μ3- vs. μ-hydroxido bridges- peripheral function controls the nuclearity of hydroxido-bridged copper(II) complexes

Article abstract of DOI:10.1002/ejic.201200819

The synthesis and characterization of three monohydroxido-bridged dinuclear Cu^II complexes, [{Cu₂(L¹)₂(μ-OH) }(ClO₄){Cu₂(L¹)₂(μ-OH)}](ClO ₄)·5.5H₂O (1), Cu₂(L²) ₂(OH)(ClO₄)(H₂O)₂ (2), [{Cu ₂(L³)₂(μ-OH)(ClO₄)} ₂]·5H₂O (3) and three monohydroxido-bridged trinuclear Cu^II complexes, [Cu₃(L⁴) ₃(μ₃-OMe)(ClO₄)₂] (4), [Cu ₃(L⁵)₃(μ₃-OH) (ClO ₄)₂]·0.5{(CH₃)₂CO}·0. 5(H₂O) (5), [Cu₃(L⁶) ₃(μ₃-OH)(ClO₄)₂] (6), is reported (HL¹-HL⁶ = tridentate N₂O ligands resulting from the Schiff base condensation of salicylaldehyde or 3-methoxysalicylaldehyde with ethylenediamine, 1,2-diaminopropane, 2-methylpropane-1,2-diamine or 1,2-diaminocyclohexane). The six new complexes are obtained through aggregation of the corresponding mononuclear precursor Cu(Lⁱ)(py)(ClO ₄) in the presence of HO^- (MeO^-) anions in methanol. The nuclearity of the resulting complex depends on the steric requirement of Lⁱ: in the presence of an OMe substituent α to the O_phenoxido atom (L¹-L³), μ-HO ^- single-bridged dinuclear compounds are obtained (1-3); in the absence of an α-substituent (L⁴-L⁶), the aggregation proceeds to the μ₃-OH^- (μ₃-OMe ^-) bridged trinuclear compounds 4-6. The molecular structures of 1, 3 and 4-6 have been established through single-crystal X-ray studies. Antiferromagnetic interactions operate in 1 (J =-57 cm^-1,-2JS ₁S₂ formalism), 2 is paramagnetic, and 3 is the first example of a ferromagnetic monohydroxido-bridged dinuclear Cu^II complex (J = 16 cm^-1). The structural and magnetic parameters for the 22 μ-HO^- single-bridged dinuclear Cu^II compounds reported to date are compared. Antiferromagnetic interactions operate in all three monohydroxido-bridged trinuclear Cu^II complexes 4, 5, and 6 (J =-16.9,-6.3, and-6.7 cm^-1, respectively). The structural and magnetic parameters for the seven μ₃-hydroxido (methoxido) bridged trinuclear Cu^II compounds presently known in the series based on tridentate Schiff bases that result from condensation of diamines with salicylaldehydes are compared.

Full text of DOI:10.1002/ejic.201200819

FULL PAPER
DOI: 10.1002/ejic.201200819
μ3- vs. μ-Hydroxido Bridges – Peripheral Function Controls the Nuclearity of
Hydroxido-Bridged Copper(II) Complexes
Fatima Zohra Chiboub Fellah,[a,b,c] Jean-Pierre Costes,*[a,b] Laure Vendier,[a,b]
Carine Duhayon,[a,b] Sonia Ladeira,[a,b] and Jean-Pierre Tuchagues*[a,b]
Keywords: Copper / N,O ligands / Bridging ligands / X-ray structures / Magnetic properties
The synthesis and characterization of three monohydroxido-
bridged dinuclear CuII complexes, [{Cu2(L1)2(μ-OH)}(ClO4)- the aggregation proceeds to the μ3-OH– (μ3-OMe–) bridged
are obtained (1–3); in the absence of an α-substituent (L4–L6),
{Cu2(L1)2(μ-OH)}](ClO4)·5.5H2O (1), Cu2(L2)2(OH)(ClO4)-
(H2O)2 (2), [{Cu2(L3)2(μ-OH)(ClO4)}2]·5H2O (3) and three
monohydroxido-bridged trinuclear CuII complexes, [Cu3(L4)3-
(μ3-OMe)(ClO4)2] (4), [Cu3(L5)3(μ3-OH) (ClO4)2]·0.5{(CH3)2-
CO}·0.5(H2O) (5), [Cu3(L6)3(μ3-OH)(ClO4)2] (6), is reported
(HL1–HL6 = tridentate N2O ligands resulting from the Schiff
base condensation of salicylaldehyde or 3-methoxysalicylal-
dehyde with ethylenediamine, 1,2-diaminopropane, 2-meth-
ylpropane-1,2-diamine or 1,2-diaminocyclohexane). The six
new complexes are obtained through aggregation of the cor-
responding mononuclear precursor Cu(Li)(py)(ClO4) in the
presence of HO– (MeO–) anions in methanol. The nuclearity
of the resulting complex depends on the steric requirement
of Li: in the presence of an OMe substituent α to the Ophenoxido
atom (L1–L3), μ-HO– single-bridged dinuclear compounds
trinuclear compounds 4–6. The molecular structures of 1, 3
and 4–6 have been established through single-crystal X-ray
studies. Antiferromagnetic interactions operate in 1 (J =
–57 cm–1, –2JS1S2 formalism), 2 is paramagnetic, and 3 is the
first example of a ferromagnetic monohydroxido-bridged di-
nuclear CuII complex (J = 16 cm–1). The structural and mag-
netic parameters for the 22 μ-HO– single-bridged dinuclear
CuII compounds reported to date are compared. Antiferro-
magnetic interactions operate in all three monohydroxido-
bridgedtrinuclearCuII complexes4,5,and6(J=–16.9,–6.3,and
–6.7 cm–1, respectively). The structural and magnetic param-
eters for the seven μ3-hydroxido (methoxido) bridged trinu-
clear CuII compounds presently known in the series based
on tridentate Schiff bases that result from condensation of
diamines with salicylaldehydes are compared.
terials science,[14–17] we decided to further explore the ability
of tridentate Schiff base ligands with the N2O donor set to
yield hydroxido-bridged CuII species with different nucleari-
ties. Considering the literature,[1–11] we noticed that most
tridentate Schiff base ligands in the reported complexes are
based on acetylacetone and its derivatives, in which the en-
amine function of the tridentate ligand deprotonates, in
contrast with the ligands based on salicylaldehyde, the
phenol function of which deprotonates. These ligands also
differ in their preparation, the former is isolated before
complexation, and the latter is obtained through a copper
template effect. It is worth noting that nonsymmetrical
salen-type ligands prepared by a nontemplate route are re-
ported in a Microreview.[5] Also, the fine-tuning of steric
and electronic properties is facilitated by tridentate Schiff
base ligands based on salicylaldehyde, as a consequence of
a larger choice in substituents. We also noticed that the two
cases that yield dinuclear species involve either phenoxido[8]
or aqua[10,11] bridges, and there is no previous report of
hydroxido-bridged dinuclear complexes. With these land-
marks in mind, we investigated the CuII complexes obtained
from six tridentate N2O ligands resulting from the Schiff
base condensation of salicylaldehyde or 3-methoxysalicyl-
aldehyde with ethylenediamine, 1,2-diaminopropane, 2-
Introduction
Tridentate Schiff base ligands that include the N2O do-
nor set with an easily deprotonatable phenol oxygen donor
atom or an enamine nitrogen donor atom have a strong
propensity to yield hydroxido- or methoxido-centered cyclic
trinuclear cationic complexes upon reaction with copper(II)
salts in a basic medium.[1–9] In fewer cases, they may yield
dimeric[8,10,11] CuII species bridged through the phenoxido
oxygen atom of the tridentate ligands,[8] or through an aqua
bridge reinforced by two hydrogen bonds.[10,11] However, we
are not aware of any hydroxido-bridged CuII dinuclear spe-
cies including such tridentate Schiff base ligands. In view of
the interest in exchange-coupled di- and polynuclear CuII
complexes in biological processes[12,13] and in inorganic ma-
[a] CNRS; LCC (Laboratoire de Chimie de Coordination),
205, route de Narbonne, 31077 Toulouse, France
Fax: +33-561-553003
E-mail: jean-pierre.costes@lcc-toulouse.fr
jean-pierre.tuchagues@lcc-toulouse.fr
Homepage: http://www.lcc-toulouse.fr/lcc/spip.php?article19
[b] Université de Toulouse; UPS, INPT, LCC,
31077 Toulouse, France
[c] Université Aboubaker Belkaid, Faculté des Sciences,
Département de Chimie,
BP 119, 13000 Tlemcen, Algérie
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5729

