Towards the design of linear homo-trinuclear metal complexes based on a new phenol-functionalised diazamesocyclic ligand: Structural analysis and magnetism

Article abstract of DOI:10.1002/ejic.200400618

The formation of two novel phenoxo-bridged linear trimetallic Cu ^II and Ni^II complexes, [Cu₃(μ-L) ₂](CH₃OH)₂(ClO₄)₂ (1) and [Ni₃(μ-L)₂(CH₃OH)₂](ClO ₄)₂ (2), with a new diazamesocyclic ligand functionalised by additional phenol donor pendants (H₂L = N,N′-bis(2- hydroxybenzyl)-1,4-diazacycloheptane), has been achieved and both complexes have been characterised by IR spectroscopy, elemental analyses, conductivity measurements, thermal analyses and UV/Vis techniques. Single-crystal X-ray diffraction analyses revealed that both 1 and 2 have the similar phenoxo-bridged linear trinuclear cores. For 1, two terminal and the central Cu^II ion are in square-planar environments with the adjacent intramolecular Cu...Cu separation of 2.9376(9) A. For 2, however, two terminal Ni^II ions are in square-planar environments and the central Ni ^II ion assumes an octahedral geometry by axial coordination of two methanol ligands, the adjacent intramolecular Ni...Ni distance being 3.007(3) A. In both cases, the 1,4-diazacycloheptane (DACH) ring adopts the normal boat configuration. The magnetic properties of complexes 1 and 2 have been investigated by variable-temperature magnetic susceptibility measurements in the solid state. Complex 1 displays a very strong antiferromagnetic coupling interaction between the neighbouring μ-phenoxo Cu^II centres with a J parameter of -314 cm^-1 and the magneto-structural correlation for such complexes is discussed in detail. For complex 2, the result indicates that both terminal Ni^II ions are diamagnetic and the central Ni ^II shows typical paramagnetic behaviour. The presence of zero-field splitting for Ni^II with a D parameter of 11 cm^-1, being active at low temperature, is further corroborated by the magnetisation measurement at 2 K. Additionally, the interesting ESR spectra of 1 at different temperatures (from 298 K to 8 K) were investigated and interpreted. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2005.

Full text of DOI:10.1002/ejic.200400618

FULL PAPER
Towards the Design of Linear Homo-Trinuclear Metal Complexes Based on a
New Phenol-Functionalised Diazamesocyclic Ligand: Structural Analysis and
Magnetism
Miao Du,*[a] Xiao-Jun Zhao,[a,b] Jian-Hua Guo,[a] Xian-He Bu,*[b] and Joan Ribas*[c]
Keywords: Functionalised diazamesocyclic ligand / Copper(ii) and Nickel(ii) / Crystal structures / Magnetochemistry /
Trinuclear complexes
The formation of two novel phenoxo-bridged linear trimetal-
lic CuII and NiII complexes, [Cu3(µ-L)2](CH3OH)2(ClO4)2 (1)
configuration. The magnetic properties of complexes 1 and
2 have been investigated by variable-temperature magnetic
and [Ni3(µ-L)2(CH3OH)2](ClO4)2 (2), with a new diazameso- susceptibility measurements in the solid state. Complex 1
cyclic ligand functionalised by additional phenol donor pen- displays a very strong antiferromagnetic coupling interaction
dants (H2L = N,NЈ- bis(2-hydroxybenzyl)-1,4-diazacyclohep- between the neighbouring µ-phenoxo CuII centres with a J
tane), has been achieved and both complexes have been
characterised by IR spectroscopy, elemental analyses, con-
ductivity measurements, thermal analyses and UV/Vis tech-
niques. Single-crystal X-ray diffraction analyses revealed
that both 1 and 2 have the similar phenoxo-bridged linear
parameter of −314 cm−1 and the magneto-structural correla-
tion for such complexes is discussed in detail. For complex 2,
the result indicates that both terminal NiII ions are diamag-
netic and the central NiII shows typical paramagnetic behavi-
our. The presence of zero-field splitting for NiII with a D para-
trinuclear cores. For 1, two terminal and the central CuII ion meter of 11 cm−1, being active at low temperature, is further
are in square-planar environments with the adjacent intram-
corroborated by the magnetisation measurement at 2 K. Ad-
ditionally, the interesting ESR spectra of 1 at different tem-
peratures (from 298 K to 8 K) were investigated and inter-
preted.
˚
olecular Cu···Cu separation of 2.9376(9) A. For 2, however,
two terminal NiII ions are in square-planar environments and
the central NiII ion assumes an octahedral geometry by axial
coordination of two methanol ligands, the adjacent intramo-
˚
lecular Ni···Ni distance being 3.007(3) A. In both cases, the
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
1,4-diazacycloheptane (DACH) ring adopts the normal boat
Germany, 2005)
Introduction
simple and/or intricate bridging ligands between the metal
centres which usually contain simple anions such as chlo-
The development of routes and strategies for the design ride,[3] hydroxide,[4] acetate,[5] methoxide[6] or azide[7] and
and preparation of polynuclear complexes of 3d metals in only a minority of them have the M3 core bridged by just
moderate oxidation states is of great importance in bioinor- a single organic ligand. Thus, the rational design and prep-
ganic chemistry, magnetochemistry, materials chemistry aration of such trinuclear complexes still remains a difficult
and solid-state chemistry.[1] Trinuclear compounds can be challenge and attracts continuing interest.
classified as linear or nonlinear according to the arrange-
As described in many other publications by us, we have
ments of the metal centres and a number of linear trinuclear
been exploring the coordination chemistry of diazameso-
compounds have been structurally and magnetically charac-
cyclic systems, especially the typical 1,5-diazacyclooctane
terised to date.[2] Of further importance is the fact that most
(DACO) which exhibits an exceptionally strong ligand field,
of the known trinuclear complexes have two, three or four
a unique configuration and the potential for further
functionalisation.[8Ϫ11] It has been well demonstrated by us
[a]
and others[12] that DACO modified by suitable functional
College of Chemistry and Life Science,
Tianjin Normal University,
pendants, such as heterocyclic, phenolic, thiol and car-
boxylic groups could be used as good building blocks for
constructing polymeric systems with unique structures and
properties. Based on our experimental findings, a general
rule for the coordination chemistry of functionalised
DACO compounds has been established according to the
types and numbers of the additional pendant arms. One of
Tianjin 300074, P. R. China
Fax: (internat.) ϩ86-22-23540315
E-mail: dumiao@public.tpt.tj.cn
Department of Chemistry, Nankai University,
Tianjin 300071, P. R. China
[b]
[c]
E-mail: buxh@nankai.edu.cn
´
`
Departament de Quımica Inorganica, Universitat de Barcelona,
Diagonal, 647, 08028 Barcelona, Spain
E-mail: joan.ribas@qi.ub.es
294
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200400618
Eur. J. Inorg. Chem. 2005, 294Ϫ304

Towards the Design of Linear Homo-Trinuclear Metal Complexes
FULL PAPER
Scheme 1
the main reasons that justifies the continuous interest in
In the IR spectra of H2L and both complexes, the ab-
such attractive system is the construction of other novel me- sorption bands resulting from the skeletal vibrations of the
tal complexes with special frameworks and properties by aromatic rings appear in the 1400Ϫ1600 cmϪ1 region. The
modifying the backbone of the diazamesocycle, for example broad band centred at ca. 3400 cmϪ1 results from the OϪH
choosing 1,4-diazacycloheptane (DACH) as the initial ma- stretching of the phenol groups (for H2L) or the solvent
terial. Our preliminary results indicate that the metal com- methanol molecules (for 1 and 2). The very strong bands at
plexes of the heterocycle-functionalised DACH derivatives 1256 cmϪ1 (for H2L), 1264 and 1247 cmϪ1 (for 1) and 1261
usually exhibit similar structures to those of the corre- cmϪ1 (for 2) are the characteristic absorptions of the CϪO
sponding DACO compounds but with a change in the coor- groups of the phenol pendant arms. Additionally, in the IR
dination environments of the metal centres due to their dif- spectra of both complexes, the strong bands that appear at
ferent configurations (normally, boat-chair for DACO ring, ca. 1100 and 624 cmϪ1 suggest the presence of the perchlor-
and boat for DACH).[13] It should be noted that almost all ate anions.[15]
metal centres chelated to DACO derivatives assume penta-
The UV/Vis spectra of complexes 1 and 2 (CH3CN and
coordinate geometries[8Ϫ12] but this is not the case for
H2O, respectively) have been investigated at room tempera-
DACH.[13] For metal complexes with the DACO ligand
ture. It is important to know the structures of the com-
bearing two phenol arms, phenoxo-bridged linear trimetal-
lic CuII and NiII frameworks were obtained. The former
exhibits a square pyramid/square plane/square pyramid Cu3
core[10a] and the latter is the first square pyramid/octa-
hedral/square pyramid Ni3 complex.[10b] Thus, as a continu-
ation of our work, we report herein the design and prep-
aration of a new phenol-functionalised DACH ligand N,NЈ-
plexes in solution prior to spectroscopic discussions. Thus,
the conductivities of both complexes were measured and
the molar conductivities of 1 and 2 indicate both complexes
behave as the typical 2:1 electrolytes in the corresponding
solution, this being consistent with their solid structures.
