Structures and magnetic properties of imidazolate-bridged tetranuclear and polymeric copper(II) complexes

Article abstract of DOI:10.1016/j.poly.2008.12.004

Eight kinds of imidazolate-bridged copper(II) complexes were found to be classified into two categories from the magnetic properties. The crystal structures of [Cu(L)(μ-im)]_n (Him = imidazole; L = nonane-4,6-dionate, 2,6-dimethylheptane-3,5-dio

Full text of DOI:10.1016/j.poly.2008.12.004

Polyhedron 28 (2009) 2001–2009
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Structures and magnetic properties of imidazolate-bridged tetranuclear
and polymeric copper(II) complexes
*
*
Nobumasa Koyama, Ryo Watanabe, Takayuki Ishida , Takashi Nogami, Tamizo Kogane
Department of Applied Physics and Chemistry, The University of Electro-Communications, Chofugaoka 1-5-1, Chofu, Tokyo 182-8585, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 12 January 2009
Eight kinds of imidazolate-bridged copper(II) complexes were found to be classified into two categories
from the magnetic properties. The crystal structures of [Cu(L)( -im)]n (Him = imidazole; L = nonane-4,6-
dionate, 2,6-dimethylheptane-3,5-dionate) and [Cu(L)( -im)]4 (L = nonane-4,6-dionate, 1-phenylbutane-
1,3-dionate) were determined, to reveal that they consist of polymeric chains and tetranuclear cycles,
respectively. Note that the nonane-4,6-dionate derivative gave the two phases. The Bonner–Fisher model
l
l
Keywords:
Imidazole
Copper(II) ion
Supramolecular chemistry
Magnetic property
Molecular grid
(a one-dimensional antiferromagnetic chain model) was plausibly applied to [Cu(L)(
l-im)]n for the best
fit, while a square model was to [Cu(L)( -im)]4. The complexes with unknown crystal structures were
l
also subjected to magnetic measurements, and the tetra- and polymeric structures could be clearly dis-
tinguished from each other by fitting the magnetic data to appropriate models. The exchange parameters
were comparable for both series (2J/kB = ꢀ78 to ꢀ97 K) because the structurally common bridges Cu–
N(eq)–N(eq)–Cu afford comparable magnitudes of couplings.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
When the crystal structure analysis is unsuccessful owing to the
poor crystallinity, their magnetic measurements would help pre-
Oligonuclear and polymeric copper(II) coordination compounds
attract much attention in materials chemistry; for example, Kahn’s
magnetic material building-blocks [1], Yaghi and co-workers’ me-
tal-organic frameworks [2], and Lehn’s supramolecular chemistry
approach [3]. Some peculiar enzyme functions are based on oligo-
nuclear copper clusters in laccase and ascorbate oxidase [4], where
imidazolates and histidinates play a role of bridging ligands. Cop-
per(II) imidazolate systems afford a variety of structures of self-
assembled supramolecular coordination complexes, e.g., trinuclear
cycles [5], tetranuclear cycles or [2 ꢁ 2] grids [6–10], discrete
linear systems [11–16], polymeric chains [17,18], layers [19–21],
and 3-D networks [19,22].
One of the authors (T.K.) reported imidazolate-bridged cop-
per(II) complexes [Cu(b-diketonato)(l-im)]x (Him stands for imid-
azole) [23], but their crystal structures have not yet been reported.
Generally, there are difficulties in preparing single crystals of poly-
meric coordination compounds, and also in characterizing com-
pounds as infinite polymers or cyclic oligomers solely from the
spectroscopic and elemental analyses. We will report here the
diction of crystal structures. The magnetic properties of the com-
plexes whose crystal structures were unknown were analyzed
according to the model of a polymer- or tetramer-based spin
system.
2. Results
2.1. Preparation
The b-diketones (HL) employed are pentane-2,4-dione (Hacac),
heptane-3,5-dione (Hdprm), nonane-4,6-dione (Hdnbm), 2,6-
dimethylheptane-3,5-dione (Hdibm), 2,2,6,6-tetramethylheptane-
3,5-dione (Hdpm), (1-phenylbutane-1,3-dione (Hbzac) and 1-(2-
thienyl)butane-1,3-dione (Htac), and their copper(II) complexes
are represented as [Cu(L)2] (Scheme 1).
Imidazolate-bridged copper(II) complexes [Cu(L)(l-im)]n have
been prepared in ethanol [23], and they were efficiently obtained
in methanol as well. The reactions of [Cu(L)2] with two molar
equiv. of imidazole gave polymeric complexes [Cu(L)(
([Cu(acac)( -im)]n (1p), [Cu(dprm)( -im)]n (2p), [Cu(dibm)(
im)]n (4p) and [Cu(dpm)( -im)]n (5p)), and tetranuclear [Cu(b-
zac)( -im)]4 (6t). However, [Cu(dnbm)2] gave tetranuclear
[Cu(dnbm)( -im)]4 (3t) instead of polymeric [Cu(dnbm)( -im)]n
(3p) under these conditions. Complex [Cu(dnbm)( -im)]n (3p)
was obtained by the reaction of [Cu(dnbm)2] with two molar equiv.
of imidazole in ethanol. Complex [Cu(tac)( -im)]4 (7t) prepared
l-im)]n
crystal structures of [Cu(dnbm)(l-im)]x as two phases, infinite
l
l
l-
polymer and cyclic tetramer (for structural formula, see below).
Preliminary results have been reported in the previous letter
[24], and the present report will further discuss the following issue.
l
l
l
l
l
* Corresponding authors. Tel.: +81 42 443 5490; fax: +81 42 443 5501 (T. Ishida).