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
methylpropane-1,2-diamine or 1,2-diaminocyclohexane 3), μ3-MeO– (4) or μ3-HO– (5, 6) bridges. These reactions
(Scheme 1). We report the synthesis, characterization, crys- are quite rapid: the dark mononuclear precursor dissolves
tal structure, and magnetic properties of three monohydrox- upon dropwise addition of Et3N, and a purple (1–3) or
ido-bridged dinuclear CuII complexes, [{Cu2(L1)2(μ- green (4–6) precipitate forms within minutes. The nuclearity
OH)}(ClO4){Cu2(L1)2(μ-OH)}](ClO4)·5.5H2O(1),Cu2(L2)2- of the resulting complex depends on the steric requirement
(OH)(ClO4)(H2O)2 (2), [{Cu2(L3)2(μ-OH)(ClO4)}2]·5H2O of Li: in the presence of an OMe substituent α to the
(3) and three monohydroxido-bridged trinuclear CuII com- Ophenoxido atom (L1–L3), dinuclear compounds form (1–3);
plexes, [Cu3(L4)3(μ3-OMe)(ClO4)2] (4), [Cu3(L5)3(μ3- on the other hand, in the absence of an α-substituent (L4–
OH)(ClO4)2]·0.5{(CH3)2CO}·0.5(H2O) (5), [Cu3(L6)3(μ3- L6), the aggregation process proceeds to the trinuclear com-
OH)(ClO4)2] (6). The parameters that participate in the pounds (4–6). The CuII centers of complexes 1–3, bridged
control of their nuclearity are examined, and their magnetic through a single μ-HO– anion, are additional examples of
properties are discussed, based on their specific structural μ-OH– single-bridged dinuclear CuII compounds.[22–34] Al-
features. Antiferromagnetic interactions operate in 1, 2 is though cubane compounds with a Cu4O4 core have been
paramagnetic, and 3 is the first example of a ferromagnetic extensively reported,[35] fewer defective cubane compounds
monohydroxido-bridged dinuclear CuII complex. Antiferro- with a Cu3O4 core have been reported; most are based on
magnetic interactions operate in the monohydroxido- oxime,[36–41] N,N-pyrazole-type,[42–44] or N,N-triazole-
bridged trinuclear CuII complexes 4, 5, and 6.
type[45–47] ligands or tridentate Schiff bases resulting from
the condensation of diamines with β-diketones[1–7] or sali-
cylaldehydes.[3,8,9] Complexes 4–6 extend the scarce μ-OH–-
bridged trinuclear CuII compounds based on tridentate
NNO ligands with the salicylaldimine motif.[3,8,9]
Description of the Structures
[{Cu2(L1)2(μ-OH)}(ClO4){Cu2(L1)2(μ-OH)}](ClO4)·
5.5H2O (1)
The asymmetric unit of 1 includes two distinct dinuclear
[Cu2(L1)2(μ-OH)]+ complex cations loosely bonded through
two oxygen atoms of an intercalated perchlorato anion, one
out-of-sphere perchlorate anion and 5.5 water molecules.
This unusual structural arrangement may be described as
bis(dinuclear) (Figure 1).
The CuII centers of each dinuclear subunit are bridged
through one μ-hydroxido oxygen atom with a Cu–O–Cu an-
gle close to 109° and Cu1···Cu2 (Cu3···Cu4) distances of
3.145(2) [3.139(2)] Å. The next shortest Cu···Cu distances
are 6.739(4) and 7.415(4) Å. Each CuII center is in a square-
pyramidal N2O3 ligand environment that includes three
equatorial donors from the polydentate ligand, one amino
and one imino nitrogen atom, and one phenoxido oxygen
atom; the fourth equatorial donor is the hydroxido oxygen
atom. Equatorial bond lengths and angles (caption to Fig-
ure 1) are in the usual ranges for CuII in similar ligand envi-
ronments.[22–34] The apical coordination site of each copper
Scheme 1. Schematic drawings of the polydentate ligands HL1–
HL6.
Results and Discussion
Syntheses
The dissymmetric potentially tri- or tetradentate Schiff atom is occupied by one oxygen atom of the intercalated
base ligands HLi (i = 1–6) considered in this study were perchlorato anion, O11 for Cu1 and Cu2 and O12 for Cu3
not isolated as independent molecules; they were generated, and Cu4 with distances in the semibonding 2.69–2.95 Å
through a template effect in the presence of an ancillary range. Together with O–Cu–O(N) angles in the 78.6(4)–
ligand (pyridine), as mononuclear copper(II) complexes of 98.4(4)° range, this leads to the description of the coordina-
general formula Cu(Li)(py)(ClO4) according to published tion sphere of each CuII center as an elongated square pyra-
procedures.[19–21]
mid.
Owing to the coordination of the intercalated perchlor-
The dinuclear (1–3) and trinuclear (4–6) complexes were
obtained in the same way, by adding a 3:1 molar excess of ato anion to the four copper centers of each bis(dinuclear)
triethylamine to the corresponding Cu(Li)(py)(ClO4) mono- unit, the cyclohexyl rings of the four L1 polydentate ligands
nuclear precursor suspended in methanol: the presence of a are held in determined conformations with respect to each
significant concentration of HO– (MeO–) anions in the other; it is remarkable that all four cyclohexyl rings in an
slurry allows an aggregation process involving μ-HO– (1– asymmetric unit are (R,R) or (S,S). However, as indicated
5730
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740

μ3- vs. μ-Hydroxido Bridges in Copper(II) Complexes
interacts with O8(methoxy) with N···O distances of 3.47,
3.35, and 3.42 Å, respectively. No significant intermolecular
hydrogen interactions have been detected.
[{Cu2(L3)2(μ-OH)(ClO4)}2]·5H2O (3)
In contrast to that of 1, the asymmetric unit of 3 includes
two independent but very similar neutral [Cu2(L3)2(μ-
OH)(ClO4)] dinuclear complex molecules as a result of the
coordination of one perchlorato anion to one CuII center
in each dinuclear unit. In addition, the absence of bonding
interactions between the independent [Cu2(L3)2(μ-
OH)(ClO4)] dinuclear molecules of an asymmetric unit al-
lows 3 to be described as a dinuclear compound (Figure 2).
Figure 2. Diamond drawing of the asymmetric unit [{Cu2(L3)2(μ-
OH)(ClO4)}2] of 3. Water molecules and hydrogen atoms are omit-
ted for clarity. Thermal ellipsoids are drawn at the 50% probability
level. Selected bond lengths [Å] and angles [°]: Cu1–N1 2.015(6),
Cu1–N2 1.949(6), Cu1–O1 1.903(5), Cu1–O5 1.947(5), Cu1–O9
2.510(8), Cu2–N3 2.006(6), Cu2–N4 1.921(6), Cu2–O3 1.915(5),
Cu2–O5 1.923(4), Cu3–N5 2.006(6), Cu3–N6 1.942(6), Cu3–O10
1.910(5), Cu3–O14 1.911(5), Cu4–N7 2.011(6), Cu4–N8 1.926(6),
Cu4–O12 1.901(5), Cu4–O14 1.949(5), Cu4–O17 2.53(1); O1–Cu1–
N1 169.2(2), O1–Cu1–N2 92.4(2), O1–Cu1–O5 89.7(2), O5–Cu1–
N1 93.3(2), O5–Cu1–N2 175.8(2), N2–Cu1–N1 83.9(3), O9–Cu1–
N1 90.9(3), O9–Cu1–N2 81.2(3), O9–Cu1–O1 98.5(3), O9–Cu1–O5
102.0(3), O3–Cu2–N3 171.4(2), O3–Cu2–N4 92.4(2), O3–Cu2–O5
89.75(19), O5–Cu2–N3 93.6(2), N4–Cu2–N3 83.8(2), N4–Cu2–O5
176.1(2), O10–Cu3–N5 168.7(2), O10–Cu3–N6 91.5(2), O10–Cu3–
O14 89.7(2), O14–Cu3–N5 94.1(2), O14–Cu3–N6 178.3(2), N6–
Cu3–N5 84.5(2), O12–Cu4–N7 168.3(2), O12–Cu4–N8 92.6(2),
O12–Cu4–O14 88.7(2), O14–Cu4–N7 94.3(2), N8–Cu4–N7 83.8(3),
N8–Cu4–O14 176.6(2), O17–Cu4–N7 88.3(3), O17–Cu4–N8
94.9(3), O17–Cu4–O12 103.0(3), O17–Cu4–O14 87.9(3), Cu1–O5–
Cu2 106.9(3), Cu3–O14–Cu4 106.99(5).
Figure 1. Diamond drawings of the asymmetric unit [{Cu2(L1)2(μ-
OH)}(ClO4){Cu2(L1)2(μ-OH)}](ClO4) of 1, top, and of one
[Cu2(L1)2(μ-OH)]+ complex cation, bottom. Water molecules and
hydrogen atoms are omitted for clarity. Thermal ellipsoids are
drawn at the 40% probability level. Selected bond lengths [Å] and
angles [°]: Cu1–N1 2.006(10), Cu1–N2 1.930(11), Cu1–O1 1.879(8),
Cu1–O9 1.941(9), Cu1–O11 2.696(9), Cu2–N3 1.994(10), Cu2–N4
1.920(11), Cu2–O3 1.879(8), Cu2–O9 1.917(8), Cu2–O11 2.834(10),
Cu3–N5 2.003(10), Cu3–N6 1.911(10), Cu3–O5 1.892(8), Cu3–O10
1.937(8), Cu3–O12 2.757(10), Cu4–N7 1.968(11), Cu4–N8
1.918(12), Cu4–O7 1.905(9), Cu4–O10 1.914(9), Cu4–O12 2.95(2);
N1–Cu1–N2 85.1(4), N2–Cu1–O1 93.8(4), O1–Cu1–O9 88.1(4),
O9–Cu1–N1 94.7(4), O11–Cu1–N1 86.1(4), O11–Cu1–N2 96.0(4),
O11–Cu1–O1 82.8(3), O11–Cu1–O9 92.2(4), N1–Cu1–O1 168.6(4),
N2–Cu1–O9 171.7(4), N3–Cu2–N4 84.4(4), N4–Cu2–O3 94.2(4),
O3–Cu2–O9 89.9(3), O9–Cu2–N3 91.8(4), O11–Cu2–N3 79.5(4),
O11–Cu2–N4 94.6(4), O11–Cu2–O3 98.4(3), O11–Cu2–O9 88.6(3),
N4–Cu2–O9 174.4(4), N3–Cu2–O3 177.3(4), N5–Cu3–N6 84.3(4),
N5–Cu3–O10 94.6(4), O10–Cu3–O5 88.2(3), O5–Cu3–N6 94.4(4),
O12–Cu3–N5 85.5(4), O12–Cu3–N6 91.9(4), O12–Cu3–O5 81.5(3),
O12–Cu3–O10 94.7(3), N6–Cu3–O10 173.2(4), N5–Cu3–O5
166.9(4), N7–Cu4–N8 83.9(5), N8–Cu4–O7 94.5(4), O7–Cu4–O10
88.6(4), O10–Cu4–N7 93.3(4), O12–Cu4–N7 78.6(4), O12–Cu4–N8
92.8(4), O12–Cu4–O10 89.0(4), O12–Cu4–O7 95.0(4), N8–Cu4–
O10 176.2(4), N7–Cu4–O7 173.3(4), Cu1–O9–Cu2 109.2(5), Cu3–
O10–Cu4 109.2(5).
Although similar to that of 1, the molecular structure
of 3 has some important distinctive features. Inside each
dinuclear unit, the ligand environment is square-planar for
one CuII atom (Cu2, Cu3), whereas it is square-pyramidal
for the other CuII center (Cu1, Cu4). The ligand environ-
ment of the square-planar CuII center includes the Namino
Nimino, and Ophenoxido donor atoms from the polydentate
,
by the centrosymmetric arrangement of asymmetric units in
¯
the crystal (P1 space group) a conglomerate is not formed. ligand, and the fourth donor is the hydroxido oxygen atom.
Three intramolecular hydrogen interactions operate In addition to these four equatorial donors, the square-py-
within each bis(dinuclear) unit: N1H(amino) interacts with ramidal ligand environment of Cu1 (Cu4) includes one oxy-
O4(methoxy) and O14(perchlorato), whereas N5H(amino) gen atom from the coordinated perchlorato anion at
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5731