Evidently the phenoxo bridges are retained in solution. The
visible spectrum of 1 shows a moderate shoulder band at
ca. 558 nm and a more intense absorption band at 386 nm
which can be assigned to the dϪd component of CuII ions
with a square-planar geometry and the charge-transfer
(CT) transition band from the filled pπ orbital of the phe-
nolic oxygen to the vacant d orbitals of CuII, respectively.[16]
For complex 2, two such types of absorptions appear at
498 nm and 362/324 nm (one is due to the terminal NiII CT
and the other to the central NiII CT with different coordi-
nation geometries[2c]), respectively. In addition, the intense
bands observed in the 200Ϫ300 nm region in the spectra of
both complexes can be attributed to the πǞπ* ligand tran-
sitions.
bis(2-hydroxybenzyl)-1,4-diazacycloheptane
(H2L,
see
Scheme 1) and its CuII and NiII complexes [Cu3(µ-
L)2](CH3OH)2(ClO4)2 (1) and [Ni3(µ-L)2(CH3OH)2](ClO4)2
(2). The magnetic properties of both complexes have been
investigated and the magneto-structural correlation be-
tween 1 and related compounds has been discussed. Of
further interest is the fact that relatively few trinuclear CuII
compounds have been studied by ESR techniques[14] and,
thus, the ESR properties of 1 at different temperatures
(from 298 K to 8 K) were analysed in detail in this study.
Results and Discussion
Both 1 and 2 are air stable at room temperature but de-
composition occurs at elevated temperature, however. Ther-
mogravimetric analyses (from room temperature to 600 °C)
indicate that for 1, the first weight loss of 6.44% from 118
Preparation, Characterisation and Spectra of Compounds 1
and 2
The new ligand H2L was synthesised using an excess of to 164 °C (peak: 132 °C) corresponds to the loss of two
2-bromomethyl phenylacetate in order to largely obtain the CH3OH solvent molecules (calculated: 5.96%). The remain-
doubly substituted product as described in detail in our pre- ing substance does not lose weight upon further heating
vious report.[10a] The yield was moderate. All analytical and until a rapid weight loss in the 244Ϫ264 °C region (peak:
spectroscopic data are in good agreement with the theoreti- 258 °C). Further heating to 600 °C induces a continuous
cal requirements. The preparations of 1 and 2 were achieved and slow weight loss. The thermal behaviour of 2 is similar
directly by treatment of the ligand with the corresponding to 1, the first weight loss of 5.84% from 85 to 160 °C (peak:
metal perchlorate and the ligand becomes doubly depro- 138 °C) corresponds to the release of two coordinated
tonated when forming the metal complex.
methanol ligands (calculated: 6.04%). The remaining coor-
Eur. J. Inorg. Chem. 2005, 294Ϫ304
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
295

M. Du, X.-J. Zhao, J.-H. Guo, X.-H. Bu, J. Ribas
FULL PAPER
Figure 1. ORTEP view of the molecular structure of 1 including atomic labelling of the asymmetric unit (hydrogen atoms, perchlorate
anions and methanol solvents are omitted for clarity) with displacement ellipsoids drawn at the 30% probability level
˚
dinated framework does not lose weight upon further heat- 1.956(3) and 1.953(3) A) which further link the other two
ing until two consecutive weight losses occur in the CuII ions to form the linear Cu3 array, resulting in a CuO4
270Ϫ325 °C region (peak: 288 and 312 °C). As for 1, plane. Additionally, weak interactions between the central
further heating to 600 °C only shows a continuous and slow Cu(2) and the perchlorate anions (located in the axial posi-
weight loss.
tion of the CuII coordination sphere) can be observed with
a Cu(2)ϪO(3) distance of 2.589(3) A which is significantly
˚
Single-Crystal X-ray Structures of Complexes 1 and 2
longer than the normal CuϪO coordination bond. The ter-
˚
minal Cu(1) ion is only ca. 0.035 A above the basal least-
Structure Description of [Cu3(µ-L)2](CH3OH)2(ClO4)2
(1)
squares plane and the central Cu(2) is located exactly on its
coordination plane due to the site symmetry. The folding
angle between the perfect square plane O(1)ϪO(1A)Ϫ
O(2)ϪO(2A) and the best calculated plane N(1)Ϫ
N(2)ϪO(1)ϪO(2) is 10.8° and the dihedral angle between
the two Cu(1)ϪO(1)ϪCu(2) and Cu(1)ϪO(2)ϪCu(2)
planes is 10.9°. The intramolecular Cu···Cu separation is
An ORTEP diagram of complex 1 is shown in Figure 1
and selected bond lengths and angles are listed in Table 1.
The structure consists of a discrete [Cu3(µ-L)2]2ϩ cation in
which the ligands are doubly deprotonated, two perchlorate
counter anions and two lattice methanol solvent molecules.
The asymmetric unit is formed by half a molecule which
exhibits inversion in the centre of the rhombohedral which
itself consists of four phenol oxygen atoms [Cu(2) is just on
this inversion centre]. Two terminal Cu(1) and Cu(1A)
atoms, located in general positions, are centrosymmetrically
related to each other and, hence, three CuII ions linked by
the phenoxo bridges lie in a straight line. Each terminal
CuII ion is coordinated to two nitrogen atoms from the
˚
2.9376(9) A whereas the Cu(1)ϪOϪCu(2) bridging angles
are 98.31(12) and 98.38(11)° for O(1) and O(2), respectively.
For the ligand, the DACH ring takes the normal boat con-
figuration and the dihedral angle between two phenol
planes is 117.6°. A pair of nitrogen atoms from DACH and
two oxygen atoms of phenolate pendant arms are in cis po-
sitions in the coordination plane of the chelated Cu(1)
atom.
