focused on the synthesis of symmetrical tri-linked species
ligands, similar to 3 and 4 but incorporating 16-membered,
N2S2-donor macrocyclic rings, and their interaction with soft
metal ions such as silver(),11 palladium()12 and platinum().12
The results now presented represent an extension of these prior
studies.
Ni(1)–N(1)
1.958(4)
1.936(6)
2.900(5)
1.974(5)
1.920(7)
3.264(14)
1.945(5)
1.949(7)
3.235(6)
Ni(1)–N(2)
Ni(1)–N(4)
Ni(1)–O(16)
Ni(2)–N(6)
Ni(2)–N(8)
Ni(2)–O(17)
Ni(3)–N(10)
Ni(3)–N(12)
Ni(3)–O(19)
1.987(7)
1.904(6)
3.109(9)
1.939(8)
1.952(7)
2.599(9)
1.975(8)
1.918(7)
2.959(11)
Ni(1)–N(3)
Ni(1)–O(1)
Ni(2)–N(5)
Ni(2)–N(7)
Ni(2)–O(14)
Ni(3)–N(9)
Ni(3)–N(11)
Ni(3)–O(4)
Experimental
Where available, all commercial reagents and solvents were of
analytical (or HPLC) grade.
oxygen site. In general, the non-hydrogen atoms were modelled
with anisotropic displacement parameters and a riding atom
model was used for the hydrogen atoms. No hydrogens were
included in the model for the water molecule. There was a
significant residual peak close to Ni(1) and another close
to Ni(3), and these peaks were treated as minor occupancy
alternative locations for the metal ions. It was not possible to
locate ligand sites associated with the alternative metal sites,
and it may be that the residual peaks are in fact artifacts
of the poor quality data. Similarly there appears to be a second
location for the Cl(2) perchlorate counterion. An ORTEP19
depiction of the molecule with 20% displacement ellipsoids
is provided in Fig. 1, and pertinent geometry details are listed
in Table 1. The structure is of very poor quality, and this is a
consequence of the high level of crystal decay and significant
molecular disorder. Distance and displacement parameter
restraints were required for some of the atoms.
Physical methods
1H and 13C NMR spectra were recorded on a Bruker Avance
DPX300 spectrometer. Low resolution mass spectra (ESI) were
obtained on a Finnigan LCQ-8 spectrometer. The EPR spectra
were obtained for samples as DMF glasses at 77 K using a
Bruker EMX EPR spectrometer at 9.464 GHz (X-band).
Solid state UV-VIS spectra were measured on a CARY 1E
UV-Visible Spectrophotometer (samples spread on filter paper),
solution UV-VIS spectra were measured on a CARY 5E
UV-VIS-NIR spectrophotometer.
Electrochemistry
Cyclic voltammetry studies were performed using a con-
ventional three-electrode configuration with iR compensation
and a BAS Model 100B electrochemical system, operated by a
computer using the BAS software. The working electrode was
glassy carbon. The reference electrode was Ag/AgCl and was
separated from the working and Pt wire auxiliary electrodes by
a glass sleeve fitted with a Vycor frit. Solutions were prepared
in HPLC-grade acetonitrile, 0.1 mol dmϪ3 in (n-Bu4N)(ClO4),
and were purged with argon gas. Complex concentrations in the
millimolar range were used throughout. Scans were measured
over the range 10–1000 mV sϪ1, with reported results being
mainly those recorded at 100 mV sϪ1. The ferrocenium–ferro-
cene reference couple had E1/2 at ϩ0.52 V (acetonitrile) under
these conditions.
Crystal structure determination
The relatively small single crystals of [Ni3(3)](ClO4)6 grown for
X-ray diffraction analysis diffracted too weakly to undertake
a data collection with a laboratory diffractometer system.
Consequently a synchrotron data collection was undertaken at
the ChemMat CARS facility at the Advanced Photon Source
of the Argonne National Laboratory, Argonne USA.
Double diamond (111) reflections were used to obtain mono-
chromated 0.56356 Å radiation from the synchrotron source,
and harmonics were eliminated with mirrors. An orange plate-
like crystal was attached with Exxon Paratone N, to a short
length of fibre supported on a thin piece of copper wire inserted
in a copper mounting pin. The pin and crystal were mounted on
a Bruker Kappa diffractometer equipped with a 4 chip mosaic
CCD detector, and an Oxford Scientific Cryojet. Data were
collected at 100(2) K with ꢀ scans to 2θ 44Њ, and cell constants
were obtained from a least squares refinement against 1020
reflections located between 2θ 4 and 34Њ. The intensities of 1019
standard reflections recollected at the end of the experiment
changed by 31.0% during the data collection and a correction
was accordingly applied to the data with SADABS.13 The
data integration and reduction were undertaken with SAINT
and XPREP,14 and subsequent computations were carried
out with the XShell,14 teXsan15 and WinGX16 graphical user
interfaces.
Fig. 1 An ORTEP19 depiction of [Ni3(3)](ClO4)6, with 20% displace-
ment ellipsoids.
Crystal data
[Ni3(3)](ClO4)6, model formulation C39H78Cl6Ni3N12O25, M =
1503.96, monoclinic, space group P21/n (no. 14), a = 20.797(3),
b = 13.440(3), c = 22.316(4) Å, β = 109.400(8)Њ, V = 5883.4(18)
Å3, Z = 4, µ(synchrotron) = 0.688 mmϪ1, λ(synchrotron) =
0.56356 Å, crystal dimensions 0.100 × 0.090 × 0.016 mm,
2θmax = 44.12, N = 393078, Nind = 14513 (Rmerge 0.090), Nobs
=
9159, R1(F) = 0.1446, wR2(F 2) = 0.4341.