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
the apical coordination site with shorter bonding distances
[Cu1–O9 2.510(8), Cu4–O17 2.53(1) Å] than the
Cu–Operchlorato distances reported above for 1.
The Cu–O–Cu angle of the μ-hydroxido bridge is smaller
[Cu1–O5–Cu2 106.9(2)°, Cu3–O14–Cu4 107.0(4)°] for both
dinuclear subunits of 3 than those of 1, and the Cu1···Cu2
(Cu3···Cu4) distances [3.10(1) Å] are also slightly shorter
than the corresponding distances in 1. As a result of the
absence of bonding interactions between the independent
[Cu2(L3)2(μ-OH)(ClO4)] dinuclear molecules, the next
shortest Cu···Cu distances are longer [10.521(2) and
10.749(2) Å] than the corresponding distances in 1, which
allows us to consider that the dinuclear units of 3 are mag-
netically isolated from each other.
Figure 3. Diamond drawing of the asymmetric unit [Cu3(L4)3(μ3-
OMe)(ClO4)2] of 4. Hydrogen atoms are omitted for clarity. Ther-
mal ellipsoids are drawn at the 30% probability level. Selected bond
lengths [Å] and angles [°]: Cu1–N1 2.013(12), Cu1–N2 1.961(13),
Five intramolecular hydrogen interactions operate within
each dinuclear unit: N1H(amino) interacts with
O3(phenoxido), O4(methoxy), and O6(perchlorato),
whereas N3H(amino) interacts with O1(phenoxido) and Cu1–O1 1.891(9), Cu1–O3 2.77(2), Cu1–O4 1.965(10), Cu1–O9
2.72(2), Cu2–N3 1.986(13), Cu2–N4 1.915(11), Cu2–O1 2.372(9),
Cu2–O2 1.932(9), Cu2–O4 1.961(9), Cu2–O6 2.95(2), Cu3–N5
O2(methoxy) with N···O distances of 3.05, 3.35, 3.12, 2.98
and 3.34 Å, respectively. A similar network is present in the
1.957(12), Cu3–N6 1.919(13), Cu3–O2 1.997(8), Cu3–O3 1.884(10),
other dinuclear complex molecule of the asymmetric unit.
Cu3–O4 2.346(10); O1–Cu1–N2 93.5(4), O1–Cu1–O4 87.0(4), N2–
These dense networks with stronger intramolecular hydro-
gen interactions than those in 1 allow an understanding of
why the Cu–O–Cu angles of the μ-hydroxido bridges are
smaller and the Cu···Cu distances are slightly shorter in 3
than those in 1. The dinuclear complex molecules of neigh-
Cu1–O4 178.1(5), O1–Cu1–N1 175.3(5), O1–Cu1–O3 83.9(4), O1–
Cu1–O9 86.1(4), N1–Cu1–O3 100.6(5), N1–Cu1–O9 89.8(5), N2–
Cu1–N1 83.9(5), N2–Cu1–O3 104.0(5), N2–Cu1–O9 86.7(5), O3–
Cu1–O4 74.2(4), O3–Cu1–O9 165.7(4), O4–Cu1–N1 95.7(4), O4–
Cu1–O9 95.2(5), N4–Cu2–O2 92.9(4), N4–Cu2–O4 176.1(5), N4–
Cu2–O6 78.0(5), O2–Cu2–O4 85.5(4), O2–Cu2–O6 85.4(5), N4–
boring asymmetric units are associated in infinite zigzag Cu2–N3 84.9(5), O2–Cu2–N3 175.2(5), O4–Cu2–N3 96.4(4), N4–
Cu2–O1 108.9(4), O2–Cu2–O1 96.1(4), N3–Cu2–O6 90.0(5), O4–
chains through hydrogen interactions involving the poly-
Cu2–O1 74.9(3), O4–Cu2–O6 98.3(5), N3–Cu2–O1 88.6(4), O1–
dentate ligands, perchlorato anions, and water molecules.
Cu2–O6 172.8(4), O3–Cu3–N6 92.9(4), O3–Cu3–N5 173.3(5), N6–
Cu3–N5 83.9(5), O3–Cu3–O2 91.4(4), N6–Cu3–O2 174.6(5), N5–
Cu3–O2 91.4(4), O3–Cu3–O4 87.1(4), N6–Cu3–O4 109.0(6), N5–
[Cu3(L4)3(μ3-OMe)(ClO4)2] (4)
Cu3–O4 99.5(4), O2–Cu3–O4 74.5(3), Cu1–O1–Cu2 93.0(4), Cu2–
O2–Cu3 106.4(4), Cu2–O4–Cu1 104.7(4), Cu2–O4–Cu3 93.4(4),
Cu1–O4–Cu3 104.9(4).
The asymmetric unit of 4 is a neutral trinuclear complex
molecule, in which the three CuII centers are bridged
through the Ophenoxido donor atoms of the three L4 triden-
tate ligands and the μ3-OMe anion; the two perchlorato
anions are both coordinated to one CuII center through one
of their oxygen atoms. The copper atoms are located at
three opposite apices of a defective cubane structure, and
four of the remaining apices are occupied by the bridging
oxygen donor atoms (Figure 3). The Cu1···Cu2, Cu2···Cu3,
and Cu1···Cu3 distances [3.109(3), 3.146(3), and 3.427(3) Å,
respectively] and Cu2–Cu1–Cu3, Cu1–Cu2–Cu3, and Cu1–
Cu3–Cu2 angles [57.30(6), 66.44(6), and 56.26(6)°, respec-
tively] indicate that the Cu1Cu2Cu3 triangle is isosceles
rather than equilateral.
The pentacoordinate Cu3 atom is equatorially bonded to
the Namino, Nimino, and Ophenoxido donor atoms of one L4
tridentate ligand, and the fourth equatorial donor is the
Ophenoxido atom of another L4 ligand; the apical donor is
the μ3-OMe anion. Bond lengths and angles (caption to
Figure 3) are in the usual ranges for CuII in similar Jahn–
bonded to one Operchlorato atom but again differ in their
other apical coordination, an Ophenoxido donor atom of one
L4 tridentate ligand for Cu1 and the μ3-OMe anion for Cu2.
Both Cu1 and Cu2 are in Jahn–Teller elongated octahe-
dral ligand environments. The bond lengths and angles
(caption to Figure 3) are in the usual ranges for CuII ions
in similar Jahn–Teller-elongated coordination octahe-
dra.[1–9]
In the absence of significant hydrogen contacts, the crys-
tal packing results from π–π stacking between the
C1C2C3C4C5C6 and C34C35C36C37C38C39 phenoxido
aromatic rings of adjacent asymmetric units with a distance
of 3.98 Å and a dihedral angle of 28°.
[Cu3(L5)3(μ3-OH)(ClO4)2]·0.5(CH3)2CO·0.5(H2O) (5)
The asymmetric unit of 5 is a neutral trinuclear complex
Teller-elongated square-pyramidal ligand environments.[1–9] molecule similar to that of 4, in which the three CuII centers
Similarly, the hexacoordinate Cu1 and Cu2 are both equa- are bridged through the Ophenoxido donors of the three L5
torially bonded to the Namino, Nimino, and Ophenoxido donor tridentate ligands and the μ3-OH anion; the two perchlor-
atoms of one L4 tridentate ligand; although the fourth ato anions are loosely coordinated: Cl(1)O4 through one of
equatorial donor is the Ophenoxido atom of another L4 ligand its oxygen atoms to Cu1, Cl(2)O4 through two of its oxygen
for Cu2, at variance with Cu2 and Cu3, for Cu1 it is the atoms to Cu2 and Cu3#1 (#1: x, 1 – y, 0.5 + z) to yield
μ3-OMe anion. Cu1 and Cu2 are similarly apically loosely zigzag chains of trinuclear complex molecules that develop
5732
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740

