Tailoring antibacteria agents: Sulfonamide-based dinuclear and 1D polymer Cu(II) complexes

Article abstract of DOI:10.1016/j.poly.2012.01.025

Starting from two bioactive sulfonamide-based ligands (HL1: 4-amino-N-(2,6-dimethoxy-4-pyrimidinyl)benzenesulfonamide; HL2: 4-amino-N-(4-methyl-2-pyrimidinyl)benzenesulfonamide) and Cu(II) salts ([Cu ₂(Ac)₄] and CuCl₂, Ac = acetate), the synthesis of [Cu₂(Ac)₂(L1)₂] (1) and [Cu ₂(L2)₄] (2) and {Cu(L1)₂(bipy)} _n·2nH₂O·n(CH₃OH) (1′) (bipy = 4,4′-bipy) architectures with potential antibiotic and antiseptic activities is reported. To confirm the possibility to fulfill specific criteria and ultimately combine both properties, the bioactive sulfonamide-based complexes are structurally, electrochemically and magnetically characterized. Depending on the synthesis conditions, the Cu:sulfonamide stoichiometry and Cu···Cu communication are first varied by controlling the chemical nature of the ancillary ligand (Ac or bipy). Then, electrochemistry data support the stability of 1 and 2 dinuclear complexes, and 1′ 1D polymer, a prerequisite for their bioactivity in solution. Interestingly, the synthesis leads to architectures where the (SO₂-Ph-NH₂) moiety which is responsible for the antibacterial activity remains non-coordinated in the vicinity of Cu(II) antiseptic ions. Magnetic susceptibility measurements combined to multireference wavefunction ab initio calculations evidence a rather strong antiferromagnetic behavior in the dinuclear compounds (H = -2JS₁S₂, 2J₁ = -307.8 cm^-1 in 1, 2J₂ = -63.2 cm^-1 in 2) whereas chain 1′ is paramagnetic. The cooperativity quantified by the hopping integral which is available from the ab initio calculations of the exchange coupling constant is reduced by a factor of two when the number of sulfonamide ligands increases in complexes 1 and 2. In contrast, it is negligibly small in the 1D polymer 1′. These characterized bioactive sulfonamide-based Cu(II) compounds appear as promising targets, complying with the structural and electronic expectations for antibacterial and antiseptic purposes. Finally, the antibacterial activity studies question the prerequisite for cooperative metal centers to rationally design antibacterial agents since the minimum inhibitory concentration in the paramagnetic chain 1′ is greatly reduced as compared to antiferromagnetic complexes 1 and 2.

Full text of DOI:10.1016/j.poly.2012.01.025

Polyhedron 37 (2012) 27–34
Contents lists available at SciVerse ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Tailoring antibacteria agents: Sulfonamide-based dinuclear and 1D polymer
Cu(II) complexes
a
b
a
Jean-Bernard Tommasino a, , Guillaume Pilet , François N.R. Renaud , Ghénadie Novitchi ,
⇑
Vincent Robert c, , Dominique Luneau a
⇑
a Université de Lyon, Laboratoire des multiMatériaux et Interfaces, UMR 5615 CNRS, Université Claude Bernard Lyon 1, avenue du 11 novembre 1918, 69622 Villeurbanne Cedex,
France
b Université de Lyon, UMR 5510 CNRS MATEIS, I2B Nosoco.techÒ, Université Claude Bernard Lyon 1, avenue du 11 novembre 1918, 69622 Villeurbanne Cedex, France
c Laboratoire de Chimie Quantique, Institut de Chimie, UMR 7177 CNRS, Université de Strasbourg, 4 rue Blaise Pascal, B.P. 1032, F-67070 Strasbourg Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Starting from two bioactive sulfonamide-based ligands (HL1: 4-amino-N-(2,6-dimethoxy-4-pyrimidinyl)
benzenesulfonamide; HL2: 4-amino-N-(4-methyl-2-pyrimidinyl)benzenesulfonamide) and Cu(II) salts
([Cu2(Ac)4] and CuCl2, Ac = acetate), the synthesis of [Cu2(Ac)2(L1)2] (1) and [Cu2(L2)4] (2) and {Cu(L1)2
(bipy)}nꢀ2nH2Oꢀn(CH3OH) (10) (bipy = 4,40-bipy) architectures with potential antibiotic and antiseptic
activities is reported. To confirm the possibility to fulfill specific criteria and ultimately combine both
properties, the bioactive sulfonamide-based complexes are structurally, electrochemically and magneti-
cally characterized. Depending on the synthesis conditions, the Cu:sulfonamide stoichiometry and
CuꢀꢀꢀCu communication are first varied by controlling the chemical nature of the ancillary ligand (Ac or
bipy). Then, electrochemistry data support the stability of 1 and 2 dinuclear complexes, and 10 1D poly-
mer, a prerequisite for their bioactivity in solution. Interestingly, the synthesis leads to architectures
where the (SO2-Ph-NH2) moiety which is responsible for the antibacterial activity remains non-coordi-
nated in the vicinity of Cu(II) antiseptic ions. Magnetic susceptibility measurements combined to multire-
ference wavefunction ab initio calculations evidence a rather strong antiferromagnetic behavior in the
dinuclear compounds (H = ꢁ2JS1S2, 2J1 = ꢁ307.8 cmꢁ1 in 1, 2J2 = ꢁ63.2 cmꢁ1 in 2) whereas chain 10 is
paramagnetic. The cooperativity quantified by the hopping integral which is available from the ab initio
calculations of the exchange coupling constant is reduced by a factor of two when the number of sulfon-
amide ligands increases in complexes 1 and 2. In contrast, it is negligibly small in the 1D polymer 10.