CCDC reference number 211305.
See
http://www.rsc.org/suppdata/dt/b4/b401254m/
for
crystallographic data in CIF or other electronic format.
Ligand and metal complex synthesis
The structure was solved in the space group P21/n (no. 14)
by direct methods with SIR97,17 and extended and refined with
SHELXL-97.18 The asymmetric unit contains the trinuclear
complex, six perchlorate counterions and a site treated as a water
The synthesis and characterization of 1, 3 and 4 have been
described elsewhere.20,21 All complexes were dried over P4O10
in a vacuum before microanalysis.
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1265

View Online
Scheme 1
1-(2-Phenoxyacetyl)-1,4,8,11-tetracyclotetradecane 2. This
55.70, 56.32, 57.31, 57.67, 62.38, 62.59, 62.87, 115.36, 115.42,
115.57, 121.92, 122.12, 130.37 and 159.23.
compound was prepared (see Scheme 1) in an analogous
manner to that described previously for 420 to yield the
product as a colourless viscous oil (yield 95%) [Found: Mϩ,
[Cd(2)](NO3)PF6ؒ0.75H2O. Using a similar procedure to that
described above for [Zn(2)](NO3)PF6, cadmium() nitrate tetra-
hydrate (0.087 g, 0.281 mmol) and 1 (0.090 g, 0.281 mmol)
yielded the required product as colourless needle-like crystals
(0.103 g, 56%) after recrystallization from methanol (Found:
C, 32.95; H, 5.06; N, 10.93. Calc. for C18H33.5CdF6N5O4.75P:
C, 33.09; H, 5.17; N, 10.72%). 1H NMR (CD3COCD3): δ 1.73–
2.02 (m, NCH2CH2, 3 H), 2.18–2.31 (m, NCH2CH2, 1 H),
2.87–3.46 (m, NCH2, 16 H; NCH2CH2O, 2 H), 3.55–3.80 (m,
NH, 3 H), 4.31–4.34 (t, NCH2CH2O, 2 H), 6.93–6.98 (m, ArH,
3 H), 7.26–7.33 (m, ArH, 2 H). 13C NMR (CD3COCD3):
δ 26.11, 29.47, 46.99, 47.11, 47.74, 47.86, 48.15, 48.27, 52.16,
52.28, 52.59, 52.71, 52.98, 53.25, 53.38, 56.10, 56.61, 63.27,
115.41, 121.98, 130.39 and 159.15.
1
320.2576 (EI). C18H32N4O requires Mϩ, 320.2571]. H NMR
(CDCl3): δ 1.65–1.82 (m, NCH2CH2, 4 H), 2.59–2.75 (m,
NCH2, 16 H), 2.87–2.91 (t, PhOCH2CH2N, 2 H), 4.07–4.11
(t, PhOCH2, 2 H), 6.89–7.34 (m, ArH, 5 H). 13C NMR (CDCl3):
δ 26.54, 28.75, 47.80, 47.91, 48.91, 49.32, 49.42, 50.74,
50.84, 51.77, 54.00, 55.86, 65.66, 114.44, 120.54, 129.35 and
158.86.
[Ni(2)](ClO4)2. Nickel() perchlorate hexahydrate (0.183 g,
0.500 mmol) in methanol (5 cm3) was added to 2 (0.160 g,
0.500 mmol) in methanol (15 cm3). The orange solution was
stirred and heated for 0.5 h and then left to stand at room
temperature overnight. The golden yellow crystals that formed
were filtered off and recrystallized from methanol (0.173 g,
60%) (Found: C, 37.30; H, 5.60; N, 9.65. Calc. for C18H32Cl2-
NiN4O9: C, 37.40; H, 5.58; N, 9.69%).
[Ni3(3)](ClO4)6. Using a similar procedure to that described
for the mononuclear complex of [Ni(2)](ClO4)2, three molar
equivalents of nickel() perchlorate hexahydrate (0.219 g,
0.600 mmol) were reacted with 3 (0.143 g, 0.200 mmol) to yield
the required product as golden-orange crystals (0.178 g, 60%)
after recrystallization from water (Found: C, 31.53; H, 5.25;
N, 11.37. Calc. for C39H78Cl6Ni3N12O24: C, 31.48; H, 5.28; N,
11.30%).
CAUTION: perchlorate-containing complexes are poten-
tially explosive and appropriate precautions should be in place
for their preparation, handling and storage.
[Cu(2)](ClO4)2. This complex was prepared using a similar
procedure to that described for [Ni(2)](ClO4)2, copper() per-
chlorate hexahydrate (0.185 g, 0.500 mmol) and 2 (0.160 g,
0.500 mmol) yielded the required product as purple crystals
(0.120 g, 65%) after recrystallization from methanol (Found:
C, 37.02; H, 5.57; N, 9.46. Calc. for C18H32Cl2CuN4O9: C, 37.09;
H, 5.53; N, 9.61%).
[Cu3(3)](ClO4)6. Using a similar procedure to that described
for the mononuclear complex of [Ni(2)](ClO4)2, three molar
equivalents of copper() perchlorate hexahydrate (0.222 g,
0.600 mmol) were reacted with 3 (0.143 g, 0.200 mmol) to yield
the required product as a blue–purple powder (0.162 g, 54%)
after recrystallization from water (Found: C, 31.01; H, 5.15;
N, 11.07. Calc. for C39H78Cl6Cu3N12O24: C, 31.18; H, 5.23; N,
11.19%).
