Magnetostructural correlation in tetracopper(II) cubanes

Article abstract of DOI:10.1016/j.poly.2013.12.010

The tetradendate Schiff base ligand N-2-hydroxy-3-methoxybenzaldehyde- α-amino-iso-butyrate (H₂3-MeOSAIB) has been complexed with copper(II) giving rise to a tetranuclear complex [Cu₄(3-MeOSAIB) ₄]·H₂O with cubane structure classified, according to the number of Cu-Cu contacts and bridging angles Cu-O-Cu, as 4+2 (type II) cubane. Magnetic data is reconstructed with two exchange coupling constants. The open angle Cu1-O1-Cu2 (α = 114) existing four times facilitates a strong antiferromagnetic coupling that is in competition with ferromagnetic interactions through the path Cu1-O2-Cu2 (α^- = 88) occurring four times; the resulting coupling constant is J₄/hc = -80.5 cm ^-1 (4×). The second exchange coupling constant occurring twice is J₂/hc = +6.2 cm^-1 (2×) as a result of weak ferromagnetic interaction via bridges Cu1-O1-Cu1 and Cu2-O2-Cu2 (β¹ = 92 and β² = 94). The analysis of available experimental data reveals a transparent magnetostructural correlation of J ₂ versus β (mean value), according to a straight line with negative slope; J₄ weakly correlates with the angle α. A deep chemometric analysis has been applied in order to unhide latent magnetostructural correlations.

Full text of DOI:10.1016/j.poly.2013.12.010

Polyhedron 70 (2014) 52–58
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Magnetostructural correlation in tetracopper(II) cubanes
a
b
c
a
a,
Z. Puterová-Tokárová a, , V. Mrázová , J. Kozíšek , J. Valentová , B. Vranovicová , R. Boca
⇑
⇑
ˇ
ˇ
ˇ
a Department of Chemistry, Faculty of Natural Sciences, University of SS Cyril and Methodius, 91701 Trnava, Slovakia
b Department of Physical Chemistry, FCHPT, Slovak University of Technology, 81237 Bratislava, Slovakia
c Department of Chemical Theory of Drugs, Faculty of Pharmacy, Comenius University, Bratislava, Slovakia
a r t i c l e i n f o
a b s t r a c t
Article history:
The tetradendate Schiff base ligand N-2-hydroxy-3-methoxybenzaldehyde-a-amino-iso-butyrate (H23-
Received 6 November 2013
Accepted 11 December 2013
Available online 18 December 2013
MeOSAIB) has been complexed with copper(II) giving rise to a tetranuclear complex [Cu4(3-MeOSAIB)4]-
ꢀH2O with cubane structure classified, according to the number of Cuꢀ ꢀ ꢀCu contacts and bridging angles
Cu–O–Cu, as 4+2 (type II) cubane. Magnetic data is reconstructed with two exchange coupling constants.
The open angle Cu1–O1–Cu2 (a = 114°) existing four times facilitates a strong antiferromagnetic coupling
that is in competition with ferromagnetic interactions through the path Cu1–O2–Cu2 (a⁄ = 88°) occurring
four times; the resulting coupling constant is J4/hc = ꢁ80.5 cmꢁ1 (4ꢂ). The second exchange coupling con-
stant occurring twice is J2/hc = +6.2 cmꢁ1 (2ꢂ) as a result of weak ferromagnetic interaction via bridges
Cu1–O1–Cu1 and Cu2–O2–Cu2 (b1 = 92 and b2 = 94°). The analysis of available experimental data reveals
a transparent magnetostructural correlation of J2 versus b (mean value), according to a straight line with
Keywords:
Copper(II) complex
Cubane
Schiff base
Amino-acid
X-ray crystallography
Magnetostructural correlations
negative slope; J4 weakly correlates with the angle
a
. A deep chemometric analysis has been applied in
order to unhide latent magnetostructural correlations.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
bipyridine [16], phenanthroline [17], urea [18] or thiocyanato
group [19]. However, resulting square-pyramidal, distorted
square-bipyramidal or mixed square-planar/square-pyramidal
polyhedra consists mostly of monomeric or dimeric units. From a
magnetic behavior point of view, tetracopper(II) cubanes with
{Cu4O4} core are of interest and their magnetic properties are tun-
able by making a small variations in their structural parameters
[20–25]. Depending on the Cu–O–Cu bridging angles and the pla-
nar/pyramidal nature of the bridging oxygen in the {Cu4O4} core,
copper(II) cubanes can show dominant ferro- as well as antiferro-
magnetic exchange interactions [26]. A value of 97.5° for the Cu–
O–Cu angle marks the border between ferro- (Cu–O–Cu < 97.5°)
and antiferromagnetic (Cu–O–Cu > 97.5°) behavior of the com-
plexes [24,27]. To the best of our knowledge, among the several
copper(II) cubanes with Schiff base ligands reported to date
[23,28–36], there is no report regarding magneto-structural corre-
lations of copper(II) cubane containing amino acid as a part of a
Schiff base ligand. This is probably due to the fact, that the majority
of Schiff base ligands in such complexes are produced by conden-
sation reaction of salicylaldehyde and amino acids and those pref-
The coordination chemistry of copper(II) complexes with Schiff
base ligands has been gaining importance in current years because
of their structural diversity [1], possible utility in modeling of the
active sites of metalloenzymes [2] and molecular magnetism, with
special emphasis to single-molecule magnets [3]. Among them, the
N-salicylidene-aminoacidato copper(II) complexes containing ami-
no-acids are of particular interest [4–12]. The copper(II) ion in such
complexes adopts a square-pyramidal (4+1) geometry which is
formed by an O, N,O0-tridentate N-salicylidene-aminoacidato dian-
ion and the aqua-ligand O-atom in the basal plane. The apical posi-
tion is occupied by oxygen from the second aqua ligand. This
unique distorted square-pyramidal arrangement around the Cu(II)
ion gives Schiff base copper(II) chelates the potential to act as valu-
able non-enzymatic models for more complicated copper-contain-
ing metalloproteins, such as Cu,Zn-superoxide-dismutaze [13] or
catechol-oxidaze [14]. On the other hand, multinuclear copper(II)
complexes are not only relevant to the active-site properties of
copper oxidases, but also in the field of molecular magnetism
[15]. One of the strategies to obtain multicopper(II) Schiff base che-
lates containing amino acids is based on the synthesis with addi-
tional molecular ligands instead of water, derived mostly from
erably exhibit
a tridentate coordination mode around the
copper(II) centre resulting in mononuclear or dinuclear species
[37].