˚
DACH ring (CuϪN distances: 1.973(3) and 1.981(3) A) and
Analysis of the crystal packing of complex 1 indicates
two phenolate oxygen donors (CuϪO lengths: 1.927(3) and that the trimers are well isolated in the unit cell, with the
1.928(3) A), adopting an approximately square-planar ge- intermolecular contacts between the CuII centres above 5.8
˚
˚
ometry (CuN2O2). The central Cu(2) atom is coordinated A. There exists only an intramolecular O(7)ϪH(7)···O(3)
to four bridging phenolate oxygens (CuϪO lengths: hydrogen bond between the methanol molecule and the per-
296
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 294Ϫ304

Towards the Design of Linear Homo-Trinuclear Metal Complexes
FULL PAPER
˚
dination plane of the terminal Ni(1) atom with a derivation
from planarity of 0.022 A. However, the central Ni(2) is six-
Table 1. Selective bond lengths (A) and angles (°) for complex 1
˚
Bond lengths
coordinate: four phenol oxygen atoms are in the equatorial
plane with NiϪO distances of 2.040(3) and 2.050(3) A with
˚
Cu(1)؊O(1)
Cu(1)؊N(1)
Cu(2)؊O(1)
C(12)؊O(1)
Cu(1)؊Cu(2)
1.927(3) Cu(1)؊O(2)
1.973(3) Cu(1)؊N(2)
1.956(3) Cu(2)؊O(2)
1.364(5) C(19)؊O(2)
2.9376(9)
1.928(3)
1.981(3)
1.953(3)
1.379(4)
two methanol ligands occupying the axial positions with
˚
NiϪO lengths of 2.067(3) A. Thus, the coordination
environment of Ni(2) can be described as a slightly dis-
torted elongated octahedron with the cis OϪNiϪO angles
being in the range of 75.33(11)Ϫ104.67(11)°. In 2, the
folding angle between the perfect square plane
O(1)ϪO(1A)ϪO(2)ϪO(2A) and the best calculated plane
N(1)ϪN(2)ϪO(1)ϪO(2) is 6.3° and the dihedral angle be-
Bond angles
O(1)ϪCu(1)ϪO(2)
O(2)ϪCu(1)ϪN(1) 178.79(13) O(1)ϪCu(1)ϪN(2) 176.92(13)
O(2)ϪCu(1)ϪN(2)
O(2)ϪCu(2)ϪO(1)
Cu(1)ϪO(1)ϪCu(2) 98.31(12) Cu(1)ϪO(2)ϪCu(2) 98.38(11)
81.89(11) O(1)ϪCu(1)ϪN(1)
98.73(13)
97.83(13) N(1)ϪCu(1)ϪN(2)
81.49(14) tween the two Ni(1)ϪO(1)ϪNi(2) and Ni(1)ϪO(2)ϪNi(2)
80.52(11) O(2)ϪCu(2)ϪO(1)i
99.48(11)
planes is 8.7°. The intramolecular Ni···Ni separation is
˚
3.007(2) A and the Ni(1)ϪOϪNi(2) bridging angles are
100.39(13) and 98.98(14)° for O(1) and O(2), respectively.
For the doubly deprotonated ligand, the DACH ring as-
sumes the normal boat configuration and the dihedral angle
between two phenol planes is 56°. Similar to 1, the trinu-
clear subunits are well isolated in the unit cell with the inter-
Symmetry code: (i) 1 Ϫ x, 1 Ϫ y, 1 Ϫ z
˚
chlorate anion. The O···O separation is 2.995 A with an
˚
H···O distance of 2.242 A and the bond angle is 152.7°
molecular contacts between the NiII centres above 6.0 A.
˚
which is in the normal range for such a weak interaction.[17]
An intramolecular O(3)ϪH(3C)···O(4) hydrogen-bonding
interaction between the coordinated methanol molecule
and the perchlorate anion can be observed and this further
Structure Description of [Ni3(µ-L)2(CH3OH)2](ClO4)2
(2)
˚
stabilises the structure. The O···O separation is 2.861 A with
˚
A single-crystal X-ray determination indicates that the
structure of complex 2 consists of a [Ni3(µ-L)2(CH3OH)2]2ϩ
cation exhibiting a similar M3 framework to 1 and two per-
chlorate anions. Selected bond lengths and angles for 2 are
listed in Table 2. As depicted in Figure 2, the asymmetric
unit is formed by half of the complex which shows an inver-
sion centre [Ni(2)] in the middle of the rhombohedral plane.
Since Ni(2) occupies the inversion centre, the resultant
Ni(1)ϪNi(2)ϪNi(1A) array is linear. As in 1, a pair of ni-
trogen atoms from DACH (NiϪN distances: 1.903(4) and
an H···O distance of 2.074 A and the bond angle is 160.8°,
which is in the normal range for such weak interactions.[17]
ESR Spectra of Complex 1
For the interpretation of the ESR spectra of complex 1,
it is convenient to consider the relative energies of the states
deduced by coupling the three S ϭ 1/2 ions. A demon-
stration of these states and the corresponding energies can
be found in the book of O. Kahn.[18] The spin diagram is
given in Figure 3 in which the occurrence of ZFS in the
state |3/2, 1/2Ͼ is considered, giving two doublets corre-
sponding to ms ϭ Ϯ 3/2 and Ϯ 1/2. As shown, ms ϭ Ϯ 1/2
is strongly anisotropic.[19]
˚
1.897(4) A) and two oxygen atoms of the phenolate pendant
˚
arms (NiϪO lengths: 1.872(3) and 1.873(3) A) are in a cis
position. These positions constitute the approximate coor-
˚
Table 2. Selective bond lengths (A) and angles (°) for complex 2
The ESR spectra of 1, as a function of temperature, are
given in Figure 4. The easiest spectrum to interpret is that
corresponding to the lowest temperature measured (8 K).
The spectrum shows a very narrow almost isotropic band
centred at g ϭ 2.05. When the temperature is raised this
signal broadens and at close to liquid nitrogen temperature
it begins to appear as broad band near 1500 Gauss. The
intensity of the spectrum diminishes as the temperature in-
creases and at 298 K (final temperature we measured) the
spectrum shows two clear and very broad signals, the first
one centred at 3000 G (g ϭ 2.08) and a new band centred
at 1200Ϫ1300 G which corresponds to a g value of approxi-
Bond lengths
Ni(1)؊O(1)
Ni(1)؊N(1)
Ni(2)؊O(1)
Ni(2)؊O(3)
C(12)؊O(1)
1.872(3) Ni(1)؊O(2)
1.903(4) Ni(1)؊N(2)
2.040(3) Ni(2)؊O(2)
2.067(3) Ni(1)؊Ni(2)
1.357(5) C(19)؊O(2)
1.873(3)
1.897(4)
2.050(3)
3.007(2)
1.361(5)
Bond angles
O(1)ϪNi(1)ϪO(2)
O(2)ϪNi(1)ϪN(1) 178.27(15) O(1)ϪNi(1)ϪN(2) 178.93(15)
O(2)ϪNi(1)ϪN(2)
O(2)ϪNi(2)ϪO(1)
O(3)ϪNi(2)ϪO(1)
O(3)ϪNi(2)ϪO(2)
83.75(12) O(1)ϪNi(1)ϪN(1)
96.61(14)
96.48(14) N(1)ϪNi(1)ϪN(2)
83.12(16)
75.33(11) O(2)ϪNi(2)ϪO(1)i 104.67(11) mately 5.15. It must be pointed out that the difference in
87.43(11) O(3)ϪNi(2)ϪO(1)i
88.29(12) O(3)ϪNi(2)ϪO(2)i
92.57(11)
91.71(13)
98.98(14)
the intensities of the limiting spectra (room temperature
and 8 K) is in the region of five orders of magnitudes (105).
At low temperature, only the |1/2, 1Ͼ ground state is
populated (J ϭ Ϫ314 cmϪ1 according to the susceptibility
measurements as described below). Thus the strong signal
Ni(1)ϪO(1)ϪNi(2) 100.39(13) Ni(1)ϪO(2)ϪNi(2)
Symmetry code: (i) 1 Ϫ x, Ϫy, 1 Ϫ z
Eur. J. Inorg. Chem. 2005, 294Ϫ304
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
297

M. Du, X.-J. Zhao, J.-H. Guo, X.-H. Bu, J. Ribas
FULL PAPER
Figure 2. ORTEP structure of the molecular structure of 2 including atomic labelling of the asymmetric unit (hydrogen atoms and
perchlorate anions are omitted for clarity) with displacement ellipsoids drawn at the 30% probability level
Figure 3. Spin levels, energies and g values in a trinuclear Cu3 complex (J Ͻ 0)
at g ϭ 2.05 is due to this ground state. This g value is an solid state, the trinuclear cations are not perfectly isolated
average of the g values of the three CuII ions (g ϭ 1/3 [2gT1 from a magnetic point of view. Effectively, we have found a
Ϫ gC ϩ2gT2] where C and T are the central and terminal nonnegligible JЈ value (molecular-field approach, see be-
Cu ions).[19,20] The g values for the |1/2, 0Ͼ and |3/2, 1Ͼ low). As has been pointed out,[22] even an extremely weak
states are different, provided that the individual gi values value for the intermolecular exchange interaction leads to
are different.[20] If we assume that the rate of transition the exchange averaging effect which is responsible for the
from the lowest to the excited multiplets is fast on the ESR isotropic or quasi-isotropic spectra observed (such as in
time scale, the observed spectrum at a given temperature complex 1).