[Zn(2)](NO3)PF6ؒ2H2O. Zinc() nitrate tetrahydrate (0.045 g,
0.172 mmol) in methanol (5 cm3) was added to a solution of 2
(0.055 g, 0.172 mmol) in methanol (15 cm3). The solution was
stirred and heated for 0.5 h and excess saturated ammonium
hexafluorophosphate in methanol was added. The solvent
was evaporated under reduced pressure and the residue was
recrystallised from acetone and isopropanol to give a white
powder (0.037 g, 35%) (Found: C, 34.17; H, 5.56; N, 11.35.
Calc. for C18H36F6N5O6PZn: C, 34.38; H, 5.77; N, 11.14%).
1H NMR (CD3COCD3): δ 1.86–1.96 (m, NCH2CH2, 2 H),
2.32–2.45 (m, NCH2CH2, 1 H), 2.82–2.92 (m, NCH2CH2, 1 H),
3.01–3.31 (m, NCH2, 16 H; NCH2CH2O, 2 H), 3.80 (br s, NH,
2 H), 3.98 (br s, NH, 1 H), 4.31–4.35 (m, NCH2CH2O, 2 H),
6.93–6.97 (m, ArH, 3 H), 7.26–7.32 (m, ArH, 2 H). 13C NMR
(CD3COCD3): δ 24.61, 25.33, 25.48, 28.53, 45.66, 46.12, 46.22,
46.75, 46.87, 47.09, 47.69, 47.91, 48.40, 49.37, 49.68, 49.96,
50.68, 50.88, 51.22, 51.49, 51.72, 52.47, 52.55, 52.72, 52.86,
[Zn3(3)](NO3)2(PF6)4ؒ3MeOH. Using a similar procedure to
that described for the mononuclear complex of [Zn(2)](NO3)-
PF6ؒH2O, three molar equivalents of zinc() nitrate tetra-
hydrate (0.157 g, 0.600 mmol) were reacted with 3 (0.143 g,
0.200 mmol) to yield the required product as a white powder
after recrystallization from methanol–water (0.103 g, 30%)
(Found: C, 29.66; H, 5.10; N, 11.29. Calc. for C42H90F24N14-
O9P4Zn3: C, 29.48; H, 5.30; N, 11.46%). 1H NMR (CD3-
COCD3–D2O): δ 1.81–1.98 (m, NCH2CH2, 12 H), 2.53–3.30
(m, NCH2, 48 H), 4.12–4.44 (m, PhCH2N, 6 H), 7.27–7.33 (m,
ArH, 3H). 13C NMR (CD3COCD3–D2O): δ 24.97, 24.99, 25.01,
28.69, 29.20, ∼45–56 (overlapping signals), ∼131–138 (over-
lapping signals).
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1266

View Online
[Cd3(3)](NO3)3(PF6)3ؒ5MeOH. Using a similar procedure
to that described for the mononuclear complex of [Zn(2)]-
(NO3)PF6ؒH2O, three molar equivalents of cadmium() nitrate
tetrahydrate (0.185 g, 0.600 mmol) were reacted with 3 (0.143 g,
0.200 mmol) to yield the required product as a white powder
after recrystallization from methanol–water (0.128 g, 35%)
(Found: C, 29.00; H, 5.19; N, 11.36. Calc. for C44H98Cd3F18-
N15O14P3: C, 28.82; H, 5.39; N, 11.46%). 1H NMR (CD3-
COCD3–D2O): δ 1.78–1.99 (m, NCH2CH2, 6 H), 2.25 (br m,
NCH2CH2, 6 H), 2.77–3.46 (m, NCH2, 48 H), 4.10–4.54
(m, PhCH2N, 6 H), 7.26–7.36 (m, ArH, 3 H). 13C NMR
(CD3COCD3–D2O): δ 22.44, 25.30, 25.51, 25.64, 26.11, 44.67,
45.19, 45.30, 45.61, 46.50, 46.84, 47.26, 47.43, 47.70, 48.01,
49.19, 49.50, 49.74, 50.11, 50.35, 50.53, 50.73, 51.31, 51.41,
52.69, 53.16, 53.47, 55.55, 56.94, 57.68, ∼131–136 (overlapping
signals).
[Ni(2)](ClO4)2, [Cu(2)](ClO4)2, [Zn(2)](NO3)PF6ؒ2H2O, [Cd(2)]-
(NO3)PF6ؒ0.75H2O, [Ni3(3)](ClO4)6, [Cu3(3)](ClO4)6, [Zn3(3)]-
(NO3)2(PF6)4ؒ3MeOH, [Cd3(4)](NO3)3(PF6)3ؒ5MeOH, [Ni3(4)]-
(ClO4)6ؒH2O, [Cu3(4)](ClO4)6ؒH2O, [Zn3(4)](NO3)2(PF6)4 and
[Cd3(4)](NO3)2(PF6)4. The synthesis of these species was
generally straight-forward by mixing hot methanol solutions
of the appropriate metal salt and ligand in the required
stoichiometries. In some instances addition of NH4PF6 proved
necessary to induce crystallisation of the complex as its hexa-
fluorophosphate salt. However, unfortunately, in no case were
suitable crystals obtained for structural analysis employing a
conventional laboratory X-ray diffractometer. Nevertheless,
in the case of [Ni3(3)](ClO4)6, small crystals were isolated that
proved suitable for structural analysis using a synchrotron
radiation source (see Experimental section).