In order to obtain a new polynuclear cluster, 2-hydroxy-3
-metyloxybenzaldehyde was used because of its unique chelating
and bridging capabilities. The new Schiff base – N-2-hydroxy-3
⇑
Corresponding authors. Tel.: +421 905763967.
-methoxybenzaldehyde-a-amino-iso-butyrate (H23-MeOSAIB, Fig. 1),
ˇ
E-mail address: roman.boca@stuba.sk (R. Boca).
0277-5387/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.poly.2013.12.010

Z. Puterová-Tokárová et al. / Polyhedron 70 (2014) 52–58
53
V = 4874.72(7) Å3; Z = 8; Dcalc = 1.678 gꢀcmꢁ3
; l ;
= 1.803 mmꢁ1
F(000) = 2528; reflections collected, 29669; independent reflec-
tions,
4957;
r
Rint = 0.0685;
goodness
of
fit = 1.061;
O
R1[F2 > 2 (F2)] = 0.0433; R1(F2) = 0.0584; wR2(F2) = 0.1068. CCDC
N
deposit No. 916294.
OH
The temperature and field dependence of the magnetic moment
for A have been taken using a SQUID apparatus (MPMS-XL7, Quan-
tum Design) in the RSO mode of detection. The data at the applied
field of B = 0.1 T was converted to the molar magnetic susceptibil-
ity and then to the effective magnetic moment.
OH
O
Fig. 1. Sketch of the tetradentate Schiff base ligand H23-MeOSAIB (left) and a
truncated structure of the {Cu4O4} core of [Cu4(3-MeOSAIB)4]ꢀH2O with its donor
set. Red (bold) marks the bridging (tetrabonding) oxygen atom linking three
copper(II) units of {Cu4O4} core in the cubane.
3. Results and discussion
contrary to its tridentate analoque N-salicylidene-a-amino-iso-
3.1. Structural data
butyrate (H2SAIB) [13], acts as a tetradentate ligand that facilitates
formation of a tetracopper(II) complex [Cu4(3-MeOSAIB)4]ꢀH2O
with cubane structure (hereafter A).
Since a tetradentate Schiff base with four donor atoms was
used, the final compound A is a tetranuclear species with the phe-
nolic-oxygen (OH) linking two copper(II) centers in an axial-equa-
torial fashion (Fig. 2). The final cubane structure is accomplished
via bridge of alkoxo-oxygen (OCH3) with neighboring copper(II)
ion. In the unit cell there are four C12H13O4NCu molecules, con-
nected with water molecules via strong hydrogen bonds (Table 1).
The coordination polyhedron is a distorted tetragonal bipyramid
with N1, O1, O2 and O4 (for Cu1) and N2, O1, O2⁄ and O7 (for
Cu2) donor atoms in the equatorial plane, with O3, resp. O6 donor
atom in the apical position and completed by weak interaction in
the sixth position by oxygen O1⁄ and O2 as well.
2. Experimental
A tetranuclear copper(II) complex was synthesized by a general
route using Schiff base ligand which results from condensation of
a-amino-iso-butyric acid and 2-hydroxy-3-methoxybenzaldehyde
[38]. All reagents were used as received form commercial sources.
Reagents used for the physical measurements were of spectro-
scopic grade. The Schiff-condensation of 2-hydroxy-3-metoxy-
benzaldehyde with
a-amino-iso-butyric acid at a 1:1 ratio in
The strongest bonds are from nitrogen atoms (Cu1–N1, Cu2–N2
of 1.930(3) and 1.924(3) Å), and oxygen atoms of carboxylato oxy-
gen (Cu1–O4, Cu2–O7 of 1.932(2) and 1.934(2) Å). The equatorial
plane is completed by Schiff base oxygen atoms. Phenolato oxygen
atoms are bonded similarly (Cu1–O3, Cu2–O6 of 2.792(2) and
2.792(2) Å). Four molecules form a unit of composition [Cu4(3-
MeOSAIB)4]ꢀH2O similar to a cubane molecule (eight carbon atoms
are replaced by Cu4O4). According to Mergehem and Haase [44]]
classification of this tetrameric cubane core refers to type II
water–ethanol solution for 3 h at room temperature resulted in
corresponding tetradentate ligand H23-MeOSAIB. To a water/etha-
nol solution (1:1, 60 cm3) of H23-MeOSAIB (10 mmol), an aqueous
solution of copper acetate dihydrate (10 mmol in 60 cm3) was
added and stirred for 1 h at 50 °C accompanied by a color change
to dark green. The resultant reaction mixture was filtered and
the filtrate was left to crystallize spontaneously for several days
at room temperature. Dark green needle-shaped crystals of A of
good quality were formed, these were separated by filtration and
washed with cold ethanol. Yield 1.82 g (63.6%).
(class
4+2)
complex
with
long
Cu–O
distances
(O3ACu1@O6ACu2 = 2.279 Å) by apical coordination and eight
short ones ranging from 1.924 to 2.006 Å bonding in four basal
planes.
IR spectra were measured as KBr pellets (Schimadzu Impact
400) in the range 4000–400 cmꢁ1. Electronic spectra of water
and DMSO solutions of compounds (c = 3 ꢂ 10ꢁ4 mol dmꢁ3) were
measured in the 200–800 nm region (Shimadzu UV-3101PC spec-
trometer). Elemental analysis was performed using FlashEA 1112,
ThermoFinnigan. Content of copper was determined by electrolysis
(Pt-cathode, U = 2 V, I = 2–3 A) after mineralization of the com-
plexes in conc. H2SO4. Anal. Calc. for C24H28Cu2N2O6 (615.56): C,
46.8; H, 4.58; N, 4.55; Cu, 20.6. Found C, 46.5; H, 4.42; N, 4.94;
Cu, 20.8%. IR(KBr)
as.), 1360 (COOꢁsym), 1246 (OCH3), 1211 (OH, phenol), 772, 542,
440 cmꢁ1. The difference between [ as(COOꢁ)] and [ s(COOꢁ)] in
m
= 3445 (OH, water), 1625 (C@N), 1558 (COOꢁ-
m
m
frequencies indicates the monodentate coordination of carboxylato
group (
D
m
ꢃ 200 cmꢁ1). UV/Vis (DMSO, c = 3 ꢂ 10ꢁ3 mol dmꢁ3):
kmax = 639 (d–d), 370.5 (LMCT), 272, 234 nm.
