A [2 × 2] nickel(II) grid and a copper(II) square result from differing binding modes of a pyrazine-based diamide ligand

Article abstract of DOI:10.1039/b614796h

The potentially bis-terdentate diamide ligand N,N′-bis[2-(2-pyridyl) ethyl]pyrazine-2,3-dicarboxamide (H₂L^Et) was structurally characterised. Potentiometric titrations revealed rather low pK_a values for the deprotonation of the first amide group of H₂L ^Et (14.2) and N,N′-bis(2-pyridylmethyl)pyrazine-2,3- dicarboxamide (H₂L^Me, 13.1). Two tetranuclear copper(ii) square complexes of H₂L^Et with a paddle-wheel appearance, in which each ligand strand acts as a linear N₃-NO hybrid terdentate-bidentate chelate, have been isolated and structurally characterised. Complex [Cu^II₄(H₂L^Et) ₂(HL^Et)₂](BF₄)₆· 3MeCN·0.5H₂O (3·3MeCN·0.5H₂O), with two nondeprotonated zwitterionic ligand strands and two monodeprotonated ligand strands, is formed in the 1: 1 reaction of H₂L^Et and Cu(BF₄)₂·4H₂O. It has a polymeric chain structure of tetranuclear subunits connected by N-H...N hydrogen bonds. The same reaction carried out with one equivalent of base gives the related compound [Cu^II₄(HL^Et)₄](BF ₄)₄ (4), with all four ligand strands monodeprotonated. It consists of isolated tetranuclear units. In both 3·3MeCN·0.5 H₂O and 4 the copper(ii) ions are in five-coordinate N₄O environments but the degree of trigonality (τ) differs [3· 3MeCN·0.5H₂O 0.14 ≤ τ ≤ 0.26; 4 τ = 0.45]. Under the same reaction conditions as for 4 but using Ni(BF₄) ₂·6H₂O a tetranuclear [2 × 2] grid-type complex, [Ni^II₄(HL^Et)₄](BF ₄)·10H₂O (5·10H₂O), is formed. The structure determination showed that the nickel(ii) ions have N₆ distorted octahedral coordination spheres and all four ligand strands are monodeprotonated and act as N₃-N₃ bis-terdentate chelates. Magnetic susceptibility data show that the complexes 3·4H₂O, 4 and 5·10H₂O exhibit very weak antiferromagnetic spin coupling. The energies and multiplicities of the spin states of 3·4H ₂O and 4 were determined from the temperature dependence of the magnetic susceptibility and indicate that a singlet state is lowest and the quintet state highest. This is consistent with the X-band EPR spectra of polycrystalline powders of 3·4H₂O and 4, measured down to 2.3 K, where the resonances observed at the lowest temperature are due to a triplet state. The g-values of the individual ions of 4 are consistent with the expected d_{x 2 -y 2} ground state for five-coordinate copper(ii) in an approximately square pyramidal configuration. The Royal Society of Chemistry 2007.

Full text of DOI:10.1039/b614796h

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
A [2 × 2] nickel(II) grid and a copper(II) square result from differing binding
modes of a pyrazine-based diamide ligand†
Julia Klingele (ne´e Hausmann),a John F. Boas,b John R. Pilbrow,b Boujemaa Moubaraki,c Keith S. Murray,c
Kevin J. Berry,d Keith A. Hunter,a Geoffrey B. Jameson,e Peter D. W. Boyd f and Sally Brooker*a
Received 11th October 2006, Accepted 28th November 2006
First published as an Advance Article on the web 14th December 2006
DOI: 10.1039/b614796h
The potentially bis-terdentate diamide ligand N,Nꢀ-bis[2-(2-pyridyl)ethyl]pyrazine-2,3-dicarboxamide
(H2LEt) was structurally characterised. Potentiometric titrations revealed rather low pKa values for the
deprotonation of the first amide group of H2LEt (14.2) and N,Nꢀ-bis(2-pyridylmethyl)pyrazine-2,3-
dicarboxamide (H2LMe, 13.1). Two tetranuclear copper(II) square complexes of H2LEt with a
paddle-wheel appearance, in which each ligand strand acts as a linear N3–NO hybrid
terdentate–bidentate chelate, have been isolated and structurally characterised. Complex
[CuII4(H2LEt)2(HLEt)2](BF4)6·3MeCN·0.5H2O (3·3MeCN·0.5H2O), with two nondeprotonated
zwitterionic ligand strands and two monodeprotonated ligand strands, is formed in the 1 : 1 reaction of
H2LEt and Cu(BF4)2·4H2O. It has a polymeric chain structure of tetranuclear subunits connected by
N–H · · · N hydrogen bonds. The same reaction carried out with one equivalent of base gives the related
compound [CuII4(HLEt)4](BF4)4 (4), with all four ligand strands monodeprotonated. It consists of
isolated tetranuclear units. In both 3·3MeCN·0.5 H2O and 4 the copper(II) ions are in five-coordinate
N4O environments but the degree of trigonality (s) differs [3·3MeCN·0.5H2O 0.14 ≤ s ≤ 0.26; 4 s =
0.45]. Under the same reaction conditions as for 4 but using Ni(BF4)2·6H2O a tetranuclear [2 × 2]
grid-type complex, [NiII4(HLEt)4](BF4)·10H2O (5·10H2O), is formed. The structure determination
showed that the nickel(II) ions have N6 distorted octahedral coordination spheres and all four ligand
strands are monodeprotonated and act as N3–N3 bis-terdentate chelates. Magnetic susceptibility data
show that the complexes 3·4H2O, 4 and 5·10H2O exhibit very weak antiferromagnetic spin coupling.
The energies and multiplicities of the spin states of 3·4H2O and 4 were determined from the
temperature dependence of the magnetic susceptibility and indicate that a singlet state is lowest and the
quintet state highest. This is consistent with the X-band EPR spectra of polycrystalline powders of
3·4H2O and 4, measured down to 2.3 K, where the resonances observed at the lowest temperature are
2
2
due to a triplet state. The g-values of the individual ions of 4 are consistent with the expected dx
−y
ground state for five-coordinate copper(II) in an approximately square pyramidal configuration.
with intriguing electronic or magnetic properties have been
reported.1 For the formation of such grid-type complexes with
octahedrally coordinated metal ions relatively rigid ligand systems
with a repeating linear array of terdentate binding pockets are
necessary. The use of amide-based ligands for the formation of
grid-type complexes offers advantages over other systems since
amides can usually be synthesised relatively easily. In addition to
their good coordinating properties, amide groups also reduce the
flexibility of a given ligand system.5
Introduction
In recent years grid-type complexes and the self-assembly pro-
cesses involved in their formation have attracted a considerable
amount of interest.1–4 With typical octahedral metal ions like
cobalt(II), iron(II), iron(III) or manganese(II) grid-type complexes
aDepartment of Chemistry, University of Otago, PO Box 56, Dunedin,
New Zealand. E-mail: sbrooker@alkali.otago.ac.nz; Fax: +64-3-479-7906;
Tel: +64-3-479-7919
We recently reported that the use of a carefully designed
amide-based ligand can also lead to pH control of product
architecture.3 Specifically, that in MeCN solution the complex-
ation of the potentially bis-terdentate diamide ligand N,Nꢀ-
bSchool of Physics, Building 27, Monash University, Clayton, Victoria, 3800,
Australia
cSchool of Chemistry, Building 23, Monash University, Clayton, Victoria,
3800, Australia
dWesternport Secondary College, Hastings, Victoria, 3195, Australia
bis(2-pyridylmethyl)pyrazine-2,3-dicarboxamide (H2LMe
) with
eInstitute of Fundamental Sciences, Chemistry, Massey University, PO Box
11222, Palmerston North, New Zealand
Cu(BF4)2·4H2O in a 1 : 1 molar ratio leads to the formation of
a navy-blue cyclic dimeric complex [CuII2(H2LMe)2(MeCN)2](BF4)4
(A) of a zwitterionic form of H2LMe (Scheme 1).3 The ligand was
thus shown to act as a hybrid N3–O terdentate–monodentate
ligand, where one binding pocket is deactivated as a terdentate
chelate by protonation of the pyridine nitrogen atom. It was
also shown that under identical reaction conditions but with the
f
Department of Chemistry, The University of Auckland, Private Bag 92019,
Auckland, New Zealand
† Electronic supplementary information (ESI) available: A discussion of
the IR data obtained for the complexes; magnetic data for complex 4
over the range 300–4.2 K (Fig. S1); crystallographic tables of selected
bond lengths and angles and hydrogen-bonds (Tables S1–S3). See DOI:
10.1039/b614796h
This journal is The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 633
©