The 1H and 13C NMR spectra of the zinc() and cadmium()
complexes of 2–4 listed above have been recorded in deutero-
acetone or D2O–deutero-acetone. While broadly similar to the
spectra of the corresponding metal-free ligands in terms of the
chemical shift regions for which signals appear, each spectrum
is considerably more complex than that of its free ligand – with
significantly more signals and additional splitting (in the case of
[Ni3(4)](ClO4)6ؒH2O. Using a similar procedure to that
described for the mononuclear complex of [Ni(2)](ClO4)2,
three molar equivalents of nickel() perchlorate hexahydrate
(0.219 g, 0.600 mmol) were reacted with 4 (0.161 g, 0.200 mmol)
to yield the product as an orange powder after recrystallization
from water (0.128 g, 40%) (Found: C, 31.66; H, 5.19; N, 10.63.
Calc. for C42H86Cl6Ni3N12O28: C, 31.61; H, 5.43; N, 10.53%).
1
the H NMR spectra) being present. This was not unexpected
given that previous NMR studies of 1 : 1 complexes of
cyclam with a number of metal ions, including zinc()22 and
cadmium(),23 also give rise to considerably more complex
spectra than cyclam itself. The additional complexity typically
reflects the simultaneous presence of two or three of the five
(trans-I–trans-V)1 possible configurations for complexed
cyclam in solution;22,23 the latter differing in the pattern of
chiralities adopted by its amine donors. The complexity of the
spectra can be both solvent and anion-dependent, with non-
equivalence of individual methylene protons also occurring
in particular instances due to macrocycle conformational
flexibility being reduced on metal coordination. It is also noted
that in some instances further complication arises from the fact
that the spectra are time dependant, reflecting the presence of
slow conformational interchange.
Similar NMR complexity to that discussed above has been
recorded for [Ni(tmc)]2ϩ (where tmc is the tetra-N-methyl
derivative of cyclam) for which three configurations for the
bound macrocycle were assigned in DMF solution.24 More
recently, the dinuclear zinc complex of xylyl-linked bis-cyclam
was also demonstrated to give rise to very complex spectra.
Detailed analysis of the latter using 2D [1H,15N] HSQC NMR
and of its aliphatic region employing 2D [1H,13C]HSCQC
NMR showed that, at equilibrium, this complex exists in
aqueous solution as a mixture of three conformational forms:
trans-I (45%), trans-III (34%) and cis-V (21%).25
While no detailed analysis of the present spectra of the
zinc and cadmium complexes of 1–4 was attempted during the
present study, it seems very likely that the observed complexity
is contributed to by the presence of a mixture of (bound)
ligand configurations in solution, present under slow exchange
conditions. Further, since the present zinc and cadmium
systems contain unsymmetrically substituted cyclam rings –
this will also contribute to increased NMR complexity over
that observed for the corresponding (more symmetrical) cyclam
systems.
The UV-VIS absorption maxima (461–477 nm) of the
nickel() complexes of 1–4 in both acetonitrile and the solid
state, together with the reported values for the complex of 1,
are listed in Table 2 and are in accord with the presence of
low-spin, square-planar nickel() in each case. Additional
weak peaks were present in the case of the solution spectra;
these are assigned to the existence of a ‘yellow-to-blue’ (low-
spin square planar/high-spin octahedral) equilibrium in this
solvent. Similar behaviour in solution has now been well
documented for nickel cyclam and related derivatives in prior
studies.6,8,21
[Cu3(4)](ClO4)6ؒH2O. Using a similar procedure to that
described for the mononuclear complex of [Ni(2)](ClO4)2,
copper() perchlorate hexahydrate (0.222 g, 0.600 mmol) and
4 (0.161 g, 0.200 mmol) yielded the required product as a
purple powder (0.145 g, 45%) after recrystallization from
water (Found: C, 31.33; H, 5.48; N, 10.35. Calc. for C42H86-
Cl6Cu3N12O28: C, 31.32; H, 5.38; N, 10.44%).
[Zn3(4)](NO3)2(PF6)4. Using a similar procedure to that
described for the mononuclear complex of [Zn(2)](NO3)PF6ؒ
H2O, zinc() nitrate tetrahydrate (0.157 g, 0.600 mmol) and 4
(0.161 g, 0.200 mmol) yielded the required product as a white
powder after recrystallization from methanol–water (0.143 g,
42%) (Found: C, 29.57; H, 5.03; N, 11.27. Calc. for C42H84F24-
1
N14O9P4Zn3: C, 29.58; H, 4.97; N, 11.50%). H NMR (D2O–
CD3COCD3): δ 1.76–2.22 (m, NCH2CH2, 12 H), 2.56–3.37
(m, NCH2, 48; NCH2CH2O, 6 H), 4.27 (br s, NCH2CH2O, 6 H),
6.26–6.32 (m, ArH, 3 H). 13C NMR (D2O–CD3COCD3):
δ 24.11, 24.41, 28.10, 28.51, 45.15, 45.58, 46.08, 46.62, 47.12,
47.37, 47.54, 47.74, 48.04, 49.19, 49.94, 50.22, 50.58, 50.67,
50.73, 51.65, 52.14, 52.21, 52.46, 53.16, 55.11, 55.40, 56.53,
95.27, 95.71, 95.96, 160.89 and 161.17.