E-mail addresses: ishi@pc.uec.ac.jp (T. Ishida), kogane@crc.uec.ac.jp (T. Kogane).
l
0277-5387/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2008.12.004

2002
N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
Cu–O distances as well as the N–Cu–N and O–Cu–O angles of 4p
are in typical ranges and just comparable to those of 3p. Since
the Cu–N bonds can rotate in solution, the geometry seems to be
fixed in a suitable packing motif on crystallization. Thus, in com-
parison between 3p and 4p, the mutual geometry between neigh-
boring units is different, which is caused by torsion around the Cu–
N bonds.
We found that [Cu(dnbm)(l-im)] has two phases, 3p and 3t. As
Fig. 3 shows the molecular structure of 3t, four Cu ions construct a
rectangle arrangement affording a [2 ꢁ 2]-molecular grid. A half of
the molecule of 3t is crystallographically independent owing to
ꢀ
centrosymmetry in a space group P1. The intramolecular Cu1ꢂꢂꢂCu2
and Cu1ꢂꢂꢂCu2 distances are 5.8694(7) and 5.8933(10) Å, respec-
tively. The Cu2 ꢂꢂꢂCu1ꢂꢂꢂCu2 and Cu1ꢂꢂꢂCu2ꢂꢂꢂCu1 angles are
89.951(12)° and 90.049(12)°, respectively.
As for the CuN2O2 coordination environments in 3t, the O3–
Cu2–N3 and O4–Cu2–N2 are much smaller than the angle of
180°, and, giving saddle-type distortion. Such an out-of-plane
deformation has also been reported in the imidazolate-bridged
Cu4 macrocyle [8]. The coordination around the Cu1 ion is n-SP,
while Cu2 is d-SP. This finding seems to be similar to that of the
Scheme 1. Abbreviations and formulas of copper(II) b-diketonates.
according to the previous report [23] was used for the magnetic
measurements. Here, the notation p and t denote polymeric and
tetranuclear, respectively.
bisimidazolato copper(II) compound possessing
a d-SP/n-SP–
mixed tetranuclear framework. [26]. The dislocations of Cu1 and
Cu2 ions are small (0.049(3) and 0.082(3) Å, respectively) from
the mean basal N2O2 planes, owing to the saddle structures. As
the side view (Fig. 3b) shows, the two basal planes are slightly
canted with each other by 7.7(1)°.
2.2. Molecular structures
Fig. 1 shows the molecular structure of 3p. The complex con-
sists of a helix, which has a dinuclear crystallographically indepen-
Fig. 4 shows the molecular structure of 6t. The refinement was
unsatisfactory because 6t gave only tiny crystals. Large standard
deviations remained and thermal displacement factors were some-
what unrealistic for several carbon atoms. However, the R factors
were decreased to be relatively small (ca. 0.06), and we can safely
conclude that 6t has a tetranuclear macrocyclic structure, or a
dent unit [{Cu(dnbm)(l-im)}2] as
a
repeating motif. The
*
intramolecular nearest-neighboring Cu1ꢂꢂꢂCu2 and Cu2ꢂꢂꢂCu1 dis-
tances are 5.9016(10) and 5.8973(11) Å, respectively. The Cu ions
have a CuN2O2 coordination environment with two nitrogen atoms
from two imidazolates and two oxygen atoms from a dnbm ligand.
Selected geometrical parameters are summarized in Table 1.
Although the geometry about the copper ions Cu1 and Cu2 is
approximately square-planar, the O1–Cu1–N1, O2–Cu1–N4#, O3–
Cu2–N3, and O4–Cu2–N2 bond angles are considerably smaller
than ideal square-planar angle of 180°, being bent toward opposite
directions to form a saddle structure. Namely, the coordination
around the copper ions is considerably distorted from square-pla-
nar toward tetrahedral (d-SP). The Cu1 ion is located slightly above
the mean basal O1–O2–N1–N4# plane by 0.018(3) Å, and the Cu2
ion above the mean basal O3–O4–N2–N3 plane by 0.053(3) Å.
There are eight copper ions in a pitch along a crystallographic 42
screw axis in 3p, contrary to the expectation from the cis configu-
ration of the N–Cu–N coordination, which seems suitable for a
fourfold axis involving four copper ions. It may be related to the
d-SP structures at Cu1 and Cu2. The Cu1ꢂꢂꢂCu2ꢂꢂꢂCu1* angle
(134.66(2)°) is considerably larger than, while the Cu2ꢂꢂꢂCu1*ꢂꢂꢂCu2*
angle (89.64(2)°) is close to, the ideal right angle. The alternating
irregular angles are responsible for the formation of a Cu8-based
chain and not a Cu4-based helix or a tetrameric macrocycle.
Fig. 2 shows the molecular structure of 4p. A dinuclear unit is
crystallographically independent. A zig-zag chain structure is
formed in a space group C2/c, which contains four copper units
in a pitch. As the N–Cu–O angles of 150.37(15)–173.13(15)° indi-
cate, Cu1 has an almost square planer structure (n-SP), while Cu2
has a d-SP one. The dislocations from the mean N2O2 planes are
small (0.053(12) and 0.031(2) Å, respectively), owing to the saddle
structures.
ꢀ
molecular [2 ꢁ 2] grid, in a space group P1. A dinuclear unit is crys-
tallographically independent. The nearest-neighboring CuꢂꢂꢂCuꢂꢂꢂCu
angles are nearly 92° and 88°. The molecular structure of 6t is al-
most identical to that of 3t except for the peripheral substituents,
and furthermore the molecular packings are also similar to each
other in the comparable crystal lattices.
