Cis-trans isomerism in diphenoxido bridged dicopper complexes: Role of crystallized water to stabilize the cis isomer, variation in magnetic properties and conversion of both into a trinuclear species

Article abstract of DOI:10.1039/c2dt31277h

The trans-[Cu₂L₂Cl₂] (1), and cis-[Cu ₂L₂Cl₂]·H₂O (2) isomers of a diphenoxido bridged Cu₂O₂ core have been synthesized using a tridentate reduced Schiff ba

Full text of DOI:10.1039/c2dt31277h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 12200
www.rsc.org/dalton
PAPER
Cis–trans isomerism in diphenoxido bridged dicopper complexes: role of
crystallized water to stabilize the cis isomer, variation in magnetic properties
and conversion of both into a trinuclear species†
Apurba Biswas,a Michael G. B. Drew,b Carmen Diaz,c Antonio Bauzá,d Antonio Frontera*d and
Ashutosh Ghosh*a
Received 13th June 2012, Accepted 19th July 2012
DOI: 10.1039/c2dt31277h
The trans-[Cu2L2Cl2] (1), and cis-[Cu2L2Cl2]·H2O (2) isomers of a diphenoxido bridged Cu2O2 core
have been synthesized using a tridentate reduced Schiff base ligand 2-[(2-dimethylamino-ethylamino)-
methyl]-phenol. The geometry around Cu(II) is intermediate between square pyramid and trigonal
bipyramid (Addison parameter, τ = 0.463) in 1 but nearly square pyramidal (τ = 0.049) in 2. The chloride
ions are coordinated to Cu(II) and are trans oriented in 1 but cis oriented in 2. Both isomers have been
optimized using density functional theory (DFT) calculations and it is found that the trans isomer is
7.2 kcal mol−1 more favorable than the cis isomer. However, the hydrogen bonding interaction of
crystallized water molecule with chloride ions compensates for the energy difference and stabilizes the cis
isomer. Both complexes have been converted to a very rare phenoxido-azido bridged trinuclear species,
[Cu3L2(μ1,1-N3)2(H2O)2(ClO4)2] (3) which has also been characterized structurally. All the complexes are
antiferromagnetically coupled but the magnitude of the coupling constants are significantly different
(J = −156.60, −652.31, and −31.54 cm−1 for 1, 2, and 3 respectively). Density functional theory (DFT)
calculations have also been performed to gain further insight into the qualitative theoretical interpretation
on the overall magnetic behavior of the complexes.
important from the point of view of their application. For
example, although several groups reported that, upon a careful
Introduction
Geometrical isomerism (cis–trans) has played a pivotal role in
the development of coordination chemistry. By isolating the cis
and trans isomers Alfred Werner proved conclusively that the
then well accepted Jorgensen’s chain theory was conceptually
wrong and that the six-coordinate Co(III) possessed octahedral
configuration rather than hexagonal planar or trigonal prismatic
geometry.1 Since then, isolation of all possible isomers and
studying their properties has become a fascinating area of
research among coordination chemists, not only for academic
interest but also for their applications in various fields.2 In fact,
the very different chemical and physical properties of cis and
trans isomers made their isolation simple and identification
straightforward. These differences in properties are extremely
choice of the non-leaving ligands, the trans isomer can gain
cytotoxicity,3 cis-platin has still been retained as a necessary
requisite for active antitumor drugs.4 It is interesting to note that
although numerous examples of cis and trans isomers are known
in coordination chemistry, most of them are concerned only
around a single metal centre i.e. in mononuclear complexes.
In recent years polynuclear transition metal complexes are of
considerable interest for their possible utility in the field of
structural chemistry,5–7 magnetism,5,8 gas storage,9 catalysis,10–12
luminescence,7,13 etc. Copper(II) complexes especially with
respect to binuclear systems have been investigated frequently
because of the interest in new inorganic materials showing
molecular ferro- or antiferromagnetic interactions14–16 and their
relevance for bioinorganic model systems in copper proteins.17
Copper(II) dimers involving a Cu2O2 moiety, owing to the rather
unique variety of structural types and spin–spin interactions that
characterize their crystal chemistry, have provided a rich oppor-
tunity for the investigation of structural effects on magnetic
coupling.15 Most of the di-phenoxido bridged Cu(II) complexes
show the metal to be five-coordinate but stereochemically
flexible as they can be square pyramidal or trigonal bipyramidal,
or almost anything in between.16 In such type of complex, there
is a possibility that the compound may exist in cis or trans form
with respect to the coordination of the monodentate anions
aDepartment of Chemistry, University College of Science, University of
Calcutta, 92 A.P.C. Road, Kolkata-700 009, India.
E-mail: ghosh_59@yahoo.com
bSchool of Chemistry, The University of Reading, P.O BOX 224,
Whiteknights, Reading RG 6 6AD, UK
cDepartament de Química Inorgànica, Universitat de Barcelona, Marti i
Franques 1-11, 08028-Barcelona, Spain
dDepartment of Chemistry, Universitat de les Illes Balears, Crta. de
Valldemossa km 7.5, 07122 Palma de Mallorca (Baleares), Spain
†Electronic supplementary information (ESI) available: CCDC reference
numbers 886052–886054. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c2dt31277h
12200 | Dalton Trans., 2012, 41, 12200–12212
This journal is © The Royal Society of Chemistry 2012

View Article Online
the solid mass with dry methanol and this methanol solution
(ca. 20 mL) was used for preparation of complexes.
Synthesis of the complex [Cu2L2Cl2] (1). An extracted solu-
tion of HL as prepared above was added to a solution of
CuCl2·2H2O (0.850 g, 5 mmol) in dry methanol (20 mL). The
mixture was stirred for 1 h and filtered. The filtrate was kept
undisturbed in a desiccator. The resultant solution, on keeping
overnight, yielded green single crystals suitable for X-ray analy-
sis. Yield: (1.048 g, 74%), Anal. calc. for C22H34Cl2Cu2N4O2:
C, 45.20; H, 5.86; N, 9.58%. found: C, 44.80; H, 5.54;
Scheme 1 Cis and trans isomers of Cu2O2 core containing monoden-
tate anion X.
(Scheme 1). The literature data shows that most of such reported
complexes are trans18 but a very limited number of complexes
with cis arrangement of monodentate anions are also reported.19
However, to the best of our knowledge there has been no pre-
vious report either on the isolation of both cis and trans isomers
of di-phenoxido bridged dinuclear copper(II) complexes or theo-
retical calculations that deal with the relative energies of such
isomers.
N, 9.25%. IR (KBr): ν(N–H), 3134 cm−1, ν(C–N), 1593 cm−1
;
λmax (nm), [εmax (dm3 mol−1 cm−1)] (dry methanol), 655 (344),
412 (2590).
Synthesis of the complex [Cu2L2Cl2]·H2O (2). An extracted
methanol solution of HL as prepared above was added to a solu-
tion of CuCl2·2H2O (0.850 g, 5 mmol) in methanol–water
mixture (9 : 1 v/v). The mixture was stirred for 1 h and filtered.
The filtrate was kept undisturbed at room temperature. The resul-
tant solution on keeping at room temperature yielded deep green
single crystals suitable for X-ray analysis. Yield: (1.069 g, 71%),
Anal. calc. for C22H36Cl2Cu2N4O3: C, 43.85; H, 6.02; N, 9.30%.
found: C, 43.55; H, 5.74; N, 9.05%. IR (KBr): ν(N–H), 3241 cm−1,
ν(C–N), 1595 cm−1; λmax (nm), [εmax (dm3 mol−1 cm−1)]
(methanol), 655 (344), 412 (2590).
In this paper we report the synthesis, crystal structures and
magnetic properties of two double phenoxido bridged dinuclear
copper(II) compounds, [Cu2L2Cl2] (1), and [Cu2L2Cl2]·H2O
(2) derived from a tridentate reduced Schiff base ligand
2-[(2-dimethylamino-ethylamino)-methyl]-phenol. Compound 1
is trans while 2 is cis with respect to the coordinated chloride
ions. Both isomers have been optimized using DFT calculations
which show that the trans isomer is more stable but the
H-bonding interaction of a crystallized water molecule with the
chloride ions can compensate for this difference and stabilize
the cis isomer. We also succeeded in converting both these
di-phenoxido bridged dicopper compounds to a very rare
phenoxido-azido bridged trinuclear species, [Cu3L2(μ1,1-N3)2-
(H2O)2(ClO4)2] (3) which has also been characterized struc-
turally and magnetically. The Cu(II) ions are antiferromagneti-
cally coupled in all three complexes but the magnitude of the
coupling constants are significantly different. Density functional
theory (DFT) combined with the broken symmetry approach has
been used to determine the coupling constants, which are in
good agreement with the experimentally determined values.