[Cd3(4)](NO3)2(PF6)4. Using a similar procedure to that
described for the mononuclear complex of [Zn(2)](NO3)PF6ؒ
H2O, cadmium() nitrate tetrahydrate (0.185 g, 0.600 mmol)
and 4 (0.161 g, 0.200 mmol) yielded the product as a white
powder after recrystallization from methanol–water (0.140 g,
38%) (Found: C, 27.17; H, 4.62; N, 10.49. Calc. for C42H84Cd3-
1
F24N14O9P4: C, 27.32; H, 4.59; N, 10.62%). H NMR (D2O–
CD3COCD3): δ 1.82–1.97 (m, NCH2CH2, 12 H), 2.71–3.26
(m, NCH2, 48 H), 3.35 (m, NCH2CH2O, 6 H), 4.21–4.30
(m, NCH2CH2O, 6 H), 6.26–6.32 (m, ArH, 3 H). 13C NMR
(D2O–CD3COCD3): δ 22.59, 25.38, 25.74, 26.20, 28.85, 44.63,
45.15, 45.55, 45.97, 46.59, 47.05, 47.29, 47.89, 47.97, 49.02,
49.71, 50.04, 50.22, 50.50, 51.40, 52.21, 53.39, 53.46, 54.24,
55.33, 56.55, 57.62, 63.67, 63.77, 63.85, ∼95.0–96.5 (overlapping
signals), 160.69, 160.80 and 161.11.
Results and discussion
Metal complex formation
In an attempt to obtain crystalline products suitable for X-ray
diffraction studies, the following solid complexes were isolated:
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1267

View Online
Table 2 UV-VIS spectrophotometric data for the nickel() and
copper() complexes of 1–4
Table 3 EPR spectroscopic parameters for copper() complexes of
cyclam and 2–4
Complex
λmax a/nm
λmax b/nm (ε/MϪ1 cmϪ1
)
Complex
g||
g⊥
A||/G
A⊥/G
a
[Ni(1)](ClO4)2
[Ni(2)](ClO4)2
[Ni3(3)](ClO4)6
[Ni3(4)](ClO4)6
[Cu(1)](ClO4)2
[Cu(2)](ClO4)2
[Cu3(3)](ClO4)6
[Cu3(4)](ClO4)6
473
463
463
461
524
504
531
511
480 (15), 704 (vw), 995 (vw)
477 (16), 330 (sh), 711 (vw), 958 (vw)
463 (78), 712 (vw), 950 (vw)
466 (80), 700 (vw), 971 (vw)
520 (90)
516 (109)
535 (353)
[Cu(cyclam)](NO3)2
[Cu(2)](ClO4)2
[Cu(3)](ClO4)2
species 1
species 2
[Cu(4)](ClO4)2
2.173
2.180
2.048
2.053
200
203
49
2.174
2.102
2.182
198
198
202
2.083
2.057
50
a Values from ref. 6.
520 (345)
a Solid state UV-VIS spectra. b Data were obtained in acetonitrile.
analogue containing 2 and the absence of low intensity (forbid-
den) half-field ∆Ms = 2 transitions at ∼1000–1500 G29 strongly
suggests that metal–metal interaction does not occur in the
former species.
Under similar conditions to the above, a relatively broader
spectrum was obtained for the remaining trinuclear complex,
Spectrophotometric titrations in dimethyl sulfoxide were
employed to follow nickel() binding by the tri-linked ligands
3 and 4. These experiments were designed to probe whether
three metal ions per ligand are taken up in this donating solvent
and, if so, whether the complexation takes place in a step-wise
fashion. The plot for the addition of nickel() nitrate to 3
shows a sharp 3 : 1 metal to ligand endpoint, consistent with the
formation of the expected trinuclear complex. A very similar
plot was obtained for 4. For each of 3 and 4, the absorbance
steadily increases at 460 and 464 nm, respectively – consistent
with the affinity of the three binding sites for nickel() being
very similar for each system.
[Cu3(3)](ClO4)6. Attempts to assign this spectrum to
a
single species proved difficult and a more satisfactory ‘fit’ was
obtained when two complex species were assumed to be present
(see Table 3); the latter may reflect the influence of con-
figurational isomers in this case. It is noted that an analysis
of peak intensities suggested that the complexity reflects the
presence of more than one isomer and, taken together with the
absence of a low field, low intensity signal (see above), appears
less likely to be a consequence of copper–copper interaction.30
The UV-VIS absorption maxima for the copper() com-
plexes of 1–4 in acetonitrile together with the maxima for the
corresponding solid-reflectance spectra are listed in Table 2.
Both the solid state and solution spectra show a broad envelope
of bands in the visible region which is consistent with the
presence of a copper()–N4 chromophore (for which solvent or
anion may occupy axial positions);26 however, the featureless
nature of the spectra results in them being of little use for
assignment of detailed coordination geometries. As expected,
the relative positions of the maxima suggest that the ligand
fields are relatively close to each other in these complexes.
UV-VIS spectrophotometric titrations of copper() nitrate
with 3 and 4 were also carried out under similar conditions to
those employed for the corresponding nickel() nitrate systems.
As again expected, the plots for 3 and 4 clearly show sharp
metal to ligand endpoints of 3 : 1 with a steady increment in the
absorption at the particular wavelength employed in each case,
again suggesting that the affinity of the three binding sites for
copper() is very similar in each system. The fact that the
absorption maxima for the copper() complexes of 3 and 4
(535 and 532 nm, respectively) are quite close to each other, as
before, suggests that the ligand fields are very similar.