2.3. Magnetic properties
We measured magnetic susceptibilities (vmol) of polycrystalline
specimens of the present compounds on a SQUID magnetometer.
The X-ray diffraction study revealed the crystal structures of 3p,
4p, 3t, and 6t, and we will check the consistency between the
structure and magnetic properties. Fig. 5a and 5b show the vmol
versus T plots for polymeric 3p and 4p, respectively, on the basis
of a copper(II) ion.
The vmol value of 3p was increased on cooling from 300 K,
reached a broad maximum around 60 K, and gradually decreased.
This behavior clearly indicates the presence of dominant antiferro-
magnetic coupling. An extrapolation of the vmol curvature to 0 K
seems to give a non-zero vm intercept. Usually such behavior is re-
lated with the infinite Heisenberg S = 1/2 antiferromagnetic chain,
as originally predicted by Bonner and Fisher [27], and so far con-
firmed by many experiments [28]. Cyclic S = 1/2 antiferromagnetic
clusters having even number of the spin sources would show null
vmol at T ? 0 K. Thus, the non-zero vmol at T ? 0 K is regarded as a
fingerprint of infinite chain structures. On further cooling, the vmol
value of 3p was further decreased with a turning point around
10 K. The origin of this anomaly is not clear at present, but might
be explained in terms of a spin-Peierls transition [29]. Namely, a
strong singlet–dimer formation takes place to reduce drastically
the magnetic susceptibility. To estimate the exchange coupling
within a chain, we used the data above 15 K. Assuming the
The CuꢂꢂꢂCu distances (5.8865(7) and 5.9239(7) Å) of 4p are
comparable with those of 3p. The nearest-neighboring CuꢂꢂꢂCuꢂꢂꢂCu
angles are 102.925(10)° and 132.817(11)°, being suitable for such a
zig-zag chain and quite different from that of 3p. The steric bulki-
ness of peripheral substituents (n-propyl versus isopropyl) may af-
fect the molecular packing in the crystal. Basically, the Cu–N and

N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
2003
Fig. 1. X-ray crystal structure of [Cu(dnbm)(l-im)]n (3p). (a) Crystallographically independent unit. Only a major conformation is shown for the C23–C24 portion. Thermal
ellipsoids are drawn at the 50% probability level. Hydrogen atoms are omitted. (b) Top and (c) side views of a chain. Eight repeating units are drawn. Symmetry operation
codes of ꢃ and # are ꢀy, ꢀ1/2 + x, ꢀ1/2 + z and 1/2 + y, ꢀx, 1/2 + z, respectively.
peak position around 60 K for 3p agrees well with the calculated J
value:
equally-spaced S = 1/2 magnetic chain and the Heisenberg spin
Hamiltonian as H = ꢀ 2JSi ꢂ Si+1, we applied the following expres-
sion [27] with x = |J|/kBT:
R
kBTmax= j J j¼ ca: 1:28:
ð2Þ
NA
l
2Bg2
0:25 þ 0:14995x þ 0:30094x2
vmol
¼
:
ð1Þ
Similarly, 4p exhibited a Bonner–Fisher-type behavior (Fig. 5b).
However, in sharp contrast to the case of 3p, the vmol value of 4p
turned to increase again below 10 K, suggesting a slight contami-
nation with Curie impurity. We analyzed the experimental data
using the Bonner–Fisher model plus a Curie contribution (Eq. (3)).
kBT 1:0 þ 1:9862x þ 0:68854x2 þ 6:0626x3
The best fit gave ꢀ89.8(7) K with g = 2.04(2) for 3p. According to the
Bonner–Fisher analysis, Tmax at which vmol shows a rounded maxi-
mum is simply related with J as expressed by Eq. (2) [27]. The vm

2004
N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
ꢀ96.9(3) K, g = 2.09(2), and Cimp = 0.0086(2) cm3 K molꢀ1 for 2p,
Table 1
Selected bond lengths (Å) and angles (°).
and
2J/kB = ꢀ85(1) K,
g = 2.13(2),
and
Cimp = 0.00232(8)
cm3 K molꢀ1 for 5p. The calculated curves well reproduced the
experimental data (solid lines in Fig. 6). Thus, we assume that they
would be chains.
The content of the Curie impurity varied from 0.6% to 2%
depending on the specimens investigated here, but the coupling
parameters are close to each other (from ꢀ85 to ꢀ97 K for 1p–
5p). This finding supports the assumption that the structurally
common bridges Cu–N(eq)–N(eq)–Cu afford comparable magni-
tudes of couplings. They largely depend on the coordination sites,
equatorial and axial, where the nitrogen atoms are coordinated
[30].
Compounds
3p
4p
3t
Cu1–O1
Cu1–O2
Cu1–N1
Cu1–N4
Cu2–O3
Cu2–O4
Cu2–N2
Cu2–N3
O1–Cu1–N1
O2–Cu1–N4
O3–Cu2–N3
O4–Cu2–N2
1.918(3)
1.932(4)
1.954(5)
1.954(5)a
1.923(4)
1.920(4)
1.981(5)
1.944(5)
158.46(19)
156.73(18)a
159.08(19)
161.3(2)
1.938(3)
1.935(3)
1.913(3)
1.929(4)
1.935(4)
1.932(5)c
1.919(4)
1.918(3)
1.957(3)
1.967(4)
154.9(2)
152.2(2)c
162.6(2)
171.5(2)
1.958(3)
1.977(3)b
1.950(3)
1.926(3)
1.941(3)
1.954(3)
173.13(15)
167.59(16)b
150.37(15)
152.82(15)
The vmol versus T pots of tetrameric 3t and 6t (Fig. 7) exhibited
a
N4 is a symmetry-related equivalence with ꢀy, ꢀ1/2 + x, ꢀ1/2 + z.