These characterized bioactive sulfonamide-based Cu(II) compounds appear as promising targets, comply-
ing with the structural and electronic expectations for antibacterial and antiseptic purposes. Finally, the
antibacterial activity studies question the prerequisite for cooperative metal centers to rationally design
antibacterial agents since the minimum inhibitory concentration in the paramagnetic chain 10 is greatly
reduced as compared to antiferromagnetic complexes 1 and 2.
Received 2 December 2011
Accepted 27 January 2012
Available online 21 February 2012
Keywords:
Sulfonamide ligands
Cu(II) complexes
Antibacterial agents
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
fulfilled in the light of recent experimental observations [5–8].
First, it has been shown that the nuclearity of Cu(II) complexes is
Since the need for commercial bioactive mixed organic–metal
compounds continuously increases, the synthesis of such systems
has become of prime importance [1]. In this respect, hybrid
metal–sulfonamide complexes have received much attention,
knowing that the pharmacological activity of the organic moieties
is often enhanced by complexation with metal ions [2–4].
Evidently, one would like to combine the ligand and metal ion
properties (antibiotic and antiseptic, respectively) to produce a sin-
gle entity exhibiting simultaneously both therapeutic activities. To
reach this challenging goal, some specific conditions must be
determinant in the oxidative DNA cleavage [6]. Polynuclear com-
plexes are indeed more efficient than their monomer analogs, a
possible consequence of metal–metal interactions. Then, any or-
ganic modification of the R moiety (see Fig. 1) of the sulfonamide
skeleton is generally incompatible with the presence of a p-amino
group which is responsible for the therapeutic activity. Indeed, the
use of bacterial enzyme inhibitors such as sulfonamide as bridging
ligands has not been much reported in the literature. In order to
facilitate the coordination to metal ions centers, it is necessary to
functionalize the sulfonamide molecule. By varying the nature of
the R substituent (see Fig. 1), the coordination character of the sul-
fonamide is enhanced as a consequence of the greater lability of
the R-NH-SO2– proton.
⇑
Corresponding authors.
E-mail addresses: jbtomasi@univ-lyon1.fr (J.-B. Tommasino), vrobert@unistra.fr
(V. Robert).
0277-5387/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2012.01.025

28
J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
known to effectively communicate as probed by a strong exchange
coupling constant ca. 2J = ꢁ300 cmꢁ1 [12], sulfonamide ligands
(HL1 = sulfadimethoxine and HL2 = sulfamerazine, see Fig. 2) were
introduced to fabricate two new dinuclear complexes with Cu:sul-
fonamide stoichiometries 1:1 (1) and 1:2 (2) and the 1D polymer
analog 10. The use of bipyridine as a spacer allows one to reduce
the cooperativity between the Cu(II) ions and to assess the re-
ported bioactivity/cooperativity correlation [6c].
Electrochemical studies were performed to control the com-
plexes stability in solution. The accessibility of the bioactive organ-
ic part (see Fig. 1) within the complexes was evidenced by X-ray
crystal structures, while the magnetic susceptibility was measured
and used to probe the metal–metal interaction through the ex-
change coupling constant modulation. Finally, multireference
wavefunction-based ab initio calculations were carried out to ex-
tract the effective resonance integral between the Cu(II) centers
and to highlight the expected cooperativity. This work is to be con-
sidered as a first step in the preparation and investigation of multi-
site bioactive complexes.
Fig. 1. Schematic representation of
(SO2-Ph-NH2) moiety.
a sulfonamide involving the bioactive
This synthetic strategy has contributed to the fast development
of N-sulfinyl-based ligands involved in highly efficient asymmetric
Cu(II) catalysts [9]. Therefore, to comply with these specifications
(i.e., polymetallic and bioactive sulfonamide-type ligand), we felt
that a new synthetic route to bind sulfonamide bioactive ligand
to metal centers would be desirable. In particular, one would like
to possibly control the metal:bio-ligand ratio within the complex,
assuming that any ligand enrichment may modulate the antibiotic
activity (Minimum Inhibitory Concentration, MIC). For a synergetic
use, the latter should be comparable to the one of the bare sulfon-
amide species. One of the central issue is that the accumulation of
sulfonamide ligands onto antiseptic dinuclear complexes is likely
to alter the interactions between the metal ions. In order to meet
these requirements, sulfonamide-based dinuclear Cu(II) complexes
were considered as natural targets. The Cu(II) ion was chosen con-
sidering its recognized antiseptic activity and strong enough Lewis
acid character to bind antibacteria soft sulfonamide bases. Mono-
mers combining metal 3d have been recently characterized [10]
by electrochemical measurements and single-crystal X-ray diffrac-
tion, a rather unexpected result considering that bioactive sulfon-
amide-based complexes are usually difficult to crystallize [11].
We herein report the synthesis and structure determinations of
bioactive sulfonamide-based dinuclear complexes. Starting from
the dinuclear precursor [Cu2(Ac)4]ꢀ2H2O where the Cu(II) ions are
2. Experimental details
2.1. Synthesis
2.1.1. [Cu2Ac2(L1)2]ꢀ2CH3CN (1)
Copper acetate salt (0.5 mmol) and the HL1 bioactive sulfon-
amide ligand (1 mmol) were dissolved separately in 5 mL of
CH3CN. Then, the blue copper acetate solution was introduced drop
wise to the colorless HL1 solution and a green solution was ob-
tained. The latter was stirred for 10 min and after 1 week of slow
evaporation of the solvent at room temperature, green single-
crystals (suitable for X-ray data characterizations) were grown.