View Online
addition of one equivalent of base, deprotonation of the pyri-
dinium group occurs and that as a consequence the now monode-
protonated ligand (HLMe)− acts as a bis-terdentate N3–N3 chelate.
This results in the formation of a grass-green [2 × 2] grid-type com-
plex [CuII4(HLMe)4](BF4)4·3.5MeCN (B) (Scheme 1).3 These two
complexes, one a dimer, the other a tetranuclear grid, form due to
a difference in the protonation state of the ligand, providing a nice
illustration of the concept of using pH to control supramolecular
architecture.3,6 Recently the synthesis4 and the crystal structures
of an orthorhombic7 and a triclinic4 polymorph of H2LMe were
published by Stoeckli-Evans and co-workers, as were the syntheses,
structures and magnetic properties of a copper(II) and a nickel(II)
[2 × 2] grid-type complex {[CuII4(HLMe)4](ClO4)4·5MeOH·4H2O
and [NiII4(HLMe)](Cl)4·5MeCN·13H2O, respectively} of (HLMe)−.4
In contrast to the synthesis of B (Scheme 1), it is interesting to note
that Stoeckli-Evans and co-workers used both a different anion
(ClO4−) and a different solvent mixture (CH2Cl2–CH3OH–H2O)
to prepare [CuII4(HLMe)4](ClO4)4·5MeOH·4H2O and that under
these conditions they did not require added base in order to form
the [2 × 2] grid complex. This is a clear indication that a number of
factors, including anion, solvent and probably also relative product
solubilities, can influence the outcome of such complexations.
synthesis of this ethylene-linked, potentially bis-terdentate, ligand
(H2LEt), we report on its 1 : 1 reactions with Cu(BF4)2·4H2O in
MeCN, with and without base, for direct comparison with the
results obtained for the methylene-linked ligand (H2LMe). It is
clearly demonstrated that the change from methylene (H2LMe) to
ethylene (H2LEt) ‘arms’ has a profound effect on the resulting
copper(II) coordination environments as well as on the molecular
architecture of the product formed. The synthesis and character-
isation of a nickel(II) [2 × 2] grid-type complex of (HLEt)− is also
described. Detailed magnetic susceptibility and EPR studies are
also discussed.
Results and discussion
Ligand synthesis and pKa studies
Dimethyl pyrazine-2,3-dicarboxylate (2) was conveniently pre-
pared in a large scale in a one-pot two-step procedure by
modification of the literature procedure (Scheme 2).10 The first
step consisted of the oxidative cleavage of quinoxaline and
the formation of pyrazine-2,3-dicarboxylic acid (1). In order to
maximise the yield of diester 2, the intermediate diacid 1 was
not isolated. Instead the subsequent esterification with MeOH
and SOCl2 was carried out using the crude reaction product.
The resulting diester 2 was easily isolated by extraction with
CH2Cl2 and was obtained in very good yields as a slightly tan
analytically pure crystalline solid. The ligand H2LEt was prepared,
by a modification of the literature procedure,8 by reacting dimethyl
pyrazine-2,3-dicarboxylate (2) with 2-(2-aminoethyl)pyridine in a
1 : 2 molar ratio in MeOH in an open flask, thus allowing the
MeOH formed in the reaction to evaporate. Redissolution of the
remaining solid in CH2Cl2–Me2CO (2 : 1) and reduction of the
volume of the solution afforded H2LEt in very good yield as an
analytically pure colourless microcrystalline solid (Scheme 2).
Scheme 2 Synthesis of H2LEt: Reagents and conditions: (i) KMnO4,
H2O, 80–90 ◦C; (ii) HCl (conc.), RT; (iii) SOCl2, MeOH, reflux; (iv)
2-(2-aminoethyl)pyridine, MeOH, open flask.
Scheme 1 Synthesis of A of the zwitterionic form of H2LMe and of B, the
[2 × 2] grid-type complex of (HLMe)−. Reagents and conditions: (i) 1 eq.
Cu(BF4)2, MeCN, RT; (ii) Et2O (vapour diffusion); (iii) 1 eq. NEt3.
Owing to the apparent symmetry of H2LEt the IR spectrum
of the compound was expected to show only one band for the
N–H and C–O stretching (mNH and mCO) vibrations as well as
for the C–N stretching/N–H bending vibration (mC–N/dNH). In
reality, two absorption bands for mCO (m¯ = 1674 and 1651 cm−1)
and mC–N/dNH (m¯ = 1541 and 1509 cm−1), were observed in the
IR spectrum taken from a KBr disk. A similar observation was
made by Stoeckli-Evans and co-workers for the IR spectrum of
In this paper we report the results of our investigation into
the effects of employing N,Nꢀ-bis[2-(2-pyridyl)ethyl]pyrazine-2,3-
dicarboxamide (H2LEt),8,9 the higher homologue of H2LMe, on the
molecular architecture of the resulting copper(II) and nickel(II)
complexes. Specifically, as well as detailing a convenient new
634 | Dalton Trans., 2007, 633–645
This journal is
The Royal Society of Chemistry 2007
©

View Online
the lower ligand homologue H2LMe (mCO : m¯ = 1676, 1665; mC–N/dNH
:
showed evidence of such a seven-membered ring in the solid state,
a motif like ii in Fig. 2 is possible for the free ligand in solution.
The acid dissociation constants of both ligands, H2LMe and
H2LEt, were measured by potentiometric titration with standard
alkali solution using an established method.13 The method in-
volved back titration of an aqueous solution containing known
amounts of the ligand and standard aqueous hydrochloric acid
with standard aqueous sodium hydroxide. It was assumed that
each neutral ligand molecule contained two pyridine-N groups
capable of accepting protons and a third functional group (amide)
capable of losing a proton. The average computed first, second and
third pKa values, obtained from the non-linear fitting procedure,
are pK1 = 4.4 and 5.3, pK2 = 5.2 and 5.6 and pK3 = 13.1 and
14.2 for H2LMe and H2LEt, respectively. The general magnitude
of the pK1 and pK2 values for both ligands is consistent with
the dissociation of isolated pyridinium protons (pKa = 5.23 for
the pyridinium ion at zero ionic strength).14 The much higher
pK3 values are likely to be due to the dissociation of an amide
proton. Compared to the pKa values of mono-amide compounds,
which usually range from 16–19 (benzamide: pK > 19; acetanilide:
pK = 17.6; benzanilide: pK = 16.5),15 the pK3 values of 13.1 and
14.2 for the first amide group of H2LMe and H2LEt, respectively,
are remarkably low. However, it is important to note that it is
difficult to obtain accurate pKa measurements at values greater
m¯ = 1534 and 1515 cm−1).4 The authors attributed this observation
to different hydrogen bonding patterns exhibited by the two amide
groups that were observed in the crystal structure of the triclinic
polymorph of the free ligand, namely an intramolecular bifurcated
hydrogen bond to a pyrazine and a pyridine nitrogen atom and
an intermolecular hydrogen bond to the amide oxygen atom of a
neighbouring ligand molecule.4
A single crystal X-ray structure analysis showed, however, that
the molecular structure of H2LEt (Fig. 1) does not exhibit strong
intramolecular hydrogen bonds but, more like the orthorhombic
form of H2LMe,7 that the two amide groups of H2LEt exhibit two
different intermolecular hydrogen bonds [N(2) · · · N(3A) 3.165(3);
N(5) · · · O(2B) 2.825(3)] (Fig. 1). Owing to steric reasons, the ortho-
positioned amide groups are not coplanar with each other or with
the pyrazine ring. The mean planes defined by the amide functions
[O(1)–C(5)–N(2) and O(2)–C(13)–N(5)] form angles of 19.5◦ and
52.6◦ to the pyrazine ring.
than about 12 in aqueous media. The resulting monodeprotonated
−
ligands (HLMe
)
and (HLEt)− may be stabilised by the formation
of equilibria involving motifs iii–v (Fig. 2). In [2 × 2] grid-type
complexes of (HLMe)− seven-membered rings like motif v in Fig. 2
have been reported.3,4
Fig. 1 Molecular structure of H2LEt. Hydrogen atoms not involved in
hydrogen bonds have been omitted for clarity. Symmetry operation used
to generate equivalent atoms: (a) −x − 2, −y + 1, −z − 1; (b) x + 1, y, z;
(c) −x − 1, −y + 1, −z − 1.
Synthesis and structures of tetranuclear copper(II) complexes
By analogy to the synthesis of [CuII2(H2LMe)2(MeCN)2](BF4)4
(A, Scheme 1)3 the 1 : 1 reaction of the higher ligand ho-
mologue H2LEt and Cu(BF4)2·4H2O was carried out in MeCN
solution (Scheme 3). Vapour diffusion of Et2O into the re-
sulting navy-blue solution afforded a compound formulated as
[CuII2(H2LEt)(HLEt)]n(BF4)3n·2nH2O, in ca. 85% yield, in the form
of a turquoise-green crystalline material. In contrast to the zwitte-
rionic form of H2LMe observed in the dimeric navy-blue complex A,
in this tetranuclear turquoise-green complex only two of the ligand
strands are in the zwitterionic form (H2LEt) whilst the other two
are in a monodeprotonated form (HLEt)−. The vapour diffusion
of Et2O into a dilute solution of the compound in MeCN afforded
single crystals of [CuII4(H2LEt)2(HLEt)2](BF4)6·3MeCN·0.5H2O
(3·3MeCN·0.5H2O) suitable for X-ray crystal structure analysis
which revealed the tetranuclear nature of the compound (n = 2)
and confirmed the form in which the ligands are present (Fig. 3,
Tables 1 and S1–S3†).
For diamide derivatives of imidazole-4,5-dicarboxylic acid it
is known that in the solid state generally an intramolecular N–
H · · · O hydrogen bond motif is formed which leads to a seven-
membered ring (motif i in Fig. 2).11 In the IR spectra of some
N,Nꢀ-disubstituted imidazole-4,5-dicarboxamides two mCO and
two mC–N/dNH vibration absorption bands have been observed.12
Although neither of the crystal structures of H2LMe 4,7 or H2LEt
Compound 3·3MeCN·0.5H2O exhibits a square-like paddle-
wheel structure built up by four copper(II) centres binding to
the perpendicularly arranged portions of four ligand strands. The
square encapsulates one tetrafluoroborate anion (Fig. 3). Each
ligand acts as a linear N3–NO hybrid terdentate–bidentate chelate
which results in an N4O coordination sphere for each copper(II)
centre with an average degree of trigonality16 of sav = 0.20 (0.14 ≤
s ≤ 0.26). In the following discussion the coordination sphere
will therefore be referred to as distorted square pyramidal. The
central pyrazine rings of the ligands bridge adjacent copper(II)
Fig. 2 i: Seven-ring motif found in diamide derivatives of imida-
zole-4,5-dicarboxylic acid; ii: seven-ring motif proposed for H2LMe and
H2LEt in solution; iii–v: equilibria proposed for the monodeprotonated
ligands (HLMe
)
and (HLEt)− in solution.
−
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 635
©