would be the thermal average of the spectra of three differ-
Only on a few occasions for this kind of trinuclear CuII
ent multiplets. Therefore, the liquid helium temperature complex, has any evidence of signals attributable to a spin
spectra should be those of the lowest doublet, while those quadruplet been found. In contrast, this spin quadruplet is
at higher temperatures should also bear the contributions clearly visible at 4 K when the coupling is ferromagnetic
of the excited multiplets. For this reason, the anisotropy of because this is the ground state in this case.[5,23] A broad
the molecular g value is usually not well defined. Only in a transition between g ϭ 4.25Ϫ5.5 occurs and can be as-
few trinuclear CuII complexes is this anisotropy mani- signed to the transition between the Ϯ 1/2 levels of the S ϭ
fested.[21] In complex 1, the observed g value is an average 3/2 system. Thus, this is undoubtedly the origin of the
of two different CuII environments. On the other hand, in broad signal centred at g ഠ 5 in complex 1.
298
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 294Ϫ304

Towards the Design of Linear Homo-Trinuclear Metal Complexes
FULL PAPER
being active only at very low temperature. Thus, in the light
of the above structural discussion of complex 1, we can in-
terpret its magnetic behaviour through a simple model with
only one intramolecular J parameter using the following
spin Hamiltonian:
H ϭ ϪJ1 (S1S2 ϩ S2S3)
From this Hamiltonian it is possible to deduce a simple
equation for χMT vs. T:[18]
This equation has been completed by introducing a JЈ
parameter and considering the possible small interactions
between the trinuclear entities using the molecular-field ap-
proach:[18]
χM ϭ Ng2β2F(J,T)/(kT Ϫ zJЈF(J,T))
The fit of the magnetic data, assuming that the g factors
are identical for Cu(1) and Cu(2), leads to J ϭ Ϫ314.0 Ϯ
0.8 cm؊1, JЈ ϭ Ϫ0.5 Ϯ 0.01 cm؊1, g ϭ 2.09 and R ϭ 1.5
ϫ 10؊6 (R is the agreement factor defined as Σi[(χMT)obs
(χMT)calc]2 / Σ[(χMT)obs]2).
Ϫ
Figure 4. X-band powder ESR spectra at different temperatures for
complex 1; the ratio of the intensities at 8 K and 298 K is 105
Magneto-Structural Correlations in Cu2O2 Cores
Magneto-structural correlations in dinuclear CuII com-
plexes bridged equatorially by pairs of hydroxo[24] or al-
koxo[25] groups show that the major factor controlling the
spin coupling between the S ϭ 1/2 metal centres is the
CuϪO(R)ϪCu angle. Hatfield and Hodgson found a linear
correlation between the experimentally determined ex-
change coupling constant and the CuϪOϪCu bond angle
(ϕ) (see a in Figure 6).[24] Antiferromagnetic behaviour is
observed for complexes with ϕ larger than 97.6°, while
ferromagnetism appears for those with smaller values of ϕ.
An apparently similar linear relationship for the alkoxo
cases[25] shows that at angles around 95.6°, the exchange
integral approaches zero, i.e. the point of experimental ‘‘ac-
cidental orthogonality’’. For such dimeric CuII complexes,
Nieminen[26] described correlations between the singlet-trip-
let splitting energy and the CuϪOϪCu angle ϕ, the devi-
ation dc of the carbon atom connected to the bridging oxy-
gen atoms from the CuϪOϪCu plane and the sum of the
angles around the bridging oxygen atoms. Walz et al.[27] re-
ported the influence of the coordination geometry around
the CuII atom on the exchange interaction in alkoxo-
Figure 5. Plot of the temperature dependence of χMT for 1; the
solid line corresponds to the best fit (see text for parameters calcu-
lated)
Magnetic Properties of 1 and 2
Complex 1
The magnetic properties of complex 1 in the form of a bridged complexes. These experimental observations are
χMT vs. T plot are shown in Figure 5 (magnetic suscepti- consistent with studies using the angular overlap approach
bility per three CuII ions). The χMT value is 0.65 cm3 molϪ1 by Bencini and Gatteschi[28] as well as extended Hückel MO
K at 300 K which is far from the value of 1.2 cm3 molϪ1 K treatments by Hoffmann[29] and Kahn.[30] More recent ab
corresponding to three isolated spin doublets. The χMT initio and DFT calculations on simple model dimers exam-
value decreases to 0.40 cm3 molϪ1 K upon cooling to ca. ined the effects of the bridging angles, CuϪO distances, fold
80 K. At this point there is a clear plateau from 80 K to (hinge) angle along the OϪO axis and degree of tetrahedral
10 K and, finally, there is a slight decrease to 0.38 cm3 (for square-planar geometry) or sp Ǟ bpt (for square-pyr-
molϪ1 K at 2 K. This curve suggests strong antiferromag- amidal geometry) distortions at the CuII centres.[25,31Ϫ33]
netic coupling between the CuII centres with the presence These theoretical studies showed that although the major
of very small intermolecular antiferromagnetic interactions factor affecting the magnetic exchange is the CuϪOϪCu
Eur. J. Inorg. Chem. 2005, 294Ϫ304
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
299

M. Du, X.-J. Zhao, J.-H. Guo, X.-H. Bu, J. Ribas
FULL PAPER
effect of replacing the hydrogen of the hydroxo with a more
electronegative carbon group which reduces the electron
density on the oxygen bridge.[36]
Figure 7. Schematic plot of the 2J values vs. CuϪOϪCu angles
given by Thompson et al. in ref.[42] together with the point indicat-
ing the experimental value for 1; in this figure the symbolism of 2J
has been kept, although in the text we have used the Hamiltonian
ϪJ instead of Ϫ2J
Similar correlations can be established for comparable
macrocyclic dicopper(ii) systems, such as ‘‘Robson’’ type
macrocyclic complexes, formed by CuII ions acting as tem-
plates for the condensation of diformylphenols and various
Figure 6. The three main factors in the Cu2O2 entities that tune
the antiferromagnetic coupling in µ-phenoxo derivatives
bridge, other features can also significantly lead to the in- diamines. These systems contain phenoxo oxygen atoms
crease in the ferromagnetic contributions which effectively bridging the two CuII centres and represent ideal systems
reduces any antiferromagnetic term associated with the for examining the exchange coupling via the phenoxo oxy-
hydroxo/alkoxo bridges. Among these features, we can gen. The degree of distortional flexibility available to simple
underline the distortions of the coordination geometry of dimeric CuII complexes with hydroxo or alkoxo bridges can
the CuII centre,[34,35] the variation in CuϪO bond be expected to be reduced when the CuϪ(O)2ϪCu entity is
lengths,[32] the hinge (roof shape) distortion of the core (δ, enclosed within the cavity of a macrocyclic ligand and, thus,
see b in Figure 6),[31Ϫ33,36] the torsion angle created by the one might expect a more straightforward relationship be-
H (or R) group with regard to the Cu2O2 plane (τ angle, tween the bridging angle and the exchange with respect to
see
tors.[31,32,37,38] Some of these geometric factors are shown this sort of complex have been reported in which the linker
in Figure 6. group between the azomethanide nitrogens is two-mem-
c
in Figure 6)[32] and different electronic fac- electronic perturbations. Only a few structural examples of
Roughly the same features discussed for the hydroxo- bered, thereby creating a five-membered chelate ring.[39]
bridged complexes also seem apparent in alkoxo-bridged The CuϪOphenoxoϪCu angles fall in the range 98.8Ϫ100.9°,
complexes. Thompson et al. found significant differences in this being at the lower end of the range of angles typical
magnetic coupling between µ-hydroxo (or µ-alkoxo) and µ- for phenoxo systems. In all cases, the molecules themselves
phenoxo complexes.[39] From this work, it is apparent that and the dinuclear centres are almost flat. There are no un-
the slopes of the graphs (2J vs. CuϪOϪCu angle) of the usual structural or electronic perturbations which might be
hydroxo and alkoxo cases are comparable but the absolute expected to influence the exchange in a significant way in
values of Ϫ2J are larger for the alkoxo species. However, those complexes. Thus, the phenoxo bridging angle can be
the slope of the graph for the phenoxo system is smaller considered to be the principal factor controlling the spin
and the absolute values of Ϫ2J are inherently larger. For coupling.