Synthesis of the complex [Cu3L2(μ1,1-N3)2(H2O)2(ClO4)2]
(3). Compound 1 (0. 584 g, 1 mmol) was dissolved in methanol
(10 mL). To this solution Cu(ClO4)2·6H2O (0.370 g, 1 mmol)
dissolved in methanol (10 mL) was added, followed by an
aqueous solution (1 mL) of sodium azide (0.130 g, 2 mmol)
with stirring. The mixture was filtered and the filtrate was kept
undisturbed at room temperature. By slow evaporation of the
resulting solution, deep brown coloured X-ray quality, rect-
angular shaped single crystals were obtained in three days.
Yield: (0.627 g, 70%), Anal. calc. for C22H38Cl2Cu3N10O12:
C, 29.49; H, 4.27; N, 15.63%. found: C, 29.22; H, 4.02; N,
15.38%. IR (KBr): ν(N–H), 3229 cm−1, ν(C–N), 1600 cm−1
;
λmax (nm), [εmax (dm3 mol−1 cm−1)] (methanol), 647 (341),
398 (3539).
Complex 3 was also prepared in the same way by using com-
pound 2 instead of compound 1. Yield: (0.645 g, 72%), Anal.
calc. for C22H38Cl2Cu3N10O12: C, 29.49; H, 4.27; N, 15.63%.
found: C, 29.19; H, 4.08; N, 15.25%. IR (KBr): ν(N–H),
3229 cm−1, ν(C–N), 1600 cm−1; λmax (nm), [εmax (dm3 mol−1
cm−1)] (methanol), 647 (341), 398 (3539).
Experimental
Materials
The reagents and solvents used were of commercially available
reagent quality.
Caution! Perchlorate salts of metal complexes are potentially
explosive; and caution should be exercised when dealing with
such derivatives.
Physical measurements
Synthesis of the reduced Schiff base ligand (HL) 2-[(2-
dimethylamino-ethylamino)-methyl]-phenol. The Schiff base
ligand was synthesized by refluxing a solution of salicylaldehyde
(0.52 mL, 5 mmol) and N,N-dimethylethylenediamine (0.54 mL,
5 mmol) in dry methanol (30 mL) for one hour.20 The solution
was cooled to 0 °C and solid sodium borohydride (210 mg,
6 mmol) was added slowly to this methanolic solution with stir-
ring. After completion of the addition, the resulting solution was
acidified with concentrated HCl (5 mL) and then evaporated to
dryness.21The reduced Schiff base ligand HL was extracted from
Elemental analyses (C, H and N) were performed using a
Perkin-Elmer 240C elemental analyzer. IR spectra in KBr pellets
(4500–500 cm−1) were recorded using a Perkin-Elmer RXI
FT-IR spectrophotometer. Electronic spectra in methanol
(1200–350 nm) were recorded in a Hitachi U-3501 spectro-
photometer. Thermal analysis (TG-DTA) were carried out on a
Mettler Toledo TGA/SDTA 851 thermal analyzer in a dynamic
atmosphere of dinitrogen (flow rate = 30 cm3 min−1). The
samples were heated in an alumina crucible at a rate of 5 °C min−1.
Powder X-ray diffraction patterns were recorded on a Bruker D-8
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12201

View Article Online
Advance 124 diffractometer operated at 40 kV voltage and
40 mA current and calibrated with a standard 125 silicon
sample, using Ni-filtered Cu-Ka (a = 0.15406 nm) radiation. The
magnetic measurements were carried out in the “Unitat de
mesures magnètiques dels SCT (Universitat de Barcelona)” on
by means of the Gaussian-09 program.28 The level of theory
used in the calculations is B3LYP/6-31+G*, which is a good
compromise between the size of the system and the accuracy of
the results. For these calculations we have used the crystallo-
graphic coordinates. The magnetic coupling constants of the
systems are described below using the Heisenberg model (mag-
netic properties subsection). It should be mentioned that the
widely and successfully used29–31 broken symmetry DFT
approach is not the unique methodology to compute and inter-
pret magnetic properties in quantum chemistry. For instance
ab initio methods based on Difference Dedicated Configuration
Interaction32 (e.g. CASSCF/DDCI) give excellent results and
offers the possibility to finely analyze the mechanisms and
origin of magnetic properties taking advantage of the access to
the wavefunction of all spin states of interest. This approach has
been recently used in the study of cis/trans isomeric effects on
magnetic properties of copper complexes.33
polycrystalline samples with
a Quantum Design SQUID
MPMSXL magnetometer in an applied field of 10 000 G and
500 G in the temperature ranges of 2–300 K and 2–30 K,
respectively. The diamagnetic corrections were evaluated from
Pascal’s constants.
Crystal data collection and refinement
Crystal data for the three crystals are given in Table 1. 3541,
3626, 4657 independent reflection data were collected with
Mo Kα radiation at 150 K using the Oxford Diffraction
X-Calibur CCD System. The crystals were positioned at 50 mm
from the CCD. 321 Frames were measured with a counting time
of 10 s. Data analyses were carried out with the CrysAlis
program.22 The structures were solved using direct methods with
the Shelxs 97 program.23 The non-hydrogen atoms were refined
with anisotropic thermal parameters. The hydrogen atoms
bonded to carbon were included in geometric positions and
given thermal parameters equivalent to 1.2 times those of the
atom to which they were attached. Absorption corrections were
carried out using the ABSPACK program.24 The structures were
refined on F2 to 0.0416, 0.0661, 0.0812; wR2 0.1037, 0.1129,
0.1678 for 2552, 2136, 2413 data with I > 2σ(I) respectively.
Results and discussion
Synthesis of the complexes
The condensation of N,N-dimethylethylenediamine in 1 : 1
molar ratio with salicylaldehyde afforded the Schiff base, 2-[(2-
dimethylamino-ethylimino)-methyl]-phenol which on reduction
with sodium borohydride readily produced the reduced Schiff
base, HL (Scheme 2). HL on reaction with copper(II) chloride in
1 : 1 molar ratios in dry methanol solvent yielded the trans
isomer 1 but when the solvent was methanol–water mixture
(9 : 1 v/v) cis isomer 2 resulted. 1 can be converted into 2 by
recrystallization from methanol–water mixture (9 : 1 v/v)
whereas recrystallization of 2 from dry methanol affords 1. The
conversions are understandable as one molecule of water of crys-
tallization is required to stabilize the cis isomer, 2. Compounds 1
and 2 are also interconvertable in the solid state. The trans
isomer 1 on keeping in humid atmosphere (relative humidity
70–90%) for several days absorbs one molecule of water and
transforms into the cis isomer 2 as is evident from IR spectra
(Fig. S1 and S2†) and PXRD patterns (Fig. S3–S5†). On the
other hand, thermogravimetric analysis of isomer 2 shows that
on heating it loses one molecule of water (obs. 2.92%, calc.
2.99%) in the temperature range 72–100 °C (Fig. S6†); the
PXRD patterns (Fig. S7†) of the dehydrated product is identical
to that of 1 indicating a solid state transformation of 2 into 1
upon dehydration.
Compound 3 was obtained by reacting compound 1 or 2 with
copper(II) perchlorate and sodium azide in 1 : 1 : 2 molar ratios in
methanol containing water medium. It is to be noted that triden-
tate NNO donor Schiff base or reduced Schiff base ligand on
reaction with Cu(II) salts usually produces di-phenoxido bridged
dinuclear complexes and there are more than twenty of such
complexes reported in the literature. Formation of mixed phe-
noxido/azido bridged trinuclear Cu(II) complexes with such
ligands are very rare; only one such compound is known.34
In the present work, we show for the first time that the very
common di-phenoxido bridged di-copper complex can be con-
verted to the very rare mixed bridged trinuclear complex, simply
by reacting with Cu(II) salt and azide ion.