N4 is a symmetry-related equivalence with 1/2 ꢀ x, ꢀ1/2 + y, 3/2 ꢀ z.
N4 is a symmetry-related equivalence with 1 ꢀ x, 2 ꢀ y, ꢀz.
a distinctly different feature from those of 3p and 4p. Namely, zero
vmol values are suggested at T ? 0 K. As for 3t, the vmol value in-
creased on lowering temperature and reached a maximum around
55 K. After that the vmol value turned to decrease somewhat stee-
b
c
ply toward
v
mol = 0 cm3 molꢀ1. A Curie impurity was included, as
NA
l
2Bg2
0:25 þ 0:14995x þ 0:30094x2
Cimp
T
indicated with a sharp upturn at the lowest temperature region.
A square model involving four S = 1/2 spins leads to the Heisen-
berg spin Hamiltonian as H = ꢀ2J(S1 ꢂ S2 + S2 ꢂ S3 + S3 ꢂ S4 + S4 ꢂ S1).
The molar magnetic susceptibility can be written as Eq. (4)
[6,7,10] by applying Kambe’s vector coupling method [31] and
the van Vleck equation:
vmol
¼
þ
:
ð3Þ
kBT 1:0 þ 1:9862x þ 0:68854x2 þ 6:062x3
The best fit gave 2J/kB = ꢀ93.1(6) K with g = 2.129(5) and
Cimp = 0.00695(4) cm3 molꢀ1 for 4p. The obtained J is very close to
that of 3p, despite of the different helix motives.
Although the crystal structures of 1p, 2p, and 5p are unknown,
we measured their magnetic susceptibility to predict the crystal
structures. Fig. 6 shows the results of them. They showed a broad
maximum and non-zero vmol at T ? 0 K. This behavior strongly
suggests that they are antiferromagnetically correlated infinite
chains, just like 3p and 4p. There also were Curie tails. The Bon-
ner–Fisher model (i.e., Eq. (3)) was applied, giving 2J/kB = ꢀ96(2) K,
g = 2.01(4), and Cimp = 0.00725(9) cm3 K molꢀ1 for 1p, 2J/kB =
NA
l
2Bg2
2 þ 5 expð2J=kBÞ þ expðꢀ2J=kBÞ
vmol
¼
kBT 7 þ 5 expð2J=kBÞ þ 3 expðꢀ2J=kBÞ þ expðꢀ4J=kBÞ
Cimp
þ
:
ð4Þ
T
A Curie impurity term was added. The experimental data of 3t were
fitted to Eq. (4) to give 2J/kB = ꢀ81.0(2) K, g = 2.061(2), and
Fig. 2. X-ray crystal structure of [Cu(dibm)(l-im)]n (4p). (a) Crystallographically independent unit. Thermal ellipsoids are drawn at the 50% probability level. Hydrogen atoms
are omitted. (b) Side views of a chain. Six repeating units are drawn. Symmetry operation codes of § and are 1/2 ꢀ x, ꢀ1/2 + y, 3/2 ꢀ z, and 1/2 ꢀ x, 1/2 + y, 3/2 ꢀ z,
respectively.

N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
2005
Fig. 3. X-ray crystal structure of [Cu(dnbm)(l-im)]4 (3t). (a) Top and (b) side views. Thermal ellipsoids are drawn at the 30% probability level. Hydrogen atoms are omitted.
Symmetry operation code of is 1 ꢀ x, 2 ꢀ y, ꢀz.
Cimp = 0.0203(2) cm3 K molꢀ1. The dotted line in Fig. 7a implies the
first term in Eq. (3) after the subtraction of the Curie term. This line
approaches to 0 cm3 molꢀ1 at T ? 0 K.
The data of 6t were similarly analyzed (Fig. 7b). The data on 6t
were fitted to Eq. (4) and the parameters were optimized as: 2J/
kB = ꢀ78(1) K, g = 2.13(2), and Cimp = 0.00662(8) cm3 K molꢀ1. The
J value is very close to that of 3t.
for aliphatic b-diketones, aroyl- or hetaroylacetones and fluori-
nated b-diketones, respectively).
The reactions of [Cu(L1)2] and [Cu(L2)2] with two molar equiv. of
imidazole in methanol or ethanol gave polymeric and tetranuclear
complexes ([Cu(L1)( -im)]n and [Cu(L2)(
l l-im)]4), respectively. The
[Cu(dnbm)2] case is exceptional. A mixture of [Cu(dnbm)2] with
two molar equiv. of imidazole gave 3t under the same conditions.
On the other hand, when a mixture solution was concentrated
gradually on heating, yielding 3p (see Section 5).
The crystal structure of 7t is unknown, but we measured its
magnetic susceptibility (Fig. 7b). The data on 7t were also well fit-
ted to Eq. (4) rather than Eq. (3). The molecular structure was sug-
gested to be a square tetramer. The parameters were optimized as:
Complexation reactions of [Cu(L)2] with pyrazole gave different
products. The complexes [Cu(L1)2] afforded di- or tetranuclear
complexes, while [Cu(L2)2] afforded pyrazolate-bridged complexes,
1:1 and/or 1:2 adducts [32]. Furthermore, [Cu(L1)2] afforded differ-
ent products depending upon the solvents used. By successive
addition of imidazole, [Cu(acac)2] gave no isolable imidazolate
complex in chloroform [25], while gave 1p instead of the imidazo-
late complex in ethanol or methanol [23]. It has been reported that
[Cu(L3)2] such as [Cu(hfac)2] (Hhfac stands for 1,1,1,5,5,5-hexafluo-
ropentane-2,4-dione) afforded 1:1 and/or 1:2 adducts in chloro-
form because of their stronger coordination ability [25].