Electrochemistry
The redox behaviour of the nickel() complexes of the tri-
linked species of 3 and 4 has been studied electrochemically
in acetonitrile and the results compared with those for the
mononuclear complexes of cyclam as well as of its substituted
derivatives 1 and 2. Each of these complexes displays waves
in its cyclic voltammogram corresponding to the presence of
nickel()/() and nickel()/() couples. The E (or Ep, for
½
irreversible cases) values were estimated as the average of the
anodic and cathodic peak potentials (Epa, Epc) obtained for
each complex in acetonitrile using a glassy carbon working
electrode (Table 4). The E values are in each case slightly more
½
positive [oxidation to nickel() more difficult] and less negative
[reduction to nickel() facilitated] than the corresponding values
obtained for the mononuclear reference complexes mentioned
above. This may, in part, reflect a weakening of metal ion
binding in the trinuclear species due to steric effects; while the
additional positive charge associated with the presence of
three nickel ions is also expected to inhibit the Ni() to Ni()
oxidation step.31
The room-temperature magnetic moments µ (µB) of the blue–
purple copper() complexes of type [Cu(2)]2ϩ, [Cu3(3)]6ϩ and
[Cu3(4)]6ϩ are 1.91, 2.00 and 2.06 per copper atom, respectively,
and hence fall in the range 1.8–2.2 µB that is typical for
copper() species with S = ½.
The anodically shifted nickel()/() couples for the trimer
complexes (relative to those of the mononuclear analogues)
display quasi-reversible or irreversible behaviour, which appears
likely to be associated with complicated adsorption–desorption
phenomena. However, for the nickel()/() couples one or
two sets of reversible waves were observed. For example,
[Ni(2)](ClO4)2 gave rise to two pairs of redox waves under
the present conditions in which the anodic and cathodic peak
currents (Ipa and Ipc) were observed to be proportional to
The X-band EPR spectra of the copper() complexes of
2–4 in dimethylformamide glasses were recorded at 77 K and
the parameters obtained are shown in Table 3. The spectra of
[Cu3(4)](ClO4)6 and of its mononuclear analogue, [Cu(2)]-
(ClO4)2 are split into four equally spaced peaks reflecting
coupling with the copper() nucleus (I = 3/2); the g|| values at
ca. 2.18 and g⊥at ca. 2.05 are quite similar to the values reported
for other related square-planar copper() systems27 and are
typical of axial symmetric copper() complexes with the
½
ν (ν = potential scan rate) in the range ν = 10–1000 mV sϪ1 and
the ratio of Ipa and Ipc at a given ν is unity. This behaviour
establishes the simple nature of the electron transfer reaction
and is consistent with electrochemical reversibility in this
case.32,33 Very weak redox peaks observed around 0.80 and
Ϫ0.80 V are assigned as arising from adsorbed species.
The nickel()/() couples for the complexes of 3 and 4 at
1.164 and 1.129 V, respectively, also clearly arise from single
reversible redox processes; in accordance with ‘simple’ redox
behaviour again occurring in these cases. These results are in
keeping with the absence of communication between metal
centres in these complexes.34
2
2
unpaired electron in the dx Ϫy orbital. Both are also quite
similar to those for the copper() complex of cyclam (Table 3),
suggesting essentially square-planar coordination geometries.6
As observed for the present complexes, square copper() species
typically have g|| > g⊥> 2.00 and A|| in the range (150–210) ×
10Ϫ4 cmϪ1 28 The A|| coupling constant for the trinuclear species
.
incorporating 4 is very close to that for the monomeric
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1268

View Online
Table 4 Observed Ni()/() and Ni()/() and Cu()/() and Cu()/() couples from cyclic voltammograms of the corresponding complexes
of cyclam and 1–4 [glassy carbon working electrode, 100 mV sϪ1 scan rate, acetonitrile–0.1 mol dmϪ3 Bu4N(ClO4), 3 × 10Ϫ4 mol dmϪ3 for nickel
complexes and 1 × 10Ϫ3 mol dmϪ3 for copper complexes, E1/2 (or Ep for irreversible couples) vs. Ag/AgCl]
Complex
E1/2(M()/())/V (∆E/mV)
E1/2(M()/())/V (∆E/mV)
[Ni(cyclam)](ClO4)2
[Ni(1)](ClO4)2
ϩ1.100 (75), ϩ0.583 (70)a
ϩ1.074 (205), ϩ0.734 (100)a
ϩ1.137 (59)
ϩ1.164 (69)
ϩ1.129 (51)
Ϫ1.447 (225)a
Ϫ1.310 (110)a
Ϫ1.261 (67)
Ϫ0.867b
[Ni(2)](ClO4)2
[Ni3(3)](ClO4)6
[Ni3(4)](ClO4)6
Ϫ1.255 (160)
[Cu(cyclam)](NO3)2
[Cu(1)](NO3)2
[Cu(2)](ClO4)2
[Cu3(3)](ClO4)6
[Cu3(4)](ClO4)6
ϩ1.377 (irrev.)a
ϩ1.442 (irrev.)a
ϩ1.539 (irrev.)
ϩ1.580 (irrev.)
ϩ1.560 (irrev.)
Ϫ0.853 (85)a
Ϫ0.640 (110)a
Ϫ0.668 (154)
Ϫ0.849c
Ϫ0.780c
a From ref 6. b Back wave displaced by >300 mV; Ep cited. c Reversibility limited or masked by adsorption phenomena.
Cyclic voltammograms of the copper() complexes of 1–4 in
acetonitrile display waves that are assigned to copper()/()
and copper()/() processes. As observed for the corresponding
nickel() complexes, the copper() complexes of the above
groups (trans-III isomer).1 The Ni–N bond lengths (1.904(6)–
1.987(7) Å) fall in the normal range for bonds of this type.36
Concluding remarks
type also yield E (or Ep for irreversible process) values for
½
oxidation that are slightly more positive, while less negative
values were obtained for reduction relative to the corresponding
values observed for the ‘reference’ complexes of cyclam, 1 and
It is noted that interest in linked dinucleating azamacrocyclic
ligands has increased considerably by the finding that
bis(cyclam) ligands (and selected metal-ion derivatives), incor-
porating either alkyl or aromatic linking groups between the
rings, are effective in inhibiting several strains of human
immunodeficiency virus type 1 (HIV-1) and type 2 (HIV-2).37
Such systems are structurally quite closely related to those
investigated in the present study and hence the present
systems may also prove to be of interest in this context.