Computational details
The study of the energetic differences between 1 and 2 and the
hydrogen bonding interaction in 2 has been performed using
ab initio calculations on the optimized structures at the RI-MP2/
def2-TZVP level of theory using the Turbomole program.25 The
magnetic calculations were carried out using density functional
theory (DFT) combined with the broken symmetry approach26,27
Table 1 Crystal data and structure refinement of complexes 1, 2 and 3
1
2
3
Formula
C
22H34Cl2-
C22H36Cl2-
Cu2N4O3
602.52
Monoclinic
C2/c
25.359(6)
7.7866(12)
17.511(5)
130.06(4)
2646.7(11)
4
C22H38Cl2-
Cu3N10O12
896.14
Monoclinic
P21/c
8.676(3)
15.090(4)
13.092(4)
99.24(3)
1691.8(9)
2
Cu2N4O2
584.51
Monoclinic
C2/c
11.8586(15)
11.8952(7)
18.7920(15)
107.420(14)
2529.2(4)
4
1.535
1.919
1208
0.038
M
Crystal system
Space group
a/Å
b/Å
c/Å
β/°
V/Å3
Z
Dc/g cm−3
μ/mm−1
F (000)
R(int)
1.512
1.839
1248
0.078
1.759
2.099
914
0.085
Total reflections
Unique reflections 3541
7618
7249
3626
9188
4657
I > 2σ(I)
R1, wR2
Temp (K)
2552
2136
2413
0.0416, 0.1083 0.0659, 0.1359 0.0812, 0.2076
150 150 150
12202 | Dalton Trans., 2012, 41, 12200–12212
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Formation of the complexes.
IR and electronic spectra. A moderately strong, sharp peak
due to a N–H stretching vibration, at 3134, 3241, and 3229 cm−1
for complexes 1, 2, and 3 respectively, shows that the imine group
of the Schiff base is reduced. Reduction of the imine group is also
very clearly indicated by the absence of the strong band due to
imine vibration which appears in the region 1620–1650 cm−1 for
the complexes of the corresponding unreduced Schiff bases.5c
The appearance of a broad band near 3450 cm−1 for complex 2
indicates the presence of a water molecule. In complex 3, the
sharp peak at 2089 cm−1 is due to ν(N3), suggesting the end-on
(μ1,1) bridging mode of the azide ion35 and the single band at
1101 cm−1 indicate the presence of anionic perchlorate group.
Electronic spectra were recorded in methanol solution, which
show a single absorption band at 655, 655 and 647 nm due to
d–d transition for compounds 1, 2, and 3 respectively. At the
higher energy region, the ligand to metal charge transfer bands
were located at 412 nm for compounds 1 and 2 and 398 nm for
compound 3.
Crystal structures of the complexes 1, 2 and 3
The dimeric structure of [Cu2L2Cl2] (1) which contains a crystal-
lographic centre of symmetry is shown in Fig. 1 together with
the atomic numbering scheme. Dimensions in the metal co-
ordination sphere are given in Table 2.
The metal atoms are five coordinate being bonded to three
donor atoms from one ligand L together with an oxygen atom
Fig. 1 The structure of 1 with ellipsoids at 30% probability.
O(11)a (a = 0.5 − x, 1.5 − y, −z) of a second ligand L and a
chloride ion. Both oxygens of the two ligands, L, bridge the two
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12203

View Article Online
Table 2 Bond distances (Å) and angles (°) for complexes 1, 2 and 3
Complex 1
Complex 2
Complex 3
Cu(1)–O(11)a
Cu(1)–N(19)
Cu(1)–N(22)
Cu(1)–O(11)
1.912(2)
1.953(2)
2.122(2)
2.177(2)
2.360(1)
3.147(1)
94.83(8)
83.76(8)
168.43(8)
79.55(8)
90.74(8)
116.24(8)
97.98(6)
90.24(6)
140.61(6)
102.68(5)
100.45(8)
1.961(3)
2.017(4)
2.069(4)
1.977(3)
2.481(2)
Cu(1)–O(11)
Cu(1)–N(1)
Cu(1)–N(19)
Cu(1)–N(22)
Cu(1)–O(1W)
Cu(2)–O(11)
Cu(2)–N(1)
Cu(2)–O84A
1.968(4)
1.992(4)
1.999(4)
2.031(5)
2.269(6)
1.921(4)
2.004(5)
2.682(6)
2.999(1)
80.3(2)
Cu(1)–Cl(2)
Cu(1)–Cu(1)a
3.066(1)
O(11)a–Cu(1)–N(22)
N(19)–Cu(1)–N(22)
O(11)a–Cu(1)–N(19)
O(11)a–Cu(1)–O(11)
N(19)–Cu(1)–O(11)
N(22)–Cu(1)–O(11)
O(11)a–Cu(1)–Cl(2)
N(19)–Cu(1)–Cl(2)
N(22)–Cu(1)–Cl(2)
O(11)–Cu(1)–Cl(2)
Cu(1)–O(11)–Cu(1)a
99.22(15)
86.47(15)
155.14(14)
75.22(13)
90.75(14)
158.86(16)
109.01(10)
93.88(12)
98.42(13)
102.68(11)
102.30(19)
Cu(1)–Cu(2)
O(11)–Cu(1)–N(1)
O(11)–Cu(1)–N(19)
N(1)–Cu(1)–N(19)
O(11)–Cu(1)–N(22)
N(1)–Cu(1)–N(22)
N(19)–Cu(1)–N(22)
O(11)–Cu(1)–O(1W)
N(1)–Cu(1)–O(1W)
N(19)–Cu(1)–O(1W)
N(22)–Cu(1)–O(1W)
O(11)–Cu(2)–N(1)
Cu(1)–O(11)–Cu(2)
Cu(1)–N(1)–Cu(2)
92.2(2)
165.6(2)
168.7(2)
98.0(2)
86.9(2)
88.9(2)
91.7(3)
100.5(2)
102.4(2)
81.17(17)
100.9(2)
97.3(2)
Symmetry element: a = (0.5 − x, 1.5 − y, −z) in 1; a = (−x, y, 0.5 − z) in 2.
Fig. 2 Hydrogen bonding polymeric structure of compound 1; hydrogen atoms except H19 have been excluded for clarity.
copper atoms. Bond lengths in the metal coordination sphere are
Cu(1)–O(11) 2.177(2) Å, Cu(1)–N(19) 1.953(2) Å, Cu(1)–N(22)
2.122(2) Å to the tridentate ligand L with Cu(1)–O(11)a 1.912(2) Å
and Cu(1)–Cl(2) 2.360(1) Å. The geometry around Cu(II)
is intermediate between square pyramid and trigonal bipyramid.
Considering as a square pyramid, the four donor atoms O(11),
O(11)a, N(19), N(22) in the equatorial plane show a r.m.s. devia-
tion of 0.479 Å with the metal atom 0.656 Å from the plane in
the direction of the axially coordinated chloride ion. However, if
the geometry around Cu(II) is considered as a trigonal bipyramid,
O(11), N(22), and Cl(1) make up the equatorial plane with
O(11)a, and N(19) occupying in the axial positions. The
Addison parameter36 (τ = 0.463) of the penta-coordinated
copper(II) clearly shows that the geometry of the coordination
sphere is intermediate between a square pyramid and a trigonal
bipyramid. The two Cu atoms are separated by 3.147(1) Å and
the Cu(1)–O(11)–Cu(1)a angle is 100.45(8)°. The hydrogen on
N(19) forms a hydrogen bond to Cl(2)b (b = 1 − x, 1 − y, −z)
with dimensions N⋯Cl 3.227(2) Å, N–H⋯Cl 158°, H⋯Cl
2.37 Å to result in a 1-D polymeric structure (Fig. 2).
The connectivity between the metal and ligand is equivalent
to that found in 1 but the geometry around the metal centre is
different. The dimer contains crystallographic C2 symmetry
rather than the centre of symmetry found in 1. Here the metal
atoms are also five-coordinate being bonded equatorially to three
donor atoms from one ligand L together with an oxygen atom
O(11)a (a = −x, y, 0.5 − z) of a second ligand L that forms a
di-phenoxido bridge between the two copper(II) ions and an
axially coordinated chloride ion. The r.m.s deviation of donor
atoms is 0.011 Å with the metal 0.381(2) Å from the plane in
the direction of the axial Cl(2). The overall geometry around
copper(II) is nearly square pyramidal with Addison parameter
(τ) = 0.049. The two Cu atoms are separated by 3.066(1) Å and
the Cu(1)–O(11)–Cu(1)a angle is 102.30(19)°. As can be seen
from Table 2, bond lengths are very different from those in 1
with Cu(1)–N(19) 2.017(4) Å, Cu(1)–N(22) 2.069(4) Å, Cu(1)–
O(11) 1.976(3) Å, Cu(1)–O(11)a 1.962(3) Å and axial Cu(1)–Cl(2)
2.481(2) Å. The tridentate ligand L coordinates to the metal
ion in a facial configuration in 1 but in a meridional configu-
ration in 2. A further distinguishing feature in 2 is that the two
axial atoms are on the same side of the Cu2O2 ring in the dimer,
rather than on opposite sides as in the centrosymetric 1. Another
The structure of [Cu2L2Cl2]·H2O (2) is shown in Fig. 3. Dimen-
sions in the metal coordination sphere are given in Table 2.