lower angles (95°) the J values approach zero for the alkoxo
In summary, the main conclusion for phenoxo systems is
species but this is not the case for phenoxo bridging com- that strong antiferromagnetic exchange will dominate in
plexes. The range of the average angles for these phenoxo these complexes. For comparison with complex 1, structural
complexes is very small (from 99 to 105°). The couplings and magnetic parameters for some related [Cu2(µ-OR)]
lie between Ϫ700 cm؊1 (for smaller angles) and Ϫ900 cm؊1 complexes are listed in Table 3. In the present case, complex
(for larger angles) as shown in Figure 7.[39] For J ϭ 0 the 1 belongs to the µ-phenoxo-bridged group of complexes.
CuϪOϪCu angle should be 77°, although this angle has The CuϪOϪCu angles are 98.31(12) and 98.38(11)°,
not been experimentally observed at present. The obser- respectively, and the intramolecular Cu···Cu separation is
˚
vation of greater J values for ϪOR than ϪOH is due to the 2.9376(9) A. The hinge angle (δ) through the O(1)ϪO(2)
300
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 294Ϫ304

Towards the Design of Linear Homo-Trinuclear Metal Complexes
Table 3. Structural and magnetic parameters for some relevant [Cu2(µ-OR)2] complexes (basal-basal coordination)
FULL PAPER
Compound
ϕ[a]
τ[a]
δ[a]
J
Ref.
[40]
[41]
[42]
[Cu(L)(CH3COO)2](O-acetato)
[Cu(AE)(CH3COO)]2(O-acetato)
[Cu(bpy)2{µ-Ph(OCH2)Npy}2]-
(ClO4)2 (alkoxo)
[Cu(phen)2{µ-Ph(OCH2)Npy}2]-
(ClO4)2 (alkoxo)
95.7/102.6
95.34
97.8
Ϫ
Ϫ
62
Ϫ
180
180
ഠ ϩ1
ഠ Ϫ1
ϩ10 (2J)
[42]
98.3
48.7
180
Ϫ98 (2J)
[43]
[43]
[48]
[44]
[44]
[45]
[46]
[34]
[39]
[39]
[Cu2(b2b)2(OR)2](ClO4)2 (alkoxo)
[Cu2(b3b)2(ORЈ)2](ClO4)2 (alkoxo)
[Cu(papen)]2·2H2O (alkoxo)
130.21
102.9
98.3
102.8/103.1
103.81/101.8
100.1
101.1
103.6
103.65
105 (av)
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
180
180
164.1
141.1
Ϫ
diamag.
diamag.
Ϫ128 (2J)
Ϫ714
[Cu2(HL)2](NO3)2 (phenoxo no-macro)
[Cu2(L)(HL)](ClO4) (phenoxo no-macro)
[Cu2(fsa2en)] (phenoxo no-macro)
[Cu(etsal)NO3]2 (phenoxo no-macro)
[Cu(ips)Cl]2 (phenoxo no-macro)
[Cu2(L1)(H2O)2]F2 (phenoxo macro)
[Cu2(L2)Cl2][Cu2(L2)(H2O)2]Cl·ClO4
(phenoxo macro)
Ϫ277
Ϫ
Ϫ650 (2J)
Ϫ166 (2J)
Ϫ145
Ϫ790
Ϫ800
180
180
180
Ϫ
[39]
[47]
[Cu2(L3)(H2O)2]BF4 (phenoxo macro)
[Cu2L1(ClO4)2] (phenoxo macro)
98.8
102.8
Ϫ
Ϫ
180
180
Ϫ690
Ϫ740
[a]
The meaning of Greek symbols is given in Figure 6.
axis is 170°, i.e. very close to 180° which corresponds to an susceptibility per NiII ion (the terminal ones are
exactly planar structure. The torsion angles (τ) created by diamagnetic). The χMT value is 1.32 cm3 molϪ1 K at 300 K
C(19) and C(12) (phenoxo ligands) from the Cu2O2 plane which is a typical value for an isolated NiII ion with g Ͼ
are 140/147.2 and 138.9/145.6°, respectively. Thus, we can 2.00. χMT is practically constant to 25 K and then decreases
expect strong antiferromagnetic coupling as experimentally to 0.3 cm3 molϪ1 K at 2 K. This feature suggests typical
observed. Plotting the 2J values indicated by Thompson et paramagnetic behaviour for a NiII ion in which the D para-
al. in reference 42 and adding the CuϪOϪCu angles and meter and/or intermolecular interactions (usually antiferro-
the 2J parameter for complex 1, a rather good agreement magnetic) are active at low temperature. In an attempt to
can be observed (Figure 7). The difference could be due to calculate the D value, we have fitted the experimental χMT
the hinge angle (very small in this case) and, mainly, to the values to the formula given by Kahn for a mononuclear
τ distortion (considerable in complex 1 as stated above).
NiII ion taking into consideration the ZFS of the S ϭ 1
ground state.[18] The best fit values are |D| ϭ 11 cmϪ1 and
g (average) ϭ 2.30. The D value is higher than that typical
for NiII ions (close to 5Ϫ8 cmϪ1), indicating that the pos-
sible intermolecular interactions are superimposed on this
D value. The magnetisation measurement at 2 K corrobor-
ates the effect of the D parameter, i.e. the reduced magneti-
sation (M/Nβ) at 5 T is only consistent with 1.4 electrons
instead of two electrons which should be the theoretical
value for an isolated NiII ion with g ϭ 2.00 (see inset in Fig-
ure 8).
Complex 2
The magnetic properties of 2 in the form of χMT vs. T
plots are shown in Figure 8, these representing the magnetic
Conclusions and Perspectives
Two novel linear phenoxo-bridged trinuclear CuII and
NiII complexes were rationally designed and prepared based
on a new diazamesocyclic ligand bearing two pendant phe-
nol donors. From a magnetic point of view, the data for
complex 1 show two features: the very strong antiferromag-
netic coupling interaction between the neighbouring µ-
phenoxo CuII centres and the small possible interactions
between the trinuclear entities. The magneto-structural cor-
relations for related complexes containing Cu2O2 cores have
been analysed in detail. For 2, both terminal NiII ions are
diamagnetic and the central NiII shows typical paramag-
netic behaviour. The zero-field splitting for NiII can be ob-
Figure 8. Plot of χMT vs. T for complex 2; the solid line in the χMT
curve represents the best fit for calculating the D value (see text);
inset: plot of the reduced magnetisation (M/Nβ) vs. H at 2 K for
complex 2
Eur. J. Inorg. Chem. 2005, 294Ϫ304
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
301

M. Du, X.-J. Zhao, J.-H. Guo, X.-H. Bu, J. Ribas
FULL PAPER
[Ni3(µ-L)2(CH3OH)2](ClO4)2 (2): The same synthetic procedure
(for both polycrystalline and single-crystal preparation) for 1 was
used except that Cu(ClO4)2·6H2O was replaced by Ni(ClO4)2·6H2O.