2J/kB = ꢀ85.0(8) K, g = 2.27(1), and Cimp = 0.00230(6) cm3 K molꢀ1
.
The calculated curves well reproduced the experimental data.
Thus, we predict that 7t would consist of a cyclic tetramer.
3. Discussion
3.1. Synthesis
Complexation reactions of [Cu(L)2] with imidazole are very sen-
sitive to the nature of the b-diketonato ligand as well as to the
reaction conditions, sometimes giving different products depend-
ing upon the solvent used. For convenience, [Cu(L)2] can be classi-
fied into [Cu(L1)2], [Cu(L2)2], and [Cu(L3)2] (HL1, HL2, and HL3 stand
3.2. Crystal structures
Chaudhuri et al. prepared cyclic tri- and tetranuclear copper(II)
complexes [5,6], where non-substituted imidazolate was utilized

2006
N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
Fig. 4. X-ray crystal structure of [Cu(bzac)(l-im)]4 (6t). Thermal ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted. Symmetry
operation code of is 2 ꢀ x, 1 ꢀ y, ꢀz.
as a bridge; namely, [Cu3(tacn)3(l-im)3](ClO4)3 and [Cu4(tacn)4(l-
im)4](ClO4)4 ꢂ 2H2O (tacn = 1,4,7-triazacyclononane). The capping
ligands are tridentate and neutral. Consequently, the copper ions
were square-pyramidal. Counter ions (ClO4ꢀ) did not work as li-
gands. On the other hand, the present compounds have monoan-
ionic b-diketonate chelates as a cap. All of the copper ions are
tetra-coordinated (actually n-SP and d-SP). There is no counter an-
ion in the crystal lattice. However, the tetranuclear skelton [Cu4(l-
im)4] of 3t is similar to that of [Cu4(tacn)4(
as well as other known tetramers [7–10].
l-im)4](ClO4)4 ꢂ 2H2O [6]
In the present study, the capping ligands are simple b-diketo-
nates, which occupy two equatorial sites in a cis manner. Accord-
ingly, the imidazolate N atoms are coordinate at the rest position
in a cis manner. The N(im)–Cu–N(im) angles close to 90° favors tet-
ranuclear macrocycles. Though several imidazolate-based preorga-
nized ligands have been utilized for development of [2 ꢁ 2] grids
[6–10], the combination of commercially available copper(II)
diketonates and the unsubstituted imidazolate ion is also versatile
for this purpose.
Fig. 5. Temperature dependence of molar magnetic susceptibility (
[Cu(dnbm)( -im)]n (3p) and (b) [Cu(dibm)( -im)]n (4p). The solid and dotted lines
represent the best fit curves. See the text for the equations and parameters.
vmol) for (a)
l
l
3.3. Magnetic properties
The magnetic data of 1p–5p were analyzed by the Bonner–Fish-
er model as an equally-spaced model. We have to think possible
fittings to alternating-chain and rectangle models, because the
crystallographic analysis indicates that there are two kinds of ex-
change couplings. The present analysis brought about averaged
values from two independent parameters, J and
nation parameter. In more precise treatment using the Heisenberg
spin Hamiltonian as H = ꢀ (2JS22iꢀ1 ꢂ S2i + 2 JS2i ꢂ S2i+1), we could
adopt an alternating antiferromagnetic chain model [33]. One
may imagine the singlet–triplet model [34] in case of the = 0 lim-
it, and accordingly the vmol value approaches to null at T ? 0 K
when is close to zero. On the other hand, the mol versus T profile
appears just like the Bonner–Fisher curve when is close to unity.
aJ, where a is alter-
R
a
a
a
v
a
Fig. 6. Temperature dependence of molar magnetic susceptibility
(vmol) for
The present results on 1p–5p exhibited the non-zero vmol at the
T ? 0 K limit, which guarantees that an approximation as an
equally-spaced chain model is valid. Moreover, the possibility of
[Cu(acac)( -im)]n (1p), [Cu(dprm)( -im)]n (2p), and [Cu(dpm)( -im)]n (5p). The
l
l
l
solid lines represent the best fit curves. See the text for the equations and
parameters.

N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
2007
the most. Hence, though an alternating parameter
introduced, was found to be close to unity.
a
should be
a
The exchange parameters for the present compounds fell in a
typical range of superexchange couplings between copper(II) ions
across imidazolates [5–10]. The mechanism of the superexchange
coupling was previously discussed by Kolks et al. [7] and Chaudh-
uri et al. [6]. The
r-pathway across the nitrogen lone-pairs has
been pointed out, like other diazaaromatic-bridged systems [35].
The present exchange couplings are much larger than that of imi-
dazolate-bridged copper(II) chains having the N-equatorial–N-ax-
ial bridging motif (for example, 2J/kB = ꢀ5.2 K [18]).
4. Concluding remarks
The magnetic properties of 1p–5p, 3t, 6t, and 7t were found to
be devided into two categories: the Bonner–Fisher model was
plausibly applied to polymeric [Cu(L)(
l
-im)]n for the best fit, while
a square model to tetrameric [Cu(L)(l-im)]4. We could not deter-
mine the crystal structures of 1p, 2p, 5p, and 7t at present. It is dif-
ficult to characterize compounds as infinite polymers or cyclic
oligomers solely from the spectroscopic and elemental analyses.