The X-ray structure determination of A was carried out on Gem-
ini R CCD diffractometer (Oxford Diffraction). The diffraction data
were collected at 100(1) K by the standard method (Mo Ka,
k = 0.71073 Å); 9 omega multi-scan, 60 s exposition time. Data
reduction and empirical absorption corrections were performed
by CRYSALISRED program [39]. The structure was solved by direct
methods and refined by the full-matrix least-squares procedure
with SHELXS97 and SHELXL97 program package [40,41]. Analytical
absorption correction was applied using a multifaceted crystal
model [42]. Visualization of the structure was performed using
DIAMOND [43]. Crystal data: formula, C24H28Cu2N2O9; formula
weight, 615.56; crystal system, monoclinic; space group, C2/c;
a = 13.6876(9), b = 23.576(2), c = 15.1067(12) Å; b = 90.510(6)°;
Fig. 2. View of compound A. Thermal ellipsoids for non-hydrogen atoms are drawn
at 30% probability level. Symmetry0 code used: (⁄) 1 ꢁ x,y,3/2 ꢁ z; (⁄⁄) 1/2 + x,1/
2 ꢁ y + 2,z. Selected bond lengths (AÅ): N1–Cu1, 1.930(3); O1–Cu1, 1.967(2); O2–
Cu1, 2.006(2); O3–Cu1, 2.279(2); O4–Cu1, 1.932(2); O1a–Cu1, 2.730(8); N2–Cu2,
1.924(3); O1–Cu2, 1.985(2); O2–Cu2a, 1.968(2); O6–Cu2, 2.279(2); O7–Cu2,
1.934(2); O2–Cu2, 2.722(8); Cu1–Cu1⁄, 3.427(7); Cu2–Cu2⁄, 3.471(1); Cu1–Cu2,
3.316(7); Cu1⁄–Cu2⁄, 3.333(7).

54
Z. Puterová-Tokárová et al. / Polyhedron 70 (2014) 52–58
Table 1
(Here we applied a notation that J2 is common for two spin pairs
Bond angles (°) and hydrogen bond contacts in A.
whereas J4 is common for four spin pairs.) The molar susceptibility
was corrected for the presence of a paramagnetic impurity that
obeys the Curie law:
N1–Cu1–O4
N1–Cu1–O1
O4–Cu1–O1
N1–Cu1–O2
O4–Cu1–O2
O1–Cu1–O2
N1–Cu1–O3
O4–Cu1–O3
O1–Cu1–O3
O2–Cu1–O3
Cu1–O1–Cu2 (
84.34 (11)
92.64 (11)
N2–Cu2–O7
84.51 (11)
93.74 (10)
173.01 (11)
178.09 (11)
93.65 (10)
88.15 (9)
102.98 (11)
93.54 (10)
93.45 (9)
N2–Cu2–O2[a]
170.30 (10) O7–Cu2–O2[a]
178.42 (10) N2–Cu2–O1
vmol
¼
vtetramerð1 ꢁ xPIÞ þ 4xPIðNAl0l2B=kBÞg2PIsðs þ 1Þ=3T þ vTIM
95.17 (10)
88.08 (10)
102.78 (10) N2–Cu2–O6
O7–Cu2–O1
O2[a]–Cu2–O1
ð2Þ
The last term accounts for the uncompensated underlying dia-
magnetism and the temperature-independent paramagnetism,
along with the diamagnetic signal of the sample holder.
94.99 (10)
94.67 (9)
75.76 (9)
O7–Cu2–O6
O2[a]–Cu2–O6
O1–Cu2–O6
76.60 (9)
a
)
114.04 (14) Cu1–O1–Cu1 (b1) 92.28 (11)
The X-ray structure showed that there are two distinct crystal-
lographic centers Cu1 and Cu2 with distances Cu1–Cu1 = 3.427 and
Cu2–Cu2 = 3.471 Å, J2 = J(Cu1–Cu1) = J(Cu2–Cu2), whereas the four
remaining contacts are Cu1–Cu2 = 3.333, 3.316, 3.333, 3.316 Å, J4.
The fitting procedure converged to the following set of magnetic
Cu1–O2–Cu2 (a⁄
Hydrogen bonds and angles (Å, °)
)
87.70 (10)
Cu2–O2–Cu2 (b2) 94.10 (11)
D–Hꢀ ꢀ ꢀA
d(D–H)
0.82(5)
d(Hꢀ ꢀ ꢀA)
2.09(5)
2.23(5)
d(Dꢀ ꢀ ꢀA)
2.891(4)
2.933(4)
<(DHA)
166(4)
168(5)
O9–H9WAꢀ ꢀ ꢀO8
O9–H9WAꢀ ꢀ ꢀO5[b] 0.72(5)
parameters: J2/hc = +6.24, J4/hc = ꢁ80.5 cmꢁ1, g = 2.053,
vTIM = -
Symmetry code used [a]: ꢁx + 1,y,ꢁz; [b]:1/2 + x,1/2 ꢁ y + 2,z.
ꢁ1.8 ꢂ 10ꢁ9 m3 molꢁ1, xPI = 0.012. The quality of the fit is perfect
as is displayed in Fig. 3; the discrepancy factor R = 0.021. The re-
ported solution of the fitting procedure is stable (the error func-
tional does not change upon variation of J2).
3.2. Magnetic data
Notice, J2 is relatively small with respect to absolute value of J4.
The phen(oxido) donor atom links three Cu(II) centers: two con-
tacts exhibit usual bond lengths (1.97–2.01 Å) whereas one contact
in the base of the prism is much longer (2.7 Å). For these reasons
the fitting procedure was repeated by fixing J2 = 0. Then J4/
The temperature dependence of the effective magnetic moment
for A is displayed in Fig. 3. The room-temperature value is (leff 300-
K = 3.16 lB as compared to the spin-only value for four uncoupled
spins s = 1/2, i.e. (
eff HT = g[4s(s + 1)]1/2 = 3.46 lB when g = 2 is as-
)
l
)
hc = ꢁ81.2 cmꢁ1
,
g = 2.056,
v
TIM = ꢁ1.5 ꢂ 10ꢁ9 m3 molꢁ1 and
sumed. On cooling the effective magnetic moment decreases pro-
gressively that confirms sizable exchange coupling of an
antiferromagnetic nature. This is also confirmed by the tempera-
ture evolution of the molar magnetic susceptibility that passes
through a maximum at Tmax = 90 K. However, on further cooling
the susceptibility does not drop to zero, but it passes through a
minimum and then rises up. This is a fingerprint of some paramag-
netic impurity xPI(T) that used to be present in exchange coupled
systems. With the exchange coupling constant J ꢄ 0 the magneti-
zation at the low temperature (2 K) will reflect only the paramag-
netic impurity, therefore S = 0 ground state is separated from the
first excited state S = 1 by an amount of |J|.
xPI = 0.012 were obtained; the quality of the fit is approximately
the same (R = 0.020).