Anal. Calc. for C28H32Cu2N8O12S2: C, 38.93; H, 3.73; Cu, 14.71; N,
12.97; O, 22.23; S, 7.42. Found: C, 39.31; H, 3.43; Cu, 14.52; N,
12.80%. IR, mmax (cmꢁ1): 3371
1071 (SO2), 771 (S–N).
m mC@O, 1148 m(SO2),
as(NH2), 1571
m
m
2.1.2. [Cu2(L2)4]ꢀ(MeOH)ꢀ0.3H2O (2)
To a methanolic solution (5 mL) of copper acetate salt (0.50
mmole), a methanolic solution of the bioactive HL2 sulfonamide li-
gand (sulfamerazine, 1.26 mmole) was added at room temperature.
After 3 min of stirring, a concentrated solution of NH3 (25% volume)
was added drop wise until the apparition of an intense blue color
characteristic of the presence of the [Cu(NH3)6]2+ complex. After
few days, the solution color turned to green and single-crystal for
X-ray characterizations were obtained by slow evaporation of the
solvent. Anal. Calc. for C44H44Cu2N16O8S4: C, 44.78; H, 3.76; Cu,
10.77; N, 18.99; O, 10.84; S, 10.87. Found: C, 44.72; H, 3.81; Cu,
10.64; N, 18.92. IR, mmax (cmꢁ1): 3358,
m
as(NH2), 1594
(S–N).
mNH2/C–
C(Ph), 669 (C–S), 1269 (SO2), 1073 (SO2), 881 m
m
m
m
2.1.3. {Cu(L1)2(bipy)}nꢀ2nH2OꢀnCH3OH (10)
The same procedure was used as for
2
with Cu(ClO4)2ꢀ
2H20 (0.50 mmole, 146 mg)/3 equiv. of 4,40-bypiridin (1.5
mmole, 235 mg)/1.2 equiv. of sulfadimethoxine (L1H, 0.60 mmole,
186 mg). Single green crystals were obtained.
Anal. Calc. for C34H34Cu1N10O8S4: C, 48.71; H, 4.09; Cu, 7.58; N,
16.71; O, 15.27; S, 7.65. Found. C, 48.72; H, 4.11; Cu, 7.64; N,
16.92.IR, mmax (cmꢁ1): 3351
m
as(NH2), 1591
(SO2/C–C(Ph), 856 m
m
(NH2/C–C(Ph)), 684
(S–N).
m(C–S), 1255
m(SO2), 1070/1126
m
2.2. Electrochemistry
Solvents and reagents were obtained commercially (Aldrich)
and used without further purification. Electrochemical measure-
Fig. 2. HL1 and HL2 sulfonamide bioactive-ligands used in the elaboration of
complexes 1, 2 and 10.

J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
29
ments were performed using an AMEL 7050 all-in one potentiostat,
using a standard three-electrode setup with a glassy carbon elec-
trode, platinum wire auxiliary electrode and SCE (saturated calo-
mel electrode) as the reference electrode. The complex solutions
in DMSO were 1.0 mM, 2 mM and 0.1 M in the supporting electro-
lyte n-Bu4NPF6. Under these experimental conditions, the ferro-
cene/ferricinium couple, used as an internal reference for
potential measurements, was located at E1/2 = 0.421 V.
cannot be accurately evaluated ignoring the dynamical correlation
phenomena. For a given geometry, the dynamical polarization
and correlation effects were then included using the Difference
Dedicated Configuration Interaction (DDCI) method as imple-
mented in the CASDI code [24]. As the number of degrees of free-
dom (i.e., holes in doubly occupied MOs, particles in virtual MOs
of the CASSCF wavefunction) increases, one generates the succes-
sive DDCI-1, DDCI-2 and DDCI-3 CI spaces as discussed in the liter-
ature for related compounds [25]. To eliminate the arbitrariness of
the set of MOs in the DDCI calculations, natural orbitals were first
generated by averaging the DDCI-1 density matrices of the singlet
and triplet states. This procedure was iterated until convergence
upon the energies. DDCI-3 calculations were performed using the
resulting set of MOs.
2.3. X-ray crystallography
Single-crystal X-ray diffraction measurements were recorded
upon the structures of the three original complexes. Diffraction
data sets were collected on an Oxford diffractometer equipped
with a CCD camera and the related softwares [13]. An absorption
correction (multi-scan [14] or analytical [15]) was applied to all
the data. The structures were solved by direct methods using the
SIR97 program [16] combined to Fourier difference synthesis and
refined against F or F2 and using the CRYSTALS program [17]. In each
structure, all atomic displacements for non-hydrogen atoms were
refined anisotropically. Hydrogen atoms belonging to carbon
atoms were located theoretically while the others (belonging to
oxygen atoms) by Fourier Difference but refined using a riding
method.
2.6. Antibacterial study: determination of minimum inhibitory
concentrations (MICs) of bioactive-ligands and their complexes by
agar dilution
The bioactive ligands HL1, HL2 and complexes 1, 2, 10 have been
tested according to the European Committee for antimicrobial sus-
ceptibility testing (EUCAST: European Society of Clinical Microbiol-
ogy and Infectious Disease). The different products were dissolved
in 20% concentration of DMSO (dimethyl sulfoxide) at concentra-
tions ranging from 5 to 1280 mg Lꢁ1. Then, 18 mL of molten agar
(Mueller Hinton agar II, bioMérieux, Lyon, France) were added to
2 mL of the different concentrations. The concentrations were ad-
justed from 0.5 to 128 mg Lꢁ1 with a final DMSO concentration
of 2% which has no inhibitory effect on bacterial growth. The tested
bacterial strains is Enterococcus faecalis (clinical strain from Noso-
co. techÒ collection. Number 20.7). Plates were incubated 18–
24 h at 37 °C. The MIC is defined as the lowest agent concentration
that fully inhibits visible growth as judged by the naked eye.