This afforded a red powder in 85% yield (or block-type single crys-
tals). C40H52Cl2Ni3N4O14 (1059.83): calcd. C 45.33, H 4.95, N 5.29;
found C 45.61, H 4.58, N 5.04%. IR (cmϪ1): 3357 b, 3059 w, 3020
w, 2939 w, 2911 w, 2860 m, 1599 s, 1577 m, 1485 vs, 1457 vs, 1398
m, 1342 w, 1325 w, 1261 vs, 1208 m, 1112 vs, 1045 vs, 1027 vs, 972
m, 943 w, 927 m, 883 s, 867 m, 794 s, 767 vs, 734 m, 657 m, 624 s,
605 w, 590 m. ΛM (H2O): 182 cm2 ΩϪ1 molϪ1. λmax/nm (ε/dm3
molϪ1 cmϪ1) (H2O): 498 (340), 362 (1150) and 324 (1020).
served and further corroborated by the magnetisation meas-
urements. Additionally, the interesting ESR properties of 1
at different temperatures (from 298 K to 8 K) were investi-
gated. Further studies of the coordination chemistry of
such attractive systems are under way in our laboratory.
Experimental Section
Materials and Physical Measurements: With the exception of 2-
bromomethyl phenylacetate, which was synthesised according to
the reported literature procedure,[49] all of the starting materials
and solvents for the syntheses and the analyses were obtained com-
mercially and used as received. 1H NMR spectra were recorded on
a Bruker AC-P 400 spectrometer (400 MHz) at 25 °C with tetra-
methylsilane as the internal reference. FT-IR spectra (KBr pellets)
were recorded on an AVATAR-370 (Nicolet) spectrometer and elec-
tronic spectra were acquired with a CARY300 (Varian) spectropho-
tometer. C, H and N analyses were performed on a CE-440 (Leem-
anlabs) analyser. Conductivities of the complexes were measured
at room temperature using a DDS 11A conductometer. Thermal
stability (TGA) experiments were carried out on a Dupont thermal
analyser from room temperature to 600 °C under a nitrogen atmos-
phere at a heating rate of 10 °C/min.
Caution! Although no problems were encountered in this study,
transition metal perchlorate complexes are potentially explosive
and should be handled with proper precautions.
Magnetic Studies: The variable-temperature magnetic susceptibilit-
´
ies were measured at the ‘‘Servei de Magnetoquımica (Universitat
de Barcelona)’’ on polycrystalline samples (30 mg) with a Quantum
Design MPMS SQUID magnetometer operating at a magnetic field
of 0.1 T between 2 and 300 K. The diamagnetic corrections were
evaluated from Pascal’s constants for all the constituent atoms.
Magnetisation measurements were carried out at 2 K in the 0Ϫ5 T
range. ESR spectra were recorded on powder samples at the X-
band frequency with a Bruker 300E automatic spectrometer, vary-
ing the temperature between 8 and 298 K.
X-ray Crystallographic Data Collections and Refinements: Single-
crystal X-ray diffraction measurements of 1 (brown, 0.30 ϫ 0.25 ϫ
0.20 mm) and 2 (red, 0.30 ϫ 0.25 ϫ 0.20 mm) were carried out
with a Bruker Smart 1000 CCD diffractometer equipped with a
graphite crystal monochromator situated in the incident beam for
data collection at 293(2) K. The lattice parameters were obtained
by least-squares refinements of the diffraction data of 781 (for 1)
and 582 (for 2) reflections, respectively, and data collections were
Synthesis of the Ligand N,NЈ-Bis(2-hydroxybenzyl)-1,4-diazacy-
cloheptane (H2L): 2-Bromomethyl phenylacetate (8.12 g,
35.5 mmol) was added to a solution of DACH (1.54 g, 15.4 mmol)
in C2H5OH (80 mL) with vigorous stirring at room temperature.
Suitable portions of solid KOH were added to the above mixture
to ensure that the pH value stayed at 7Ϫ8 for ca. 3 days. After
filtration of the mixture, the solvent was removed by rotary evapor-
ation. The residue was purified by silica gel column chromatogra-
phy (CH2Cl2/CH3OH/NH3·H2O from 10:10:1 to 5:5:1). The final
product was obtained as a white solid in 55% yield (2.65 g, based
on DACH). 1H NMR (CDCl3): δ ϭ 1.91Ϫ1.97 (m, 2 H, CH2),
2.79Ϫ2.85 (m, 8 H, CH2), 3.80 (s, 4 H, CH2), 6.75Ϫ6.84 (m, 4 H,
Ph), 6.94 (d, J ϭ 1.5 Hz, 1 H, Ph), 6.96 (d, J ϭ 1.5 Hz, 1 H, Ph),
7.14Ϫ7.26 (m, 2 H, Ph). IR [KBr pellet (cmϪ1)]: 3432 b, 3043 m,
2934 m, 2912 w, 2876 w, 2841 m, 2717 w, 2627 w, 1933 w, 1897 w,
1784 w, 1613 s, 1590 vs, 1479 vs, 1458 vs, 1421s, 1383 w, 1357 s,
1349 s, 1334 m, 1307 w, 1256 vs, 1183 m, 1151 s, 1090 s, 1057 m,
1034 s, 1016 m, 983 s, 936 m, 917 m, 871 m, 850 w, 811 s, 755 vs,
720 m, 625 m, 595 w, 547 m, 459 s. C19H24N2O2 (312.41): calcd. C
73.05, H 7.74, N 8.97; found C 73.19, H 7.51, N 8.84.
˚
performed with Mo-Kα radiation (λ ϭ 0.71073 A) by the ω scan
Table 4. Crystallographic data and structural refinement param-
eters of complexes 1 and 2
1
2
Empirical formula
Mr
C40H52Cl2Cu3N4O14
1074.41
monoclinic
P21/n
10.915(4)
14.208(5)
14.114(5)
90
98.412(6)
90
2165.3(13)
2
C40H52Cl2Ni3N4O14
1059.83
Crystal system
Space group
triclinic
¯
P1
˚
a (A)
9.682(4)
11.182(4)
11.278(5)
76.624(7)
81.561(7)
68.404(7)
1101.9(8)
1
˚
b (A)
˚
c (A)
α (°)
β (°)
γ (°)
[Cu3(µ-L)2](CH3OH)2(ClO4)2 (1): To a solution of H2L (63 mg,
0.2 mmol) in CHCl3 (20 mL) was slowly added a CH3OH solution
(15 mL) of Cu(ClO4)2·6H2O (110 mg, 0.3 mmol) with stirring and
a brown powder of complex 1 precipitated gradually. The powder
was filtered off, washed with methanol/water and dried under vac-
uum. Yield: 97 mg (90%). Single crystals of 1 suitable for X-ray
analysis were obtained as follows: a solution of Cu(ClO4)2·6H2O
3
˚
V (A )
Z
Dc [g cmϪ3
]
1.648
1.655
1106
1.597
1.460
550
µ [mmϪ1
F(000)
]
in CH3OH was carefully layered onto a solution of H2L in CHCl3 Total data
10022
4392
2860
0.0448
0.0470
0.1011
1.008
0.575, Ϫ0.393
4702
3850
2377
0.0322
0.0447
0.0842
1.011
Unique data
Data [I Ͼ 2σ(I)]
Rint
in a 3:2 molar ratio in a straight glass tube and rhombic brown
crystals were observed after ca. two weeks. C40H52Cl2Cu3N4O14
(1074.41): calcd. C 44.72, H 4.88, N 5.21; found C 44.61, H 4.74,
N 5.17. IR (cmϪ1): 3446 b, 3061 w, 2940 w, 2905 w, 2860 m, 1598
s, 1574 w, 1482 vs, 1445 s, 1395 m, 1340 w, 1327 w, 1264 s, 1247 vs,
1212 m, 1198 m, 1102 vs, 1061 vs, 970 m, 923 m, 881 m, 864 m,
785 m, 773 s, 763 vs, 761 w, 642 m, 623 s, 577m. ΛM (CH3CN):
266 cm2 ΩϪ1 molϪ1. λmax/nm (ε/dm3 molϪ1 cmϪ1) (CH3CN): 558
(220) and 386 (520).
R1[a]
wR2[b]
GOF[c]
Ϫ3
˚
Residuals [e·A
]
0.340, Ϫ0.357
[a]
[b]
2
2 2
2 2 1/2 [c]
R1 ϭ ΣFo Ϫ Fc /ΣFo.
wR2 ϭ [Σw(Fo Ϫ Fc ) /Σw(Fo ) ] .