A
rather high negative exchange coupling constant J4 -
ꢆ ꢁ80 cmꢁ1 could originate in the open angle Cu1–O–Cu2
(a = 114°) occurring four times; such a coupling path prefers strong
antiferromagnetic exchange that is in competition with the ferro-
magnetic one proceeding via four Cu1–O–Cu2 (a⁄ = 88°) linkages;
the average angle aav = 101°. The critical angle when antiferromag-
netic coupling alters to ferromagnetic coupling has been reported
as Cu–O–Cu = 97.5° based upon the data for bis-hydroxido bridged
Cu(II) complexes [45]. However, this result is not fully transferable
to our phenoxido-bridged {Cu4O4} unit where the oxygen atom
links three Cu(II) centers and thus a different value of the critical
angle could exist.
Table 2 shows magnetic and structural data of selected com-
plexes with the cubane-type {Cu4O4} core; the CCDC offers 131
hits, however, the sources about reliable magnetic data is much
less numerous. These complexes exhibit a great variety of six
Cuꢀ ꢀ ꢀCu contacts and twelve bond angles Cu–O–Cu along the
The magnetic data was fitted by using the spin Hamiltonian that
accounts for the isotropic exchange of four Cu(II) centers
4
X
ꢀꢁ1
ꢀꢁ2
^
^
~
~
~
~
0
0
H ¼
l
BB gSAzh ꢁ J2½ðSCu1 ꢀ SCu1 Þ þ ðSCu2 ꢀ SCu2 Þꢅh
A
ꢀꢁ2
~
~
~
~
~
~
~
~
0
0
0
0
ꢁJ4½ðSCu1 ꢀ SCu2Þ þ ðSCu1 ꢀ SCu2 Þ þ ðSCu1 ꢀ SCu2Þ þ ðSCu1 ꢀ SCu2 Þꢅh
ð1Þ
0.15
0.10
6
5
Cu1
4
μ
3
μ
Cu1'
0.05
2
χ
1
Cu2'
0.00
0
0
50 100 150 200 250 300
T/K
Cu2
Fig. 3. Temperature dependence of the molar magnetic susceptibility (left axis) and the effective magnetic moment (right axis) for A. Circles – experimental data, lines –
fitted. Right – simplified structural motif.

Table 2
Magnetostructural relationships for selected Cu4-type complexes.a
No
Compound
Chromophore
g
J2/cmꢁ1 J4/cmꢁ1 Cu1–Cu1 Cu1–Cu2
Cu-O(bases) Cu-O(walls) b1 (Cu1–O–Cu1),
(8ꢂ) (4ꢂ)
b2(Cu2–O–Cu2)
a
(Cu1–O–Cu2),
bav
aav
Ground Refs.
state
R (%)
(2ꢂ)
(4ꢂ)
Cu2-Cu2
(4ꢂ)
a⁄ (Cu1⁄–O-Cu2⁄)
1
[{Cu(OCH2CH2NH2)}4] {CuO3N2}
[W(CN)8].2H2O
2.096 ꢁ3.4
+82.8
3.365
3.395
3.116, 3.119, 1.922, 1.958 1.942, 1.960 99.0, 101.4 (A)
3.129, 3.190 2.408, 2.451 1.953, 1.973 97.4, 101.5 (0, A) 89.3, 105.7 (F, A)
90.2, 107.4 (F, A)
99.8
98.6
S = 2
[21]b
3.7
1.928, 1.949
2.419, 2.557
89.2, 109.3 (F, A)
90.6, 106.9 (F, A)
89.1, 105.7 (F, A)
2
[{Cu(OCH2CH2NH2)}4] {CuO3N2}
[Mo(CN)8].2H2O
2.101 ꢁ2.2
+76.6
3.359
3.399
3.110, 3.118, 1.920, 1.964 1.963, 1.969 98.7, 101.0 (A)
99.7
98.5
S = 2
[21]b
8.1
3.133, 3.187 2.413, 2.444 1.940, 1.949 97.5, 101.5 (0, A) 90.1, 107.3 (F, A)
1.931, 1.953
2.420, 2.447
91.0, 107.3 (F, A)
89.2, 108.5 (F, A)
3
4
[{Cu(hsae)}4]ꢀ2H2O
4CH3CN
[Cu4(L)4]
{CuO3ON}
{CuO3ON}
2.02
2.18
ꢁ35.2
ꢁ11
+72.0
+61.0
3.443
3.615
3.305
3.297
3.108, 2ꢂ
3.134, 2ꢂ
1.960, 2.467 1.930, 1.956 101.5, 101.5 (A)
90.0, 104.7 (F, A), 2ꢂ 102.8 96.6
S = 2
S = 2
[46]b
[49]b
3.2
5.2
2.001, 2.561
104.2, 104.2 (A)
85.8, 106.0 (F, A), 2ꢂ
3.126, 3.122 1.955, 1.958 1.938, 1.950 96.6, 98.1
3.113, 3.139 2.400, 2.452 1.956, 1.962 97.5, 97.6
1.948, 1.954
91.0, 106.5
89.5, 106.9
90.3, 105.3
91.7, 106.6
97.4
99.5
98.5
98.5
2.413, 2.415
5
[Cu4(L)4]ꢀCH3CN
{CuO3ON}
2.09
ꢁ21.4
+54.6
3.359
3.442
3.171, 3.126, 1.957, 1.967 1.935, 1.939 98.6, 98.6 (A)
3.111, 3.170 2.454, 2.447 1.948, 1.954 100.0, 100.8 (A)
1.969, 1.969
90.2, 105.7 (F, A)
89.1, 105.1 (F, A)
89.9, 109.1 (F, A)
90.7, 108.5 (F, A)
89.7, 106.1, 2ꢂ
87.7, 102.3, 2ꢂ
S = 2
[30]
3.7
2.479, 2.502
6
[Cu4(L)4]ꢀ2H2O
{CuO3ON}
{CuO3ON}
{CuO3ON}
{CuO3ON}
{CuO3ON}
2.12
2.01
2.02
2.06
1.95
+3.0
+44.6
(av)
+38.4
3.480
3.420
3.438, 2ꢂ 3.124, 4ꢂ
3.132, 2ꢂ
3.060, 2ꢂ
1.973, 2.455 1.951, 1.969 103.1, 2ꢂ
102.2 96.4
S = 2
S = 2
S = 2
S = 0
S = 0
[36]
[23]
[23]
[23]
[48]b
5.5
6.9
3.7
9.9
6.0