2.4. Magnetic susceptibility measurements
Magnetic susceptibility data (2–300 K) were collected on pow-
dered polycrystalline samples by SQUID magnetometer on Quan-
tum Design model MPMS instrument under an applied magnetic
field of 0.1 T dependencies. Magnetization isotherm was collected
at 2 K between 0 and 5 T. All data were corrected for the contribu-
tion of the sample holder and diamagnetism of the samples esti-
mated from Pascal’s constants [18]. The analysis of the magnetic
data was carried out by simultaneous
dependencies including temperature-independent paramagnetism
(TIP), impurity contribution ( ), and intermolecular interaction (zJ).
vMT and vM(T) thermal
3. Results and discussion
q
The minimization was carried out with an adapted version of
Visualiseur-Optimiseur for MatlabÒ [19,20] using nonlinear least-
square Lavenberg–Marquard algorithm.
Complexes [Cu2Ac2(L1)2]ꢀ2(CH3CN) (1), [Cu2(L2)4]ꢀ(MeOH)ꢀ0.3
H2O (2) and {Cu(L1)2(bipy)}nꢀ2nH2OꢀnCH3OH (10) were obtained
by ligand exchange reaction between Cu(II) salts and respective
sulfonamide (HL1 for 1 and 10; HL2 for 2) in the presence of
NH3(aq) 25% in water solution [11]. Let use mention that the use
of other bases such as NaOH, KOH or triethylamine in any molar ra-
tio was not successful. This suggests the special role of ammonium
solution in the formation of complexes based on bioactive sulfon-
amide ligands. Thus, from our observations, the first step should be
the formation of a [Cu(NH3)6]2+ complex evidenced by the blue in-
tense color of the solution [11]. This intermediate complex reacts
progressively with the sulfonamide ligand (HL1 or HL2) to produce
the complexes 1, 2 and monodimensional polymer 10. Following
this original synthesis method, 1, 2 and 10 were isolated with sig-
nificant yields (ca. 70%) and fully characterized (see Tables 1A
and 1B).
Due to multiple donor atoms in the sulfonamide moiety, differ-
ent coordination modes to metal centers have been reported in the
literature (see Fig. 3) [8]. In complexes 1 and 2, the deprotonated
sulfonamide bioactive ligands L1ꢁ and L2ꢁ adopt a coordination
mode IV (see Fig. 3) once connected to the Cu(II) metal centers
resulting in dinuclear complexes (see Fig. 4).
In both complexes, the environment of the two metal Cu(II) ions
is the same with a regular X4 square plane (X = N or O; N2O2 for 1
and N4 for 2) where the Cu–X bond lengths range from 1.966(3) to
2.023(3) Å (X = N or O; average: 2.00 Å) for 1, from 1.997(4) to
2.048(4) Å (X = N; average: 2.02 Å) for 2. This metal environment
is completed by oxygen atoms belonging to SO2 ligand moiety as
second coordination sphere: one oxygen atom per Cu(II) within 1
2.5. Computational details
It is known that some care must be taken to properly define en-
ergy spectrum of open-shell systems [21]. However, spectroscopic
accuracy can be reached using complete active space self-
consistent field (CASSCF) and subsequent correlation effects treat-
ments. Indeed, the CASSCF method gives reasonable electron dis-
tribution and accounts for the leading electronic configurations.
From the d9 electronic configuration of Cu(II) ion, CAS[2,2]SCF
(i.e., two electrons in two molecular orbitals, MOs) calculations
were first performed upon a simplified structure of complexes 1
and 2. To reduce the computational cost, each p-aminobenzyl
group of the four ligands was changed into hydrogen atoms with
adapted C–H bond distances. It has been shown that chemical
changes which maintain the nature of the bridging ligands and
the polarization properties of the coordination spheres are unlikely
to deeply modify the exchange coupling intensity. These CASSCF
calculations were performed using the MOLCAS7.2 package [22]. All
atoms were described using ANO RCC-type atomic functions [23].
Carbon, nitrogen, oxygen, and sulfur atoms were described with
DZP-type contractions, whereas a (21s15p10d6f4g2h)/[5s4p2d1f]
contraction was used for the copper atom. Finally, the hydrogen
atoms were depicted using
a minimal basis set contraction
(8s4p3d1f)/[1s]. Whatever the definition of the active space, it
has been clearly demonstrated that the exchange interaction

30
J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
Table 1A
Details of the data collections and refinements.
1
2
10
Ref. formula
C32H38Cu2N10O12S2
945.9
C45H48.66Cu2N16O9.33S4
C35H42Cu1N10O11S2
906.5
monoclinic
P2/n
18.059(1)
11.0588(6)
20.509(1)
90
98.047(6)
90
FW (g molꢁ1
)
1218.3
monoclinic
C2/c
39.949(1)
12.5809(3)
20.8751(6)
90
101.352(3)
90
10286.3(5)
8
Crystal system
Space group
a (Å)
b (Å)
c (Å)
triclinic
ꢀ
P1
8.332(1)
9.911(1)
13.092(2)
98.99(1)
106.82(1)
98.06(1)
1002.4(2)
1
a
(°)
b (°)
c
(°)
V (Å3)
Z
4055.4(4)
4
T (K)
100
1.567
1.237
293
1.573
1.062
293
1.485
0.713
D (g cmꢁ3
)
l
(mmꢁ1
)
Independent reflections
Rint
4640
0.059
10163
0.039
7042
0.039
R(F)/Rw(F)
S
0.0483/0.0556
1.06
0.0611/0.0629
1.13
0.0555/0.0764
1.05
Number of Reflections
Number of parameters
3135
262
6662
675
7037
538
D
q
/
max Dqmin (e Åꢁ3
)
1.39/ꢁ1.18
1.35/ꢁ2.13
0.59/ꢁ0.54
Absorption correction
analytical
multi-scan
multi-scan
(average Cu–O bond lengths equals to 2.47 Å) and two oxygen
atoms per Cu(II) within 2 (average Cu–O bond lengths equals to
2.59 Å). Fixing z as the internuclear axis, such environment leaves
Table 1B
Selected bond lengths (Å) and bond angles (°).