View Online
Fig. 3 Molecular structure of [CuII4(H2LEt)2(HLEt)2]6+, the cation of
3·3MeCN·0.5H2O, and its guest (BF4)− anion. Hydrogen atoms except
H(5C), H(11C), H(17C), H(23C) and the two pyridinium protons have
been omitted for clarity.
˚
˚
the apical pyrazine ligand [Cu(1): 0.165 A, Cu(2): 0.162 A, Cu(3):
˚
˚
0.157 A, Cu(4): 0.127 A]. Relative to the Cu4 mean plane, the
˚
copper(II) ions of the square are shifted by ca. 0.28 A in an up–
down–up–down fashion leading to a small tetrahedral distortion
of the Cu4 square.
Scheme 3 Synthesis of 3 and 4: Reagents and conditions: (i) 1 eq.
Cu(BF4)2·4H2O, MeCN, RT; (ii) Et2O (vapour diffusion); (iii) 1 eq. NEt3,
MeCN, RT; (iv) toluene (precipitation).
Each O-coordinated amide group of the NO bidentate ligand
half is still N-protonated. Within each ligand strand this proton
is involved in an intramolecular N–H · · · O hydrogen bond to
the oxygen atom of the N-metalated amide group of the N3
terdentate half of the same ligand strand (Fig. 3). This motif
is also present in the equilibria postulated for the metal-free
monodeprotonated ligand (motif iii in Fig. 2) and is important
as it enables the ortho-positioned amide groups, the deprotonated
ions, occupying an equatorial and an apical position at each metal
centre (see Scheme 3). In the square two shorter [Cu(2) · · · Cu(3)
˚
˚
6.997(1) A and Cu(3) · · · Cu(4) 7.009(1) A] and two slightly
˚
˚
longer [Cu(1) · · · Cu(2) 7.051(1) A and Cu(1) · · · Cu(4) 7.037(1) A]
Cu · · · Cu distances are observed. In all four distorted square
pyramids the copper(II) ions are pulled out of the base towards
◦
˚
Table 1 Comparison of selected bond distances [A] and angles [ ], and selected other data, for the complexes A, B, 3·3MeCN·0.5H2O and 4
7
A7
Ba
3·3MeCN·0.5H2Oa
4a
˚
Cu–L/A
Cu–Npz
Cu–Nam
Cu–Oam
2.044(2)
1.913(2)
2.270(1)
2.017(2)
2.171(9)–2.248(9)
1.900(12)–2.002(11)
—
2.027(5)–2.316(5)
1.914(5)–1.931(5)
1.958(4)–1.997(4)
2.004(5)–2.027(5)
2.024(6), 2.226(7)
1.895(6)
1.966(5)
Cu–Npy
2.082(10)–2.194(10)
2.009(6)
L–Cu–L/◦
Npy–Cu–Nam
Npy–Cu–Oam
Npy–Cu–Npz
81.82(7)
93.70(6)
162.85(6)
106.16(6)b
—
77.2(3)–81.6(4)
—
152.5(4)–157.0(4)
—
—
—
93.7(2)–94.9(2)
91.4(2)–92.9(2)
159.3(2)–163.7(2)
171.6(2)–174.9(2)b
73.2(2)–75.0(2)a
90.9(2)–95.4(2)b
80.1(2)–80.8(2)
93.9(3)
92.0(2)
147.3(2)
174.0(3)b
76.5(2)a
94.3(2)b
80.1(2)
b
O
am–Cu–Nam
a
Oam–Cu–Npz
Oam–Cu–Npz
Nam–Cu–Npz
Other data
T/K
92.22(6)b
81.08(6)
b
74.5(4)–76.8(4)
150
Dimer
0.05
200
150
168
Structural type
Tetramer; [2 × 2] grid-type
Tetramer; square
0.14 ≤ s ≤ 0.26
Tetramer; square
0.45
s21
Six-coordinate
a Only angles within the same ligand strand are given unless indicated otherwise. b Only angles between different ligands strands are given.
636 | Dalton Trans., 2007, 633–645
This journal is
The Royal Society of Chemistry 2007
©

View Online
group of one N-metalated pocket and the nondeprotonated group
of the other O-metalated pocket, to become coplanar. The N · · · O
distances in the resulting seven-membered rings are in the range
copper(II) coordination geometry (Table 1) by providing larger
Npy–Cu–Nam bite angles (six-membered chelate ring rather than
five-membered chelate ring formed by the methylene linked ligand)
and altering the preferred H-bonding patterns, favouring the
intramolecular Namide–H · · · Oamide hydrogen bonds between the N-
protonated O-coordinated amide group in one half of the ligand
and the oxygen atom of the N-metalated amide group in the other
˚
of 2.547–2.576 A (Table S3) and are thus significantly shorter
than those observed in related diamide derivatives of imidazole-
4,5-dicarboxylic acid which usually are around 2.75–2.80 A.11 The
˚
N-metalated amide function is deprotonated and is part of the
N3 terdentate half of each ligand strand. In the tetrameric unit
two of the four ligands are monodeprotonated [(HLEt)−] whereas
the other two ligands are neutral (H2LEt). The latter are in a
zwitterionic form, as the two former N–H amide protons have
relocated from the N3-terdentate half of each of these two ligands
to protonate the uncoordinated pyridines of the NO-bidentate
halves of each of these ligands. This feature has also been observed
in the molecular structure of the related, but dimeric, complex A
(Scheme 1) of the methylene ligand homologue and in other related
systems.3,17 In 3 both relocated protons of the two H2LEt ligands
are involved in intermolecular N–H · · · N hydrogen bonds to the
two unprotonated pyridine nitrogen atoms of two (HLEt)− ligands
half of the same ligand strand as well as the intermolecular Npyridine
–
H · · · Npyridine hydrogen bonds between the pairs of protonated
and unprotonated pyridine rings in neighbouring complexes, and
associated solubility factors.
By analogy with the formation of B of the lower ligand homo-
logue (Scheme 1), the 1 : 1 reaction of H2LEt with Cu(BF4)2·4H2O
in MeCN was repeated with the addition of one equivalent of
base (Scheme 3) in order to ensure that all of the ligand strands
are monodeprotonated so that both halves of each ligand strand
are able, in principle, to act as N3 terdentate chelates. A bottle-
green amorphous solid formulated as [CuII(HLEt)]n(BF4)n could be
isolated, in excellent yield, from the resulting bottle-green solution
by precipitation with toluene. The formula, proposed on the basis
of elemental analysis results, clearly indicates that the strategy has
been successful in producing a complex of monodeprotonated
ligand strands (HLEt)− only. Vapour diffusion of Et2O into a
solution of the compound in MeCN afforded single crystals of
[CuII4(HLEt)4](BF4)4 (4) suitable for an X-ray crystal structure
determination (Fig. 5, Tables 1 and S1–S3†). This revealed the
tetrameric nature of the compound (n = 4) and confirmed the
presence of (HLEt)− ligands only, albeit binding in a very different
manner to that observed in the analogous copper(II) complex B of
(HLMe)−.
˚
˚
in neighbouring tetranuclear subunits [2.703(5) A and 2.732(5) A].
In this way a polymeric chain of tetranuclear subunits is formed
(Fig. 4). This was not a feature of the related dimeric complex A
(Scheme 1).
There are many more differences between the complex of the
ethylene linked ligand, 3·3MeCN·0.5H2O and the complex of the
methylene linked ligand, A, not least the fact that A is dimetallic
whereas 3·3MeCN·0.5H2O, is tetrametallic. Both complexes con-
tain distorted square pyramidal copper(II) centres but the basal
plane in A comprises the N3 terdentate chelate provided by H2LMe
and an MeCN solvent molecule whereas in 3·3MeCN·0.5H2O it
comprises the N3 terdentate chelate provided by the higher ligand
homologue H2LEt or (HLEt)− and an amide oxygen atom from
a different ligand strand. The apical sites are also occupied by
different groups, an amide oxygen atom from a different ligand
strand in the case of A and a pyrazine nitrogen atom from the same
ligand strand that provides the basal amide oxygen atom in the case
of 3·3MeCN·0.5H2O. In A one half of each ligand strand acts as an
N3-terdentate chelate while the other half acts as a monodentate
apical O-donor whereas in 3·3MeCN·0.5H2O one half of each
ligand strand acts as an N3-terdentate chelate while the other half
acts as an NO bidentate chelate which provides the remaining
basal donor as well as the apical donor. These differences are
presumably the result of the ethylene linked ligand modifying the
The overall molecular structure of complex 4 is very similar to
that of its precursor complex 3·3MeCN·0.5H2O, again exhibiting
¯
a paddle-wheel appearance. A special position with 4 symmetry
is located at the unoccupied centre of the tetranuclear square so
the asymmetric unit comprises just a quarter of the tetramer, i.e.
one [CuII(HLEt)](BF4) moiety. The ligand (HLEt)− in complex 4
acts as an N3–NO hybrid terdentate–bidentate chelate and, as
seen in 3·3MeCN·0.5H2O, this results in an N4O five-coordinate
copper(II) centre. The degree of trigonality (s = 0.45) is signif-
icantly higher than that of its precursor complex 3·3MeCN·0.5
H2O (sav = 0.20) and is between the values for a perfect square
pyramid (s = 0) and a perfect trigonal bipyramid (s = 1).16 The
˚
Cu · · · Cu distance in this square [6.922(2) A] is somewhat shorter
Fig. 4 Section of the polymeric chain structure of 3·3MeCN·0.5H2O. Anions, solvent molecules and hydrogen atoms not involved in hydrogen bonds
have been omitted for clarity.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 637
©