μ3- vs. μ-Hydroxido Bridges in Copper(II) Complexes
along the b axis. The copper atoms are located at three op- chlorato anions [O100···O5 2.887(2), O100···O9 3.010(2) Å]
posite apices of a defective cubane structure; four of the and with the μ3-OH anion [O1···O100 2.58(1) Å, O1–H1–
remaining apices are occupied by the bridging oxygen do- O11 167(6)°] is the keystone of this network.
nor atoms (Figure 4). The Cu1···Cu2, Cu2···Cu3, and
Compounds referenced as “3a” in reference[8] and as “1”
Cu1···Cu3 distances [3.126(1), 3.134(1), and 3.123(1) Å, in reference[9] (P21/c) and 5 (Cc, this work) are polymorphs:
respectively] and Cu2–Cu1–Cu3, Cu1–Cu2–Cu3, and Cu1– althoughtheypossesssimilar[Cu3(L5)3(μ3-OH)]2+cations,the
Cu3–Cu2 angles [60.2(3), 59.9(3), and 59.9(3)°, respectively] (ClO4)– anions are differently located and loosely coordi-
indicate that, in contrast with that of 4, the Cu1Cu2Cu3 nated to copper centers, yielding pairs of slightly different
triangle is equilateral.
[Cu3(L5)3(μ3-OH)]2+ cations in “3a”[8] and “1”[9], but 1D
chains of identical [Cu3(L5)3(μ3-OH)]2+ cations in 5. These
polymorphs also differ in their solvents of crystallization,
seven H2O molecules for “3a”[8] and “1”[9], and 0.5 (CH3)2-
CO and 0.5 H2O molecules for 5, which result in different
hydrogen bond networks and crystal packing.
[Cu3(L6)3(μ3-OH)(ClO4)2] (6)
The asymmetric unit of 6 is a neutral trinuclear complex
molecule similar to that of 4, in which the three CuII centers
are bridged through the Ophenoxido donor atoms of the three
L6 tridentate ligands and the μ3-OH anion (Figure 5). The
Figure 4. Diamond drawing of the asymmetric unit [Cu3(L5)3(μ3-
OH)(ClO4)2] of 5. Hydrogen atoms are omitted for clarity. Thermal
ellipsoids are drawn at the 30% probability level. Selected bond
lengths [Å] and angles [°]: Cu1–N1 1.998(6), Cu1–N2 1.914(6),
Cu1–O1 1.966(5), Cu1–O2 2.400(5), Cu1–O3 1.894(5), Cu1–O7
2.808(8), Cu2–N3 1.998(6), Cu2–N4 1.916(7), Cu2–O1 1.991(5),
Cu2–O2 1.900(5), Cu2–O4 2.353(6), Cu2–O10 2.744(8), Cu3–N5
2.007(6), Cu3–N6 1.925(8), Cu3–O1 2.009(6), Cu3–O3 2.264(5),
Cu3–O4 1.900(5), Cu3–O11#2 2.87(1); Cu1–O1–Cu2 104.4(2),
Cu1–O1–Cu3 103.6(2), Cu1–O2–Cu2 92.5(2), Cu1–O3–Cu3
97.0(2), Cu2–O1–Cu3 103.2(2), Cu2–O4–Cu3 94.3(2), O3–Cu1–N2
93.5(2), O3–Cu1–O1 84.4(2), N2–Cu1–N1 84.1(3), O1–Cu1–N1
97.5(2), O3–Cu1–O2–97.1(2), N2–Cu1–O2 109.0(2), O1–Cu1–O2
75.43(18), N1–Cu1–O2 88.4(2), N2–Cu1–O1 175.4(3), O3–Cu1–N1
174.5(2), O2–Cu2–N4 93.4(3), O2–Cu2–O1 87.6(2), N4–Cu2–N3
84.3(3), O1–Cu2–N3 94.5(3), O2–Cu2–O4 98.7(2), N4–Cu2–O4
105.3(3), O1–Cu2–O4 75.57(19), N3–Cu2–O4 88.6(3), N4–Cu2–O1
178.6(3), O2–Cu2–N3 172.7(3), O4–Cu3–N6 93.1(3), N6–Cu3–N5
83.9(3), O4–Cu3–O1 86.4(2), N5–Cu3–O1 95.8(3), O4–Cu3–O3
95.4(2), N6–Cu3–O3 118.0(3), N5–Cu3–O3 89.5(2), O1–Cu3–O3
74.46(19), O4–Cu3–N5 175.0(3), N6–Cu3–O1 167.5(3). Symmetry
operation used to generate equivalent atoms #2: x, –y, 1/2 + z.
Figure 5. ORTEP drawing of the asymmetric unit [Cu3(L6)3(μ3-
OH)(ClO4)2] of 6. Hydrogen atoms are omitted for clarity. Thermal
ellipsoids are drawn at the 30% probability level. Selected bond
lengths [Å] and angles [°]: Cu1–N1 1.980(4), Cu1–N2 1.939(4),
Cu1–O1 1.907(3), Cu1–O3 2.320(3), Cu1–O4 1.996(3), Cu1–O10
3.123(9), Cu2–N3 1.985(5), Cu2–N4 1.929(5), Cu2–O1 2.385(4),
Cu2–O2 1.895(3), Cu2–O4 1.987(4), Cu2–O8 2.921(5), Cu3–N5
2.001(5), Cu3–N6 1.928(4), Cu3–O2 2.404(4), Cu3–O3 1.901(3),
Cu3–O4 1.988(3); Cu1–O1–Cu2 94.23(13), Cu2–O2–Cu3 94.24(13),
Cu2–O4–Cu1 105.68(15), Cu1–O3–Cu3 95.82(15), Cu3–O4–Cu1
104.23(14), Cu2–O4–Cu3 105.68(15), O3–Cu1–N2 109.40(15), O3–
The hexacoordinate Cu1, Cu2, and Cu3 centers are each
equatorially bonded to the Namino, Nimino, and Ophenoxido
donor atoms of one L4 tridentate ligand; the fourth equato-
rial donor is the μ3-OH anion for all three CuII centers, in Cu1–O1 97.49(14), N2–Cu1–N1 84.57(18), O1–Cu1–N1
171.83(17), O3–Cu1–O4 74.71(12), N2–Cu1–O4 175.82(16), O1–
Cu1–O4 85.59(14), N1–Cu1–O4 96.30(16), N2–Cu1–O1 92.99(17),
O3–Cu1–N1 90.68(16), O2–Cu2–N4 94.8(2), O2–Cu2–O1
96.67(13), N4–Cu2–N3 83.7(2), O1–Cu2–N3 91.50(18), O2–Cu2–
contrast with 4. Cu1, Cu2, and Cu3 are similarly apically
loosely bonded to one Operchlorato and an Ophenoxido donor
atom from one L4 tridentate ligand; this again is at variance
with 4. All three CuII centers are in Jahn–Teller-elongated
O4 86.19(13), N4–Cu2–O4 178.6(2), O1–Cu2–O4 74.04(12), N3–
Cu2–O4
95.16(18),
N4–Cu2–O1
106.83(19),
O2–Cu2–
octahedral ligand environments.
N3 172.79(19), O4–Cu3–N6 178.35(16), N6–Cu3–N5 85.07(19),
O4–Cu3–O2 73.69(12), N5–Cu3–O2 90.49(16), O4–Cu3–O3
85.14(13), N6–Cu3–O3 93.31(17), N5–Cu3–O3 167.05(17), O2–
Cu3–O3 102.24(14), O4–Cu3–N5 96.32(16), N6–Cu3–O2
Several intramolecular hydrogen interactions stabilize
this trinuclear molecular structure as well as the 1D chain
arrangement of trinuclear complex molecules. The water
molecule of crystallization, which interacts with both per- 107.24(16).
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5733

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
Cu1···Cu2, Cu2···Cu3, and Cu1···Cu3 distances [3.162(1),
3.170(2), and 3.144(1) Å, respectively] and Cu2–Cu1–Cu3,
Cu1–Cu2–Cu3, and Cu1–Cu3–Cu2 angles [60.34(3),
59.56(3), and 60.10(3)°, respectively] indicate that the Cu1-
Cu2Cu3 triangle is close to equilateral. The Cu1, Cu2, and
Cu3 centers are each equatorially bonded to the Namino
,
N
imino, and Ophenoxido donor atoms of one L6 tridentate li-
gand; the fourth equatorial donor is the μ3-OH anion for
all three CuII centers, in line with 5, but at variance with 4.
Cu1 and Cu2 are similarly apically loosely bonded to one
O
perchlorato atom and an Ophenoxido donor atom from one L6
tridentate ligand and may be considered as hexacoordinate,
whereas Cu3 is pentacoordinate with the Ophenoxido donor
atom from the third L6 tridentate ligand in the apical loca-
tion.
In addition to the two weak Cu···Operchlorato bonds and
numerous van der Waals interactions, two hydrogen con-
tacts constitute the basis of the crystal packing in 6:
N1(H)···O9#1 3.07(1) Å, O4(H)···O11#1 2.94(2) Å (#1: 1 –
x, 1 – y, 1 – z).
Although similar to the ligand environments for the cop-
per centers in 4, the three CuII coordination spheres in 6
are different from both those in 4 and 5. Cu1 and Cu2 are
in Jahn–Teller-elongated octahedral ligand environments,
whereas Cu3 is in a Jahn–Teller-elongated square-pyrami-
dal ligand environment.
Figure 6. χM vs.
T and χMT vs. T
data for [{Cu2(L1)2(μ-
OH)}(ClO4){Cu2(L1)2(μ-OH)}](ClO4) (1; bottom), and [{Cu2-
(L3)2(μ-OH)(ClO4)}2] (3; top). The squares and triangles are mea-
sured data, and the lines are the best fit based on the Hamiltonian
described in the text.
Magnetic Properties
The magnetic susceptibility of all six compounds has for through the mean-field Hamiltonian[50] H = –2zjЈ͗Sz͘Sz.
been measured in the 2–300 K temperature range under an The parameter values obtained for the best fits shown in
applied magnetic field of 0.1 T. Considering the distances Figure 6 are J = –57 cm–1, zjЈ = –0.1 cm–1, paramagnetic
(6.7–7.4 Å) and nature of the bridges (–O–Cl–O–) between impurity (par) = 0.6% for g = 2.075, agreement factor R
the dinuclear units of its bis(dinuclear) moiety, we have con- {[(χMT)obs – (χMT)calc]2/[(χMT)obs]2} = 9ϫ10–5 (1) and J =
sidered the dinuclear units of 1 as magnetically isolated (as 16 cm–1, zjЈ = –0.1 cm–1, g = 2.120, R = 4ϫ10–5 (3). In line
subsequently confirmed, see below). The magnetic behavior with the crystal structure, the minute zjЈ value obtained for
of 1 may then be compared to those of the dinuclear com- 1 confirms that its dinuclear units are magnetically isolated.
pounds 2 and 3. The χMT products (χM is the molar mag- The crystal structure determination of 3 has evidenced di-
netic susceptibility and T the temperature at which χ was nuclear units loosely bridged into 1D chains through hydro-
measured) are 0.677, 0.163, and 0.061 (1), 0.868, 0.855, and gen interactions involving the polydentate ligands, perchlor-
0.859 (2), and 0.894, 0.990, and 0.975 cm3 mol–1 K (3) at ato anions and water molecules, which are unsuitable for
300, 50, and 2 K, respectively. In addition, a maximum the transmission of sizeable magnetic interactions. It is thus
value of 1.093 cm3 mol–1 K is observed for 3 at 10 K. It fol- highly unlikely that the quite large ferromagnetic interac-
lows that, although an identical unique μ-OH bridge links tion (J = 16 cm–1) obtained for 3 arises from 1D long range
the two CuII ions in the three compounds, they have dif- interactions. The minute intermolecular antiferromagnetic
ferent magnetic behavior: whereas 2 is paramagnetic, anti- interaction (zjЈ = –0.1 cm–1) may much better reflect 1D
ferromagnetic exchange interactions prevail in 1, and ferro- long-range interactions mediated through the perchlorato
ˆ
ˆ
magnetic ones prevail in 3 (Figure 6).
bridges.
The theoretical magnetic susceptibility for two magneti-
Compounds 1–3 belong to the series of compounds in
cally coupled S = 1/2 spins was computed through exact which two CuII centers are bridged through a single RO–
calculation of the energy levels associated with the appro- oxygen atom (R = H, alkyl, phenyl).[22–34] The magnetic
priate spin Hamiltonian through diagonalization of the full- behavior of these compounds is considered to be governed
matrix with a general program[48] and fitted by least-squares by the Cu–O–Cu angle (θ). The superexchange interaction
techniques[49] to the sets of experimental magnetic data ob- is ferromagnetic for θ close to 90°, at which point the anti-
tained for 1 and 3. The zero-field part of the Hamiltonian ferromagnetic contribution vanishes, and it is antiferromag-
including Zeeman and isotropic exchange terms was netic for θ values large enough to allow domination of the
–2JS1S2, and intermolecular interactions were accounted ferromagnetic contribution by the antiferromagnetic
5734
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740