The polymeric and tetranuclear structures could be clearly distin-
guished from each other by fitting the magnetic data to theoretical
curves.
We utilized a simple imidazolate anion for linking several cop-
per(II) diketonates without preorganization of the ligands. It
should be noted that the combination of simple copper(II) diketo-
nates and imidazolate ion is versatile for supramolecular
chemistry.
5. Experimental
5.1. Materials
Bis(b-diketonato)copper(II) complexes employed were [Cu(a-
cac)2], [Cu(dprm)2], [Cu(dnbm)2], [Cu(dibm)2], [Cu(dpm)2] and
[Cu(bzac)2]. They were prepared by the reported method [23].
Imidazole (Tokyo Chemical Industry Co.) was recrystallized from
light petroleum (b.p. range 30–70 °C).
The imidazolate-bridged complexes 1p–5p and 6t are previ-
ously characterized except for melting points. However, these
complexes and 3t were prepared for X-ray diffraction study and
magnetic measurements according to the following method. Com-
plex 7t prepared by the previous report [23] was applied for mag-
netic measurements.
Fig. 7. Temperature dependence of vmol for (a) [Cu(dnbm)(
l-im)]4 (3t) and (b)
[Cu(bzac)( -im)]4 (6t) and [Cu(tac)( -im)]4 (7t). The solid and dotted lines
l
l
represent the best fit curves. See the text for the equations and parameters.
the cyclic structures such as a tetramer is completely discarded.
Compounds 1p, 2p, and 5p consist of chains as well as 3p and
4p. The reports on imidazolate-bridged copper(II) chains are rare
[17,18], in which the copper ions were weekly bonded with ligands
in an axial manner. To our knowledge, the present work is the first
characterization of one-dimensional imidazolate-bridged cop-
per(II) chains showing appreciable antiferromagnetic couplings.
The magnetic data of 3t, 6t, and 7t were analyzed by the square
model as an equally-spaced model. Similarly, possibility of rectan-
gle models deserves discussion, because the crystallographic anal-
ysis indicates that there are two crystallographic independent
copper ions in a tetramer. We actually tried to fit the data of 7t
as well as 3t and 6t to the alternating antiferromagnetic chain
model [32] and the rectangle model [7]. We eventually found that
the latter analysis more likely from the better fit. Furthermore, we
also found that the analysis based on the equal spacing satisfacto-
rily reproduced the experimental results.
[Cu(acac)(l-im)]n (1p). The complex [Cu(acac)2] (103.2 mg,
0.39 mmol) was dissolved in methanol (20 ml) with two molar
equiv. of imidazole (53.3 mg, 0.78 mmol). The solution was re-
fluxed for 2 h. After standing for a few days, a violet precipitate
was formed. The product was filtered, washed with a small amount
of methanol, and dried in vacuo. The yield was 33.5 mg (37%). Mp.
182–185 °C.
[Cu(dprm)(l-im)]n (2p). This complex was prepared according
to a similar method to that described for 1p. The yield was 39%.
Mp. 183–186 °C.
[Cu(dnbm)(l-im)]n (3p). The complex [Cu(dnbm)2] (74.8 mg,
0.20 mmol) was dissolved in ethanol (15 ml) with two molar equiv.
of imidazole (27.3 mg, 0.40 mmol). The solution was refluxed for
2 h and the solvent was concentrated gradually on heating. Black
fine needles were formed during the concentration. They were fil-
tered, washed with a small amount of ethanol, and dried in vacuo,
giving 17.3 mg of 3p (30%). Mp. 153–156 °C.
[Cu(dnbm)(l-im)]4 (3t). The complex [Cu(dnbm)2] (74.8 mg,
0.20 mmol) was dissolved in ethanol (15 ml) with two molar equiv.
of imidazole (27.3 mg, 0.40 mmol). The solution was stirred at
The structurally common bridges Cu–Nim(eq)–Nim(eq)–Cu af-
ford comparable magnitudes of couplings. The 2J values were ob-
served between ꢀ78 and ꢀ97 K for both the chains and cycles
investigated here, and accordingly the
a value should be 0.80 at

2008
N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
Table 2
Selected crystallographic data for 3p, 4p, 3t, and 6t.
Compounds
3p
4p
3t
6t
Formula
C12H18CuN2O2
285.83
black block
0.36 ꢁ 0.18 ꢁ 0.09
97
C12H18CuN2O2
285.83
blue platelet
0.2 ꢁ 0.2 ꢁ 0.2
90
C48H72Cu4N8O8
1143.32
blue prism
0.51 ꢁ 0.4 ꢁ 0.24
296
C52H48Cu4N8O8
1167.18
blue prism
0.08 ꢁ 0.05 ꢁ 0.01
90
Formula weight
Habit
Dimension (mm3)
T (K)
Crystal system
Space group
a (Å)
b (Å)
c (Å)
tetragonal
P42/n
21.047(15)
21.047(15)
12.334(11)
90
90
90
5764(7)
16
1.390
monoclinic
C2/c
28.03(4)
16.91(2)
11.125(18)
90
95.03(6)
90
5254(13)
16
1.445
6012
1.655
0.0546
0.0978
triclinic
P1
triclinic
P1
ꢀ
ꢀ
8.396(6)
12.616(6)
13.738(8)
93.78(4)
99.89(5)
106.49(5)
1364.3(14)
1
1.391
5987
1.593
0.0580
0.0737
8.168(18)
11.39(3)
13.16(3)
95.23(7)
91.78(6)
103.72(7)
1183(5)
1
1.639
5353
1.840
0.0569
0.0806
a
(°)
b (°)
c
(°)
V (Å3)
Z
Dcalc (g cmꢀ3
Unique data
)
5764
1.591
0.0566
0.0588
l
(Mo K
a
) (mmꢀ1
)
R (F)a (I > 2
r(I))
Rw(F2)b (all data)
a
R =
R
||Fo| ꢀ |Fc||/
R|Fo|.