1.968, 2.439
101.3, 2ꢂ
7
[Cu4(HL)4]ꢀ4H2O
ꢁ18.0
ꢁ15.6
ꢁ18.4
ꢁ26
1.974, 2.438 1.970, 4ꢂ
101.9, 101.9 (A)
101.9, 101.9 (A)
89.6, 104.7 (F, A), 2ꢂ 101.9 97.1
89.6, 104.7 (F, A), 2ꢂ
1.974, 2.438
8
[Cu4(L)2(OMe)2]
ꢀ2H2OꢀTHF
+33.1
+14.7
ꢁ50
3.419
3.384
3.433
3.458
3.438
3.387
3.077, 2ꢂ
3.116, 2ꢂ
3.107, 2ꢂ
3.103, 2ꢂ
1.948, 2.477 1.961, 1.961 100.5, 100.5 (A)
1.991, 2.319 103.2, 103.2 (A)
1.926, 2.451 1.964, 1.965 102.7, 102.7 (A)
1.969, 2.418 103.6, 103.6 (A)
91.5, 103.8 (F, A), 2ꢂ 101.8 97.0
88.4, 104.1 (F, A), 2ꢂ
9
[Cu4(L)4(OH)2]ꢀ6H2O
88.8, 104.4 (F, A), 2ꢂ 103.2 97.2
89.5, 105.9 (F, A), 2ꢂ
10
[Cu4(L)4]
3.113, 3.102, 1.975, 1.984 1.947, 1.951 98.5, 102.9
3.169, 3.092 2.401, 2.539 1.947, 1.960 99.3, 99.6
1.977, 1.986
89.2, 104.1
89.3, 105.1
88.9, 107.7
89.9, 103.2
100.1 97.2
2.441, 2.442
11
[Cu4(L)4] NaClO4
{CuO2(O)ON} 2.19
+27.2
ꢁ69.8
ꢁ80.5
3.486
3.435
3.427
3.471
3.229, 2ꢂ
3.253, 2ꢂ
3.316, 2ꢂ
3.333, 2ꢂ
1.958, 2.697 1.964, 1.971 94.3, 94.3 (F)
1.958, 2.680 95.7, 95.7 (F)
1.966, 2.730 1.985, 2.006 92.3, 92.3 (F)
1.967, 2.722 94.1, 94.1 (F)
86.2, 110.5 (F, A), 2ꢂ 95.0
99.1
S = 0
[47]b
9.7
87.4, 112.1 (F, A), 2ꢂ
12 (A)
[{Cu(3-MeOSAIB)}4]
ꢀH2O
{CuO2(O)ON} 2.053 +6.24
88.5, 114.0 (F, A), 2ꢂ 93.2
87.7, 114.0 (F, A), 2ꢂ
101.0 S = 0
This work 4.3
a
F – ferromagnetic coupling, A – antiferromagnetic coupling, 0 – nearly zero coupling. Expected critical angle: 97.5°.
Values of J conform the definition HAB ¼ ꢁJðSA ꢀ SBÞ; rescaled from original source when needed. Distances in Å, angles in °, coupling constants in cm
b
ꢁ1
^
~ ~
.

56
Z. Puterová-Tokárová et al. / Polyhedron 70 (2014) 52–58
superexchange paths so that a simple magnetostructural correla-
tion is not straightforward. There are also twelve Cu–O bond
lengths within the {Cu4O4} core and each O-atom is involved in
both coupling paths (along bases of the prism covered by J2, and
walls of the prism covered by J4).
however, differences in slope and intercept are evident. Such a
shift has already been observed when the hydroxido bridges were
substituted for alkoxido or phenoxido ones [53]. For instance, in
dinuclear copper(II) complexes bridged by phenoxido bridges the
correlation line is J½cmꢁ1ꢅ ¼ ꢁ31:95 ꢀ
a
½ꢇꢅ þ 2462 and ꢁJ values are
In doing a step forward to a transparent magnetostructural cor-
relation, the structural parameters were classified as they belong
to a bisphenoide of the D2d symmetry (a compressed or elongated
cubane). Then the distances Cu1–Cu1 and Cu2–Cu2 occur twice
whereas those of Cu1–Cu2 four times; consequently J2 (2ꢂ) and
J4 (4ꢂ) hold true. The problem of the magnetostructural correlation
in the cubane-type complexes lies in the feature that there are
twelve bonding angles Cu–O–Cu but only two coupling constants.
This means that some angles need averaging. Four angles are of the
type b1(Cu1–O–Cu1) and b2(Cu2–O–Cu2) (each usually twice) and
they determine the sign of J2. Remaining eight angles are of the
substantially higher than for the alkoxido or hydroxido cases. A
worth noting is also the correlation quoted for [Ni4(OR)4]-type
complexes [54].
It need be noticed that for a tetranuclear spin s = 1/2 system the
total spin is S = 0 (twice), S = 1 (three times) and S = 2 (once).
Depending upon values of the coupling constants the ground state
could be either S = 0, S = 1, or S = 2 since the formula for the energy
levels is
EðS12; S34; SÞ ¼ ðJ4 ꢁ J2Þ½S12ðS12 þ 1Þ þ S34ðS34 þ 1Þꢅ=2 ꢁ J4½SðS þ 1Þꢅ=2
ð3Þ
type
a
(Cu1–O–Cu2) and a⁄(Cu1⁄–O–Cu2⁄), and after a careful
when the following spin Hamiltonian applies
assignment of them one can estimate the sign (type) of the J4
parameter (F for ferromagnetic exchange, A for the antiferromag-
netic one).