12204 | Dalton Trans., 2012, 41, 12200–12212
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 4 Hydrogen bonding supramolecular network of compound 2
(hydrogen atoms except H19, H23A and those of water molecules
omitted for clarity).
Fig. 3 The structure of 2 with ellipsoids at 30% probability. Hydrogen
bonds shown as dotted lines.
Table 3 Hydrogen bonding distances (Å) and angles (°) for the
complexes 1, 2 and 3
D–H A⋯H D⋯A
(Å) (Å) (Å)
∠D–H–A
(°)
Complex D–H⋯A
1
2
N(19)–H(19) ⋯Cl(2)b
0.91 2.37 3.227(2) 158
0.91 2.78 3.389(3) 125
0.86 2.41 3.259(3) 172
N(19)–H(19) ⋯Cl(2)c
O(1)–H(1) ⋯Cl(2)d
C(23)–H(23A) ⋯O(1)e 0.96 2.50 3.413(6) 159
3
N(19)–H(19) ⋯O(81)f
O(1W)–H(2) ⋯O(82)f
0.91 2.19 3.064(7) 162
0.85 2.40 3.110(8) 142
O(1W)–H(1) ⋯O(84A)g 0.88 2.40 3.109(8) 138
b
c
d
Symmetry element: = (1 − x, 1 − y, −z), = (−x, −y, −z), = (−x, y,
1/2 − z), e = (x, 1 + y, z), f = (−1 + x, y, z), g = (1 − x, −y, 1 − z).
interesting difference is the position of phenyl rings relative to
the Cu2O2 plane. In complex 1, one of the phenyl rings is above
and the other is below the Cu2O2 plane whereas in 2 both the
phenyl rings are on the same side of the Cu2O2 plane. The
N(19)–H(19) bond forms hydrogen bonds with Cl(2)c (c = −x,
−y, −z) with dimensions N⋯Cl 3.389(3) Å, H⋯Cl 2.78 Å,
N–H⋯Cl 125°. The solvent water molecule which is sitting on a
two-fold axis also forms hydrogen bonds to Cl(2) and Cl(2)d
(d = −x, y, 1/2 − z) with dimensions 3.259(3) Å, 2.41 Å, and
172°. The C(23)–H(23A) bond forms hydrogen bonds with
O(1)e (e = x, 1 + y, z) with dimensions C(23)⋯O(1) 3.413(6) Å,
H(23A)⋯O(1) 2.50 Å, C(23)–H(23A)⋯O(1) 159° (Table 3).
These hydrogen bonding interactions lead to the formation of a
2-D supramolecular network as shown in Fig. 4.
Fig. 5 The structure of 3 with ellipsoids at 30% probability. The two
open bonds show weak interactions of the central metal with perchlorate
anions. The perchlorate anions are disordered and the minor component
(16% occupancy) is not shown.
environments and the central Cu(2) positioned on a centre of
symmetry in a six-co-ordinate octahedral geometry. Each of the
two terminal copper(II) atoms, Cu(1), are coordinated by the
ligand L through two nitrogen atoms N(19) and N(22) and a
phenoxido oxygen atom O(11). In addition each of the two
Cu(1) ions is bonded to a nitrogen atom N(1) of an end-on
bridged azide ligand and one oxygen atom O(1) of a terminal
water molecule. Unlike the geometry of the copper atoms in 1,
Cu(1) here is in an almost ideal square pyramidal environment
The structure of [Cu3L2(μ1,1-N3)2(H2O)2(ClO4)2] (3) is shown
in Fig. 5. Dimensions of the metal coordination sphere are given
in Table 2.
The structure of 3 is centrosymmetric trinuclear with two
outer copper atoms, Cu(1), in square pyramidal five-co-ordinate
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12205

View Article Online
with a τ value of 0.049. The four atoms in the equatorial plane
consist of the three donor atoms from the ligand together with a
bridging azide. The four atoms show a r.m.s. deviation of
0.014 Å with the metal 0.211(3) Å from the plane in the direc-
tion of the axial water molecule. Bond lengths are much more
regular than in 1 with Cu(1)–N(19) 1.999(4) Å, Cu(1)–N(22)
2.031(5) Å, Cu(1)–O(11) 1.968(4) Å, Cu(1)–N(1) 1.992(4) Å
and Cu(1)–O(1) 2.269(6) Å although as expected the axial bond
is lengthened.
Cu(2) sits on a centre of symmetry, thus has two unique bonds
to O(11) at 1.921(4) Å and N(1) at 2.004(5) Å both of which
bridge to the adjacent Cu(1) atoms. Two oxygen atoms O(84A)
and O(84A)a (a = 1 − x, −y, 1 − z) of two perchlorate anions are
weakly coordinated to the axial sites at 2.682(6) Å. Cu(1) and
Cu(2) atoms are separated by 2.999(1) Å. The Cu(1)–O(11)–
Cu(2) and Cu(1)–N(1)–Cu(2) bridge angles are 100.9(2)° and
97.3(2)° respectively.
A comparison of the structure of 3 with the previously
reported unreduced Schiff base complex34 containing methyl
group at the benzylic carbon reveals that the basic structures are
very similar. The main difference is found in the Cu(1)–N(19)
bond length which is significantly longer in 3 (Table 2) than that
in the reported compound (1.936(4) Å). The lengthening of this
bond in 3 is not surprising as the π accepting ability of the imine
(CvN) functions is lost on reduction. Another interesting differ-
ence is in the Cu(1)–O(11) and Cu(2)–O(11) bond lengths. In
the reported complex Cu(1)–O(11) at 1.925(2) Å is shorter than
Cu(2)–O(11) at 1.952(2) Å whereas in 3 it is reversed probably
due to the higher flexibility of reduced Schiff base ligand. There
are also slight increase of Cu(1)–O(11)–Cu(2) and Cu(1)–N(1)–
Cu(2) bond angles in 3 than those in the reported complexes at
100.79(11)° and 96.53(15)° respectively.
The amine nitrogen N(19) forms a hydrogen bond to a per-
chlorate oxygen atom O(81)f (f = x − 1, y, z) with dimensions
N⋯O 3.064(7) Å, H⋯O 2.19 Å, N–H⋯O 162°. The water
molecule, based on O(1W) also shows hydrogen bonds to per-
chlorate oxygen atoms to O(82)f (f = x − 1, y, z) and O(84A)g
(g = 1 − x, −y, −z) with dimensions O⋯O 3.110(8), 3.109(8) Å,
H⋯O 2.40, 2.40 Å and O–H⋯O 142°, 138° (Table 3) to form a
1D polymeric structure (Fig. 6).
Energetic and geometric considerations for complexes 1 and 2
Experimentally, both cis and trans isomers of the dinuclear CuII
complex have been obtained (see Fig. 7). It should be mentioned
that the cis isomer is quite rare since a search in the CSD reveals
that there is no other dinuclear oxobridged Cu2O2N4Cl2 complex
with a cis arrangement of the Cl atoms. As a matter of fact,
when the search is extended to any transition metal, there are
Fig. 6 Hydrogen bonding 1D polymeric structure of compound 3; hydrogen atoms except H19 and those of water molecules omitted for clarity.
Fig. 7 X-Ray structures of isomers 1 and 2, with indication of the relative formation energies. Distances in Å.