Complex 1
one unpaired electron in a dx2 -type orbital upon each Cu(II) cen-
ꢁy2
Cu1–N11
Cu1–O34
N11–Cu–N21
N11–Cu1–O33
O34–Cu1–O33
2.023(3)
1.966(3)
172.2(1)
91.1(1)
170.0(1)
Cu1–N21
Cu1–O33
N11–Cu1–O34
N21–Cu1–O33
1.991(3)
1.963(3)
88.2(1)
ter. The relative orientation of the coordination spheres suggest a
d-type overlap between the magnetic Cu(II) orbitals. The CuꢀꢀꢀCu
distances are relatively short for both complexes (2.57 and 2.60 Å
for 1 and 2, respectively). The crystal packing is maintained by
hydrogen bonds between the dimers and non-coordinated solvent
molecules (acetonitrile for 1 and methanol/water for 2) forming a
compact network.
88.8(1)
Complex 10
Cu1–N8
Cu1–N31
Cu2–N24
2.727(3)
Cu1–N38
Cu1–N20
Cu2–N6
Cu2–N27
N8–Cu1–N31
N8–Cu1–N8
N8–Cu1–N20
N31–Cu1–N20
N6–Cu2–N24
N24–Cu2–N9
N9–Cu2–N6
N6–Cu2–N27
N9–Cu2–N27
2.011(4)
2.003(4)
2.621(3)
2.016(4)
54.4(1)
160.0(1)
100.02(7)
90.47(9)
98.28(7)
91.26(9)
123.6(1)
81.72(7)
88.74(9)
1.986(3)
1.999(4)
1.976(3)
79.98(7)
89.53(9)
125.4(1)
180
179.1(2)
56.0(1)
163.4(1)
177.5(2)
180
The situation is rather different in 10 which exhibits a 1D poly-
mer structure (see Fig. 5) with chains running along the b-axis of
the unit-cell. The chelating mode III (see Fig. 3) is observed once
the deprotonated L1ꢁ ligand is connected to the Cu(II) metal ion.
The Cu(II) cation is located in a N6 distorted octahedral environ-
ment with four short (2.00 Å) and two long (2.67 Å) Cu–N bond
lengths. The former involve the nitrogen atoms belonging to the
two deprotonated bioactive ligands. Such arrangement gives rise
to chains characterized by CuꢀꢀꢀCu distances equal to 11.06 Å. The
structural cohesion is obtained by hydrogen bonds between chains
and co-crystallized solvent molecules (water and/or methanol).
To complement the solid state characterization, the stability of
these three original complexes in solution was checked using elec-
trochemical measurements performed in DMSO. It has been shown
that the HL1 and HL2 bioactive sulfonamide ligands exhibit an irre-
versible oxidation step (1.23 and 1.20 V, see Table 2) [11]. Their
corresponding dinuclear complexes 1 and 2, and chain 10 display
a similar wave 1.30, 1.35 and 1.29 V, respectively (see Table 2).
The latter was assigned to the oxidation of the terminal –NH2
group of the sulfonamide ligand [11]. This result suggests that
the amine group of the organic moiety responsible for the thera-
peutic effects remains non-coordinated to metal ion within the
complexes.
Cu2–N9
N8–Cu1–N38
N38–Cu1–N31
N31–Cu1–N8
N38–Cu1–N20
N31–Cu1–N31
N6–Cu2–N9
N6–Cu2–N6
N9–Cu2–N9
N24–Cu2–N27
Complex 2
Cu1–N1
Cu1–N48
Cu1–N68
Cu2–N8
Cu2–O10
Cu2–N28
N1–Cu1–N21
N21–Cu1–N48
N21–Cu1–O51
N1–Cu1–N68
N48–Cu1–N68
N1–Cu1–O71
N48–Cu1–O71
N68–Cu1–O71
N8–Cu2–O10
N8–Cu2–N28
O10–Cu2–N28
N8–Cu2–O30
O30–Cu2–N41
N8–Cu2–N61
O10–Cu2–N61
2.008(4)
2.019(4)
2.035(4)
2.048(4)
2.539(4)
2.037(4)
178.5(2)
90.3(2)
86.3(2)
88.7(2)
161.9(2)
90.5(2)
138.3(2)
59.8(1)
Cu1–N21
Cu1–O51
Cu1–O71
Cu2–O30
Cu2–N41
Cu2–N61
2.014(4)
2.601(4)
2.633(4)
2.570(4)
1.986(4)
1.997(4)
91.1(2)
93.7(2)
60.4(2)
90.2(2)
137.6(1)
88.0(2)
77.9(1)
76.1(1)
60.7(1)
89.6(2)
90.1(2)
90.6(2)
91.1(2)
91.1(2)
178.3(2)
N1–Cu1–N48
N1–Cu1–O51
N48–Cu1–O51
N21–Cu1–N68
O51–Cu1–N68
N21–Cu1–O71
O51–Cu1–O71
O10–Cu2–O30
N28–Cu2–O30
N8–Cu2–N41
O10–Cu2–N41
N28–Cu2–N41
N28–Cu2–N61
O30–Cu2–N61
N41–Cu2–N61
61.0(2)
162.1(2)
136.9(1)
137.1(2)
89.4(2)
88.9(2)
88.4(2)
During the reduction process in Fig. 6, the voltammograms of 1
and 2 are characterized by two steps with the same relative inten-
sity: a quasi-reversible large wave (ꢁ0.19 and ꢁ0.45 V for 1 and 2,
respectively) followed by an irreversible peak (ꢁ1.80 and ꢁ2.10 V
for 1 and 2, respectively). These irreversible peaks are related to
typical redissolution oxidation peaks (Table 2 and Fig. 6). The gen-
eral patterns of the voltammograms of all complexes remained
unchanged for hours, suggesting that there is no degradation of
the dimer entities in solution. Moreover, the cyclic voltammetry

J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
31
Fig. 3. Coordination modes of deprotonated sulfonamides (M: metal ion. R = alkyl;
R0 = aromatic or alkyl groups).
of [Cu2(Ac)4] salt performed in the same conditions is different
from the ones recorded for 1 and 2 demonstrating again the com-
plexes stability in solution. The coulometry on the first reduction
step for 1 and 2 exhibits an electron apparent number value equal
to 2 F/mole. Therefore, these processes correspond to the simulta-
neous Cu(II) ? Cu(I) reduction of both metal centers within the
complexes.