P
P
2
1=2
Rw ¼ ½ wðF20 ꢀ Fo2Þ = wðF2oÞꢄ
.
b
room temperature for 2 h, and allowed to stand at room tempera-
ture for 10 days. The resultant bluish black blocks were filtered,
washed with a small amount of ethanol, and dried in vacuo. The
yield was 10.3 mg (18%). Mp. 113–114 °C.
to 1.8 K on a Quantum Design MPMS SQUID magnetometer. The
magnetic responses were corrected with diamagnetic blank data
of the sample holder measured separately. The diamagnetic contri-
bution of the sample itself was estimated from Pascal’s constants.
[Cu(dibm)(l-im)]n (4p). The complex [Cu(dibm)2] (50.8 mg,
0.14 mmol) was dissolved in methanol (10 ml) with two molar
equiv. of imidazole (18.4 mg, 0.27 mmol). The solution was re-
fluxed for 2 h. After standing for a few days, dark blue needles were
formed. They were filtered, washed with a small amount of meth-
anol, and dried in vacuo. The yield was 16.2 mg (40%). Mp. 183–
185 °C.
Acknowledgements
This work was supported by Grants-in-Aid for Scientific Re-
search (Nos. 15073101, 16550121, and 19550135) from the
Ministry of Education, Culture, Sports, Science and Technology,
Japan.
[Cu(dpm)(l-im)]n (5p). This complex was prepared by a similar
method to that described for complex 1p. The yield was 85%. Mp.
>195 °C.
Appendix A. Supplementary data
[Cu(bzac)(l-im)]4 (6t). The complex [Cu(bzac)2] (100.2 mg,
CCDC 702612, 702613, 702869, and 702870 contain the supple-
mentary crystallographic data for 3p, 3t, 4p, and 6t, respectively.
These data can be obtained free of charge via http://www.ccdc.ca-
m.ac.uk/conts/retrieving.html, or from the Cambridge Crystallo-
graphic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK;
fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk. Sup-
plementary data associated with this article can be found, in the
online version, at doi:10.1016/j.poly.2008.12.004.
0.26 mmol) was dissolved in methanol (25 ml) with two molar
equiv. of imidazole (35.3 mg, 0.52 mmol). The solution was re-
fluxed for 2 h. After standing for a few days, dark green crystals
were formed. They were filtered, washed with a small amount of
methanol and dried in vacuo. The yield was 35.1 mg (46%). Mp.
185–187 °C.
5.2. X-ray crystallographic analysis
References
X-Ray diffraction data of 3p, 4p, 3t, and 6t were collected on a
[1] O. Kahn, Molecular Magnetism, VCH, New York, 1993.
[2] M. Eddaoudi, D.B. Moler, H. Li, B. Chen, T.M. Reineke, M. O’Keefe, O.M. Yaghi,
Acc. Chem. Res. 34 (2001) 319.
[3] J.-M. Lehn, Supramolecular Chemistry: Concepts and Perspective, VCH,
Weinheim, 1995.
[4] A. Messerschmidt, R. Huber, Eur. J. Biochem. 187 (1990) 341.
[5] P. Chaudhuri, I. Karpenstein, M. Winter, C. Butzlaff, E. Bill, A.X. Trautwein, U.
Florke, H.J. Haupt, J. Chem. Soc., Chem. Commun. (1992) 321.
[6] P. Chaudhuri, I. Karpenstein, M. Winter, M. Lengen, C. Butzlaff, E. Bill, A.X.
Trautwein, U. Flörke, H.-J. Haupt, Inorg. Chem. 32 (1993) 888.
[7] G. Kolks, S.J. Lippard, J.V. Waszezak, H.R. Lilienthal, J. Am. Chem. Soc. 104
(1982) 717.
Rigaku R-axis RAPID diffractometer or a Rigaku Saturn-70 CCD dif-
fractometer with graphite monochromated MoK
a radiation
(k = 0.71069 Å). The structures were directly solved by a heavy-
atom method and expanded using Fourier techniques in the CRYS-
TALSTRUCTURE 3.8.2 program package [36]. Numerical absorption cor-
rection was used. All of the hydrogen atoms were located at
calculated positions and the parameters were refined as ‘‘riding”.
The thermal displacement parameters of non-hydrogen atoms
were refined anisotropically. Full-matrix least-squares methods
were applied using all of the unique reflection data. Selected crys-
tallographic data are listed in Table 2. Preliminary crystallographic
analysis on 6t was also performed.
[8] S. Mukherjee, T. Weyhermüller, E. Bill, P. Chaudhuri, Eur. J. Inorg. Chem. (2001)
4209.
[9] W.A. Alves, G. Cerchiaro, A. Paduan-Filho, D.M. Tomazela, M.N. Eberlin,
A.M.D.C. Ferreira, Inorg. Chim. Acta 358 (2005) 3581.
[10] T. Higa, M. Moriya, Y. Shimazaki, T. Yajima, F. Tani, S. Karasawa, M. Nakano, Y.
Naruta, O. Yamauchi, Inorg. Chim. Acta 360 (2007) 3304.
[11] C.A. Koch, C.A. Reed, G.A. Brewer, N.P. Rath, W.R. Scheidt, G. Gupta, G. Lang, J.
Am. Chem. Soc. 111 (1989) 7645.