2. The respective E values for the above complexes using a
½
glassy carbon working electrode are listed in Table 4.
Copper()/() couples for N4-macrocyclic ligand complexes
usually display quasi-reversible electrochemical behaviour,
although complicated in many cases by adsorption–desorption
phenomena, with large separation (>300 mV) of anodic (Epa)
and cathodic (Epc) maxima in their cyclic voltammograms. The
CVs of the trinuclear complexes of 3 and 4 each clearly show
one quasi-reversible wave at Ϫ0.849 and Ϫ0.780 V, respectively,
comparable with the behaviour of the mononuclear complex
of 2, and this implies that the three copper() centres in the
complexes of 3 and 4 simultaneously exchange electrons with
the electrode and that the coulombic interactions between the
three metal centres are again negligible.34,35
Acknowledgements
We thank the Australian Research Council for support and
Professor G. A. Lawrance, University of Newcastle, New South
Wales, for assistance with the electrochemical studies. We also
thank Dr M. Fainerman-Melnikova for help with the EPR
studies. Use of the ChemMatCARS Sector 15 at the Advanced
Photon Source, was supported by the Australian Synchrotron
Research Program, which is funded by the Commonwealth
of Australia under the Major National Research Facilities
Program. ChemMatCARS Sector 15 is also supported by the
National Science Foundation/Department of Energy under
grant numbers CHE9522232 and CHE0087817 and by the
Illinois board of higher education. The Advanced Photon
Source is supported by the U.S. Department of Energy, Basic
Energy Sciences, Office of Science, under Contract No. W-31-
109-Eng-38.
Crystal structure of [Ni3(3)](ClO4)6
Single crystals of [Ni3(3)](ClO4)6 suitable for X-ray diffraction
analysis were obtained by slow evaporation of an aqueous
solution of the complex. The crystals were small and a structure
determination required the use of synchrotron radiation. As
indicated in the experimental section, the diffraction of the
synchrotron radiation was hindered by a surprisingly high level
of crystal decomposition in the X-ray beam, and significant
molecular disorder. The crystal structure model is accordingly
of very poor quality, and effectively serves only to establish
the molecular connectivity of the [Ni3(3)](ClO4)6 complex. The
structure is shown in Fig. 1, and metal to ligand distances for
the trinuclear complex are listed in Table 1. The metal co-
ordination geometry is essentially square planar, although in
each case there is weak axial coordination to two adjacent
perchlorate counterions that completes a pseudo-octahedral
coordination sphere. The weak axial ligation is unlikely to
persist in solution, and indeed visible spectroscopy indicates
four-coordinate complexes in solution. The Cl(1) perchlorate
forms a bridge between the Ni(1) centre of one complex and
the Ni(3) centre of an adjacent complex. Similarly a Cl(4) per-
chlorate bridges the Ni(1) and Ni(2) metal centres of adjacent
complexes, trans to the Cl(1) perchlorate weakly coordinated
to Ni(1). Likewise a Cl(5) perchlorate bridges the Ni(2) and
Ni(3) metal centres of adjacent complexes, trans to the Cl(4)
perchlorate weakly coordinated to Ni(2).
References
1 L. F. Lindoy, The Chemistry of Macrocyclic Ligand Complexes,
Cambridge University Press, Cambridge, UK, 1989.
2 A. De Blas, G. De Santis, L. Fabbrizzi, M. Licchelli and
P. Pallavicini, Pure Appl. Chem., 1993, 65, 455.
3 L. F. Lindoy, Advances in Inorg. Chem., 1998, 45, 75; J. D. Chartres,
L. F. Lindoy and G. V. Meehan, Coord. Chem. Rev., 2001, 216–217,
249.
4 A. McAuley and S. Subramanian, Coord. Chem. Rev., 2000,
200–202, 75.
5 M. Lachkar, B. Andrioletti, B. Boitrel, R. Guilard and A. Atmani,
New J. Chem., 1995, 19, 777; F. Rabiet, F. Denat and R. Guilard,
Synth. Commun., 1997, 27, 979; S. Brandes, F. Denat, S. Lacour,
F. Rabiet, F. Barbette, P. Pullumbi and R. Guilard, Eur. J. Org.
Chem., 1998, 2349; M. Lachkar, R. Builard, A. Atmani, A. De Cian,
J. Fischer and R. Weiss, Inorg. Chem., 1998, 37, 1575.
6 Y. Dong, G. A. Lawrance, L. F. Lindoy and P. Turner, Dalton Trans.,
2003, 1567.
Each cyclam subunit resides on the same side of the plane of
the aromatic core and has an RRSS configuration of its amine
7 Y. Dong, S. Farquhar, K. Gloe, L. F. Lindoy, B. R. Rumbel,
P. Turner and K. Wichmann, Dalton Trans., 2003, 1558.
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1269

View Online
8 Y. Dong and L. F. Lindoy, Coord. Chem. Rev., 2003, 245, 11.
9 S. Sun, J. Saltmarsh, S. Mallik and K. Thomasson, Chem. Commun.,
1998, 519.