12206 | Dalton Trans., 2012, 41, 12200–12212 This journal is © The Royal Society of Chemistry 2012

View Article Online
only nine dinuclear complexes exhibiting a cis orientation of Cl
atoms, of which six contain ZnII,37 two CdII,38 and one FeIII.39
The differentiating feature of 2 is the presence of a water
molecule that forms a double hydrogen bond with both chloride
ligands. These hydrogen bonds probably compensate for the
difference in energy between the cis and trans isomers. We opti-
mized both isomers using DFT calculations starting from the
crystallographic coordinates and eliminating the water molecule
in 2 in order to be equivalent. We then compared both energies
and, as expected, the trans isomer is 7.2 kcal mol−1 more favor-
able than the cis isomer. Therefore the extra stabilization pro-
vided by the presence of the water molecule should, at least,
compensate for this energy. The presence of the water molecule
has a strong influence on the crystal packing. It is involved in the
formation of 1D infinite columns that interact with each other in
antiparallel orientation giving rise to the formation of 2D sheets
(see Fig. 8, left). It can be observed in Fig. 8 (right) that the
water molecule forms two strong O–H⋯Cl hydrogen bonds with
the cis-dichlorido ligands and two unconventional C–H⋯O
hydrogen bonds with the methyl groups of another molecule,
causing the formation of the infinite 1D column. We have com-
puted the interaction energy associated to the hydrogen bonds
present in 2 at the MP2/def2-TZVP level of theory. The inter-
action energy is −11.4 kcal mol−1 for the conventional O–H⋯Cl
hydrogen bonds that compensates for the 7.2 kcal mol−1 of
difference with compound 1 (trans isomer). The interaction
energy of these unconventional C–H⋯O hydrogen bonds that
allow the column growth is −4.0 kcal mol−1
.
In the CSD search we have found some related structure that
crystallize including a solvent molecule in the structure (CH2Cl2
or CH3OH) which are responsible for the cis orientation of the
chloride ligands due to the formation of a double hydrogen
bond. The 2D architecture is also similar (two examples are
shown in Fig. 9) forming parallel columns instead of the anti-
parallel orientation found in compound 2. Both structures
involve ZnII metals.37a,e
Magnetic properties
Temperature-dependence molar susceptibility measurements of
powdered samples of 1, 2 and 3 were carried out in an applied
field of 10 000 G and 500 G in the temperature range of
2–300 K and 2–30 K, respectively. The χmT vs. T plots of com-
plexes 1 and 2 are shown in Fig. 10, in both cases χmT being the
magnetic susceptibility per Cu2 unit. At room temperature the
χmT values are smaller than the value expected for two
uncoupled S = 1/2 spins with g = 2.00 (0.75 emu mol−1 K), and
are equal to 0.68 emu mol−1 K for 1 and 0.14 emu mol−1 K for
2. Upon cooling, χmT values decrease to zero at low temperature.
These features are characteristic of the occurrence of an anti-
ferromagnetic coupling. Assuming the isotropic Hamiltonian
Ĥ = −JŜ1Ŝ2 where Ŝ1 = Ŝ2 = 1/2, the experimental data were
fitted to the eqn (1) given in the literature for dinuclear copper
compounds.14b
2Ng2β2
Ng2β2
2kT
χ ¼
ð1 ꢀ ρÞ þ
ρ
ð1Þ
kT½3 þ expðꢀJ=kTÞꢁ
The parameters N, g, β, k, and T in the equation have their
usual meanings, J = singlet–triplet splitting and ρ is the percen-
tage of the paramagnetic impurities. The best-fit parameters for
reproducing satisfactorily the experimental data, as shown in
Fig. 10, are J = −156.60 cm−1, g = 2.15, ρ = 0.022 with R =
1.0 × 10−5 for 1 and J = −652.31 cm−1, g = 2.16, ρ = 0.021
with R = 3.3 × 10−5 for 2, (R = Σi(χTicalc − χTiexp)2/Σi(χTiexp)2).
For complex 3, the χmT vs. T plot (being the magnetic suscep-
tibility per three copper(II) ions) is shown in Fig. 10 (right). At
Fig. 8 Left: solid state structure of 2, with the formation of antiparallel
columns formed as a consequence of the presence of a water molecule.
Right: partial view of the solid state structure of 2 showing the hydrogen
bond network. Distances in Å.
room temperature the χmT value is equal to 1.16 emu mol−1
K
Fig. 9 Solid state structures of two dinuclear Zn(II) complexes retrieved from the CSD (UKESAF and DOYBOJ). The solvent molecule stabilizes
the syn arrangement of the chloride ligands.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12207

View Article Online
Fig. 10 Plot of χmT vs. T in the range 2–300 K for 1 (left) and 2 (middle) and 3 (right).
which is very near to the expected value for three uncoupled
S = 1/2 spins with g = 2.00 (1.12 emu mol−1 K). Upon cooling,
the χmT value decreases gradually to 0.37 emu mol−1. Consider-
ing that the distances between the terminal copper(II) ions and
the central one are equal (the complex is centrosymmetric), and
the interaction between the terminal copper(II) atoms is negligible,
Table 4 Selected structural parameters for complexes 1 and 2 related
to their magnetic data
Complex 1
Complex 2
J value (cm−1
τ (Addison parameter)
Distances (Cu–O) (Å)
)
−156.60
0.463
2.122(2)
1.912(2)
100.45(8)
19.43
−652.31
0.049
1.976(3)
1.962(3)
102.29(19)
13.24
the isotropic spin Hamiltonian employed is: Ĥ = −J[Ŝ1Ŝ2
+
Ŝ2Ŝ3]. The magnetic data were analyzed through the expression
Angles (Cu–O–Cu) (°)
Out-of-plane shift of the phenyl group (°)
Torsion angle (Cu–O–Cu–O) (°)
given in eqn (2).14b
0
16.98
Ng2β2χM trinuclear
kT ꢀ zJ′χM trinuclear
χM
¼
ð2Þ
complexes is the Addison parameter (τ) as an increase in τ can
be correlated with a decrease in antiferromagnetic coupling.36
Complexes 1 and 2, have different structural parameters (see
Table 4) that can explain their significantly different J values
(−156.6 and −652.3 cm−1 for 1 and 2, respectively). The lower
value of τ (0.049 for 2 compared to 0.463 for 1) the greater
value of the Cu–O–Cu angle (102.29° for 2 compared to
100.45° for 1), the smaller out-of plane shifts of the phenoxido
groups (13.24° for 2 compared to 19.43 for 1) and the shorter
Cu–O distances (1.969 Å (average) for 2 compared to 2.017 Å
(average) for 1) indicate that J must be more antiferromagnetic
for complex 2 than complex 1.
with J = intratrimeric magnetic exchange coupling constant,
zJ′ = intertrimeric magnetic exchange coupling constants and
Ng2β2 1 þ expðJ=kTÞ þ 10 expð3J=2kTÞ
χM trinuclear
¼
4kT 1 þ expðJ=kTÞ þ 2 expð3J=2kTÞ
Best fit parameters are J = −31.54 cm−1, g = 2.08, zJ′ = −0.49
and R = 5.3 × 10−4 (R = Σi(χTicalc − χTiexp)2/Σi(χTiexp)2).
Magnetostructural correlations
In complex 3 the EO (end-on coordination mode) μ-azido
group (N1) and the oxygen atom of μ-phenoxido group (O11)
bridges the copper centers in a basal-equatorial fashion. The
magnetic orbitals describing the single electron in the three
copper ions mainly consist of a dx2–y2 contribution. In the litera-
ture only one trinuclear copper complex, similar to 3 containing
EO μ-azido and the μ-phenoxido bridges has been found, and
that also showed weak antiferromagnetic coupling.34 From the
theoretical calculations on dinuclear complexes with EO azido
bridged complexes, it has been established that the J parameters
depends mainly on the Cu–N–Cu angle, the ferromagnetic coup-
ling decreases with increase in the angle (critical angle 104° or
108° according to theoretical and experimental studies, respect-
ively).35,40 In 3, the Cu(1)–N(1)–Cu(2) angle is 97.3(2)°. The
phenoxido bridged dinuclear complexes with bridging angles
greater than 98° are generally antiferromagnetic. In the present
case the Cu(1)–O(11)–Cu(2) angle is 100°. Therefore, in
complex 3 ferromagnetic and antiferromagnetic coupling are
expected through azido bridge and phenoxido bridges respec-
tively. However the net antiferromagnetic coupling indicates that
coupling through the phenoxido bridge dominates. It is worth
In Table 4 we have indicated the main structural parameters of
complexes 1 and 2 that can influence the corresponding J values.
Magnetic properties of dinuclear copper complexes containing
Cu2O2 core depend on the structural properties of the core.