GOF ϭ {Σ[w(Fo Ϫ Fc ) ]/(n Ϫ p)}1/2
.
2
2 2
302
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 294Ϫ304

Towards the Design of Linear Homo-Trinuclear Metal Complexes
FULL PAPER
R. Cortes, L. Lezama, T. Rojo, Coord. Chem. Rev. 1999, 195,
1027Ϫ1068.
mode in the range of 2.21 Ͻ θ Ͻ 26.39° (for 1) and 2.00 Ͻ θ Ͻ
25.00° (for 2). There was no evidence of crystal decay during data
collection for both complexes. All the measured independent reflec-
tions were used in the structural analyses and semi-empirical ab-
sorption corrections were applied using the SADABS program. The
program SAINT[50] was used for integration of the diffraction pro-
files. Both structures were solved by direct methods using the
SHELXS program of the SHELXTL package and refined with
SHELXL.[51] Metal atoms were located from the E-maps and other
nonhydrogen atoms were located in successive difference Fourier
syntheses. The final refinements were performed by full-matrix le-
ast-squares methods with anisotropic thermal parameters for all
the nonhydrogen atoms on F2. All hydrogen atoms were first found
in difference electron density maps and then placed in the calcu-
lated sites and included in the final refinement in the riding model
approximation with displacement parameters derived from the par-
ent atoms to which they were bonded. In the refinement of both
complexes, a disorder model of the perchlorate anion was used (two
constraint components for four oxygen atoms of the perchlorate
anion have occupancy factors of 0.5) to make the oxygen atoms
exhibit suitable displacement ellipsoids. A summary of the crystal-
lographic data and structure refinements are listed in Table 4.
CCDC-232051 (for 1) and -232052 (for 2) contains the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge via www.ccdc.cam.ac.uk/cons/ retrieving.html
(or from the CCDC, 12 Union Road, Cambridge CB2 1EZ, UK;
Fax: ϩ44-1223-336-033; E-mail: deposit@ccdc.cam.ac.uk).
[8] [8a]
M. Du, X. H. Bu, G. C. Wang, Y. M. Guo, R. H. Zhang,
Inorg. Chim. Acta 2001, 320, 190Ϫ197. [8b] M. Du, Y. M. Guo,
X. H. Bu, Inorg. Chim. Acta 2002, 335, 136Ϫ140.
[9] [9a]
X. H. Bu, M. Du, L. Zhang, Z. L. Shang, R. H. Zhang,
[9b]
M. Shionoya, J. Chem. Soc., Dalton Trans. 2001, 729Ϫ735.
X. H. Bu, M. Du, Z. L. Shang, L. Zhang, Q. H. Zhao, R. H.
Zhang, M. Shionoya, Eur. J. Inorg. Chem. 2001, 1551Ϫ1558.
[9c]
M. Du, Z. L. Shang, X. B. Leng, X. H. Bu, Polyhedron
2001, 20, 3065Ϫ3071.
[10] [10a]
X. H. Bu, M. Du, Z. L. Shang, R. H. Zhang, D. Z. Liao,
M. Shionoya, T. Clifford, Inorg. Chem. 2000, 39, 4190Ϫ4199.
[10b]
X. H. Bu, M. Du, L. Zhang, D. Z. Liao, J. K. Tang, R.
H. Zhang, M. Shionoya, J. Chem. Soc., Dalton Trans. 2001,
593Ϫ598. [10c] M. Du, Y. M. Guo, X. H. Bu, J. Ribas, M. Mon-
fort, New J. Chem. 2002, 26, 645Ϫ650.
[11] [11a]
M. Du, X. H. Bu, Y. M. Guo, L. Zhang, D. Z. Liao, J.
[11b]
Ribas, Chem. Commun. 2002, 1478Ϫ1479.
M. Du, X. H.
Bu, Y. M. Guo, J. Ribas, C. Diaz, Chem. Commun. 2002,
2550Ϫ2551. [11c] M. Du, X. H. Bu, Y. M. Guo, J. Ribas, Chem.
Eur. J. 2004, 10, 1345Ϫ1354.
C. A. Grapperhaus, M. Y. Darensbourg, Acc. Chem. Res. 1998,
31, 451Ϫ459 and references cited therein.
[12]
[13] [13a]
D. L. An, Y. M. Guo, M. Du, X. S. Shi, X. H. Bu, Chin.
[13b]
J. Struct. Chem. 2004, 23, 126Ϫ130.
M. Du, Y. M. Guo,
[13c]
S. T. Chen, X. H. Bu, J. Mol. Struct. 2002, 641, 29.
Y. M .
[13d]
Guo, M. Du, X. H. Bu, J. Mol. Struct. 2002, 610, 27Ϫ31.
M. Du, Y. M. Guo, X. H. Bu, J. Ribas, M. Monfort, New J.
[13e]
Chem. 2002, 26, 939Ϫ945.
Y. M. Guo, Doctor Disser-
tation, 2001, Nankai University.
[14] [14a]
Acknowledgments
H. Okawa, M. Koikawa, S. Kida, D. Luneau, H. Oshio, J.
[14b]
Chem. Soc., Dalton Trans. 1990, 469Ϫ475.
Nakao, K. Matsumoto, S. Takamizawa, W. Mori, S. Suzuki,
Chem. Lett. 1997, 181Ϫ182.
Finazzo, S. Gorter, D. Rehorst, J. Reedijk, M. Lutz, A. L.
F. Akagi, Y.
This work was financially supported by the National Natural
Science Foundation of China (No. 20401012) and the Key Project
of Tianjin Natural Science Foundation (to M. D.). X. H. B.
acknowledges financial support from the Outstanding Youth Foun-
dation of NSFC (No. 20225101) and J. R. from the Spanish
Government (Grant BQU2003/00539).
[14c]
W. L. Driessen, L. Chang, C.
Spek, Inorg. Chim. Acta 2003, 350, 25Ϫ31.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, Wiley, New York, 1986.
E. V. RybakAkimova, D. H. Busch, P. K. Kahol, N. Pinto, N.
W. Alcock, H. J. Clase, Inorg. Chem. 1997, 36, 510Ϫ520.
G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in Struc-
tural Chemistry and Biology, Oxford University Press, 1999.
O. Kahn, Molecular Magnetism, VCH Publishers, New York,
1993.
A. Bencini, D. Gatteschi, EPR of Exchange Coupled Systems,
Springer-Verlag, Berlin, 1989.
L. Banci, A. Bencini, D. Gatteschi, Inorg. Chem. 1983, 22,
2681Ϫ2683.
R. Veit, J. J. Girerd, O. Kahn, F. Robert, Y. Jeannin, Inorg.
Chem. 1986, 25, 4175Ϫ4180.
[1] [1a]
[1b]
D. Gatteschi, Adv. Mater. 1994, 6, 635Ϫ645.
Winpenny, Comments Inorg. Chem. 1999, 20, 233Ϫ262.
Leininger, B. Olenyuk, P. J. Stang, Chem. Rev. 2000, 100,
R. E. P.
[1c]
S.
[1d]
853Ϫ908.
E. I. Solomon, R. K. Szilagyi, S. D. George, L.
Basumallick, Chem. Rev. 2004, 104, 419Ϫ458. [1e] D. Gatteschi,
[1f]
R. Sessoli, Angew. Chem. Int. Ed. 2003, 42, 268Ϫ297.
M.
A. Halcrow, G. Christou, Chem. Rev. 1994, 94, 2421Ϫ2481.
[2]
[3]
[2a]
For examples, see:
G. A. van Albada, I. Mutikainen, U.
[2b]
Turpeinen, J. Reedijk, Eur. J. Inorg. Chem. 1998, 547Ϫ549.
L. Cronin, P. H. Walton, Inorg. Chim. Acta 1998, 269,
[2c]
241Ϫ245.
T. C. Higgs, K. Spartalian, C. J. O’Connor, B. F.
Matzanke, C. J. Carrano, Inorg. Chem. 1998, 37, 2263Ϫ2272.