View Online
(Scheme 1).3 The copper(II) ions in A are N3–NO coordinated
whereas in B they are N3–N3 coordinated. Complexes B and 4
are directly comparable to one another as they have the same
Cu(II) : ligand ratio and contain ligands in the same protonation
state (monodeprotonated). It is therefore interesting to note that,
despite these similarities, the copper(II) centres in B and 4 adopt
differing coordination environments, N3–N3 six-coordinate and
N3–NO five-coordinate respectively, as a consequence of the
difference in Npy–Cu–Nam chelate ring sizes, five-membered and
−
six-membered, offered by (HLMe
)
and (HLEt)−, respectively. In
contrast, A and 3·3MeCN·0.5H2O are not directly comparable.
In this pair of complexes the Cu(II) : ligand ratios are the same,
but it is interesting to note that 3·3MeCN·0.5H2O crystallises
out with a mixture of monodeprotonated and non-deprotonated
zwitterionic ligands whereas for the methylene linked analogue
A the product contained only non-deprotonated zwitterionic
ligands. Presumably this is a consequence of the relative
solubilities of the various possible products in each case as much
as their relative stabilities. In this context it is interesting to
recall that [2 × 2] grids resulted from all of the following 1 :
1 complexations of H2LMe (a) with copper(II) tetrafluoroborate
in MeCN–Et2O when NEt3 is added (Scheme 1, complex B)3
(b) with copper(II) perchlorate in CH2Cl2–MeOH–H2O with no
added base4 and (c) with nickel(II) chloride in MeCN–MeOH
with NEt3 added.4 It should also be noted that 1 : 2 reactions
of H2LMe or H2LEt with copper(II) salts have consistently been
reported to yield Cu2(HLMe/Et)X3 type species in the absence
of added base. In 1974 Fleischer and co-workers9 reported
the structure of [Cu2(HLEt)Cl3]∞, formed using copper(II)
chloride in MeOH. More recently, again employing no added
base, Stoeckli-Evans and co-workers4 reported the structure of
[Cu2(HLMe)Cl3(H2O)], formed using copper(II) chloride in EtOH–
H2O, and we have isolated [Cu2(HLMe)(MeCN)4](BF4)3, using
copper(II) tetrafluoroborate in MeCN–Et2O. Taken together,
these results clearly indicate that a great number of factors,
including reaction stoichiometry, choice of anion and solvent(s),
presence or otherwise of added base and the relative solubilities of
the various possible products, all contribute, along with the exact
nature of the ligand, to determining the nature of the complex
isolated from these reactions.
Fig. 5 Molecular structure of [CuII4(HLEt)4]4+, the cation of 4. Hydrogen
atoms not involved in hydrogen bonds have been omitted for clarity.
The hollow bonds denote the severely disordered free ligand arm, which
was “SQUEEZEd away”18 in the refinement of the structure. Symmetry
operations used to generate equivalent atoms: (A) x − 0.5, −y + 0.5, −z +
0.5; (B) −x + 1, −y, z; (C) −x + 0.5, y + 0.5, −z + 0.5.
˚
than it is in 3·3MeCN·0.5H2O (6.997–7.051 A). Consistent with
this, the copper(II) ions in 4 are shifted in an up–down–up–down
fashion relative to the Cu4 mean plane by a smaller amount (ca.
˚
0.17 A) than was observed in the case of 3·3MeCN·0.5H2O (ca.
˚
0.28 A) and the Cu–Npz bonds, involved in the Cu–pyrazine–
Cu bridges, are also slightly shorter in 4 (Table 1). As seen in
the precursor complex 3·3MeCN·0.5H2O, each non-deprotonated
amide function in complex 4 forms an intramolecular N–H · · · O
hydrogen bond (Table S3), which leads to a seven-membered ring
like motif iii in Fig. 2. Again, the pyridine ring of the bidentate
ligand half points away from the square and is not coordinated
to the copper(II) ion. However, in contrast to 3·3MeCN·0.5H2O,
the uncoordinated pyridine nitrogen atom in 4 is not involved in
intermolecular hydrogen bonding and is so disordered that it was
subjected to the SQUEEZE procedure of the PLATON suite18 in
order to complete the structural analysis.
Synthesis and structure of a tetranuclear nickel(II) complex
To summarise, the addition of base to the 1 : 1 reaction of
the ligand H2LEt and Cu(BF4)2·4H2O turned the dark navy-blue
solution bottle-green and led to the isolation of the bottle-green
compound [CuII4(HLEt)4](BF4)4 (4) instead of the turquoise-
green complex [CuII4(H2LEt)2(HLEt)2](BF4)6·3MeCN·0.5H2O
(3·3MeCN·0.5H2O). The structural consequences of removing
the two pyridinium protons from 3·3MeCN·0.5H2O to form 4 were
subtle rather than dramatic as neither of the resulting pyridine
nitrogen lone pairs coordinated to the copper(II) ions. Instead
both 3·3MeCN·0.5H2O and 4 contain four N3–NO coordinated
copper(II) ions in a tetrameric square (Scheme 3). In stark contrast,
in the analogous reaction with the methylene linked ligand H2LMe
the addition of base had dramatic structural consequences: the
colour of the reaction solution changed from dark navy-blue to
grass-green and the product isolated changed from the navy-blue
dimer {[CuII2(H2LMe)2(MeCN)2](BF4)4, A} to the grass-green
tetrameric [2 × 2] grid {[CuII4(HLMe)4](BF4)4·3.5MeCN, B}
Employing similar reaction conditions to those employed in the
synthesis of 3·3MeCN·0.5H2O but replacing Cu(BF4)2·4H2O with
Ni(BF4)2·6H2O led to a yellow-brown reaction mixture from
which no well-defined product could be isolated. The addition
of one equivalent of base to the reaction solution resulted in
the isolation of the red-brown nickel(II) [2 × 2] grid-type com-
plex [NiII4(HLEt)4](BF4)·10H2O (5·10H2O) of monodeprotonated
ligands (HLEt)−. Single crystals of 5·10MeCN suitable for X-ray
crystal structure determination were obtained by vapour diffusion
of Et2O into a MeCN solution of the compound (Fig. 6, Tables 2
and S1–S3†).
In contrast to the tetracopper(II) complex 4, in the tetranickel(II)
complex 5·10MeCN the monodeprotonated ligands act as N3–N3
bis-terdentate chelates. The equatorial N3 coordination of two
approximately perpendicular binding pockets results in N6 dis-
torted octahedral coordination spheres for the nickel(II) ions. The
638 | Dalton Trans., 2007, 633–645
This journal is
The Royal Society of Chemistry 2007
©

View Online
Table 2 Comparison of selected bond distances [A] and angles [◦], and selected other data, for the complexes 5·10MeCN and
[NiII4(HLMe)4](Cl)4·5MeCN·13H2O8
˚
8
5·10MeCNa
[NiII4(HLMe)4](Cl)4·5MeCN·13H2Oa
˚
Ni–L/A
Ni–Npz
Ni–Nam
2.104(5)–2.174(4)
2.017(5)–2.047(4)
2.042(4)–2.163(4)
2.107(5)–2.131(5)
1.979(4)–2.000(4)
2.065(4)–2.094(4)
Ni–Npy
L–Ni–L/◦
Npy–Ni–Nam
Npy–Ni–Npz
90.6(2)–93.1(2)
165.3(2)–169.4(2)
75.6(2)–77.4(2)
79.2(2)–79.9(2)
155.7(2)–156.9(2)
76.4(2)–77.0(2)
N
am–Ni–Npz
Other data
T/K
Structural type
150
153
Tetramer; [2 × 2] grid-type
Tetramer; [2 × 2] grid-type
a Only angles within the same ligand strand are given.
shorter than those observed in the ethylene linked analogue 5
˚
[2.017(5)–2.047(4) A]. Given that the amide nitrogen atom is the
central donor atom of the approximately planar N3-chelate, this
allows the Nam–Ni–Npy bond angle in these five-membered chel◦ate
rings in [NiII(HLMe)]4 to open up somewhat [79.2(2)–79.9(2) ],4
4+
although they are still rather small for a six-coordinate first row
transition metal ion, especially when compared to the same angle
in the six-membered chelate rings in the ethylene linked analogue
5 [90.6(2)–92.6(2)◦]. In contrast, in both of these complexes the
Npz–M–Nam angles of the N3-chelate are contained within rather
rigid five-membered chelate rings so are similar to one another,
although a slightly wider range of angles is observed for the
ethylene linked complex 5 [76.4(2)–77.0(2)◦ for [NiII(HLMe)]4
4+ 4
vs 75.6(2)–77.4(2)◦ for 5].
Magnetic and EPR studies
The temperature dependence of the magnetic susceptibility was
measured for the complexes 3·4H2O, 4 and 5·10H2O. The plots of
leff and vm, per Cu (or Ni), versus T for the compounds 3·4H2O
and 5·10H2O are given in Fig. 7 and 8 while that for 4 is given in
Fig. 6 Molecular structure of [NiII4(HLEt)4]4+, one of the two cations
present in the asymmetric unit of 5·10MeCN. Hydrogen atoms not
involved in hydrogen bonds are omitted for clarity.
asymmetric unit contains two crystallographically independent
grid cations. As in the square compounds 3·3MeCN·0.5H2O and
4 the pairs of immediately adjacent amide groups in 5·10MeCN
are arranged in a coplanar fashion. However, in contrast to
the copper(II) complexes 3·3MeCN·0.5H2O and 4 where the
coplanarity was achieved by N–H · · · O hydrogen bonding between
these amide moieties, in the case of this nickel(II) complex the
coplanarity is achieved by intramolecular O · · · H · · · O hydrogen
bonds between these amide moieties (Table S3). Similar [2 × 2]
grid-type copper(II)3,4 and nickel(II)4 complexes of the mono-
−
deprotonated lower ligand homologue (HLMe
)
also contained
O · · · H · · · O hydrogen bonds and such a species is part of the
equilibrium postulated for the monodeprotonated free ligand
(HLEt)− (motif v in Fig. 2).
The Ni–N distances in this ethylene linked [2 × 2] grid-type
complex 5·10 MeCN are broadly similar to those in the related
methylene linked [2 × 2] grid-type complex [NiII4(HLMe)4](Cl)4·
5MeCN·13H2O described earlier by Stoeckli-Evans4 (Table 2).
However, it is interesting to note that the Ni–Nam bonds in the
Fig. 7 Thermal variation of leff (᭺) and vm (ꢀ), per copper(II), for
[CuII4(H2LEt)2(HLEt)2](BF4)6·4H2O (3·4H2O). The solid lines represent the
best fits (see text). Inset: Tetranuclear arrangement of the copper(II) centres
in 3·4H2O.
4
˚
methylene linked analogue [1.979(4)–1.999(4) A] are all slightly
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 639
©