Factors, such as the coordination geometry of the copper ions,
the Cu–O–Cu angle, the Cu–O bond lengths, and Cu–O–Cu–O
torsion angle are the parameters that have been postulated in the
literature as influencing the J values of the spin coupling in
di-phenoxido bridged dinuclear copper(II) complexes.15,16 It has
been shown that the overlap between the metal and the phenox-
ido bridging oxygen is a factor that controls the spin coupling
and is correlated not only with the bridging angle, but also with
the M–O bond distances. Literature survey of dinuclear
phenoxo-bridged Cu(II) complexes reveals that when the Cu–O
bond distance is less than 1.98 Å, strong antiferromagnetic coup-
ling is observed and the strength of this coupling is linearly
dependent on the Cu–O bond distances; large out-of plane shifts
for the phenoxido groups reduce the antiferromagnetic term; large
Cu–O–Cu angles increase the antiferromagnetic contribution.17
Another important structural factor in penta-coordinated Cu(II)
12208 | Dalton Trans., 2012, 41, 12200–12212
This journal is © The Royal Society of Chemistry 2012

View Article Online
J value of 3 ) with that
(−31.54 cm−1
Table 5 The spin densities on the selected atoms for compounds 1 and
2 at the UB3LYP/6-31G* level of theory
comparing the
(−9.86 cm−1) of the only other similar complex formed by the
unreduced Schiff base ligand.34 The bridging angles or the dis-
tances are very similar in the two complexes but the τ value
(0.05) of 3 is significantly lower than that in the reported
complex (0.15) and this seems to be responsible for the stronger
antiferromagnetic coupling in complex 3.
Compound 1
Compound 2
Atoms
Spin density
Atoms
Spin density
Cu1
Cu1a
Cl1
0.581
−0.581
0.102
−0.102
0.080
−0.080
0.123
0.078
−0.124
−0.078
Cu1
Cu1a
Cl1
−0.594
0.594
0.002
−0.002
0.022
−0.022
−0.106
−0.094
0.106
Cl1a
O11
O11a
N19
N22
N19a
N22a
Cl1a
O11
O11a
N19
N22
N19a
N22a
Magnetic calculations
The determination of J values. In the present study, all the
calculations were carried out with the density functional theory
(DFT) combined with the broken symmetry approach. The mag-
netic coupling constants J of the systems are described by the
Heisenberg model, and the Heisenberg spin Hamiltonian used
for the magnetic fitting was Ĥ = −JŜ1Ŝ2 where Ŝ1 and Ŝ2 present
the spins at sites 1 and 2, respectively. For the copper(II) dinuc-
lear complexes, calculation of the J values has been performed
computing the difference between the energy values of the
highest spin state and the broken-symmetry state.
0.094
Symmetry element: a = (0.5 − x, 1.5 − y, −z) in 1; a = (−x, y, 0.5 − z) in 2.
The calculated magnetic constant J of compound 1 with the
BS-DFT/6-31+G* method is −115.1 cm−1, which shows that
there is an antiferromagnetic coupling interaction in this com-
pound, in agreement with experimental results, however the DFT
method underestimates the J coupling constant. For compound 2
we have also found a strong antiferromagnetic coupling with J =
−726.1 cm−1. In this case the J value is to some extent over-
estimated; however an acceptable agreement is obtained. In any
case the experimental trend is reproduced by the DFT calcu-
lations, since the antiferromagnetic coupling is stronger in com-
pound 2 than in compound 1. In order to analyze the influence
of the hydrogen bonding on the calculated J values, we have
computed them in a dimeric unit of 1 (see Fig. 2) and in 2 in the
presence of the water molecule. For both compounds the com-
puted J value remains the same (<5% variation) with and
without the presence of H-bonding. Le Guennic and Robert’s
research group have reported interesting and important work
regarding magnetic interactions involving hydrogen bonding-
mediated chains in dinulear copper systems.33,41 In our case, the
H-bonds do not seem to play a preeminent role of in the main
pathways for the magnetic exchange observed by magnetic sus-
ceptibility measurements.
Fig. 11 The SOMO and SOMO−1 for compounds 1 and 2.
Spin density distribution analysis. In order to further analyze
the magnetic coupling mechanism, the spin density distribution
is analyzed. According to the molecular orbital theory, spin delo-
calization is the result of electron transfer from the magnetic
centers to the ligand atoms. The spin densities of compounds 1
and 2 in the broken-symmetry state are summarized in Table 5,
where positive and negative signs denote α and β spin states,
respectively. In Table 5 it is shown that the spin densities on the
two Cu(II) ions in compounds 1 and 2 have similar absolute
values but opposite signs. For instance in 1, the Cu1 atom is
mainly populated by the unpaired electrons with α spin (0.581)
and Cu1a with β spin (−0.581). It is obvious that the unpaired
electrons are mainly localized on each Cu(II) ions, which indi-
cates that the Cu(II) ions are indeed the magnetic centers. The
spin densities on the ligand (O11, N19 and N22) atoms have the
same signs as that of the Cu(II) atoms to which they are bonded,
which reveals that there is spin delocalization from the Cu(II)
ions to the ligands. In addition, the spin delocalization is strong
enough that >40% of the spin for the unpaired electrons on the
Cu(II) centers are delocalized to the ligand atoms.
The spin densities on the Cu(II) centers in compound 1
(0.581) are comparable to those of compound 2 (0.594), which
indicates that the differences observed in the magnetic coupling
constants J cannot be explained in terms of the spin density on
the metal centers.
Molecular orbital analysis. According to the active-electron
approximation, it can be considered that in the bridged magnetic
coupling systems only the overlaps between the singly occupied
molecular orbitals (SOMO) of the metal centers and the p orbi-
tals of the ligands contribute to the magnetic coupling
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12209

View Article Online
interaction. As an example, the SOMO diagrams of compound 1
are shown in Fig. 11. The X-ray crystal geometries of the com-
plexes anticipate the higher antiferromagnetic character of com-
pound 2, since in this structure the Cu(II) atoms are in an almost
ideal four-coordinate square planar geometry with the O/N
ligands. Therefore the magnetic orbitals on the Cu(II) ions are
mainly formed by the dx2−y2 orbitals. In compound 1 the geo-
metry of the Cu(II) atoms is clearly different and the overlap of
the magnetic d orbitals on the Cu(II) ions with the p orbitals of
the dioxo bridge is more difficult. The SOMO diagrams of 2
clearly show that the two lone electrons of Cu(II) centers mainly
occupy the dx2−y2 orbitals. The Cu(II) centers have large inter-
actions with the px and py orbitals of the bridging ligands that
lead to the magnetic coupling contribution. The magnetic orbi-
tals on the Cu(II) ions for compound 1 are more difficult to
describe and they are probably a combination of two d orbitals
of the Cu(II) centers. The participation px and py orbitals of the
bridging ligands that lead to the magnetic coupling contribution
is also clear in compound 1 (see Fig. 11). A differentiating
feature of both compounds is the participation of the chloride
atoms. In 1 there is some spin delocalization on the Cl ligands
(see Table 5) and this is reflected in the SOMOs (see Fig. 11). In
contrast, such delocalization of spin in the Cl ligands is not
observed in compound 2. The results from the orbital analysis
are useful in explaining that the antiferromagnetic coupling inter-
action of 2 is stronger than that of 1, as observed experimentally.
Analysis of compound 3. We have extended our theoretical cal-
culations to compound 3, which is a trinuclear Cu(II) complex (see
Fig. 5). The magnetic coupling constant J of this system is defined
as Ĥ = −J[Ŝ1Ŝ2 + Ŝ2Ŝ3] for a symmetrical linear trinuclear Cu(II)
complex, assuming that the exchange integrals between the neigh-
boring copper ions are identical (J12 = J23 = J) and Ŝ1, Ŝ2 and Ŝ3
represents the spins at sites 1, 2 and 3, respectively. We also assume
that the integral between the terminal copper ions is zero (J13 = 0).
The calculated magnetic constant J of compound 3 by means
of the BS-DFT/6-31+G* method is −48.0 cm−1, which shows
that there is an antiferromagnetic coupling interaction in agree-
ment with experimental results. In this case the DFT method over-
estimates the J coupling constant, however it is able to reproduce
the experimental trend, since the coupling constant in the trinuc-
lear complex is lower than those in the dinuclear complexes 1 and
2. The spin densities of compound 3 in the broken-symmetry state
are summarized in Table 6, where positive and negative signs
denote α and β spin states, respectively. In Table 6 it is shown that
the spin densities on the two equivalent Cu(II) ions have lower
absolute values than the central Cu(II) ion and opposite sign, indi-
cating more spin delocalization to the ligands. The Cu1 and Cu1a
atoms are mainly populated by the unpaired electrons with α spin
(0.66) and Cu2 with α spin (−0.72). The spin densities on the
organic ligand (O11, N19 and N22) atoms have the same signs as
the Cu1 and Cu1a to which they are bonded, which reveals that
there is spin delocalization from the Cu(II) ions to the ligands. In
addition, the spin delocalization to the perchlorate from the central
metal ion and to water ligands from Cu1 and Cu1a is negligible.