For examples, see:
´
R. Costa, A. Garcıa, J. Ribas, T. Mallah, Y. Journaux, X. Sol-
[3a]
R. Prins, M. Biagini-Cingi, M. Drillon,
´
ans, V. Rodrıguez, Inorg. Chem. 1993, 32, 3733Ϫ3742 and refer-
ences cited therein.
R. A. G. de Graaff, J. G. Haasnoot, A. M. Manotti-Lanfredi,
P. Rabu, J. Reedijk, F. Uggozzoli, Inorg. Chim. Acta 1996, 248,
[23]
[24]
´
´
L. Gutierrez, G. Alzuet, J. A. Real, J. Cano, J. Borras, A. Casti-
n˜eiras, Eur. J. Inorg. Chem. 2002, 2094Ϫ2102.
[3b]
35Ϫ44.
P. A. Angaridis, P. Baran, R. Boca, F. Cervantes-
Lee, W. Haase, G. Mezei, R. G. Raptis, R. Werner, Inorg. Chem.
2002, 41, 2219Ϫ2228.
V. H. Crawford, H. W. Richardson, J. R. Wason, D. J. Hodg-
son, W. E. Hatfield, Inorg. Chem. 1976, 15, 2107Ϫ2110.
[4]
[5]
For examples, see: S. Ferrer, J. G. Haasnoot, J. Reedijk, E.
Muller, M. B. Cingi, M. Lanfranchi, A. M. M. Lanfredi, J.
Ribas, Inorg. Chem. 2000, 39, 1859Ϫ1867.
[25] [25a]
L. Merz, W. Haase, J. Chem. Soc., Dalton Trans. 1980,
[25b]
875Ϫ879.
M. Handa, N. Koga, S. Kida, Bull. Chem. Soc.,
Jpn. 1988, 61, 3853Ϫ3857.
For examples, see: [5a] R. L. Beddoes, J. A. Conner, D. Dubow-
[26]
[27]
K. Nieminen, Ann. Acad. Sci. Fennicae A II 1982, 197, 1Ϫ60.
L. Walz, H. Paulus, W. Haase, J. Chem. Soc., Dalton Trans.
1985, 913Ϫ920.
ski, A. C. Jones, O. S. Mills, R. Price, J. Chem. Soc., Dalton
[5b]
´
L. Gutierrez, G. Alzuet, J. A.
Trans. 1981, 2119Ϫ2126.
´
Real, J. Cano, J. Borras, A. Castin˜eiras, Inorg. Chem. 2000,
[28]
[29]
A. Bencini, D. Gatteschi, Inorg. Chim. Acta 1978, 31, 11Ϫ18.
P. J. Hay, J. C. Thibault, R. Hoffman, J. Am. Chem. Soc. 1975,
97, 4884Ϫ4899.
39, 3608Ϫ3614.
[6]
[7]
For examples, see: X. M. Liu, M. P. de Miranda, E. J. L.
McInnes, C. A. Kilner, M. A. Halcrow, Dalton Trans. 2004,
59Ϫ64.
[30] [30a]
[30b]
O. Kahn, Inorg. Chim. Acta 1982, 62, 3Ϫ14.
J. Com-
For examples, see: J. Ribas, A. Escuer, M. Monfort, R. Vicente,
marmond, P. Plumere, J. M. Lehn, Y. Agnus, P. Louis, R. Weiss,
Eur. J. Inorg. Chem. 2005, 294Ϫ304
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
303

M. Du, X.-J. Zhao, J.-H. Guo, X.-H. Bu, J. Ribas
FULL PAPER
[40]
[41]
[42]
[43]
[44]
O. Kahn, I. Morgenstern-Badarau, J. Am. Chem. Soc. 1982,
104, 6330Ϫ6340.
A. M. Greenaway, C. J. O’Connor, J. W. Overman, E. Sinn,
Inorg. Chem. 1981, 20, 1508Ϫ1513.
[31]
H. Astheimer, W. Haase, J. Chem. Phys. 1986, 85, 1427Ϫ1432.
J. P. Costes, F. Dahan, J. P. Laurent, Inorg. Chem. 1985, 24,
1018Ϫ1022.
[32] [32a] E. Ruiz, P. Alemany, S. Alvarez, J. Cano, J. Am. Chem. Soc.
[32b]
´
1997, 119, 1297Ϫ1303.
E. Ruiz, P. Alemany, S. Alvarez, J.
J. M. Seco, U. Amador, M. J. Gonzalez-Garmendia, Inorg.
Cano, Inorg. Chem. 1997, 36, 3683Ϫ3688.
C. Blanchet-Boiteux, J. M. Muesca, J. Phys. Chem. A 2000,
Chim. Acta 2000, 303, 256Ϫ264.
[33]
[34]
[35]
G. A. van Albada, I. Mutikainen, I. Riggio, U. Turpeinen, J.
Reedijk, Polyhedron 2002, 21, 141Ϫ146.
H. Saimiya, Y. Sunatsuki, M. Kojima, S. Kashino, T. Kambe,
M. Hirotsu, H. Akashi, K. Nakajima, T. Tokii, Dalton Trans.
2002, 3737Ϫ3742.
104, 2091Ϫ2097.
R. J. Butcher, E. Sinn, Inorg. Chem. 1976, 15, 1604Ϫ1608 and
references cited therein.
A. W. Addison, T. N. Rao, J. Reedijk, J. V. Rijn, G. C. Ver-
schoor, J. Chem. Soc., Dalton Trans. 1984, 1349Ϫ1356.
[45]
J. Galy, J. Jaud, O. Kahn, P. Tola, Inorg. Chim. Acta 1979, 36,
229Ϫ236.
E. Sinn, Inorg. Chem. 1976, 15, 366Ϫ369.
S. K. Mandal, L. K. Thompson, K. Nag, J. P. Charland, E. J.
Gabe, Inorg. Chem. 1987, 26, 1391Ϫ1395.
J. A. Bertrand, E. Fujita, P. G. Eller, Inorg. Chem. 1974, 13,
2067Ϫ2071.
R. Range, Tetrahedron 1971, 27, 1499Ϫ1502.
Bruker AXS, SAINT Software Reference Manual, Madison,
WI, 1998.
[36] [36a] M. F. Charlot, S. Jeannin, Y. Jeannin, O. Kahn, J. Lucrece-
Abaul, J. Martin-Frere, Inorg. Chem. 1979, 18, 1675Ϫ1681. [36b]
M. F. Charlot, S. Jeannin, Y. Jeannin, O. Kahn, J. Lucrece-
Abaul, J. Martin-Frere, Inorg. Chem. 1980, 19, 1410Ϫ1411.
[37] [37a] S. Youngme, G. A van Albada, H. Kooijman, O. Roubeau,
[46]
[47]
[48]
W. Somjitsripunya, A. L. Spek, C. Pakawatchai, J. Reedijk, Eur.
[37b]
[49]
[50]
J. Inorg. Chem. 2002, 2367Ϫ2374.
I. Castro, M. Julve, G.
de Munno, G. Bruno, J. A. Real, F. Lloret, J. Faus, J. Chem.
Soc., Dalton Trans. 1992, 1739Ϫ1744.
[38] [38a]
[51]
S. Youngme, W. Somjitsripunya, K. Chinnakali, S. Chan-
G. M. Sheldrick, SHELXTL NT Version 5.1. Program for Solu-
tion and Refinement of Crystal Structures, University of
Göttingen, Germany, 1997.
[38b]
traproma, H. K. Fun, Polyhedron 1999, 18, 857Ϫ862.
L.
P. Wu, M. E. Keniry, B. J. Hathaway, Acta Crystallogr., Sect.
C 1992, 48, 35Ϫ40.
L. K. Thompson, S. K. Mandal, S. S. Tandon, J. N. Bridson,
M. K. Park, Inorg. Chem. 1996, 35, 3117Ϫ3125.
Received July 16, 2004
Early View Article
Published Online November 24, 2004
[39]
304
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 294Ϫ304