View Online
For the complexes 4 and 5·10H2O, it was found that J12 = J14
=
J, in agreement with the very similar M · · · M distance obtained
from crystallography, and thus one value for J (−3.02 0.02 and
−0.80 0.02 cm−1, respectively) yielded the best fit, together with
the following parameters: g = 2.07 0.01, 2.14 0.01; TIP = (65
5) × 10−6, (100 5) × 10−6 cm3 mol−1, per M(II), % monomer =
0.1, 0 for the complexes 4 and 5·10H2O, respectively. An S = 1 cycle
square model was used for the nickel(II) complex 5·10H2O20,21 and
the zero-field splitting term D was assumed to be zero. Perusal of
the value of vm at the maximum (0.114 cm3 mol−1), and comparing
it to the plateau value expected in distorted octahedral nickel(II)
species having zero field splitting with no exchange coupling,22
suggests an approximate value of |D| of ∼2 cm−1. We note
that J values of −2.93 cm−1 and −1.32 cm−1 were obtained
for the related complexes [CuII4(HLMe)4](ClO4)4·5MeOH·4H2O
and [NiII4(HLMe)](Cl)4·5MeCN·13H2O, respectively,4 in very good
agreement with the present results for the HLEt compounds.
The multiplicities and relative energies of the spin states in zero
field can be obtained directly through matrix diagonalization of
the exchange Hamiltonian (eqn (1)) or from published equations.23
For the “magnetically rectangular” complex 3·4H2O (J12 = J14,
J13 = 0) we obtain a ground singlet state, triplet states at 4.5 cm−1
and 8.6 cm−1, a singlet state at 9.0 cm−1, a triplet state at 9.3 cm−1
and a quintet state at 13.4 cm−1. For the “square” complex 4
(J12 = J14, J13 = 0) we obtain a ground singlet state, a triplet state
at 6.04 cm−1, two triplet states and a singlet state at 12.08 cm−1
and a quintet state at 18.12 cm−1. These spin state multiplicities
and energies do not take account of the admixture of states
from closely spaced multiplets in applied magnetic fields or of
small zero field splittings within triplet and quintet states due to
second order exchange (possibly ∼0.01 cm−1) and dipole–dipole
Fig. 8 Thermal variation of leff (᭺) and vm (ꢀ), per nickel(II), for
[NiII4(HLEt)4](BF4)4·10H2O (5·10H2O). The solid lines represent the best
fits.
Fig. S1.† Observations of maxima in the magnetic susceptibility
measurements at very low temperatures for the tetranuclear
copper(II) complexes 3·4H2O and 4 and the nickel(II) complex,
5·10H2O, indicated very weak antiferromagnetic coupling.
The magnetic data for the two copper(II) tetranuclear clusters
were fitted to a model19,20 that uses the numbering scheme shown
in Fig. 3 and 7 (inset) and the exchange Hamiltonian (eqn (1)), in
which the 2J13 cross (diagonal) term was set at zero;
H = −2J12 (S1·S2 + S3·S4) − 2J14(S1·S4 + S2·S3)
− 2J13(S1·S3 + S2·S4)
(1)
A full matrix diagonalisation of the 16 × 16 matrix was also
carried out that confirmed the J values and yielded the energies
of the coupled states, S, used below in the EPR discussion. The
best-fit to this tetranuclear S = 1/2 model was obtained with
˚
interactions (for copper(II) ions approximately 7 A apart estimated
as ∼0.005 cm−1).
J12 = −1.9
0.1 cm−1, J14 = −2.4
0.1 cm−1, g = 1.99
The EPR spectrum of a polycrystalline powder of 3·4H2O was
recorded at X-band frequencies (∼9.5 GHz) at temperatures from
295 K down to 2.5 K. As shown in Fig. 9 it exhibited an intense
almost isotropic resonance at g ∼ 2.10 whose derivative peak to
peak width of around 150 G was independent of temperature.
0.01, TIP = (65
5) × 10−6 cm3 mol−1 [per copper(II)] and %
monomer = 0.1 for compound 3·4H2O. This was a markedly
improved fit compared to using a single J value. The g value
is a little low for copper(II) but it essentially reflects the room
temperature value of leff, per copper(II), being close to the spin-
only value for d9. It is lower than the value obtained from the
powder EPR spectrum (vide infra), but this is not uncommon in
fitting of susceptibilities when allowing parameters to refine to the
best fit. The small J values obtained from the fitting procedure
are anticipated for a tetranuclear system having pyrazine bridging
along each side of the distorted square since pyrazine is known to
yield a poor antiferromagnetic exchange pathway when bridging
copper(II) ions. Furthermore, each pyrazine moiety binds in an
equatorial–axial manner to adjacent copper(II) atoms (Fig. 3).
2
2
Thus the mutual orthogonality of the dx
‘magnetic’ orbitals
−y
on adjacent copper(II) centres will provide a ferromagnetic con-
tribution to the nett coupling and lead to the small negative J
values obtained. The distortions of the copper(II) coordination
geometries towards trigonal bipyramidal will further cause a
diminution in resultant J values. The H-bonded amide pathways
are not likely to contribute significantly to the J values. The small
differences in J12 and J14 probably reflect differences noted in
Cu · · · Cu edge separations even though the adjacent edges are of
rather similar separation, i.e. Cu(1) · · · Cu(2) ≈ Cu(1) · · · Cu(4) >
Cu(3) · · · Cu(4) ≈ Cu(2) · · · Cu(3).
Fig. 9 EPR spectrum of 3·4H2O at 9.435 GHz and 110 K. Main spectrum
(g = 2 region): Spectrometer gain 4.0 × 103, microwave power 1.05 mW,
100 kHz modulation amplitude 1.0 G, scan time 84 s, time constant 41 ms.
Inset of low field region (including g = 4 region): Spectrometer gain 2.0 ×
105, microwave power 5.26 mW, 100 kHz modulation amplitude 5.0 G,
scan time 84 s, time constant 41 ms.
640 | Dalton Trans., 2007, 633–645
This journal is
The Royal Society of Chemistry 2007
©

View Online
Table 3 Spin Hamiltonian parameters: g values, spin state multiplicities
(S) and zero-field splitting parameters (D) and line widths for the simulated
spectra of 4 shown in Fig. 10. The copper(II) hyperfine interaction has been
set to zero, as it is assumed to be averaged out by the exchange interactions
Another isotropic resonance of width 120 G was observed at g ∼
4.2 at 1000 times the spectrometer gain at 295 K and 110 K
(inset, Fig. 9), but there was no clear evidence for its presence
at 2.5 K. No other resonances were observed at any temperature
over the field range from 0 to 8000 G. The X-band EPR spectra
of a polycrystalline powder of 4 at 295 K and 2.3 K are shown
in Fig. 10 (solid lines), where the changes in appearance of the
spectra are due to an increased linewidth and a decreased g value
for the higher field region at 2.3 K. No other resonances were
observed in the range from zero field to 8000 G, specifically near
g = 4, at any temperature.
T/K gꢂ
g⊥
S
D/10−4 cm−1 rꢂ/10−4 cm−1 r⊥/10−4 cm−1
295.0 2.048 2.150
295.0 2.048 2.150
2.3 2.042 2.143
1
2
1
15
7
15
37
37
37
42
42
47
transitions between levels within the triplet (S = 1) and quintet
(S = 2) states. The present complexes are similar to some other
tetranuclear grid-like copper(II) complexes,26–28 although in the
present case, the exchange interactions are an order of magnitude
or more smaller. This means that the zero-field splittings of the
triplet and quintet states due to second order exchange and dipolar
interactions between the copper(II) ions will be quite small and
not resolved if they are less than the component line widths.
Furthermore, it can be shown that since the directions of the x, y
and z axes of the g matrices of the individual ions are the same,
the components of the g matrices of each of the multiplets will be
the same.26,27 Thus the question of whether the resonances arise
from transitions within triplet and/or quintet states cannot be
resolved from the appearance of the spectra alone, but requires a
detailed analysis of the temperature dependence of the resonance
intensities. Although a detailed analysis was not performed, a
comparison of the spectral intensities at 4.3 K and 2.5 K for
3·4H2O and at 3.3 K and 2.3 K for 4, combined with the relative
population of each of the spin multiplets and the application of the
Boltzmann 1/T factor to the states within each multiplet, leads to
the conclusions, consistent with the magnetic susceptibility results,
that both complexes have a S = 0 (singlet) state lowest in energy
with the resonances observed in the 4 K region arising almost
completely from the lowest triplet level. At temperatures much
above 50 K, the spin states are populated very closely according
to their multiplicity and presumably the other two triplet states
and the quintet state contribute to the EPR spectral intensity. As
shown by the broken lines in Fig. 10a and 10b, the spectra of
4 can be adequately simulated assuming either triplet or quintet
states and the parameters listed in Table 3. The simulations of
Fig. 10 included a small zero-field splitting term, D, in the spin
Hamiltonian, where the value of D given in Table 3 was the largest
that did not have a significant effect on the spectrum.
Fig. 10 EPR spectra of 4. Full line: experimental spectra; broken lines:
simulated spectra using parameters as in Table 3. (a) Experimental
spectrum at 295 K, microwave frequency 9.623 GHz, spectrometer gain
1.0 × 103, microwave power 1.05 mW, 100 kHz modulation amplitude
0.5 G, scan time 84 s, time constant 41 ms. Simulated spectrum: S =
2. The simulated spectrum for S = 1 is identical. (b) 2.3 K, microwave
frequency 9.695 GHz, spectrometer gain 4.0 × 103, microwave power
1.05 mW, 100 kHz modulation amplitude 1.0 G, scan time 84 s, time
constant 41 ms. Simulated spectrum: S = 1.
The relationship between the EPR spectra of the tetranuclear
cluster as a whole and that expected from the individual copper(II)
ions can be understood by an extension of the argument of Mann
et al. as follows.28 Each copper(II) ion in both 3·4H2O and 4 may be
visualised as lying near the centre of the base of an approximately
square pyramidal arrangement. Using Cu(1) in 3·4H2O as an
example, we see that O(4), N(1), N(2) and N(3) form the base for
Cu(1) and N(10) is in the apical position. Although the g matrix
axes of Cu(1) need not closely coincide with the bond directions,
it is reasonable to propose that gz lies along the Cu(1) · · · N(10)
On first inspection, both spectra have the appearance of arising
from a simple S = 1/2 system. Indeed, EPR spectra with a
superficial similarity to that of 4 have been attributed to copper(II)
ions subjected to fluxional Jahn–Teller distortions24 or in a dz
2
ground state.25 In the present case, the essentially temperature
independent appearance of both spectra from room temperature
down to below 3 K contra-indicates these interpretations. More
importantly, interpretations in terms of total spin S = 1/2 systems
are clearly untenable given the tetranuclear configuration of
these complexes and the EPR spectral features must arise from
direction (the longest bond distance 2.316 A) and that gx and gy are
˚
closely similar in magnitude and lie close to the plane containing
˚
O(4), N(1), N(2) and N(3) with bond distances between 1.91 A
˚
and 2.032 A respectively. Further, we may define the direction
of gx as being close to that of the bond Cu(1) · · · O(4), leaving
the bond direction of Cu(1) · · · N(1) to define the direction of
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 641
©