μ3- vs. μ-Hydroxido Bridges in Copper(II) Complexes
Table 1. Selected structural and magnetic parameters of CuII dinuclear compounds with a unique equatorial μ-OH bridge.
Compound[ref]
θ [°]
Cu···Cu [Å]
φ [°]
J[a] [cm–1]
3
106.8
124.9
109.2
101.3
102.9
100.1
101.8
141.6
113.8
125.7
145.7
141.6
131.1
113.8
137.9
115.3
115.6
128.1
119.3
126.2
132.2
116.4
118.9
143.7
3.10
3.420
3.14
102
–
113
–
–
–
–
–
–
–
–
–
–
–
–
–
17
0
–57
–93
[Cu2L7(OH)Br3(H2O)]·H2O[30][b]
1
[Cu2L8(OH)Br3]·1.5H2O[26][c]
3.010
2.986
3.001
2.972
3.570
3.165
3.370
3.642
3.645
3.437
3.165
3.663
3.138
3.211
3.435
3.338
3.425
3.384
3.458
3.413
3.642
–93
[Cu2L8(OH)Cl3]·1.5H2O[26][c]
–100.5
–100.5
–120
–145
–148
–151.5
–161
–167
–167.5
–180
–248.5
–266
–374
–375
–400
–410
–449
–457.5
–500
[Cu2(L9)(OH)(ClO4)(OH2)](ClO4)2
[Cu2L8(OH)IO3]·4H2O[27][c]
[25][d]
K[Cu2(L102)(OH)·2H2O[34][e]
[Cu2(terpy)2(OH)(OH2)(ClO4)3][31][f]
[22][g]
[Cu2(bpy)2(OH)](ClO4)3
Na[Cu2(L102)(OH)·2H2O[34][e]
[Cu2L8(OH)IO3]·4H2O[26][c]
[Cu2(dpm)2(L11)(OH)](ClO4)3
[33][h]
[28][i]
[Cu2L12(OH)(H2O)2(NO3)2]NO3
[Cu2L8(OH)ClSO4]·2H2O[28][c]
–
[Cu2(dien)2(OH)(ClO4)3][32][j]
123.3
–
–
–
–
–
–
[Cu2L13(OH)(H2O)(NO3)2]NO3·H2O[29][k]
[Cu2L7(OH)Cl3(H2O)]·H2O[27][b]
[23][l]
[Cu2(L14)(OH)](BF4)3
[Cu2L15(OH)Cl3(H2O)]·0.8H2O[30][m]
[Cu2L15(OH)Br3(H2O)]·0.6H2O[30][m]
[Cu2(L16)(OH)(ClO4)](ClO4)2
[24][n]
[a] –2JS1S2 formalism. [b] L7 = 1,4-di(1Ј-methylimidazol-2-yl)phthalazine. [c] L8 = 1,4-di(2Ј-pyridylamino)phthalazine. [d] L9 = 1,13-
dioxa-4,7,10,16,19,29-tetraazacyclotetracosane. [e] L10 = 2,6-bis(N-phenylcarbamoyl)pyridine. [f] terpy = 2,2Ј;6Ј,2"-terpyridine. [g] bpy =
2,2Ј-bipyridine. [h] dpm = di(2-pyridyl)methane, L11 = 1,1,2,2-tetrakis(2-pyridyl)ethylene. [i] L12 = 1,4-di(4Ј-methyl-2Ј-pyridylamino)-
phthalazine. [j] dien = diethylenetriamine. [k] L13 = 3,6-bis(3,5-dimethylpyrazol-1-yl)pyridazine. [l] L14 = 1,4-bis[(1-oxa-4,10-dithia-7-
azacyclododecan-7-yl)methyl]benzene. [m] L15 = 3,6-bis(pyrazol-1-yl)pyridazine. [n] L16 = decadentate N6O4 macrocyclic ligand from the
reaction of 2,6-diacetylpyridine with 3,6-dioxaoctane-1,8-diamine.
one.[18,51,52] The examples of overall ferromagnetic ex- and 0.378 cm3 mol–1 K (6). The χMT values at 2 K suggest
change interaction are scarce for the series of compounds an ST value of 1/2, and the overall χMT variation with T
in which two CuII centers are bridged through a single RO– suggests the operation of dominant antiferromagnetic inter-
oxygen atom (R = alkyl, phenyl),[51,52] and to the best of actions for all three compounds (Figure 6). Consequently,
our knowledge there is no previous example of ferromag- we have computed the theoretical magnetic susceptibility
netic exchange interaction for a dinuclear CuII compound for three exchange-coupled CuII centers through exact cal-
with a single μ-OH bridge. Table 1 collates the J values and culation of the energy levels associated with the appropriate
structural parameters that may play a significant role in the spin Hamiltonian through diagonalization of the full-ma-
magnetic behavior of dinuclear CuII compounds with a sin- trix with a general program[48] and fitted by least-squares
gle equatorial μ-OH bridge, namely, the θ angle, Cu···Cu techniques[49] to the sets of experimental magnetic data ob-
distance, and φ angle between the mean coordination planes tained for 4, 5, and 6. The zero-field part of the Hamilto-
of the Cu centers that do not include the OH bridge.
nian including Zeeman and isotropic exchange terms was
Among the μ-OH bridged dinuclear CuII compounds –2J (S1·S2 + S2·S3 + S1·S3), and intermolecular interactions
collated in Table 1, complex 3 is the only one characterized were taken into account through the mean-field Hamilto-
by a ferromagnetic exchange interaction. A comparison of nian[50] H = –2zjЈ͗Sz͘Sz. The parameter values obtained for
the structural parameters of 1 and 3 shows that a small the best fits shown in Figure 7 are: J = –16.9 cm–1, zjЈ =
decrease in the Cu–O(H)–Cu angle (θ), 106.8 vs. 109.2°, 0 cm–1, paramagnetic impurity (par) = 0%, g = 2.044,
Cu···Cu distance, 3.10 vs. 3.14 Å, and φ angle, 102 vs. 113°, agreement factor R {[(χMT)obs – (χMT)calc]2/[(χMT)obs]2} =
is sufficient to switch the exchange interaction from antifer- 3ϫ10–4 (4), J = –6.3 cm–1, zjЈ = –0.15 cm–1, par = 0%, g =
romagnetic to ferromagnetic, 17 vs. –57 cm–1. This trend is 2.095, R = 2ϫ10–4 (5), and J = –6.7 cm–1, zjЈ = –0.08 cm–1,
confirmed when the comparison is extended to [Cu2- par = 0%, g = 2.05, R = 3.3ϫ10–4 (6).
ˆ
ˆ
(dien)2(OH)(ClO4)3].[32] However, the lack of φ angle values
In all three compounds, the three CuII centers are doubly
for all other examples precludes a more conclusive analysis. bridged pairwise through one hydroxido (methoxido) and
The magnetic susceptibilities of 4–6 have been measured one phenoxido oxygen atom. However, in view of the longer
in the 300–2 K temperature range under an applied mag- bonds and smaller bond angles of the asymmetric
netic field of 0.1 T. The χMT products of the trinuclear com- Cu–Ophenoxido–Cu bridges compared to those of the Cu–
pounds at 300, 50, and 2 K, respectively, are 1.091, 0.684,
Ohydroxido–Cu or Cu–Omethoxido–Cu ones, superexchange in-
and 0.412 (4), 1.191, 1.022, and 0.363 (5), and 1.111, 0.972, teractions are essentially mediated through the hydroxido
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5735

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
the Cu3O(H) core from a regular tetrahedron in 4 compared
to those of 5 and 6, the antiferromagnetic interaction J is
larger in 4 (–16.9 cm–1) than in 5 (–6.3 cm–1) and 6
(–6.7 cm–1). The similar Cu1···Cu2, Cu2···Cu3, and
Cu1···Cu3 distances in 5 (3.126, 3.134, and 3.123 Å, respec-
tively) and 6 (3.162, 3.170, and 3.144 Å, respectively) advo-
cate for three similar pairwise magnetic interactions in both
compounds. This trend is corroborated by the similar Cu–
O(H)–Cu angles in 5 (104.4, 103.6, 103.2°) and 6 (105.68,
104.23, 105.68°). Both observations support weighting the
three pairwise magnetic interactions with a common J value
in 5 and 6.
In the case of 4, the Cu1···Cu2, Cu2···Cu3, and
Cu1···Cu3 distances (3.109, 3.146, and 3.427 Å, respec-
tively) would advocate for a slightly less antiferromagnetic
Cu1···Cu3 interaction compared to the Cu1···Cu2 and
Cu2···Cu3 ones. On the other hand, the Cu1–O4–Cu2,
Cu1–O4–Cu3, and Cu2–O4–Cu3 angles, 104.7, 104.9, and
93.4°, respectively, would advocate for a more antiferro-
magnetic Cu2···Cu3 interaction compared to the Cu1···Cu2
and Cu1···Cu3 ones. A combination of these effects would
result in a Cu2···Cu3 Ͼ Cu1···Cu2 Ն Cu1···Cu3 trend in
antiferromagnetic interactions. However, considering the
possibility of two different J values led to J(Cu1···Cu2,
Cu2···Cu3) = –14.4 cm–1 and JЈ(Cu1···Cu3) = –22.5 cm–1,
zjЈ = 0 cm–1, par = 1%, g = 2.04, agreement factor R =
1.3ϫ10–4. The fit is thus not significantly improved, a
JЈ(Cu1···Cu3) more antiferromagnetic than J(Cu1···Cu2,
Cu2···Cu3) does not seem to bear much physical meaning,
and an additional parameter is involved (par = 1%) which
suggests overparametrization. We thus consider that the
three pairwise magnetic interactions in 4 are satisfactorily
Figure 7. χM vs. T and χMT vs. T data for [Cu3(L4)3(μ3-OMe)-
(ClO4)2], 4, bottom, [Cu3(L5)3(μ3-OH)(ClO4)2]·0.5{(CH3)2CO}·
0.5(H2O), 5, middle, and [Cu3(L6)3(μ3-OH)(ClO4)2], 6, top. The
squares and triangles are measured data and the lines are the best
fit based on the Hamiltonian described in the text.
(methoxido) oxygen bridge of the Cu3O(R) core.[53] The pa- taken into account with a unique exchange integral J.
rameters considered as determining the nature and effec- The χMT product of the trinuclear compounds at 2 K,
tiveness of such magnetic interactions are the Cu–O(R) dis- 0.412 (4), 0.363 (5), and 0.378 cm3 mol–1 K (6) is either
tances, the Cu–O(R)–Cu angles, the geometry of the slightly above (4) or slightly below (5, 6) the 0.4 cm3 mol–1
Cu3O(R) core, and the angles between the equatorial coor- value expected for an S = 1/2 ground state. The marginally
dination planes of the copper centers. lower than expected values obtained for 5 and 6 seem to
K
In all three trinuclear compounds, the Cu3O(R) core is result from weak intermolecular interactions, in line with
approximately tetrahedral with the equatorial coordination the structural features and as suggested by the zjЈ values
plane of each copper atom virtually perpendicular to those from the best fits, –0.15 cm–1 (5) and –0.08 cm–1 (6). On the
of the other two copper centers, and the ligand environment other hand, the zjЈ = 0 cm–1 value resulting from the best
of each copper atom is either Jahn–Teller-elongated square- fit obtained for 4 indicates the absence of intermolecular
pyramidal (Cu3 in 4 and 6) or Jahn–Teller-elongated octa- interactions in line with the structural features and in agree-
hedral (Cu1 and Cu2 in 4 and 6, and Cu1, Cu2, and Cu3 ment with the 0.412 cm3 mol–1 K experimental value mea-
in 5). In the case of a strictly tetrahedral Cu3O(H) core, the sured for the χMT product of 4 at 2 K. Satisfactorily fitting
orthogonal sp3 functions lead to x2 – y2 ʈ sp3 Ќ sp3 ʈ x2 – the magnetic data of 4, 5, and 6 with an isotropic Hamilto-
y2 Cu–O–Cu superexchange pathways (ʈ and Ќ stand for nian suggests that antisymmetric exchange (AE)[17] is either
overlap and orthogonality, respectively) and according to ineffective or not discernible in the case of this type of tri-
Anderson’s theory[54] the resulting superexchange interac- nuclear copper compound, in line with previous re-
tions would be ferromagnetic or weakly antiferromagnetic. ports.[4,5,7–9,17,47]
The J values obtained by fitting the experimental magnetic
Comparisons between trinuclear CuII complexes includ-
data collected for 4–6 are thus in qualitative agreement with ing the Cu3O(R) core have been drawn in view of evidenc-
this rationale inasmuch as the average Cu–O–Cu angles ing magnetostructural correlations.[4–9] Unfortunately, no
(101.0, 103.7, and 105.2° for 4, 5, and 6, respectively) imply clear-cut relationship could be established with any of the
a larger s weight in the sp3 hybrid orbitals, which prevents considered structural parameters. It seems that an intricate
a strict sp3–sp3 orthogonality and thus favors weak antifer- interplay of several structural and electronic parameters op-
romagnetic interactions. In line with a larger departure of erates that does not allow extension of the comparisons to
5736
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740