10 P. V. Bernhardt and E. J. Hayes, J. Chem. Soc., Dalton Trans., 1998,
3539.
11 J. D. Chartres, A. M. Groth, L. F. Lindoy and G. V. Meehan,
J. Chem. Soc., Dalton Trans., 2002, 371.
12 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, 2801.
13 (a) R. H. Blessing, Acta Crystallogr., Sect. A, 1995, 51, 33;
(b) G. M. Sheldrick, SADABS, Empirical absorption correction
program for area detector data, University of Göttingen, Germany,
1996.
14 (a) Bruker (1995). SMART, SAINT and XPREP. Area detector
control and data integration and reduction software; (b) Bruker
(2000) XShell. Bruker Analytical X-ray Instruments Inc., Madison,
Wisconsin, USA.
15 Molecular Structure Corporation (1997–1998). teXsan for
Windows: Single Crystal Structure Analysis Software, MSC, 3200
Research Forest Drive, The Woodlands, TX 77381, USA.
16 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
17 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano,
C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and
R. Spagna, J. Appl. Crystallogr., 1998, 32, 115.
18 G. M. Sheldrick, SHELXL97, Program for crystal structure
refinement, University of Göttingen, Germany, 1997.
19 (a) C. K. Johnson, ORTEPII, Report ORNL-5138, Oak Ridge
National Laboratory, Oak Ridge, TN, 1976; (b) Xtal3.6 System,
ed. S. R. Hall, D. J. du Boulay and R. Olthof-Hazekamp, University
of Western Australia, 1999.
R. W. Hay, B. Kinsman and C. I. Smith, Polyhedron, 1995, 14,
1245.
22 X. Liang, M. Weishaüpl, J. A. Parkinson, S. Parsons,
P. A. McGregor and P. J. Sadler, Chem. Eur. J., 2003, 9, 4709.
23 X. Liang, J. A. Parkinson, S. Parsons, M. Weishaüpl and P. J. Sadler,
Inorg. Chem., 2002, 41, 4539.
24 S. F. Lincoln, J. H. Coates, D. A. Hadi and D. L. Pisaniello, Inorg.
Chim. Acta, 1984, 81, L9.
25 X. Liang, J. A. Parkinson, M. Weishaüpl, R. O. Gould, S. J. Paisey,
H-s. Park, T. M. Hunter, C. A. Blindauer, S. Parsons and P. J. Sadler,
J. Am. Chem. Soc., 2002, 124, 9105.
26 R. Buxtrof and T. A. Kaden, Helv. Chim. Acta, 1974, 57, 1035;
M. Hediger and T. A. Kaden, Helv. Chim. Acta, 1983, 66, 86;
L. Fabbrizzi, L. Montagna, A. Poggi, T. A. Kaden and L. C.
Siegfried, J. Chem. Soc., Dalton Trans., 1987, 2631.
27 K. A. Beveridge, A. McAuley and C. Xu, Inorg. Chem., 1991, 30,
2074.
28 S. V. Rosokha, Y. D. Lampeka and I. M. Maloshtan, J. Chem. Soc.,
Dalton Trans., 1993, 631; M. Lachkar, R. Guilard, A. Atmani,
A. De Cian, J. Fischer and R. Weiss, Inorg. Chem., 1998, 37, 1575.
29 J. D. Smith and A. E. Martell, J. Am. Chem. Soc., 1972, 94, 4123.
30 T. D. Smith and J. R. Pilbrow, Coord. Chem. Rev., 1974, 13, 173.
31 R. Schneider, A. Riesen and T. A. Kaden, Helv. Chim. Acta, 1985,
68, 53.
32 B. W. Rossiter and J. F. Hamilton Physical Methods of Chemistry,
Wiley, New York, vol. 2, 1986.
33 K. V. Gobi and T. Ohsaka, Electrochem. Acta, 1998, 44, 269.
34 M. Ciampolini, L. Fabbrizzi, A. Perotti, A. Poggi, B. Seghi and
F. Zanobini, Inorg. Chem., 1987, 26, 3527.
35 C. de Alwis, J. A. Crayston, T. Cromie, T. Eisenblatter, R. W. Hay,
Ya. D. Lampeka and L. V. Tsymbal, Electrochem. Acta, 2000, 45,
2061.
36 V. J. Thom, C. C. Fox, J. C. A. Boeyens and R. D. Hancock, J. Am.
Chem. Soc., 1984, 106, 5947.
37 G. J. Bridger, R. T. Skerlj, S. Padmanabhan, S. A. Martellucci,
G. W. Henson, M. J. Abrams, H. C. Joao, M. Witvrouw,
K. D. Vreese, R. Pauwels and E. De Clercq, J. Med. Chem., 1996, 39,
109, and references therein.
20 Y. Dong and L. F. Lindoy, Aust. J. Chem., 2001, 54, 291.
21 See, for example: C. K. Jørgensen, Acta Chem. Scand., 1957, 11, 399;
A. Anichini, L. Fabbrizzi, P. Paoletti and R. M. Clay, Inorg. Chim.
Acta, 1977, 24, L21; L. Sabatini and L. Fabbrizzi, Inorg. Chem.,
1979, 18, 438; L. Fabbrizzi, M. Micheloni and P. Paoletti, Inorg.
Chem., 1980, 18, 535; M. Ciampolini, L. Fabbrizzi, M. Licchelli,
A. Perotti, F. Pezzini and A. Poggi, Inorg. Chem., 1986, 25, 4131;
D a l t o n T r a n s . , 2 0 0 4 , 1 2 6 4 – 1 2 7 0
1270