The scenario is contrasted in 10 where both mono-electronic
reduction steps (Cu(II) ? Cu(I) and Cu(I) ? Cu(0)) are irreversible,
and followed by two peaks assigned to the 4,40-bipy bridge reduc-
tion (Table 2). In conclusion, the ligand moiety (SO2-Ph-NH2) which
is responsible for the antibacterial activity remains non-coordi-
nated whatever the architecture. In both solid state and solution,
such structural arrangement is a prerequisite to maintain the bioac-
tivity of the ligands in the vicinity of the Cu(II) antiseptic ions.
The magnetic properties of compounds 1, 2 and 10 were then
investigated by DC susceptibility measurements in the 2–300 K
temperature range under a field of 0.1 T. Thermal evolution of mag-
netic susceptibility for both dinuclear complexes is given in Fig. 4. At
room temperature, the
is smaller than the theoretical expected
for two non-interacting Cu(II) ions (d9, g = 2.0, S = 1/2). With
v
T values for 1 is 0.421 cm3 K molꢁ1 which
T value (0.75 cm3 K molꢁ1
v
)
Fig. 4. Crystal structures of 1 (a) and 2 (b). For clarity, hydrogen atoms and non-
coordinated solvent molecules have been removed; (c) simplified view of metal ion
dimers with their organic bridges within 1 and 2 (X = O or N).
decreasing temperature, the
vT product continuously decreases
and finally reaches a value close to zero (0.011 cm3 K molꢁ1) at
60 K. This behavior indicates the presence of a strong antiferromag-
netic interaction. The value of
Upon cooling, the
v .
T at 300 K for 2 is 0.665 cm3 K molꢁ1
v
T product continuously decreases and reaches
the value of 0.014 cm3 K molꢁ1 at 7 K, suggesting antiferromagnetic
interactions in 2. In contrast, 10 is essentially paramagnetic as a con-
sequence of the introduction of the bipyridine spacer (see Fig. 7).
Using a S = 1/2 chain model for 10 [26–30]:
X
H ¼ ꢁ2J
SiSiþ1
NAg2l2B
1 þ 0:08516aꢁ1 þ 0:23351aꢁ2
vð
aÞ ¼
4j2JjT 1 þ 0:73382aꢁ1 þ 0:13696aꢁ2 þ 0:53568aꢁ3
Fig. 5. Crystal structure of 10 with chains running along the b-axis of the unit-cell.
For clarity, hydrogen atoms and solvent molecules have been removed.
kT
a
¼
j2Jj
a negligible J1 = 0.067 cmꢁ1 value (g = 2.15) was obtained, a feature
values are consistent with previously reported dinuclear and poly-
mer Cu(II) connected by 4,40-bipy-bridged [31,32]. According to
0
of the absence of communication between the Cu(II) ions. These

32
J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
Table 2
Table 3
Electrochemical data for HL1 and HL2 sulfonamide ligands, 1 and 2 dimers, and 10
Exchange coupling constants (cmꢁ1) and Cu–Cu distances (Å) for tetrakis aminoben-
zyl bridging dinuclear Cu(II) compounds.
polymer. 4,4’-bipy data are introduced for comparison.
Anodic
Cathodic
Compound
Cu–Cu distance (Å)
ꢁ2J (cmꢁ1
–
)
Reference
HL1d
1.23a
1.20a
ꢁ2.45a
ꢁ2.95a
[Cu2(TzTs)4]
[Cu2(tz-tol)4]
[Cu2(stz)4]
2.7859(5)
2.722(1)
2.671(2)
2.629(2)
2.626(1)
2.569(9)
2,598(7)
2.6143
[33]
[34]
[35]
[5]
HL2d
ꢁ121
ꢁ61.5
ꢁ114.1
ꢁ104
ꢁ63.18
ꢁ307.8
ꢁ30
[Cu2Ac2]
4,40-bipy
1
2
10
ꢁ0.40b
ꢁ1.30b
ꢁ1.92
ꢁ2.32
[Cu2(tz-ben)4]
[Cu2(st-naf)4]
[Cu2(L2)4] (2)
[Cu2Ac2(L1)2] (1)
Cu2Ac4(H2O)2
[Cu2(sulfameter)4]
[Cu2(PyBp)4]
[Cu2Ac2(L)2]
1.30a
1.35a
1.29a
ꢁ0.19b
ꢁ0.45b
ꢁ0.22b
ꢁ1.80
ꢁ2.10
ꢁ1.66
[34]
This work
This work
[12]
[36]
[37]
ꢁ1.88c
ꢁ2.37c
Peak potentials (V) recorded in DMSO at 293 K with a glassy carbon electrode, 0.1 M
2.556
2.5162(9)
2.5412(6)
–
n-Bu4NPF6 as supporting electrolyte; all potentials are vs. SCE, scan rate 0.1 V sꢁ1
.
ꢁ284
a
Irreversible system.
Quasi-reversible.
Reduction of 4,40-bipy in the complex.