μ3- vs. μ-Hydroxido Bridges in Copper(II) Complexes
different series of compounds.[6] To test this hypothesis, we
consider in the following the unique series of trinuclear CuII
complexes that include the Cu3O(R) core and are based on
Conclusions
We have reported the synthesis, characterization, crystal
tridentate Schiff bases that result from the condensation of structure, and magnetic properties of three monohydroxido-
diamines with salicylaldehydes. Table 2 collates the Cu– bridged dinuclear CuII complexes (1–3) and three hy-
O(R), Cu–O(R)–Cu, and Cu···Cu structural data together droxido-bridged trinuclear CuII complexes 4–6. Regardless
with the J values resulting from fitting the experimental of their nuclearity, all six complexes were obtained through
magnetic data for the seven presently known complexes in the same experimental procedure from the corresponding
the above-mentioned series (“3a”[8] and “1”[9] are two struc- Cu(Li)(py)(ClO4) mononuclear precursor suspended in
tural determinations of the same P21/c polymorph, but 5 is methanol; the presence of a significant concentration of
a Cc polymorph of the same complex molecule [Cu3- HO– (MeO–) anions in the slurry (from added triethyl-
(L5)3(μ3-OH)(ClO4)2]). The distance of O(R) to the Cu1,C- amine) allows an aggregation process. The nuclearity of the
u2,Cu3 plane depends on Cu–O(R), Cu–O(R)–Cu, and resulting complex depends on the steric requirement of Li:
Cu···Cu, and is thus not considered as an independent pa- in the presence of an OMe substituent α to the Ophenoxido
rameter. In all eight structures, the three metallacyclic units atom (L1–L3), μ-HO– single-bridged dinuclear compounds
that constitute the trinuclear complex cation are virtually are obtained (1–3); on the other hand, in the absence
perpendicular with respect to each other, and the dihedral of an α-substituent (L4–L6), the aggregation proceeds
angles between the least-squares planes of the six-mem- to the μ3-OH– (μ3-OMe–) bridged trinuclear compounds
bered rings that contain the Cu atoms are thus not included 4–6.
in Table 2.
Antiferromagnetic interactions operate in 1, whereas 2 is
paramagnetic. Compound 3 is the first example of a ferro-
magnetic monohydroxido-bridged dinuclear CuII complex.
A comparison of the usually considered structural [Cu–
O(H)–Cu angle, θ, and Cu···Cu distance, d] and magnetic
(J) parameters for the 22 μ-HO– single-bridged dinuclear
CuII compounds reported to date does not evidence any
clear magnetostructural correlation (J/θ combined or not
with J/d). From the three CuII compounds for which it has
been reported, φ, the angle between the mean coordination
planes of the Cu centers that do not include the OH bridge,
seems to be an additional parameter required for establish-
ing a magnetostructural correlation in this series of dinu-
clear CuII compounds.
Table 2. Selected structural and magnetic parameters of trinuclear
CuII compounds with the Cu3O(R) core and based on tridentate
Schiff bases resulting from condensation of diamines with salicylal-
dehydes.
Com-
Cu–O [Å] Cu–O–Cu [°] Cu···Cu [Å] J[a] [cm–1]
pound[ref.]
“2”[3]
“3A”[9]
“3B”[9]
“1”[9]
“2”[9]
5
2.078
2.024
2.039
1.985
2.006
1.889
1.985
1.996
1.990
2.091
101.4
101.3
99.4
105.4
103.5
103.5
105.5
105.1
105.2
101.0
3.210
–
–
3.153
3.150
3.128
3.155
3.178
3.165
3.227
7.8
5.8 or 2.8
2.8 or 5.8
–2.3
–3.9
–6.3
–6.5
–10.1
–6.7
“3a”[8]
“3c”[8]
6
The formulation of the [Cu3(L5)3(μ3-OH)]2+ cations in 5
is identical to those of the complex cations in “3a”[8] and
“1”[9], and all three structures include (ClO4)– anions. How-
ever, 5 (this work) crystallizes in the Cc space group,
whereas “3a”[8] and “1”[9] belong to the P21/c space group,
which leads to the conclusion that these three complexes are
polymorphs of the same trinuclear CuII chemical species.
Indeed, the (ClO4)– anions are differently located and
4
–16.9
[a] –2JS1S2 formalism. Cu–O: average values of all Cu–O(H) dis-
tances; Cu–O–Cu: average values of all Cu–O(H)–Cu angles;
Cu···Cu: average values of all Cu···Cu distances.
The most significant observations resulting from exami- loosely coordinated to the copper centers, yielding pairs of
nation of Table 2 include: (a) “2”[3], “3A,B”[9] (ferromag- slightly different [Cu3(L5)3(μ3-OH)]2+ cations in “3a”[8] and
netic), and 4 (antiferromagnetic) have opposite magnetic “1”[9], but 1D chains of identical [Cu3(L5)3(μ3-OH)]2+ cat-
behavior in spite of their structurally similar Cu3O(R) cores, ions in 5.
(b) the “3a”[8] and 5 polymorphs have very similar (if not
Antiferromagnetic interactions operate in all three
identical) magnetic behavior in spite of structurally different monohydroxido-bridged trinuclear CuII complexes 4, 5, and
Cu3O(R) cores, (c) “3c”[8] and 6 have structurally similar 6. In line with a larger departure of the Cu3O(R) core from
Cu3O(R) cores but significantly different magnetic behav- a regular tetrahedron in 4 compared to 5 and 6, the antifer-
ior. Even if we take into account that the Cu3O(R) cores romagnetic interaction J is larger in 4 (–16.9 cm–1) than in 5
differ in the nature of R (hydroxido for “2”[3] and methox- (–6.3 cm–1) and 6 (–6.7 cm–1). Comparison of the structural
ido for 4) it should be concluded as a whole that magnetic [Cu–O(R), Cu–O(R)–Cu and Cu···Cu] and magnetic (J) pa-
properties for the compounds in this series depend not only rameters for the seven μ3-hydroxido (methoxido) bridged
on structural characteristics, but also on subtle differences trinuclear CuII compounds presently known in the series
in the electronic properties among the Cu3O(R) cores, prob- based on tridentate Schiff bases that result from the con-
ably as a result of their belonging to different [Cu3(L)3(μ3- densation of diamines with salicylaldehydes does not evi-
OR)]2+ complex cations, not to mention the role of complex dence any clear magnetostructural correlation. It should be
counteranions and crystal packing.
concluded that magnetic properties for the compounds in
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5737