View Online
gy. An analogous argument applies to the other three copper(II)
ions of 3·4H2O and to the four copper(II) ions of 4. Since the +
and − directions are indistinguishable by EPR, directions such
as those of Cu(1). . .O(4) and Cu(1). . .N(2) can be regarded as
magnetically equivalent. Similar considerations apply to 4.
Conclusions
The potentially bis-terdentate diamide ligand N,Nꢀ-bis[2-
(2-pyridyl)ethyl]pyrazine-2,3-dicarboxamide (H2LEt) has been
structurally characterised. Potentiometric titrations carried
out on N,Nꢀ-bis(2-pyridylmethyl)pyrazine-2,3-dicarboxamide
(H2LMe) and H2LEt, revealed rather low pKa values for the
deprotonation of the first amide group. This was tentatively
attributed to the possible formation of equilibrium structures,
which could stabilise the monodeprotonated species and at the
same time enable the amide groups to become coplanar.
Considering the tetranuclear structure as a whole, we see that
the bond directions equivalent to Cu(1) · · · O(4) and hence the
directions of gx of the individual ions are approximately parallel
for all four copper(II) ions, point out of the plane of the copper(II)
grid and become the de-facto gꢂ principal direction. Furthermore,
the directions of gy and gz are rotated by 90◦ around gx as we
progress around the sides of the grid, establishing the conditions
for anti-symmetric exchange coupling in two dimensions between
magnetically inequivalent sites, analogous to that occurring in
Tetracopper(II) complexes with paddle-wheel square-like struc-
tures, [CuII4(H2LEt)2(HLEt)2](BF4)6 (3) and [CuII4(HLEt)4](BF4)4 (4),
have been obtained by the 1 : 1 reaction of ligand H2LEt and
Cu(BF4)2·4H2O, without and with the addition of one equivalent
of base, respectively. In both complexes the ligands act as hybrid
NO–N3 bidentate–terdentate chelates, resulting in N4O coordi-
nation spheres about the copper(II) ions. The most pronounced
differences between the tetrameric cations of 3 and 4 are observed
in the five-coordinate copper(II) coordination spheres (sav = 0.20
versus s = 0.45) and the fact that in compound 3 the ligand exists as
monodeprotonated ligand (HLEt)− as well as neutral zwitterionic
ligand H2LEt, whereas in complex 4 solely monodeprotonated
some CuCl4 complexes.29
2−
In the case of 3·4H2O the absence of features due to anisotropic
g values, copper(II) hyperfine structure and splittings due to second
order exchange and dipolar interactions suggests that the exchange
interactions are such that all anisotropic interactions are collapsed
into a single isotropic line. The exchange averaged g value of
2.10 ( 0.01) can be obtained from g values of the individual
2
2
ions consistent with the range of expectation for a dx
ground
−y
state for five-coordinate copper(II), namely gꢂ ∼ 2.24 and g⊥
∼
2.03. One dimensional intermolecular exchange evidenced by the
observation of the g ∼ 4 resonance and attributed to the N–
H · · · N hydrogen bonds linking the polymeric chain structure of
tetranuclear subunits (see above and Table S3) may well make a
contribution to the exchange averaging at higher temperatures.26
The disappearance of the resonance at g ∼ 4.2 at very low
temperatures could be due to the depopulation of the higher lying
multiplets and consequent reduction in this contribution.
ligand (HLEt)− is present. The change from methylene (H2LMe
)
to ethylene (H2LEt) ligand ‘arms’ has had a profound effect on the
resulting copper(II) coordination environments (square pyramidal
and octahedral in the analogous methylene linked complexes A
and B, respectively3) as well as on the molecular architecture of
the product formed in the absence of added base (the analogous
methylene linked complex A is a dimer3 not a tetramer).
A nickel(II) [2 × 2] grid-type complex [NiII4(HLEt)4](BF4)4 (5) has
been obtained by the 1 : 1 reaction of ligand and Ni(BF4)2·6H2O,
employing one equivalent of base. In contrast to the copper
chemistry described above, no defined product could be isolated
in the absence of added base. In complex 5 the monodeproto-
nated ligands (HLEt)− act as bis-terdentate chelates, resulting in
distorted N6 octahedral coordination spheres about the nickel(II)
ions.
Weak antiferromagnetic spin coupling has been observed for all
three tetranuclear complexes 3–5. For the copper(II) complexes, the
magnetic susceptibility measurements have been used to establish
the energies and multiplicities of the spin states. A spin singlet
state (S = 0) is lowest in both cases, with a spin quintet state (S =
2) highest at energies of 13.4 cm−1 and 18.12 cm−1 for 3·4H2O
and 4 respectively. The X-band EPR spectra of polycrystalline
powders of the copper(II) complexes show evidence for exchange
interactions consistent with the J values obtained from the
magnetic susceptibility measurements. At temperatures below
4 K, the EPR signals for these complexes are primarily due to
transitions within the lowest lying triplet state. For 4, the apparent
g values of gꢂ = 2.05 and g⊥ = 2.15 can be explained as resulting
from the alignment of the g matrix axes of the individual copper(II)
ions and intramolecular exchange between the copper(II) ions
of the square. Due to solute aggregation and disintegration of
the complexes in solution, frozen solution spectra are unable to
confirm the relative roles of inter- and intra-molecular exchange.
However, the g values of the individual ions are consistent with
In the case of 4, the apparent reversal of the g values can be
explained if exchange averaging occurs between the ions of the
grid along the y and z axes to give an exchange averaged peak with
a g value of 1/2(gy + gz). Since this peak is the accumulation of
intensity when the magnetic field is in the y − z plane, it will have a
greater intensity than the peak corresponding to gx and thus give
a perpendicular appearance. We then can identify the “parallel”
peak at g ≈ 2.05 with gx and the “perpendicular” peak at g ≈
2.15 with 1/2(gy + gz). If gx = gy = 2.05, we obtain gz = 2.25,
leading to single copper(II) ion values of gꢂ = 2.25 and g⊥ = 2.05,
2
2
consistent with the expected dx
ground state for five-coordinate
−y
copper(II) in an approximately square pyramidal configuration.
This is also consistent with coordination to N3O1 inasquare planar
environment.30 A further effect of the exchange averaging is to
smear out both the hyperfine interactions of each of the copper(II)
ions and the small zero-field splittings arising from second order
exchange and dipolar interactions between the copper(II) ions.
Although inter- and intra-molecular interactions may be de-
lineated through a comparison of the spectra of powders and
frozen solutions, both these tetranuclear complexes appear to
form solute aggregates in the more concentrated frozen solutions
of less aggressively coordinating solvents such as MeCN and
MeOH and disintegrate in dilute solution and in the more strongly
coordinating solvents such as DMF. A frozen solution of 4 in DMF
exhibited a monomeric copper(II) spectrum with gꢂ = 2.240 and
g⊥ = 2.059, together with superhyperfine structure consistent with
coordination by three N atoms. These g values are consistent with
the analysis of the powder spectrum described above.
2
2
the expected dx
an approximately square pyramidal configuration.
ground state for five-coordinate copper(II) in
−y
642 | Dalton Trans., 2007, 633–645
This journal is The Royal Society of Chemistry 2007
©

View Online
E was taken when variations over 1–2 min were no greater than the
0.1 mV resolution of the pH meter. About 50 titration points were
taken for each case. Both ligands were titrated in duplicate. The
results of the titrations were processed using a custom computer
application to calculate the required pKa values.
Experimental
General
The magnetic susceptibility measurements were carried out using
a Quantum Design MPMS5 Squid Magnetometer as described
previously.31 Continuous wave (CW) EPR spectra were obtained
with a Bruker ESP380FT/CW X-band spectrometer at Monash
University. A standard rectangular TE102 cavity was used with a
Bruker nitrogen gas flow system for temperatures between 295 K
and 120 K and with a quartz finger dewar for measurements at
77 K. Measurements at temperatures below 77 K down to 2.3 K
were performed with an ER4118 cylindrical cavity and an Oxford
Instruments CF935 cryostat. The temperature was calibrated
against a germanium thermometer, using a carbon resistor as a
transfer standard. The microwave frequency was measured with
an EIP Microwave 548A frequency counter and the g values de-
termined with reference to the F+ line in CaO (2.0001 0.0001).32
Spectrum simulations used the SOPHE software described by
Hanson and co-workers.33 The potentiometric titrations of H2LMe
and H2LEt were conducted in aqueous solutions with the addition
of known amounts of standard 0.1 M HCl. Electrode potentials
were measured using a high-precision Metrohm pH meter in mV
mode with a resolution of 0.1 mV. Titrant (CO2-free 0.1 M NaOH)
was introduced manually for each point using a Metrohm digital
burette. Titrations were carried out in the presence of 0.1 M
NaClO4 as an inert electrolyte. For each point the potential reading
X-Ray crystallography
Single crystal X-ray data were collected with a Bruker SMART
˚
CCD area detector diffractometer (k = 0.71073 A). The struc-
tures were solved by direct methods using SHELXS-9734 and
refined against F2 using full-matrix least-squares techniques
with SHELXL-97.35 A summary of the crystallographic data
and refinement parameters for H2LEt, 3·3MeCN·0.5H2O, 4 and
5·10MeCN is given in Table 4.
CCDC reference numbers 296868 (4), 296869 (3·3MeCN·
0.5H2O), 296870 (H2LEt) and 296871 (5·10MeCN).
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b614796h
Syntheses
Dimethyl pyrazine-2,3-dicarboxylate (2). Solid KMnO4 (174 g,
1.10 mol) was slowly added to a solution of quinoxaline (24.0 g,
◦
184 mmol) in H2O (2 L) at 80–90 C. The resulting mixture was
maintained at this temperature for 1 h, before it was allowed to
stir at RT for another 8 h. Then EtOH (40 mL) was added to
Table 4 Crystallographic data for H2LEt and the tetranuclear complexes 3·3MeCN·0.5H2O, 4 and 5·10MeCN
H2LEt
3·3MeCN·0.5H2O
4a
5·10MeCN
Empirical formula
Mr/g mol−1
T/K
C20H20N6O2
376.42
83(2)
Monoclinic
P21/n
5.0400(2)
20.6432(6)
17.8496(2)
90
93.204(2)
90
1854.20(9)
4
1.348
C21.5H22B1.5CuF6N6.75O2.13
602.71
150(2)
Triclinic
¯
P1
13.9483(2)
14.1741(2)
26.0046(2)
82.3380(10)
86.1350(10)
87.7230(10)
5081.28(11)
8
C80H76B4Cu4F16 N24O8
2103.04
168(2)
Tetragonal
¯
I4
16.152(3)
16.152(3)
17.855(8)
90
90
90
4658(2)
2
1.499
C100H106B4F16N34Ni4O8
1912.12
150(2)
Triclinic
¯
P1
21.7134(3)
21.8434(3)
26.24680(10)
108.8230(10)
103.0400(10)
90.7210(10)
11429.8(2)
4
Crystal system
Space group
˚
a/A
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
3
˚
V/A
Z
qcalcd/g cm−3
l/mm−1
1.576
0.939
1.449
0.744
0.092
0.999
F(000)
792
2446
2136
5136
Crystal colour and shape
Crystal size/mm
h range/◦
Colourless needle
0.26 × 0.06 × 0.05
1.51/25.69
−5/6
Green prism
0.40 × 0.22 × 0.20
0.79/25.11
−16/16
Aqua-green plate
0.37 × 0.31 × 0.11
5.11/24.69
−17/8
Red-brown block
0.40 × 0.20 × 0.10
2.72/24.71
−25/24
h
min/hmax
min/kmax
min/lmax
k
l
−25/24
−20/21
10187
−16/16
−18/18
−20/20
9277
−25/24
0/30
0/30
Reflections collected
Independent reflections
Completeness to h (%)
43733
93504
3494 [Rint = 0.0765]
17828 [Rint = 0.0331]
3910 [Rint = 0.0883]
38305 [Rint = 0.0996]
99.2
98.5
98.9
98.3
Data/restraints/parameters 3494/0/261
17828/339/1480
1.049
3910/12/195
0.859
38305/0/2270
0.750
GOF on F2
1.063
Final R indices [I > 2r(I)]
R indices (all data)
R1 = 0.0604, wR2 = 0.1066 R1 = 0.0673, wR2 = 0.1976 R1 = 0.0613, wR2 = 0.1286 R1 = 0.0629, wR2 = 0.1392
R1 = 0.1226, wR2 = 0.1315 R1 = 0.0840, wR2 = 0.2126 R1 = 0.1252, wR2 = 0.1479 R1 = 0.1484, wR2 = 0.1629
Largest diff. max, min/e A−3 0.256, −0.279
2.101, −0.968
0.281, −0.373
0.485, −0.310
a The structure of 4 has disordered solvate and anion species that are not included in the structural model but were taken care of with SQUEEZE (see the
CIF file for more details).
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 643
©