ꢀꢁ2
~ ~
^
~
~
~
~
~
~
~
~
~
~
H ¼ ½ꢁJ2ðS1 ꢀ S2 þ S3 ꢀ S4Þ ꢁ J4ðS1 ꢀ S3 þ S1 ꢀ S4 þ S2 ꢀ S3 þ S2 ꢀ S4Þꢅh
ð4Þ
The magnetic and structural data from Table 2 served as a basis
for the magnetostructural correlations as displayed in Fig. 4. It can
be concluded that the coupling constant J2 occurring twice corre-
lates with the averaged value of bav according to a straight line
with negative slope J2½cmꢁ1ꢅ ¼ ꢁ3:61 ꢀ bav½ꢇꢅ þ 351 (|r| = 0.65). The
second correlation concerns the coupling constant J4 (occurring
4ꢂ) that (weakly) correlates with the averaged bonding angle
The intermediate spins are S12 = 0, 1 and S34 = 0, 1 and then
Eð0; 0; 0Þ ¼ 0
ð5Þ
ð6Þ
ð7Þ
ð8Þ
ð9Þ
Eð0; 1; 1Þ ¼ Eð1; 0; 1Þ ¼ ꢁJ2
Eð1; 1; 0Þ ¼ 2ðJ4 ꢁ J2Þ
Eð1; 1; 1Þ ¼ J4 ꢁ 2J2
Eð1; 1; 2Þ ¼ ꢁJ4 ꢁ 2J2
Cu–O–Cu
a
; J4½cmꢁ1ꢅ ¼ ꢁ19:93 ꢀ aav½ꢇꢅ þ 1976 (|r| = 0.45). Relatively
low correlation coefficient (r) in the aforementioned correlations
results from the fit according to a straight line, though there are
no physical arguments that such a correlation is necessarily linear
in the given group of compounds. Last but not least complication is
that the magnetic data was taken by different groups using differ-
ent hardware operating at different conditions and the data was
also analyzed differently. A possible reason for weakness of the
second correlation J4–aav (|r| = 0.45 < rcrit = 0.576) could originate
Magnetic data in Table 2 shows that the competition of the neg-
ative J2 and positive J4 coupling constants leads to the magnetic
ground state S = 2 for chromophores {CuO3N2} and {CuO3ON}. In
two cases of the chromophore {CuO3ON} the ground state is S = 0
and the diamagnetic ground state also applies for {CuO3O2N} chro-
mophore, including also the complex under study.
In order to perform a more deep statistical analysis, the struc-
tural and magnetic data was treated by contemporary chemomet-
ric software by which three kinds of analysis has been performed.
For such a reason 10 variables have been selected: the coupling
constants J2 and J4, the magnetogyric factor g, the average angles
aav (abbr. Aav) and bav (abbr. Bav) as well as averages of distances
at the bases (abbr. Cuav) and walls of the prism (Cu12av); in addi-
tion three Cu–O distances were involved: average of 4 Cu–O dis-
tances along walls of the prism (Rav), average of 4 shorter and
average of 4 longer distance within bases (abbr. P and RLav).
1. The correlation analysis (Pearson method – Table 3) identifies
these significant (the correlation coefficients |r| > rcrit = 0.576) or
nearly significant correlations.
in averaging of very different angles
a ranging between 86° and
114°.
Correlations suggested in the present publication are qualita-
tively analogous to that outlined by Hatfield [50,51] and theoreti-
cally studied by Ruiz [52] in dinuclear complexes of [Cu–(OH)2–Cu]
type where J½cmꢁ1ꢅ ¼ ꢁ77:96 ꢀ
a
½ꢇꢅ þ 7615 (|r| = 0.99) holds true;
average of 4 angles β
regression for J2 vs βav
100
50
confidence interval
prediction interval
average of 8 angles α
regression for J4 vs αav
(a) By inspecting the first column, the coupling constant J2 cor-
relates positively with g, Cu12av and RLav; it correlates neg-
atively with Bav in conformity with Fig. 4. (A negative
correlation means that with increase of the first parameter
the second one is decreasing and vice versa.)
complex A
0
(b) By analyzing the second column, the coupling constant J4
correlates positively with Bav and negatively with Cu12av,
Rav and RLav; a weak negative correlation exists for J4 versus
Aav (as displayed in Fig. 4).
-50
(c) There are some more correlations among selected parame-
ters; some of them are obvious from the geometrical
reasons.
-100
92
94
96
98
100
102
104
Cu-O-Cu /o
Based upon these data, plots of additional graphs are justified,
at least significant negative correlation J4 versus Rav(Cu–O) and
weak-negative correlation J2 versus Pav(Cu–O) as displayed in
Fig. 5.
Fig. 4. Magnetostructural corelation J2–bav and J4–aav in the cubane-type [Cu4O4]
complexes. Confidence and prediction intervals (95%) are shown for J2–bav
correlation.

Z. Puterová-Tokárová et al. / Polyhedron 70 (2014) 52–58
57
Table 3
Correlation table containing pairwise correlation coefficients r for continuous variables; software SPSS 15.0, critical value rcrit = 0.576.
Variables
J2
J4
g
Cuav
Cu12av
Bav
Aav
Rav
Pav
J4
g
ꢁ0.432
0.687
0.105
Cuav
Cu12av
Bav
ꢁ0.073
ꢁ0.283
ꢁ0.698
0.616
ꢁ0.345
0.224
ꢁ0.439
0.372
0.607
0.185
0.205
ꢁ0.894
0.884
0.722
ꢁ0.669
0.570
0.543
Aav
Rav
ꢁ0.452
ꢁ0.241
ꢁ0.919
ꢁ0.536
ꢁ0.004
0.179
0.562
ꢁ0.649
Pav
RLav
ꢁ0.414
ꢁ0.29
ꢁ0.576
0.511
0.431
ꢁ0.047
0.171
ꢁ0.381
ꢁ0.028
0.591
ꢁ0.739
0.219
0.943
ꢁ0.789
0.712
0.628
0.095
a
The significant correlations are bold-typed; correlations displayed in Figs. 4 and 5 are underlined.
average of 4 Pav(bases)
average of 4 Rav(walls)
regression for J2 vs Pav(bases)
regression for J4 vs Rav(walls)
100
50
100
50
confidence interval
prediction interval
confidence interval
prediction interval
0
0
-50
-50
-100
-100
1.93 1.94 1.95 1.96 1.97 1.98 1.99 2.00
1.93 1.94 1.95 1.96 1.97 1.98 1.99 2.00
Rav(walls)/A
Pav(bases)/A
Fig. 5. Left - attempt to correlate J4 vs. Rav(Cu–O) where four Cu–O distances occurring only along walls of the prism (coupling path covered by J4) were averaged. Correlation
coefficient |r| = 0.65. Right – attempt to correlate J2 vs. Pav(Cu–O) where four shorter Cu–O distances occurring only in bases of the prism (coupling path covered by J2) were
averaged. Correlation coefficient |r| = 0.41.