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
C 42.86, H 5.96, N 7.21. IR (KBr): ν = 3498, 3300, 3241, 2934,
˜
this series depend not only on structural characteristics, but
also on subtle differences in the electronic properties among
the Cu3O(R) cores.
2854, 1636, 1602, 1546, 1472, 1450, 1325, 1239, 1219, 1107, 974,
750, 622 cm–1. Crystallization from a methanol/acetone mixture
yielded dark red to purple crystals of 1 within one week,
[{Cu2(L1)2(μ-OH)}(ClO4){Cu2(L1)2(μ-OH)}](ClO4)·5.5H2O.
Cu2(L2)2(OH)(ClO4)(H2O)2
C20H31ClCu2N4O11 (666.03): calcd. C 36.07, H 4.69, N 8.41; found
(2):
Yield
0.190 g
(57%).
Experimental Section
C 36.32, H 4.92, N 8.15. IR (KBr): ν = 3568, 3302, 2960, 1637,
˜
Materials: Solvents and chemicals were of reagent grade and used
without further purification.
1602, 1544, 1471, 1448, 1320, 1240, 1117, 1077, 748, 621 cm–1.
Cu2(L3)2(OH)(ClO4)·2H2O
C24H39ClCu2N4O11 (722.14): calcd. C 39.92, H 5.44, N 7.76; found
(3):
Yield
0.202 g
(56%).
Syntheses: Mononuclear Cu(Li)(py)(ClO4) precursor complexes (i
= 1–6) were prepared according to literature methods[19–21] and by
applying the general experimental procedure described below.
C 39.67, H 5.48, N 7.71. IR (KBr): ν = 3560, 3297, 3149, 2957,
˜
1637, 1602, 1544, 1471, 1447, 1315, 1240, 1221, 1119, 1079, 1056,
973, 747, 623 cm–1. Crystallization from a methanol/acetone mix-
ture yielded dark red to purple crystals of 3 within one week,
[{Cu2(L3)2(μ-OH)(ClO4)}2]·4H2O.
Caution! Due to their explosive character, perchlorate salts and com-
plexes should be handled with care and in very small amounts.
To a solution of salicylaldehyde (2 mmol, 0.25 g) or orthovanillin
(2 mmol, 0.30 g) in methanol (20 mL), were added copper perchlo-
rate (2 mmol, 0.74 g) dissolved in water (3 mL), an excess of pyr-
idine (4 mmol, 0.32 g), and the appropriate diamine (2 mmol,
0.12 g for ethylenediamine, 0.15 g for 1,2-diaminopropane, 0.18 g
for 2-methylpropane-1,2-diamine, and 0.23 g for 1,2-diaminocyclo-
hexane) dissolved in methanol (5 mL). The reaction mixture was
stirred for 2 h to yield an abundant precipitate. The precipitate was
collected by filtration and washed with cold methanol and diethyl
ether.
Cu3(L4)3(OMe)(ClO4)2
(4):
Yield
0.197 g
(55%).
C40H54Cl2Cu3N6O12 (1072.45): calcd. C 44.80, H 5.08, N 7.84;
found C 44.55, H 5.26, N 7.79. IR (KBr): ν˜ = 3314, 3260, 2932,
2858, 1640, 1601, 1448, 1088, 910, 755, 620 cm–1. Slow concentra-
tion of the filtrate (2 d) yielded crystals of 4.
[Cu3(L5)3(μ3-OH)(ClO4)2]·(H2O):
C27H36Cl2Cu3N6O13 (914.16): calcd. C 35.48, H 3.97, N 9.19;
Yield
0.228 g
(75%).
found C 35.28, H 3.92, N 9.05. IR (KBr): ν = 3589, 3336, 3279,
˜
1639, 1600, 1540, 1467, 1448, 1310, 1149, 1088, 760, 625 cm–1.
Crystallization from a methanol/acetone mixture yielded dark
Cu(L1)(py)(ClO4): Yield 80%. C19H24ClCuN3O6 (489.37): calcd. C
green crystals of
5
within 2 d, [Cu3(L5)3(μ3-OH)(ClO4)2]·
46.63, H 4.94, N 8.59; found C 46.38, H 4.74, N 8.47. IR (KBr): ν
˜
0.5{(CH3)2CO}·0.5(H2O).
= 3294, 2931, 1637, 1601, 1473, 1446, 1321, 1247, 1223, 1094, 974,
699, 623 cm–1.
Cu3(L6)3(OH)
(ClO4)2
C30H40Cl2Cu3N6O12 (938.23): calcd. C 38.41, H 4.30, N 8.96;
(6):
Yield
0.144 g
(46%).
Cu(L2)(py)(ClO4): Yield 90%. C15H18ClCuN3O6 (435.27): calcd. C
found C 38.53, H 3.99, N 8.99. IR (KBr): ν = 3541, 3333, 3270,
˜
41.39, H 4.17, N 9.65; found C 41.34, H 4.23, N 9.62. IR (KBr): ν
˜
1632, 1602, 1538, 1445, 1309, 1195, 1071, 1028, 904, 755, 620 cm–1.
= 3320, 3259, 1641, 1605, 1446, 1295, 1248, 1227, 1098, 1082, 747,
700, 621 cm–1.
Slow concentration (4 d) of the filtrate yielded crystals of 6.
Physical Measurements: C, H, and N elemental analyses were per-
formed with a Perkin–Elmer 2400 series II device at the microana-
lytical laboratory of the Laboratoire de Chimie de Coordination in
Toulouse, France. Infrared spectra were recorded at room tempera-
ture with a 9800 FTIR spectrometer (Perkin–Elmer) with samples
as KBr pellets (%T). Magnetic susceptibilities were measured in
the 2–300 K temperature range at a sweeping rate of 2 Kmin–1 un-
der an applied magnetic field of 0.1 T with an MPMS5 Quantum
Design superconducting quantum interference device (SQUID) su-
sceptometer. The apparatus was calibrated with palladium metal.
All samples were 3 mm diameter pellets molded from ground crys-
talline samples. Diamagnetic corrections were applied by using Pas-
cal’s constants.[55] The magnetic susceptibilities were computed by
exact calculations of the energy levels associated with the spin
Hamiltonian through diagonalization of the full-matrix with a ge-
neral program.[48] Least-squares fittings were accomplished with an
adapted version of the function-minimization program MIN-
UIT.[49]
Cu(L3)(py)(ClO4): Yield 88%. C17H22ClCuN3O6 (463.33): calcd. C
44.07, H 4.79, N 9.07; found C 43.96, H 4.98, N 8.77. IR (KBr): ν =
˜
3280, 1638, 1597, 1442, 1290, 1243, 1216, 1070, 988, 697, 622 cm–1.
Cu(L4)(py)(ClO4): Yield 85%. C18H22ClCuN3O5 (459.34): calcd. C
47.07, H 4.83, N 9.15; found C 46.96, H 5.04, N 9.13. IR (KBr): ν
˜
= 3320, 3263, 2934, 2855, 1638, 1608, 1540, 1450, 1343, 1150, 1070,
910, 764, 691, 621 cm–1.
Cu(L5)(py)(ClO4): Yield 89%. C14H16ClCuN3O5 (405.25): calcd. C
41.49, H 3.98, N 10.37; found C 41.47, H 4.04, N 10.38. IR (KBr):
ν = 3323, 3111, 1631, 1607, 1446, 1091, 1050, 774, 755, 695,
˜
619 cm–1.
Cu(L6)(py)(ClO4): Yield 82%. C15H18ClCuN3O5 (419.27): calcd. C
42.97, H 4.33, N 10.02; found C 42.92, H 3.78, N 9.92. IR (KBr):
ν = 3294, 3243, 1635, 1605, 1533, 1453, 1445, 1326, 1078, 1063,
˜
758, 690, 620 cm–1.
Dinuclear complexes 1–3 and trinuclear complexes 4–6 were pre-
pared through by the general experimental procedure described be-
low.
Crystallographic Data Collection and Structure Determination:
Crystals of 1 and 3–6 were kept in the mother liquor until they
were dipped into oil. The selected crystals, on a Mitegen micro-
mount and quickly cooled to 180 K, were mounted on a Stoe Im-
aging Plate Diffractometer System (IPDS) using a graphite mono-
chromator (λ = 0.71073 Å; 1, dark purple, 0.18ϫ0.18ϫ0.08 mm;
To Cu(Li)(py)(ClO4) (1 mmol) suspended in MeOH (20 mL) was
added NEt3 (3 mmol, 0.22 g). The mixture was stirred at room tem-
perature for 2 h. A dark precipitate, blackish for 1–3, dark green
for 4–6, was collected by filtration and washed with methanol and
diethyl ether.
3,
25ϫ0.125ϫ0.10 mm), or an Oxford Diffraction XCALIBUR sys-
(55%). tem using a graphite monochromator (λ = 0.71073 Å; 5, dark
C28H44ClCu2N4O11.5 (783.22): calcd. C 42.94, H 5.7, N 7.15; found green, 0.40ϫ0.25ϫ0.07 mm), or an Oxford Diffraction GEMINI
dark
purple,
0.25ϫ0.20ϫ0.2 mm;
4,
green,
0.
Cu2(L1)2(OH)(ClO4)·0.5H2O:
Yield
0.215 g
5738
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740

μ3- vs. μ-Hydroxido Bridges in Copper(II) Complexes
Table 3. X-ray crystallographic data for complexes 1 and 3–6.
1
3
4
5
6
Empirical formula C28H44.5ClCu2N4O11.75 C24H34ClCu2N4O11 C40H54Cl2Cu3N6O12 C28.5H38Cl2Cu3N6O13 C30H40Cl2Cu3N6O12
Formula mass
Space group
a [Å]
b [Å]
c [°]
α [°]
β [°]
787.73
P1
717.08
P1
1072.41
P21/c
11.587(2)
30.420(3)
13.018(2)
90
100.86(2)
90
4507(1)
4
934.17
Cc
13.739(2)
23.896(2)
12.195(2)
90
104.54(2)
90
3875.4(7)
4
938.23
P21/c
12.013(5)
16.880(5)
22.183(5)
90
122.195(13)
90
3807(2)
4
¯
¯
14.413(5)
15.883(5)
17.089(5)
89.712(5)
88.545(5)
63.396(5)
3496.6(19)
4
12.705(2)
13.060(2)
22.032(2)
95.428(2)
101.603(2)
118.436(2)
3071.0(5)
4
γ [°]
V [Å3]
Z
T [K]
180
0.71073
1.496
180
0.71073
1.551
180
0.71073
1.581
180
0.71073
1.601
180
1.5418
1.637
λ [Å]
ρcalcd. [gcm–1]
μ [mm–1]
1.355
1.533
1.589
1.836
3.814
[a]
R1
0.0689[c]
0.0787[c]
0.1814
0.0810[d]
0.2201[d]
0.1195
0.1243[d]
0.2882[d]
0.2629
0.3637
0.0686[d]
0.1351[d]
0.1566
0.1659
0.0483[d]
0.1408[d]
0.073
[b]
wR2
R1(all data)[a]
wR2(all data)[b]
0.1303
0.2513
0.151
2
2
[a] R = Σ||Fo| – |Fc||/Σ|Fo|. [b] wR = [Σw(|Fo | – |Fc |)2/Σw|Fo2|2]1/2. [c] IϾ2.3σ(I). [d] IϾ2σ(I).
system using Cu-Kα radiation (λ
=
1.5418; 6, green,
Acknowledgments
0.16ϫ0.12ϫ0.04 mm). The diffractometers were equipped with an
Oxford Cryosystems cooler device, or a Cryojet cooler device from
Oxford Instruments. The data were collected at 180 K. The unit
cell determination and data integration were carried out using the
Xred,[56] CrysAlis RED,[57] and CrysAlisPro[58] packages for the
data recorded on the IPDS, Xcalibur, and GEMINI diffractomet-
ers, respectively. 34742 reflections were collected for 1, of which
12822 were independent [4012 with IϾ2.3σ(I)], 30512 reflections
for 3, of which 11224 were independent [6804 with IϾ2σ(I)], 37995
reflections for 4, of which 8735 were independent [2958 with
IϾ2σ(I)], 18851 reflections for 5, of which 8791 were independent
[4503 with IϾ2σ(I)], and 17386 reflections for 6, of which 5829
were independent [4033 with IϾ2σ(I)]. Except for 3, absorption
corrections were applied using Multiscan.[59] The structures were
solved by Direct Methods using SIR92,[60] and refined by least-
squares procedures on F2 using the program SHELXL97[61] in-
cluded in the software package WinGX version 1.63[62] for 3, 4, 5,
and 6, and on F with the program CRYSTALS[63] for 1. The atomic
scattering factors were taken from the International Tables for X-
ray Crystallography.[64] Hydrogen atoms were introduced in ideal-
ized positions and refined by using a riding model. Non-hydrogen
atoms were anisotropically refined, and in the last cycles of refine-
ment a weighting scheme was used. In the case of 1, due to the
average value of Rint (0.133), only the CuII centers were anisotropi-
cally refined. Several crystals of 4 originating from different prepa-
rations were measured, but all of them yielded similarly poor dif-
fraction data. Highly disordered carbon atoms C27, C29–C32, and
C40 were refined isotropically. Despite the low quality of the refine-
ment, the main structural features obtained are in line with those
reported for related monohydroxido-bridged trinuclear CuII com-
plexes[1–9] and may be safely used for comparison purpose inside
this series. The crystallographic data are summarized in Table 3.
CCDC-892483 (for 1), -892484 (for 3), -892485 (for 4), -89486 (for
5) and -892487 (for 6) contain the supplementary crystallographic
data for this paper. These data can be obtained from The Cam-
bridge Crystallographic Data Centre free of charge via
www.ccdc.cam.ac.uk/data_request/cif.
This work was supported by the Centre National de la Recherche
Scientifique (CNRS), and MAGMANet (Grant NMP3-CT-2005-
515767). The authors are grateful to Dr. A. Mari for technical as-
sistance.
[1] J. P. Costes, F. Dahan, J. P. Laurent, Inorg. Chem. 1986, 25,
413–416.
[2] M. Kwiatkowski, E. Kwiatkowski, A. Olechnowicz, D. M. Ho,
E. Deutsch, Inorg. Chim. Acta 1988, 150, 65–73.
[3] H. D. Bian, J. Y. Xu, W. Gu, S.-P. Yan, P. Cheng, D.-Z. Liao,
Z.-H. Jiang, Polyhedron 2003, 22, 2927–2932.
[4] M. S. Ray, S. Chattopadhyay, M. G. B. Drew, A. Figuerola, J.
Ribas, C. Diaz, A. Ghosh, Eur. J. Inorg. Chem. 2005, 4562–
4571.
[5] B. Sarkar, M. S. Ray, M. G. B. Drew, A. Figuerola, C. Diaz,
A. Ghosh, Polyhedron 2006, 25, 3084–3094.
[6] B. Sarkar, M. S. Ray, Y.-Z. Li, Y. Song, A. Figuerola, E. Ruiz,
J. Cirera, J. Cano, A. Ghosh, Chem. Eur. J. 2007, 13, 9297–
9309.
[7] P. Mukherjee, M. G. B. Drew, M. Estrader, C. Diaz, A. Ghosh,
Inorg. Chim. Acta 2008, 361, 161–172.
[8] L. Rigamonti, A. Cinti, A. Forni, A. Pasini, O. Piovesana, Eur.
J. Inorg. Chem. 2008, 3633–3647.
[9] C. Biswas, M. G. B. Drew, A. Figuerola, S. Gómez-Coca, E.
Ruiz, V. Tangoulis, A. Ghosh, Inorg. Chim. Acta 2010, 363,
846–854.
[10] P. Talukder, S. Sen, S. Mitra, L. Dahlenberg, C. Desplanches,
J. P. Sutter, Eur. J. Inorg. Chem. 2006, 329–333.
[11] C. Biswas, M. G. B. Drew, S. Asthana, C. Desplanches, A.
Ghosh, J. Mol. Struct. 2010, 965, 39–44.
[12] E. I. Solomon, R. K. Szilagyi, S. DeBeer George, L. Basumal-
lick, Chem. Rev. 2004, 104, 419–458.
[13] J. Yoon, E. I. Solomon, Coord. Chem. Rev. 2007, 251, 379–400.
[14] P. J. Hay, J. C. Thibeault, R. Hoffman, J. Am. Chem. Soc. 1975,
97, 4884–4899.
[15] O. Kahn, in Molecular Magnetism, VCH Publishers, New
York, 1993.
[16] E. Ruiz, P. Alemany, S. Alvarez, J. Cano, J. Am. Chem. Soc.
1997, 119, 1297–1303.
Eur. J. Inorg. Chem. 2012, 5729–5740
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5739