View Online
destroy excess KMnO4 and the resulting mixture was filtered. The
remaining MnO2 cake was extracted with H2O (5 × 400 mL). The
combined aqueous solutions were reduced in volume (to approx.
300 ml) under reduced pressure, then conc. HCl (90 mL) was added
and the resulting pale yellow suspension was stirred at RT for 1 h.
Removal of the solvent under reduced pressure gave a pale yellow
solid, which was taken up in MeOH (250 mL) and treated with
SOCl2 (10 mL). The suspension was refluxed for 5 h, was allowed
to cool to RT and was then filtered. The remaining colourless solid
was washed with MeOH (3 × 50 mL) and the combined filtrates
were evaporated to dryness under reduced pressure. The resulting
orange solid was taken up in Et2O (100 mL) and the suspension
was filtered. The remaining solid was extracted with Et2O (5 ×
100 mL). The filtrates were combined. Removal of the solvent
under reduced pressure gave 32.8 g (91%) of compound 2 in the
form of an analytically pure colourless crystalline material; mp:
for [CuII4(H2LEt)2(HLEt)2](BF4)6·4H2O, C80H86B6N24O12F24Cu4
(2350.72 g mol−1): C 40.88, H 3.69, N 14.30; found: C 40.82, H 3.71,
N 14.58. IR (KBr, disk): m¯ (cm−1) = 3422, 3076, 2924, 2854, 1636,
1607, 1569, 1542, 1484, 1446, 1423, 1376, 1338, 1310, 1253, 1226,
1132, 1083, 1062, 1037, 904, 870, 770, 745, 681, 650, 626, 591, 576,
521. ESI-MS (MeCN): m/z (fragment) = 438.0 ([CuII(HLEt)]+),
377.1 ([H3LEt]+), 189.3 ([H4LEt]2+). UV/Vis/NIR (MeCN): kmax
(e) = 262 (87300), 622 nm (434 M−1 cm−1). Km(MeCN) = 768
X−1 cm2 mol−1. Single crystals of 3·3MeCN·0.5H2O suitable for
an X-ray crystal structure determination were obtained by vapour
diffusion of Et2O into a dilute (1 mg mL−1) solution of compound
3·4H2O in MeCN.
[CuII4(HLEt)4](BF4)4 (4). A solution of ligand H2LEt (400 mg,
1.06 mmol) in MeCN (60 mL) was treated with a solution of
Cu(BF4)2·4H2O (329 mg, 1.06 mmol) in MeCN (10 mL) and a
solution of NEt3 (107 mg, 1.06 mmol) in MeCN (5 mL). Toluene
(220 mL) was added and a dark bottle green solid precipitated.
Filtration of the solid gave 480 mg (86%) of compound 4 in the
form of a bottle green amorphous solid. Elemental analysis (%)
calcd for [CuII4(HLEt)4](BF4)4, C80H76B4N24O8F16Cu4 (2103.04 g
mol−1): C 45.69, H 3.64, N 15.98; found: C 45.36, H 3.63, N 15.77.
IR (KBr, disk): m¯ (cm−1) = 3417, 3074, 2924, 2838, 1636, 1569,
1539, 1484, 1446, 1422, 1373, 1332, 1310, 1250, 1225, 1133, 1083,
1060, 881, 771, 595, 520. ESI-MS (MeCN): m/z (fragment) =
483.9, 438.0 ([CuII(HLEt)]+), 377.1 ([H3LEt]+), 189.3 ([H4LEt]2+).
UV/Vis/NIR (MeCN): kmax (e) = 263 (73200), 656 (603), 803 nm
(shoulder) (356 M−1 cm−1). Km(MeCN) = 531 X−1 cm2 mol−1.
Single crystals of complex 4 suitable for an X-ray crystal structure
determination were obtained by vapour diffusion of Et2O into a
solution of the compound in MeCN.
◦
52–55 C. Elemental analysis (%) calcd for C8H8N2O4 (196.16 g
mol−1): C 48.98, H 4.11, N 14.28; found: C 48.73, H 3.99, N 14.28.
TLC (SiO2, CH2Cl2–10% MeOH): Rf = 0.80. 1H NMR (300 MHz,
CDCl3): d (ppm) = 8.54 (s, 2 H, 2 × pzH), 3.70 (s, 6 H, 2 × CH3).
1
13C{ H} NMR (75 MHz, CDCl3): d (ppm) = 164.0 (CO2), 145.3
(pzCO2), 144.0 (pzH), 52.8 (CH3).
N,Nꢀ-Bis[2-(2-pyridyl)ethyl]pyrazine-2,3-dicarboxamide (H2LEt).
A solution of compound 2 (2.00 g, 10.2 mmol) and 2-(2-
aminoethyl)pyridine (2.49 g, 20.4 mmol) in MeOH (20 mL)
was maintained at 80–90 ◦C for 5 h in an open flask, allowing
the MeOH formed in the reaction to distil off and was then
allowed to cool to RT. Removal of the remaining methanol
under reduced pressure gave the crude product as a pale brown
foamy solid in quantitative yield. Redissolution of the crude
product in CH2Cl2–Me2CO 2 : 1 (50 mL), reduction of the solvent
under reduced pressure (to approx. 10 mL) and filtration of the
resulting colourless solid gave 3.22 g (83%) of analytically pure
H2LEt in the form of a colourless crystalline material; mp: 130–
133 ◦C. Elemental analysis (%) calcd for C20H20N6O2 (376.42 g
mol−1): C 63.82, H 5.36, N 22.33; found: C 63.78, H 5.37, N 22.35.
TLC (SiO2, CH2Cl2–10% MeOH): Rf = 0.38. 1H NMR (300 MHz,
CDCl3): d (ppm) = 8.58 (s, 2 H, 2 × pzH), 8.51 (ddd, 3J6,5 = 4.9 Hz,
4J6,4 = 1.7 Hz, 5J6,3 = 0.9 Hz, 2 H, 2 × 6-pyH), 7.83 (s, 2 H, 2 ×
[NiII4(HLEt)4](BF4)4 (5). A solution of H2LEt (80.2 mg,
213 lmol) in MeCN (7 mL) was treated with a solution of
Ni(BF4)2·6H2O (72.5 mg, 213 lmol) in MeCN (5 mL) and a
solution of NEt3 (21.6 mg, 213 lmol) in MeCN (4 mL). Vapour
diffusion of Et2O into the red-brown reaction mixture afforded
72.4 mg (60%) of compound 5·10H2O in the form of dark red-
brown crystals. Elemental analysis (%) calcd for [NiII4(HLEt)4]-
(BF4)4·10H2O, C80H96B4N24O18F16Ni4 (2263.77 g mol−1): C 42.45,
H 4.27, N 14.85; found: C 42.53, H 3.91, N 14.70. IR (KBr,
disk): m¯ (cm−1) = 3419, 1652, 1605, 1562, 1464, 1443, 1413, 1371,
1346, 1307, 1238, 1145, 1083, 1063, 1031, 976, 898, 847, 769,
735, 680, 643, 588, 521. ESI-MS (MeCN): m/z (fragment) =
3
3
4
NH), 7.61 (dt, J4,5 = J4,3 = 7.6 Hz, J4,6 = 1.7 Hz, 2 H, 2 ×
4-pyH), 7.20 (d, J = 7.8 Hz, 2 H, 2 × 3-pyH), 7.13 (ddd, 3J5,4
=
3
4
7.6 Hz, J5,6 = 4.9 Hz, J5,3 = 1.1 Hz, 2 H, 2 × 5-pyH), 3.91
(q, J = 6.2 Hz, 4 H, 2 × NHCH2), 3.14 (t, J = 6.5 Hz, 4 H,
2+
997.0 ({[NiII4(H2LEt)(HLEt)3](BF4)3} ), 973.0 ({[NiI4(H2LEt)2-
2 × pyCH2). 13C{ H} NMR (75 MHz, CDCl3): d (ppm) = 164.5
1
2+
2+
(HLEt)2(H2O)2](BF4)2} ), 953.1 ({[NiII4(HLEt)4](BF4)2} ), 909.2
(CO), 159.5 (2-py), 149.3 (6-py), 147.1 (pzCO), 144.0 (pzH), 136.6
(4-py), 123.6 (3-py), 121.6 (5-py), 39.1 (NHCH2), 36.9 (pyCH2).
IR (KBr, disk): m¯ (cm−1) = 3410, 3300, 3076, 3010, 2939, 2981,
1674, 1651, 1590, 1560, 1541, 1509, 1474, 1450, 1433, 1366, 1314,
1297, 1275, 1216, 1192, 1160, 1153, 1119, 1074, 1052, 1023, 1006,
991, 886, 774, 761, 749, 650, 612, 567, 511, 445. Single crystals of
H2LEt suitable for an X-ray crystal structure determination were
obtained by recrystallisation from Me2CO.
({[NiII4(HLEt )3(LEt )](BF4 )3 } ), 683.8 ({[NiII4(H2LEt )2(HLEt )2 -
2+
(H2O)3 ](BF4 )3 } ), 653.3, 620.0 ({[NiIINiI3(H2LEt )3(HLEt )-
3+
(H2O)2](BF4)} ), 606.7 ({[NiII4(HLEt)4](BF4)} ), 598.4, 577.5
([NiII4(HLEt)3(LEt)]3+), 433.0 ([NiII4(HLEt)4]4+), 377.2 ([H3LEt]+),
189.4 ([H4LEt]2+). UV/Vis/NIR (MeCN): kmax (e) = 260 (50000),
853 nm (89 M−1 cm−1). Km(MeCN) = 454 X−1 cm2 mol−1. Single
crystals of complex 5·10MeCN suitable for an X-ray crystal
structure determination were obtained by vapour diffusion of Et2O
into a solution of compound 5·10H2O in MeCN.
3+
3+
[CuII4(H2LEt)2(HLEt)2](BF4)6 (3). A solution of ligand H2LEt
(75.3 mg, 200 lmol) in MeCN (10 mL) was treated with a
solution of Cu(BF4)2·4H2O (61.8 mg, 200 lmol) in MeCN (5 mL).
Vapour diffusion of Et2O into the navy blue reaction mixture
afforded 108 mg (88%) of compound 3·4H2O in the form of
turquoise-green feathery crystals. Elemental analysis (%) calcd
Acknowledgements
This work was supported by grants from the University of
Otago (Postgraduate Scholarship to JK and a Bridging Grant),
644 | Dalton Trans., 2007, 633–645
This journal is
The Royal Society of Chemistry 2007
©