We have also computed the singly occupied molecular orbitals
(SOMO) of compound 3 (see Fig. 12). As expected the metal
centers and the p orbitals of the ligands contribute to the mag-
netic coupling interaction. The SOMO diagrams of 3 show that
the three lone electrons of Cu(II) centers mainly occupy the dx2–y2
orbitals. The Cu(II) centers have large interaction with the px and
py orbitals of the bridging ligands that lead to the magnetic coup-
ling contribution. The magnetic orbital on the central Cu(II) ion
does not participate in the SOMO−2 orbital. Therefore SOMO
and SOMO−1 are responsible of the antiferromagnetic coupling
interaction observed in compound 3.
Table 6 The spin densities on the selected atoms for compound 3 at
the UB3LYP/6-31G* level of theory
Compound 3
Atoms
Spin density
Cu1
0.664
0.665
−0.719
0.006
0.006
0.000
0.000
0.024
0.023
0.079
0.077
0.079
0.077
0.000
0.000
Cu1a
Cu2
N1
N1a
O1W
O1Wa
O11
O11a
N19
N22
N19a
N22a
O84
O84a
We have estimated the contribution of each bridge (phenoxido
and azido) over the antiferromagnetic interactions by calculating
the contribution of the atomic orbitals of the bridging atoms on
Symmetry element: a = (1 − x, −y, 1 − z).
Fig. 12 SOMOs of compound 3.
This journal is © The Royal Society of Chemistry 2012
12210 | Dalton Trans., 2012, 41, 12200–12212

View Article Online
and S. Wołowiec, J. Chem. Soc., Dalton Trans., 2002, 885; (c) D. Das,
N. R. Ray Choudhury and A. Ghosh, Polyhedron, 1996, 15, 3919;
(d) I. R. Laskar, A. Ghosh, G. Mostafa, D. Das, A. Mondal and N. Ray
Chaudhuri, Polyhedron, 2000, 19, 1015; (e) I. R. Laskar, T. K. Maji,
S. Chaudhuri, A. Ghosh and N. Ray Chaudhuri, Polyhedron, 2000, 19,
1803; (f) F. M. Macdonald and P. J. Sadler, Polyhedron, 1991, 1443;
(g) N. Mathew, B. R. Jagirdar and A. Ranganathan, Inorg. Chem., 2003,
42, 187; (h) F. D. Rochon and M. Fakhfakh, Inorg. Chim. Acta, 2009,
362, 1455.
3 (a) N. Farrell, L. R. Kelland, J. D. Roberts and M. van Beusichem,
Cancer Res., 1992, 52, 5065; (b) G. Natile and M. Coluccia,
Coord. Chem. Rev., 2001, 216–217, 383; (c) E. I. Montero, S. Dìaz,
A. M. Gonzalez-Vadillo, J. M. Perez, C. Alonso and C. J. Navarro-
Ranninger, J. Med. Chem., 1999, 42, 4264; (d) J. Kasparkova,
O. Novakova, N. Farrell and V. Brabec, Biochemistry, 2003, 42, 792;
(e) J. Kasparkova, V. Marini, Y. Najajreh, D. Gibson and V. Brabec,
Biochemistry, 2003, 42, 6321.
the SOMO and SOMO−1. The total contribution of the two
oxygen atom belonging to the phenoxido bridges is 20.2% and
the total contribution of the two nitrogen atoms of the azido
bridges is 15.0%. Therefore the magnetic coupling observed in
compound 3 is slightly more effective through the phenoxido
ligand, which is in agreement with the fact that coupling through
phenoxido bridge dominates the ferro/antiferromagnetic interplay
between both ligands.
Conclusions
The tridentate NNO donor ligands such as employed in the
present paper are well known to produce diphenoxido bridged
trans or cis dinuclear Cu(II) complexes and in very rare cases
mixed bridged trinuclear complexes. We have shown here that
the formation of these different compounds is not serendipitous.
The cis isomer is stabilized by the presence of a crystallized
water molecule hence methanol containing water is required for
its synthesis whereas dry methanol solvent affords the trans
isomer. The higher proportion of the copper ion in the trinuclear
complex justifies the addition of copper salt to the solution of
the dinuclear complexes for its synthesis. The DFT calculations
reveal that the cis isomer is energetically less favorable as
expected. However, the two cis-oriented chloride ions form
strong H-bonds with the two H-atoms of the water molecule and
the interaction energy associated with the hydrogen bonds is
larger than the energy difference between cis and trans isomers
and consequently the cis isomer is stabilized. The experimentally
observed magnetic coupling constants J of all three compounds
are well reproduced by DFT calculations. The molecular orbital
analyses reveal that in the cis isomer, the two unpaired electrons
of Cu(II) mainly occupy the dx2–y2 orbitals whereas in the trans
isomer they are located in the orbital which is a combination of
two d orbitals as is anticipated from the geometries around Cu(II)
ions. This difference explains the remarkably stronger antiferro-
magnetic coupling in the former. In the mixed-bridged trinuclear
complex the coupling through phenoxido dominates over the
azido bridge and consequently the overall coupling becomes
antiferromagnetic, which is substantiated by DFT calculations.
4 L. R. Kelland, Nat. Rev. Cancer, 2007, 7, 573.
5 (a) M. Albrecht, Chem. Rev., 2001, 101, 3457; (b) R. Ziessel, Coord.
Chem. Rev., 2001, 216–217, 195; (c) P. Mukherjee, M. G. B. Drew,
C. J. Gómez-García and A. Ghosh, Inorg. Chem., 2009, 48, 5848;
(d) B. Sarkar, M. S. Ray, Y.-Z. Li, Y. Song, A. Figuerola, E. Ruiz,
J. Cirera, J. Cano and A. Ghosh, Chem.–Eur. J., 2007, 13, 9297;
(e) P. Mukherjee, M. G. B. Drew, C. J. Gómez-García and A. Ghosh,
Inorg. Chem., 2009, 48, 4817; (f) P. Mukherjee, M. G. B. Drew,
M. Estrader and A. Ghosh, Inorg. Chem., 2008, 47, 7784.
6 (a) M. Dey, C. P. Rao, P. K. Saarenketo and K. Rissanen, Inorg. Chem.
Commun., 2002, 5, 924; (b) S. Mukherjee, T. Weyhermuller, E. Bothe,
K. Wieghardt and P. Chaudhuri, Eur. J. Inorg. Chem., 2003, 863.
7 D. Zhang, H. Wang, Y. Chen, Z.-H. Ni, L. Tian and J. Jiang, Inorg.
Chem., 2009, 48, 11215.
8 Q. Wang, J. Zhang, C.-F. Zhuang, Y. Tang and C.-Y. Su, Inorg. Chem.,
2009, 48, 287.
9 (a) K. C. Gupta and A. K. Sutar, Coord. Chem. Rev., 2008, 252, 1420;
(b) K. L. Gurunatha and T. K. Maji, Inorg. Chem., 2009, 48, 10886;
(c) T. K. Maji, G. Mostafa, R. Matsuda and S. Kitagawa, J. Am. Chem.
Soc., 2005, 127, 17152.
10 S. A. Serron, C. M. Haar and S. P. Nolan, Organometallics, 1997, 16,
5120.
11 (a) E. Fujita, B. S. Brunschwig, T. Ogata and S. Yanagida, Coord. Chem.
Rev., 1994, 132, 195; (b) T. Opstal and F. Verpoort, Angew. Chem., Int.
Ed., 2003, 42, 2876.
12 M. Volpe, H. Hartnett, J. W. Leeland, K. Wills, M. Ogunshun,
B. J. Duncombe, C. Wilson, A. J. Blake, J. McMaster and J. B. Love,
Inorg. Chem., 2009, 48, 5195.
13 H. Wang, D. Zhang, Z.-H. Ni, X. Li, L. Tian and J. Jiang, Inorg. Chem.,
2009, 48, 5946.