J.-P. Costes, J.-P. Tuchagues et al.
FULL PAPER
[17]
[41] F. Akagi, Y. Michihiro, Y. Nakao, K. Matsumoto, T. Sato, W.
Mori, Inorg. Chim. Acta 2004, 357, 684–688.
S. Ferrer, F. Lloret, E. Pardo, J. M. Clemente-Juan, M. Liu-
González, S. García-Granda, Inorg. Chem. 2012, 51, 985–1001.
[42] F. B. Hulsbergen, R. W. M. ten Hoedt, G. C. Verschoor, J. Re-
edijk, A. L. Spek, J. Chem. Soc., Dalton Trans. 1983, 539–545.
[43] M. Angaroni, G. A. Ardizzoia, T. Beringhelli, G. La Monica,
D. Gatteschi, N. Masciocchi, M. Moret, J. Chem. Soc., Dalton
Trans. 1990, 3305–3309.
[44] K. Sakai, Y. Yamada, T. Tsubomura, M. Yabuki, M. Yamagu-
chi, Inorg. Chem. 1996, 35, 542–544.
[18]
[19]
[20]
A. W. Kleij, Eur. J. Inorg. Chem. 2009, 193–205.
J. P. Costes, Bull. Soc. Chim. Fr. 1986, 78–82.
J. P. Costes, F. Dahan, M. B. Fernandez Fernandez, M. I. Fer-
nandez Garcia, A. M. Garcia Deibe, J. Sanmartin, Inorg. Chim.
Acta 1998, 274, 73–81.
F. Z. Chiboub Fellah, J. P. Costes, F. Dahan, C. Duhayon, G.
Novitchi, J.-P. Tuchagues, L. Vendier, Inorg. Chem. 2008, 47,
6444–6451.
[21]
[45] A. V. Virovets, N. V. Podberezskaya, L. G. Lavrenova, J. Struct.
Chem. 1997, 38, 440–445.
[46] S. Ferrer, J. G. Haasnoot, J. Reedijk, E. Müller, M. Biag-
ini Cingi, M. Lanfranchi, A. M. Manotti Lanfredi, J. Ribas, In-
org. Chem. 2000, 39, 1859–1867.
[47] S. Ferrer, F. Lloret, I. Bertomeu, G. Alzuet, J. Borrás, S.
García-Granda, M. Liu-González, J. G. Haasnoot, Inorg.
Chem. 2002, 41, 5821–5830.
[48] A. K. Boudalis, J.-M. Clemente-Juan, F. Dahan, J.-P. Tuchag-
ues, Inorg. Chem. 2004, 43, 1574–1586.
[49] Minuit – a System for Function Minimization and Analysis of
the Parameters Errors and Correlations: F. James, M. Roos,
Comput. Phys. Commun. 1975, 10, 343–367.
[50] A. P. Gingsberg, M. E. Lines, Inorg. Chem. 1972, 11, 2289–
2290.
[51] V. H. Crawford, H. W. Richardson, J. R. Wasson, D. J. Hodg-
son, W. E. Hatfield, Inorg. Chem. 1976, 15, 2107–2110.
[52] O. Kahn, Inorg. Chim. Acta 1982, 62, 3–14.
[53] E. Ruiz, S. Alvarez, J. Cano, V. Polo, J. Chem. Phys. 2005, 123,
164110.
[54] A. P. Gingsberg, Inorg. Chim. Acta 1971, 5, 45–68.
[55] P. Pascal, Ann. Chim. Phys. 1910, 19, 5–70.
[56] STOE: IPDS Manual, version 2.75, Stoe & Cie, Darmstadt,
Germany, 1996.
[57] CrysAlis RED, Version 1.171.32.5, Oxford Diffraction Ltd.,
2007.
[22]
[23]
[24]
M. S. Haddad, S. R. Wilson, D. J. Hodgson, D. N. Hendrick-
son, J. Am. Chem. Soc. 1981, 103, 384–391.
P. L. Burk, J. A. Osborn, M. T. Youinou, J. Am. Chem. Soc.
1981, 103, 1273–1274.
P. K. Coughlin, S. J. Lippard, J. Am. Chem. Soc. 1981, 103,
3228–3229; P. K. Coughlin, S. J. Lippard, J. Am. Chem. Soc.
1984, 106, 2328–2336.
M. G. B. Drew, M. McCann, S. M. Nelson, J. Chem. Soc., Dal-
ton Trans. 1981, 1868–1878.
L. K. Thompson, Can. J. Chem. 1983, 61, 579–583.
L. K. Thompson, F. W. Hartstock, P. Robichaud, W. Hanson,
Can. J. Chem. 1984, 62, 2755–2762.
L. K. Thompson, A. W. Hanson, B. S. Ramaswamy, Inorg.
Chem. 1984, 23, 2459–2465.
[25]
[26]
[27]
[28]
[29] L. K. Thompson, T. C. Woon, D. B. Murphy, E. J. Gabe, F. L.
Lee, Y. Le Page, Inorg. Chem. 1985, 24, 4719–4725.
[30] L. K. Thompson, S. K. Mandal, E. J. Gabe, F. L. Lee, A. W.
Addison, Inorg. Chem. 1987, 26, 657–664.
[31] J. V. Folgado, E. Coronado, D. Beltran-Porter, T. Rojo, A.
Fuertes, J. Chem. Soc., Dalton Trans. 1989, 237–241.
[32] I. Castro, J. Faus, M. Julve, F. Lloret, M. Verdaguer, O. Khan,
S. Jeannin, Y. Jeannin, J. Vaisserman, J. Chem. Soc., Dalton
Trans. 1990, 2207–2212.
[58] CrysAlisPro, Version 1.171.33.66, Oxford Diffraction Ltd.,
2010.
[33] E. Spodine, J. Manzur, M. T. Garland, M. Kiwi, O. Penã, D.
Grandjean, L. Toupet, J. Chem. Soc., Dalton Trans. 1991, 365–
369.
[59] R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38.
[60] SIR97 – A program for automatic solution of crystal structures
by direct methods: A. Altomare, M. C. Burla, M. Camalli, G. L.
Cascarano, C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni,
G. Polidori, R. Spagn, J. Appl. Crystallogr. 1999, 32, 115–119.
[34] A. K. Patra, M. Ray, R. Mukherjee, Polyhedron 2000, 19,
1423–1428.
[35] See, for example: a) A. Mukherjee, R. Raghunathan, M. K.
Saha, M. Nethaji, S. Ramasesha, A. R. Chakravarty, Chem.
Eur. J. 2005, 11, 3087–3096; b) L. F. Jones, C. A. Kilner, M. A.
Halcrow, Polyhedron 2007, 26, 1977–1983, and references
therein.
[61] G. M.
Sheldrick,
SHELX97
(includes
SHELXS97,
SHELXL97, CIFTAB) – Programs for Crystal Structure
Analysis, release 97–2, Universität Göttingen, 1998.
[62] WINGX 1.63 – Integrated System of Windows Programs for
the Solution, Refinement and Analysis of Single Crystal X-ray
Diffraction Data:L. Farrugia, J. Appl. Crystallogr. 1999, 32,
837–840.
[63] P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, D. J.
Watkin, J. Appl. Crystallogr. 2003, 36, 1487.
[64] International Tables for X-ray Crystallography, Kynoch Press,
Birmingham, England, 1974, vol. IV.
[36] R. Beckett, B. F. Hoskins, J. Chem. Soc., Dalton Trans. 1972,
291–295.
[37] P. F. Ross, R. K. Murmann, E. O. Schlemper, Acta Crystallogr.,
Sect. B 1974, 30, 1120–1123.
[38] S. Baral, A. Chakravorty, Inorg. Chim. Acta 1980, 39, 1–8.
[39] R. J. Butcher, C. J. O’Connor, E. Sinn, Inorg. Chem. 1981, 20,
537–545.
[40] Y. Agnus, R. Louis, B. Metz, C. Boudon, J.-P. Gisselbrecht, M.
Gross, Inorg. Chem. 1991, 30, 3155–3161.
Received: July 18, 2012
Published Online: October 5, 2012
5740
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 5729–5740