ꢁ216.7
[38]
b
c
tz-tol, N-(thiazol-2-yl)toluenesulfonamide; tz-naf, N-(thiozol-2-yl)naphthalene-
sulfonamide; stz, sulfathiazole; TzTs, N-thiazol-2-yl-(4-methylphenyl)sulfonamide;
PyBp, N-(pyridin-2-yl)biphenyl-4-sulfonamide; Htz-ben, N-(thiazol-2-yl)benzene-
sulfonamide; sulfameter, 4-amino-N-(5-methoxy-2-pyrimidinyl)benzenesulfon-
amide; L, 4-amino-N-[4,6-domethyl-2-pyrimidyl]benzenesulfonamide.
d
See Ref. [11].
As seen in Table 3 large variations (ꢁ308 to ꢁ61.5 cmꢁ1) of the
exchange coupling constants have been observed in different dinu-
clear Cu(II) compounds holding similar bridging ligands [29–38].
To further inspect the magnetic properties in complexes 1 and 2,
ab initio calculations were performed. The parameters governing
the exchange coupling constant J = K ꢁ 2t2/U are the effective direct
exchange K, resonance integral t and on-site repulsion U which can
be extracted from ab initio wavefunction calculations [39]. Clearly,
the acetate fragment is an efficient and well recognized magnetic
coupler. Since the super-exchange contributions are additive along
the different channels characterized by the different hopping inte-
grals t, the L1 ligand turns out to be as efficient as the acetate li-
Fig. 6. Cyclic voltammetry of 1 (blue full line), 2 (green dashed line) and 10 (orange
dotted line) in reduction in DMSO using a glassy carbon electrode (3 mm diameter)
at 100 mV/s. (Color online.)
gand. Indeed, the exchange coupling constants of
1 and
Cu2(Ac)4ꢀ2H20 are very similar in amplitude (see Table 3). The sit-
uation is evidently contrasted when ligand L2 is involved. Thus, to
further inspect the magnetic properties in complexes 1 and 2, ab
initio calculations were performed. From the paramagnetic behav-
ior of 10, the resonance integral is expected to be vanishingly small,
a signature of the low-cooperativity between the Cu(II) ions and
we felt that calculations would not be instructive. For 1 and 2,
the active space includes the anticipated unpaired electrons and
MOs of the system, leading to a CAS[2,2] (i.e., two electrons in
two MOs) zeroth-order description. The system under study was
simplified by leaving out atoms which belong at least to the third
coordination spheres of the metal ions (see Fig. 8a). As expected,
the active MOs correspond to the in-phase and out-of-phase linear
combinations of the d-type orbitals of the Cu atoms (see Fig. 8b).
As a consequence of the d-type overlap between the magnetic
orbitals, the antiferromagnetic behavior is much weaker than in
acetate analogs. Indeed the singlet-triplet energy difference corre-
sponding to 2J is calculated at this CAS-CI level ꢁ2.5 cmꢁ1. Using
the DDCI-1 iterated set of MOs to dispose of the arbitrariness of
the initial set (i.e., singlet versus triplet MOs), the singlet–triplet
energy differences were then computed at the DDCI-3 level to be
ꢁ98 and ꢁ21 cmꢁ1 for 1 and 2, respectively. Despite a 30% devia-
tion from the experimental values, the antiferromagnetic behavior
is clearly identified, suggesting a stronger cooperativity in complex
1 than in 2. The on-site repulsion U was extracted from our calcu-
lations and turns out to be ꢃ60 000 cmꢁ1 for both systems, in
agreement with previous calculations [40]. In contrast, the reso-
nance integrals is significantly reduced (t1 = 1.7 ꢂ 103 cmꢁ1 and
t2 = 0.8 ꢂ 103 cmꢁ1 for 1 and 2, respectively) as the number of sul-
fonamide ligands in the complex increases.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
100
200
300
T(K)
Fig. 7.
v
T vs. T plots data for 1 (green triangle), 2 (blue square) and 10 (orange
circle). The black lines correspond to the best fit according to dinuclear or chain
models with parameters indicated in the text. (Color online.)
the isolated dinuclear structures of 1 and 2, the magnetic properties
can be analyzed using the isotropic spin Heisenberg Hamiltonian
H = ꢁ2JS1S2 where S1 = S2 = 1/2, with the following analytical
expression:
2Ng2Cub2
1
vd
¼
kT
3 þ eꢁ2J=kT
The fitting procedure (see Fig. 7) leads to J1 = ꢁ153.9(7) cmꢁ1
,
q
= 0.046(4), g = 2.179(9), zJ1 = ꢁ0.001(2) cmꢁ1 and TIP = ꢁ1.84 ꢂ
10ꢁ5 cm3 molꢁ1 for 1; J2 = ꢁ31.60(2) cmꢁ1
,
q
= 0.040(2), g = 2.11
At this stage one may question how much the accumulation of
bioactive ligands is likely to compete with the metal–metal coop-
erativity probed by the magnetic interactions. X-ray diffraction and
(2), zJ2 = ꢁ0.02(2) cm-1 and TIP = ꢁ1.84 ꢂ 10ꢁ5 cm3 molꢁ1 for 2,
(R = 4.08 ꢂ 10ꢁ5 for 1; R = 2.57 ꢂ 10ꢁ5 for 2).

J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
33
Fig. 8. (a) Simplified structure of 1 used for the ab initio calculations. CH3 groups were replaced by hydrogen atoms with adapted C–H bond lengths (pink). A similar structural
simplification was used for 2; (b) triplet active orbitals from CAS[2,2] calculations performed on 1 and corresponding to the in-phase and out-of phase combinations of the d-
type orbitals localized on the Cu(II) center. A similar picture holds for 2. (Color online.)
cyclic voltammetry studies suggest that the bioactive part of the
sulfonamide ligands remains non-coordinated within the com-
plexes structures (dimers and chain). This is the particular specifi-
cation we wanted to achieve, binding bioactive and accessible
ligands to antiseptic Cu(II) metal centers. Besides, magnetic mea-
surements combined with ab initio calculations demonstrate a
much stronger metal–metal interactions in complexes 1 and 2 than
in the polymer analog 10.
was measured and reveals that the minimum inhibitory concentra-
tions in 1 and 2 are comparable to those of the isolated ligands.