14 (a) Magneto-Structural Correlations in Exchange Coupled Systems,
ed. D. D. Willett, D. Gatteschi and O. Khan, Reidel, Dordrecht,
The Netherlands, 1985; (b) O. Kahn, Molecular Magnetism, VCH,
New York, 1993; (c) S. Mukhopadhyay, D. Mandal, P. Baran Chatterjee,
C. Desplanches, J.-P. Sutter, R. J. Butcher and M. Chaudhury, Inorg.
Chem., 2004, 43, 8501; (d) L. K. Thompson, S. K. Mandal,
S. S. Tandon, J. N. Bridson and M. K. Park, Inorg. Chem., 1996, 35,
3117; (e) J. Manzur, H. Mora, A. Vega, E. Spodine, D. Venegas-Yazigi,
M. T. Garland, M. S. EI Fallah and A. Escuer, Inorg. Chem., 2007, 46,
6924; (f) J. Vanco, J. Marek, Z. Travnicek, E. Racanska, J. Muselik and
O. Svajlenova, J. Inorg. Biochem., 2008, 102, 595; (g) A. Mukherjee,
F. Lloret and R. Mukherjee, Inorg. Chem., 2008, 47, 4471.
Acknowledgements
We thank CSIR, Government of India [Senior Research Fellow-
ship to A. B., Sanction No. 09/028(0717)/2008-EMR-I], the
British Engineering and Physical Sciences Research Council
(EPSRC) and the University of Reading for funds for the
X-Calibur system. We also thank CONSOLIDER–Ingenio 2010
(projects CSD2010-0065), the MICINN of Spain (projects
CTQ2011-27512, FEDER funds), the Direcció General de
Recerca, Desenvolupament Tecnològic i Innovació del Govern
Balear (Accions Especials, 2011) and the Spanish Government
(Grant CTQ2009-07264) for their financial support. We thank
the CESCA for computational facilities.
15 D. Venegas-Yazigi, D. Aravena, E. Spodine, E. Ruiz and S. Alvarez,
Coord. Chem. Rev., 2010, 254, 2086 and references therein.
16 M. Stylianou, C. Drouza, Z. Viskadourakis, J. Giapintzakis and
A. D. Keramidas, Dalton Trans., 2008, 6188 and references therein.
17 (a) K. D. Karlin and J. Zubieta, Biological and Inorganic Copper Chem-
istry, Academic Press, New York, 1986; (b) Bioinorganic Chemistry of
Copper, ed. K. D. Karlin, Z. Tyeklar, Chapman & Hall, New York, 1993.
18 (a) D. Venegas-Yazigi, S. Cortés, V. Paredes-Garcia, O. Pena, A. Ibanez,
R. Baggio and E. Spodine, Polyhedron, 2006, 25, 2072; (b) X. Wang,
J. Ding and J. J. Vittal, Inorg. Chim. Acta, 2006, 359, 3481; (c) R. Gupta,
S. Mukherjee and R. Mukherjee, J. Chem. Soc., Dalton Trans., 1999,
4025; (d) H.-P. Zhang, J.-X. Zhu, H. Zhoua and Z.-Q. Pan, Acta Crystal-
logr., Sect. E: Struct. Rep. Online, 2008, E64, m340; (e) S. K. Dutta,
U. Florke, S. Mohanta and K. Nag, Inorg. Chem., 1998, 37, 5029;
(f) A. Biswas, M. G. B. Drew, J. Ribas, C. Diaz and A. Ghosh, Inorg.
Chim. Acta, 2011, 379, 28.
References
1 G. B. Kauffman, J. Chem. Educ., 1959, 36, 521.
2 (a) D. Cornacchia, R. Z. Pellicani, F. P. Intini, C. Pacifico and G. Natile,
Inorg. Chem., 2009, 48, 10800; (b) Z. Ciunik, J. A. Wolny, M. F. Rudolf
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 12200–12212 | 12211

View Article Online
19 (a) L. Rigamonti, A. Cinti, A. Forni, A. Pasini and O. Piovesana,
Eur. J. Inorg. Chem., 2008, 3633; (b) A. Biswas, M. G. B. Drew, Y. Song
and A. Ghosh, Inorg. Chim. Acta, 2011, 376, 422.
20 A. Biswas, M. G. B. Drew, C. J. Gómez-García and A. Ghosh, Inorg.
Chem., 2010, 49, 8155.
21 (a) A. Biswas, M. G. B. Drew and A. Ghosh, Polyhedron, 2010, 29,
1029; (b) A. Biswas, R. Saha and A. Ghosh, CrystEngComm, 2011, 13,
5342.
22 CrysAlis, Oxford Diffraction Ltd., Abingdon, U.K., 2006.
23 G. M. Sheldrick, Shelxs97 and Shelxl97, programs for crystallographic
solution and refinement, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, A64, 112.
24 ABSPACK Program for Absorption Corrections, Oxford Diffraction,
Abingdon, UK, 2007.
25 R. Ahlrichs, M. Bär, H. Hacer and C. Kömel, Chem. Phys. Lett., 1989,
162, 165.
26 (a) L. Noodleman, J. Chem. Phys., 1981, 74, 5737; (b) L. Noodleman
and E. R. Davidson, Chem. Phys., 1986, 109, 131.
27 L. Noodleman, C. Y. Peng, D. A. Case and J. M. Mouesca, Coord. Chem.
Rev., 1995, 144, 199.
28 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov,
J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
Gaussian 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009.
29 N. Onofrio and J.-M. Mouesca, Inorg. Chem., 2011, 50, 5577.
30 (a) F. Neese, Coord. Chem. Rev., 2009, 253, 526; (b) F. Neese, T. Petrenko,
D. Ganyushin and G. Olbrich, Coord. Chem. Rev., 2007, 251, 288.
31 (a) C. Adamo, V. Barone, A. Bencini, R. Broer, M. Filatov,
N. M. Harrison, F. Illas, J. P. Malrieuand and I. de P. R. Moreira,
J. Chem. Phys., 2006, 124, 107101; (b) C. Adamo and V. Barone,
J. Chem. Phys., 1999, 110, 6158; (c) C. Adamo, A. di Matteo and
V. Barone, Adv. Quantum Chem., 2000, 36, 45.
32 J. Miralles, O. Castell, R. Caballol and J. P. Malrieu, Chem. Phys., 1993,
33, 172.
33 B. Kozlevčar, N. Kitanovski, Z. Jagličić, N. A. G. Bandeira, V. Robert,
B. Le Guennic and P. Gamez, Inorg. Chem., 2012, 51, 3094.
34 C. Biswas, M. G. B. Drew, E. Ruiz, M. Estrader, C. Diaz and A. Ghosh,
Dalton Trans., 2010, 39, 7474.
35 S. S. Tandon, L. K. Thompson, M. E. Manuel and J. N. Bridson, Inorg.
Chem., 1994, 33, 5555.
36 A. W. Addison, T. N. Rao, J. Reedijk, J. Van Rijn and G. C. Verschoor,
J. Chem. Soc., Dalton Trans., 1984, 1349.
37 (a) S. Khatua, S. H. Choi, J. Lee, K. Kim, Y. Do and D. G. Churchill,
Inorg. Chem., 2009, 48, 2993; (b) E. Katsoulakou, N. Lalioti,
C. P. Raptopoulou, A. Terzis, E. Manessi-Zoupa and S. P. Perlepes, Inorg.
Chem. Commun., 2002, 5, 719; (c) Z.-L. You, D.-H. Shi and H.-L. Zhu,
Inorg. Chem. Commun., 2006, 9, 642; (d) A. Tamilselvi, M. Nethaji and
G. Mugesh, Chem.–Eur. J., 2006, 12, 7797; (e) K. Ghosh, P. Kumar,
N. Tyagi and U. P. Singh, Inorg. Chem., 2010, 49, 7614; (f) H.-Y. Chen
and C.-C. Lin, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2006, 62,
m1693.
38 (a) H.-C. Fang, X.-Y. Yi, Z.-G. Gu, G. Zhao, Q.-Y. Wen, J.-Q. Zhu,
A.-W. Xu and Y.-P. Cai, Cryst. Growth Des., 2009, 9, 3776; (b) S. Basak,
S. Sen, C. Marschner, J. Baumgartner, S. R. Batten, D. R. Turner and
S. Mitra, Polyhedron, 2008, 27, 1193.
39 J. A. K. Howard, G. Ilyashenko, H. A. Sparkes and A. Whiting, Dalton
Trans., 2007, 2108.
40 E. Ruiz, J. Cano, S. Alvarez and P. Alemany, J. Am. Chem. Soc., 1998,
120, 11122.
41 J. S. Costa, N. A. G. Bandeira, B. Le Guennic, V. Robert, P. Gamez,
G. Chastanet, L. Ortiz-Frade and L. Gasque, Inorg. Chem., 2011, 50, 5696.
12212 | Dalton Trans., 2012, 41, 12200–12212
This journal is © The Royal Society of Chemistry 2012