2. This approach clearly reduces 45 possible pair-wise correla-
100
80
60
40
20
0
tions to the very limited number of significant correlations.
3. The analysis of principal components (PCA) leads to a reduc-
tion of a number of related variables to a couple of main compo-
nents. Thus a biplot (PC2 versus PC1) relates variables and
objects; the angle between two variables is inversely proportional
to the degree of correlation between them (Fig. 6).
4. Finally, the cluster analysis allows a classification of objects
and variables into groups based upon their similarities (Fig. 7).
Important finding is that the objects (complexes) form three
groups (numbering according to Table 2). The compounds 11 and
12 (this work) possessing the chromophore {CuO3O2N} form a
3
2
Pav
120
100
80
60
40
20
0
Cuav
3
10
12
RLav
7
Rav
1
8
Cu12av
Bav
0
6
11
9
5
Aav
J2
-1
-2
-3
J4
2
1
g
4
-4
-2
0
2
4
6
PC1 (52.4%)
Fig. 6. PCA biplot showing 10 selected variables and 12 objects in the PC2–PC1
Fig. 7. Dendogram (Ward’s method, squared Euclidean) for clustering of variables
plane; software Statgraphics Centurion XV.
(top) and objects (bottom); software Statgraphics Centurion XV.

58
Z. Puterová-Tokárová et al. / Polyhedron 70 (2014) 52–58
[3] S. Miller, M. Drillon (Eds.), Magnetism, Molecules to Materials, Vols. 1–5,
WILEY-VCH, Weinheim 2001–2005.
[4] D.E. Metzler, M. Ikawa, E.E. Snell, J. Am. Chem. Soc. 76 (1954) 648.
[5] M. Kohútová, A. Valent, Chem. Pap. 59 (2002) 95.
[6] A. Valent, M. Kohútová, O. Švajlenová, D. Hudecová, P. Olejníková, M. Melník, J.
Coord. Chem. 57 (2004) 1279.
[7] K. Bauerová, J. Valentová, S. Poništ, J. Navarová, D. Komendová, D. Mihalová,
Biologia 60 (2005) 65.
[8] H.M. Parekh, M.N. Patel, Russ. J. Coord. Chem. 32 (2006) 431.
[9] J. Lv, T. Liu, S. Cai, X. Wang, L. Liu, Y. Wang, J. Inorg. Biochem. 100 (2006) 1888.
[10] Y. Li, Z. Yang, B. Wang, Transition Met. Chem. 31 (2006) 598.
[11] Y.H. Fan, S.Y. Bi, Y.Y. Li, C.F. Bi, S.T. Xie, Russ. J. Coord. Chem. 34 (2008) 772.
common group that is different from all other objects; these ob-
jects have J4 strongly negative. A distinct group is formed by com-
plexes 1 and 2 having the {CuO3N2} chromophore where also the
complexes 4 and 5 belong to; these objects have J4 strongly posi-
tive. The third group of compounds is formed by remaining six
complexes possessing the {CuO3ON} chromophore; the coupling
constants J2 and J4 vary between positive and negative values. Such
a grouping of objects is also visible at the biplot in Fig. 6.
To this end one can conclude that all three chemometric ap-
proaches yield essentially the same results being much more infor-
mative in comparison with attempts to correlate coupling
constants with the bond angles according to a straight line. How-
ever, it need be mentioned, that since the quality of the structural
data can be easily evaluated using the R-factor, no such a straight-
forward identification can be attributed to the magnetic data.
ˇ
ˇ
ˇ
[12] J. Vanco, J. Marek, Z. Trávnícek, E. Racanská, J. Muselík, J. Inorg. Biochem. 102
(2008) 595.
ˇ
[13] Z. Puterová, J. Valentová, Z. Bojková, J. Kozíšek, F. Devínsky, Dalton Trans. 40
(2011) 1484.
[14] B. Sreenivasulu, M. Vetrichelvan, F. Zhao, S. Gao, J.J. Vittal, Eur. J. Inorg. Chem.
22 (2005) 4635.
[15] M. Mannini, F. Pineider, P. Sainctavit, C. Danielli, E. Otero, C. Sciancalepore,
A.M. Talarico, M. Arrio, A. Cornia, D. Gatteschi, R. Sessoli, Nat. Mater. 8 (2009)
194.
[16] P.A.N. Reddy, M. Nethaji, A.R. Charkravarty, Inorg. Chim. Acta 337 (2002) 450.
[17] P.A.N. Reddy, M. Nethaji, A.R. Charkravarty, Eur. J. Inorg. Chem. (2004) 1440.
4. Conclusion
ˇ
[18] O. Švajlenová, J. Vanco, J. Marek, Acta Crystallogr., Sect. C 60 (2004) m275.
ˇ
ˇ
[19] J. Vanco, Z. Trávnícek, J. Marek, R. Herchel, Inorg. Chim. Acta 363 (2010) 3887.
[20] W. Plass, Coord. Chem. Rev. 253 (2009) 2286.
Using 2-hydroxy-3-methoxybenzaldehyde a Schiff base ligand
N-2-hydroxy-3-methoxybenzaldehyde-a-amino-iso-butyrate
(H23-MeOSAIB) has been synthesized. This molecule acts as a tet-
–
[21] Z.-X. Wang, X.-L. Li, B.-L. Liu, H. Tokoro, P. Zhnag, Y. Song, S.-I. Ohkoshi, K.
Hashimoto, X.-Z. You, Dalton Trans. (2008) 2103.
[22] X. Zhang, B. Li, J. Tang, J. Tian, G. Huang, J. Zhang, Dalton Trans. 42 (2013) 3308.
[23] A. Mukherjee, R. Raghunathan, M.K. Saha, M. Nethaji, S. Ramasesha, A.R.
Chakravarty, Chem. Eur. J. 11 (2005) 3087.
[24] J. Tercero, E. Ruiz, S. Alavarez, A. Rodríguez-Fortea, P. Alemany, J. Mater. Chem.
16 (2006) 2729.
[25] E. Ruiz, A. Rodríguez-Fortea, P. Alemany, S. Alvarez, Polyhedron 20 (2001)
1323.
[26] C.J. Calzado, Chem. Eur. J. 19 (2013) 1254.