View Online
the Marsden Fund (Royal Society of New Zealand), and the
Australian Research Council (Discovery Grant). JFB and JRP
acknowledge support from the School of Physics, Monash Uni-
versity. JK thanks Dr M. H. Klingele (University of Otago) for
many helpful discussions. The authors thank Prof. W. T. Robinson
and Dr J. Wikaira (University of Canterbury) and Prof. G. Clark
and T. Groutso (University of Auckland) for the X-ray data
collections, Prof. M. A. Hitchman (University of Tasmania) and
Dr E. J. L. McInnes (University of Manchester) for helpful
discussions and calculations, and S. S. Iremonger (University of
Otago) for practical support.
14 G. Aylward and T. Findlay, SI Chemical Data, John Wiley & Sons,
Singapore, 1998.
15 B. C. Challis and J. A. Challis, Comprehensive Organic Chemistry,
Pergamon Press, Liverpool, 1979.
16 A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn and G. C. Vershoor,
J. Chem. Soc., Dalton Trans., 1984, 1349.
17 D. S. Cati and H. Stoeckli-Evans, Acta Crystallogr., Sect. E, 2004, 60,
m180; T. Yano, R. Tanaka, T. Nishioka, I. Kinoshita, K. Isobe, L. J.
Wright and T. J. Collins, Chem. Commun., 2002, 1396.
18 P. van der Sluis and A. L. Spek, Acta Crystallogr,, Sect. A, 1990, 46,
194.
19 W. E. Hatfield and G. W. Inman, Jr., Inorg. Chem., 1969, 8, 1376.
20 K. S. Murray, Adv. Inorg. Chem., 1995, 43, 261.
21 J. Ribas, M. Montserrat, T. Costa and X. Solans, Inorg. Chem., 1993,
32, 695.
22 B. J. Kennedy, K. S. Murray, M. A. Hitchman and G. L. Rowbottom,
J. Chem. Soc., Dalton Trans., 1987, 825.
23 R. W. Jotham and S. F. A. Kettle, Inorg. Chim. Acta, 1970, 4, 145.
The same expressions are obtained after some algebraic manipulation
from:; J. W. Hall, W. E. Estes, E. D. Estes, R. P. Scaringe and W. E.
Hatfield, Inorg. Chem., 1977, 16, 1572.
24 A. Abragam and B. Bleaney, Electron paramagnetic resonance of
transition ions, Clarendon Press, Oxford, 1970, p. 456 and p. 459; M. A.
Halcrow, Dalton Trans., 2003, 4375.
References
1 M. Ruben, J. Rojo, F. J. Romero-Salguero, L. H. Uppadine and J.-M.
Lehn, Angew. Chem., Int. Ed., 2004, 43, 3644; L. K. Thompson, O.
Waldmann and Z. Xu, Coord. Chem. Rev., 2005, 249, 2677.
2 J. R. Nitschke, M. Hutin and G. Bernardinelli, Angew. Chem., Int. Ed.,
2004, 43, 6724; S. Brooker, S. J. Hay and P. G. Plieger, Angew. Chem.,
Int. Ed., 2000, 39, 1968; Y. Lan, D. K. Kennepohl, B. Moubaraki, K. S.
Murray, J. D. Cashion, G. B. Jameson and S. Brooker, Chem.–Eur. J.,
2003, 9, 3772; J. Hausmann and S. Brooker, Chem. Commun., 2004,
1530.
25 N. K. Solanki, M. A. Leech, E. J. L. McInnes, F. E. Mabbs, J. A. K.
Howard, C. A. Kilner, J. M. Rawson and M. A. Halcrow, J. Chem.
Soc., Dalton Trans., 2002, 1295.
3 J. Hausmann, G. B. Jameson and S. Brooker, Chem. Commun., 2003,
2992.
26 A. Bencini and D. Gatteschi, EPR of Exchange Coupled Systems,
Springer-Verlag, Berlin, 1990.
4 D. S. Cati, J. Ribas, J. Ribas-Arino and H. Stoeckli-Evans, Inorg. Chem.,
2004, 43, 1021.
5 See for example: D. L. Caulder, C. Bruckner, R. E. Powers, S. Konig,
T. N. Parac, J. A. Leary and K. N. Raymond, J. Am. Chem. Soc., 2001,
123, 8923.
6 See for example:J. R. Nitschke and J.-M. Lehn, Proc. Natl. Acad. Sci.
U. S. A., 2003, 100, 11970; J. R. Nitschke, D. Schultz, G. Bernardinelli
and D. Ge´rard, J. Am. Chem. Soc., 2004, 126, 16538.
7 D. S. Cati and H. Stoeckli-Evans, Acta Crystallogr., Sect. E, 2004, 60,
o210.
27 A. Bencini, D. Gatteschi, C. Zanchini, J. G. Haasnoot, R. Prins
and J. Reedijk, J. Am. Chem. Soc., 1987, 109, 2926; P. Chaudhuri,
I. Karpenstein, M. Winter, M. Lengen, C. Butzlaff, E. Bill, A. X.
Trautwein, U. Florke and H. J. Haupt, Inorg. Chem., 1993, 32, 888.
28 K. L. V. Mann, E. Psillakis, J. C. Jeffery, L. H. Rees, N. M. Harden,
J. A. McCleverty, M. D. Ward, D. Gatteschi, F. Totti, F. E. Mabbs,
E. J. L. McInnes, P. C. Riedi and G. M. Smith, J. Chem. Soc., Dalton
Trans., 1999, 339.
29 Z. G. Soos, K. T. McGregor, T. T. P. Cheung and A. J. Silverstein, Phys.
Rev. B, 1977, 16, 3036.
8 E. B. Fleischer and M. B. Lawson, Inorg. Chem., 1972, 11, 2772; I.
Mallik and S. Mallik, Synlett, 1996, 734.
30 J. Peisach and W. E. Blumberg, Arch. Biochem. Biophys., 1974, 165, 691.
31 J. Klingele (ne´e Hausmann), B. Moubaraki, K. S. Murray, J. F. Boas
and S. Brooker, Eur. J. Inorg. Chem., 2005, 1530.
9 E. B. Fleischer, D. Jeter and R. Florian, Inorg. Chem., 1974, 13,
1042.
10 S. Gabriel and A. Sonn, Ber. Dtsch. Chem. Ges., 1907, 40, 4850; R. G.
Jones and K. C. McLaughlin, Org. Synth., 1950, 30, 87.
11 P. W. Baures, J. R. Rush, A. V. D. Wiznycia, J. B. A. Helfrich and A. M.
Beatty, Cryst. Growth Des., 2002, 2, 653.
12 K. Nonoyama, W. Mori and M. Nonoyama, Polyhedron, 1994, 13, 891.
13 A. Braibanti, F. Dallavalle, G. Mori and B. Veroni, Talanta, 1982, 29,
725.
32 J. E. Wertz, J. W. Orton and P. Auzins, Discuss. Faraday Soc., 1961, 31,
140.
33 G. R. Hanson, K. E. Gates, C. J. Noble, M. Griffin, A. Mitchell and S.
Benson, J. Inorg. Biochem., 2004, 98, 903.
34 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467; G. M.
Sheldrick, Methods Enzymol., 1997, 276, 628.
35 G. M. Sheldrick and T. R. Schneider, Methods Enzymol., 1997, 277,
319.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 633–645 | 645
©