[12] (a) E. Colacio, M. Ghazi, R. Klevekas, M. Klinga, F. Lloret, J.M. Moreno, Inorg.
Chem. 39 (2000) 2770;
5.3. Magnetic measurements
Magnetic properties of polycrystalline specimens of 1p–5p, 3t,
6t, and 7t were measured at 5000 Oe in a temperature range down

N. Koyama et al. / Polyhedron 28 (2009) 2001–2009
2009
(b) E. Colacio, J.M. Dominguez-Vera, M. Ghazi, R. Liveskas, M. Klinga, J.M.
Moreno, Inorg. Chem. 37 (1998) 3040.
[13] N. Matsumoto, K. Inoue, M. Ohba, H. Okawa, S. Kida, Bull. Chem. Soc. Jpn. 65
(1992) 2283.
[14] J.M. Dominguez-Vera, N. Galvez, E. Colacio, R. Cuesta, J.-P. Costes, J.-P. Laurent,
J. Chem. Soc., Dalton Trans. (1996) 861.
[15] R. Boca, M. Hvastijova, J. Kozisek, M. Valko, H. Ehrenberg, H. Fuess, W. Haase, I.
Svoboda, R. Werner, Inorg. Chem. Commun. 8 (2005) 548.
[16] Y.F. Song, C. Massera, M. Quesada, I.A. Koval, Patrick Gamez, A.M.M. Lanfredi, J.
Reedijk, Eur. J. Inorg. Chem. (2004) 4566.
[17] S.H. Rahaman, H. Chowdhury, D. Bose, G. Mostafa, H.-K. Fun, B.K. Ghosh, Inorg.
Chem. Commun. 8 (2005) 1041.
[18] N. Matsumoto, T. Nozaki, H. Ushio, K. Motoda, M. Ohba, G. Mago, H. Okawa, J.
Chem. Soc., Dalton Trans. (1993) 2157.
[19] N. Masciocchi, S. Bruni, E. Cariati, F. Cariati, S. Galli, A. Sironi, Inorg. Chem. 40
(2001) 5897.
[20] N. Masciocchi, G.A. Ardizzoia, G. LaMonica, A. Maspero, S. Galli, A. Sironi, Inorg.
Chem. 40 (2001) 6983.
[25] J.T. Chen, S.M. Wang, J. Chin. Chem. Soc. 18 (1971) 213.
[26] (a) J. Peisach, P. Aisen, W.E. Blumberg, The Biochemistry of Copper, Academic
Press, 1966;
(b) J.A.J. Jarvis, A.F. Wells, Acta Crystallogr. 13 (1960) 1027.
[27] J.C. Bonner, M.E. Fisher, Phys. Rev. A 135 (1964) 640.
[28] (a) R. Feyerherm, S. Abens, D. Günther, T. Ishida, M. Meissner, M. Meschke, T.
Nogami, M. Steiner, J. Phys.: Condens. Matter 12 (2000) 8495;
(b) R. Feyerherm, T. Ishida, T. Nogami, M. Seiner, Mol. Cryst. Liq. Cryst. 335
(1999) 947.
[29] (a) J.S. Miller, Extended Linear Chain Compounds, Plenum Press, New York,
1983;
(b) K. Mukai, N. Wada, J.B. Jamali, N. Achiwa, Y. Narumi, K. Kindo, T. Kobayashi,
K. Amaya, Chem. Phys. Lett. 257 (1996) 538.
[30] (a) N. Koyama, T. Ishida, T. Nogami, T. Kogane, Polyhedron 27 (2008)
2341;
(b) M. Yasui, Y. Ishikawa, N. Akiyama, T. Ishida, T. Nogami, F. Iwasaki, Acta
Crystallogr., Sect. B 57 (2001) 288.
[31] K. Kambe, J. Phys. Soc. Jpn 5 (1950) 48.
[21] (a) S.J. Retting, A. Storr, D.A. Suumers, R.C. Thompson, J. Trotter, J. Am. Chem.
Soc. 119 (1997) 8675;
[32] T. Kogane, T. Yamamoto, M. Hayashi, R. Hirota, C.A. Horiuchi, Polyhedron 14
(1995) 2475.
(b) S.J. Retting, V. Sanchez, A. Storr, R.C. Thompson, J. Trotter, J. Chem. Soc.,
Dalton Trans. (2000) 3931.
[33] (a) W.E. Estes, D.P. Gavel, W.E. Hatfield, D. Hodgson, Inorg. Chem. 17 (1978)
1415;
[22] (a) Y.-Q. Tian, C.-X. Cai, X.-M. Ren, C.-Y. Duan, Y. Xu, S. Gao, X.-Z. You, Chem.
Eur. J. 9 (2003) 5673;
(b) J.-P. Zhang, X.-M. Chen, Chem. Commun. (2006) 1689.
[23] T. Kogane, R. Hirota, Bull. Chem. Soc. Jpn. 53 (1980) 91.
[24] R. Watanabe, N. Koyama, T. Nogami, T. Ishida, T. Kogane, Chem. Lett. 38 (2009)
30.
(b) W.E. Hatfield, J. Appl. Phys. 52 (1981) 1985.
[34] B. Bleaney, K.D. Bowers, Proc. R. Soc. London, Ser. A 214 (1952) 451.
[35] T. Ishida, T. Kawakami, S.-i. Mitsubori, T. Nogami, K. Yamaguchi, H. Iwamura, J.
Chem. Soc., Dalton Trans. (2002) 3177.
[36] CRYSTALSTRUCTURE, version 3.8.2, Rigaku/MSC, The Woodlands, TX 77381, 2003–
2007, USA.