[27] V.H. Crawford, H.W. Richardson, J.R. Wasson, D.J. Hodgson, W.E. Hatfield,
Inorg. Chem. 15 (1976) 2107.
[28] P. Bhowmik, N. Aliaga-Alcalde, V. Gómez, M. Corbella, S. Chattopadhyay,
Polyhedron 49 (2013) 269.
[29] S.G. Naik, A. Mukherjee, R. Raghunathan, M. Nethaji, S. Ramasesha, A.R.
Chakravarty, Polyhedron 25 (2006) 2135.
[30] Z. Lu, T. Fan, W. Guo, J. Lu, Ch. Fan, Inorg. Chim. Acta 400 (2013) 191.
[31] J.-W. Lu, Y.-H. Hunag, S.-I. Lo, H.-H. Wei, Inorg. Chem. Commun. 10 (2007)
1210.
radendate ligand which facilitates the formation of a novel tetra-
nuclear complex [Cu4(3-MeOSAIB)4]ꢀH2O with
a cubane-like
structure. The distances of the cube edges can be distinguished into
two types: eight short ones ranging from 1.924 to 2.006 Å occur-
ring as the bonds in four basal planes, and other four longer ones
being of value 2.279 Å by apical coordination. According to those
parameters the architecture of the tetranuclear core is classified
as a type II or class 4+2 with the intra-cubane Cuꢀ ꢀ ꢀCu distances
3.427 and 3.471 Å in the top and bottom, respectively, and 3.316
and 3.333 Å in the side faces. The magnetic behavior of A could
be reconstructed by considering two coupling constants;
J2 = J(Cu1–Cu1) = J(Cu2–Cu2) = +6.24 cmꢁ1 (consistent with the an-
gle Cu1–O–Cu1 and Cu2–O–Cu2 being 92° and 94°, respectively)
and a negative coupling constant J4 = ꢁ80.5 cmꢁ1 (four remaining
interactions) revealing antiferromagnetic exchange along the open
[32] H. Liu, H. Wang, H. Wu, D. Niu, J. Coord. Chem. 58 (2005) 1345.
[33] Y.-C.h. Shi, Ch.-X. Sui, H.-B. Song, P.-M. Jang, J. Coord. Chem. 58 (2005) 363.
[34] M. Nihei, A. Yoshida, S. Koizumi, H. Oshio, Polyhedron 26 (2007) 1997.
[35] A. Burkhardt, E.T. Spielberg, H. Gorls, W. Plass, Inorg. Chem. 47 (2008) 2485.
[36] S. Thakurta, P. Roy, R.J. Butcher, M.S. El Fallah, J. Tercero, E. Garribba, S. Mitra,
Eur. J. Inorg. Chem. (2009) 4385.
angle Cu1–O–Cu2 (a = 114°), which is in competition with the fer-
romagnetic one proceeding via linkage Cu1–O–Cu2 (a⁄ = 88°). The
analysis of available experimental data reveals a transparent mag-
netostructural correlation of J2 versus b (mean value), according to
a straight line with negative slope, whereas J4 also (weakly) corre-
ˇ
[37] Z. Puterová-Tokárová, V. Mrázová, R. Boca, Polyhedron 61 (2013) 87.
[38] Ch.T. Yang, M. Vetrichelvan, X. Yang, B. Moubaraki, K.S. Murray, J.J. Vittal,
Dalton Trans. (2004) 113.
[39] Oxford Diffraction Ltd., CrysAlis CCD, CrysAlisRED CCD RED, Abington,
Oxfordshire, England, 2010.
lates with the angle a.
[40] G.M. Sheldrick, Acta Crystallogr., Sect. A 64 (2008) 112.
[41] G.M. Sheldrick, SHELXS97 and SHELXL, University of Göttingen, Germany, 1997.
[42] R.C. Clarck, J.S. Reid, Acta Crystallogr., Sect. A 51 (1995) 887.
[43] K. Brandenburg, DIAMOND, Visual Information System for Crystal Structures,
Bonn, Germany, 1998.
[44] R. Megehenn, W. Haase, Acta Crystallogr., Sect. B 33 (1977) 1877.
[45] O. Kahn, Molecular Magnetism, VCH, New York, 1993.
[46] H. Oshio, Y. Saito, T. Ito, Angew. Chem., Int. Ed. 36 (1997) 2673.
[47] J. Chakraborty, S. Thakurta, G. Pilet, D. Luneau, S. Mitra, Polyhedron 28 (2009)
819.
Acknowledgements
Slovak grant agencies (VEGA 1/0073/13, APVV-0014-11, APVV-
0202-10) are acknowledged for the financial support.
Appendix A. Supplementary data
[48] R. Wegner, M. Gottschaldt, H. Görls, E.-G. Jäger, D. Klemm, Chem. Eur. J. 7
(2011) 2143.
CCDC 916294 contains the supplementary crystallographic data
for 1. These data can be obtained free of charge via http://
www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; fax: +44 1223 336 033; or e-mail: deposit@ccdc.cam.ac.uk.
[49] C. Aronica, Y. Chumakov, E. Jeanneau, D. Luneau, P. Neugebauer, A.-L. Barra, B.
Gillon, A. Goujon, A. Cousson, J. Tercero, E. Ruiz, Chem. Eur. J. 14 (2008) 9540.
[50] W.E. Hatfield, in: R.D. Willett, D. Gatteschi, O. Kahn (Eds.), Magneto-Structural
Correlations in Exchange Coupled Systems, NATO ASI Series, Reidel, Dordrecht,
1985, p. 555.
[51] V.H. Crawford, H.W. Richardson, D.J. Wasson, W.E. Hodgson, W.E. Hatfield,
Inorg. Chem. 15 (1976) 2107.
[52] E. Ruiz, P. Alemany, S. Alvarez, J. Cano, J. Am. Chem. Soc. 119 (1997) 1297.
[53] L.K. Thompson, S.K. Mandal, S.S. Tandon, J.N. Bridson, M.K. Park, Inorg. Chem.
35 (1996) 3117.
[54] S. Mukherjee, T. Weyhermuller, E. Bothe, K. Weighardt, P. Chaudhuri, Eur. J.
Inorg. Chem. (2003) 863.
References
[1] P.G. Cozzi, Chem. Soc. Rev. 33 (2004) 410.
[2] S.J. Lippard, J.M. Berg, Principles of Bioinorganic Chemistry, University Science
Books, Mill Valley, CA, 1994.