These preliminary bacteriological studies demonstrate that the
bioactive efficiency is maintained in these sulfonamide based
Cu(II) complexes. By preserving the structure of the bioactive li-
gand and allowing for communication between the Cu(II) centers,
the targeted dinuclear species 1 and 2 fulfill the reported specifica-
tions found in the literature. Nevertheless, our results question the
prerequisite for cooperative metal centers to rationally design anti-
bacterial agents since the minimum inhibitory concentration in the
paramagnetic chain 10 is greatly reduced.
In order to try out the correlation between metal–metal com-
munication and bioactivity, preliminary antibacterial studies were
finally undertaken.
The antibacterial activity of bioactive sulfonamide which are
structurally similar to p-aminobenzoic acid (PABA) is due to their
ability to interfere with the conversion of PABA to dihydrofolate
(DHF = folate) by an inhibitory competition with the enzyme dihy-
dropteroate synthetase (DHPS) [41]. The reaction between 10
(chain) and E. faecalis exhibits a MIC which falls down to a value
ca. 16 mg Lꢁ1 (HL1: MIC > 128 mg Lꢁ1) while complexes 1 and 2 (di-
Acknowledgment
Single-crystal X-ray diffraction studies were performed at the
‘‘Centre de Diffractométrie Henri Longchambon’’ at Université
Claude Bernard Lyon 1.
mers) are as active as the isolated ligands HL1 (MIC > 128 mg Lꢁ1
)
Appendix A. Supplementary data
and HL2 (MIC > 128 mg Lꢁ1). Therefore, the absence of communica-
tion between the Cu(II) centers in 10 questions the need for interact-
ing Cu(II) centers in polynuclear bioactive ligand-based structures
to ever enhance the antibacteriological activity. The results of the
strategies that we have developed, namely (i) increasing the num-
ber of sulfonamide ligands within the complex, and (ii) introducing
a spacer ligand such as bipyridine between the Cu(II) ions seems to
give priority to the second approach in order to tailor antibacteria
agents.
CCDC 549111, 827013 and 827014 contain the supplementary
crystallographic data for structures of complexes 1, 2 and 10,
respectively. An X-ray crystallographic file in CIF format, this mate-
rial is available free of charge via the Internet at http://www.pub-
s.acs.org. 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.
4. Conclusion
References
[1] L.-J. Ming, Med. Res. Rev. 23 (2003) 697.
Following a synthetic strategy to prepare bioactive sulfonamide
ligand-based architectures, dinuclear coordination compounds and
1D polymer using Cu(II) antiseptic ions were isolated, character-
ized and electrochemically investigated. Depending on the
synthetic conditions, one can generate different Cu(II):ligand stoi-
chiometries which form single crystals. The combination of elec-
trochemical measurements and single crystal X-ray diffraction
supports that (i) the coordination of the sulfonamide ligands to
the Cu(II) centers is not altered when dissolution occurs, and (ii)
the crucial amino group remains accessible (i.e., non-coordinated)
for bioactivity. Using magnetic susceptibility measurements in
contact with ab initio calculations, a rather strong antiferromag-
netic behavior is observed in both dinuclear compounds while
the cooperativity as measured through the effective hopping inte-
gral is reduced as the number of sulfonamide ligands increases.
Despite this apparent drawback, the bioactivity towards E. faecalis
[2] M.S. Iqhal, A.H. Khan, B.A. Loothar, I.H. Bukkari, Med. Chem. Res. 18 (2009) 31.
[3] B.M. Macias, V. Villa Maria, L. Sanchez-Mesonero, F. Sanz, J. Borras, M.Z.
Gonzalez-Alvarez, Anorg. Allg. Chem. (2007) 1937.
[4] (a) E. Kremer, G. Facchin, E. Estevez, P. Albores, E.J. Baran, J. Ellena, M.H. Torre,
J. Inorg. Biochem. 100 (2006) 1167;
(b) M. Mondelli, V. Brune, G. Borthagaray, J. Ellena, O.R. Nascimento, C. Leite,
A.A. Batista, M.H. Torre, J. Inorg. Biochem. 102 (2008) 282.
[5] R. Cejudo, G. Alzuet, M. Gonzalez-Alvarez, J.L. Garcia-Gimenez, J. Borras, M. Liu-
Gonzalez, J. Inorg. Biochem. 100 (2006) 70.
[6] (a) K.J. Humpreys, A.J. Jonhson, K.D. Karlin, S.E. Rokita, J. Biol. Inorg. Chem. 7
(2002) 835;
(b) M. Gonzalez-Alvarez, G. Alzuet, J. Borras, M. Pitié, B. Meunier, J. Biol. Inorg.
Chem. 8 (2003) 644;
(c) L. Li, K.D. Karlin, S.E. Rokita, J. Am. Chem. Soc. 127 (2005) 520.
[7] I. Beloso, J. Castro, J.A. García-Vázquez, P. Pérez-Lourido, J. Romero, A. Sousa,
Polyhedron 25 (2006) 2673.
[8] I. Beloso, J. Borras, J. Castro, J.A. Garcia-Vazquez, P. Perez-Lourido, J. Romero, A.
Sousa, Eur. J. Inorg. Chem. 3 (2004) 635.
[9] (a) T.D. Owens, F.J. Hollander, A.G. Oliver, J.A. Elmann, J. Am. Chem. Soc. 123
(2001) 1539;

34
J.-B. Tommasino et al. / Polyhedron 37 (2012) 27–34
(b) C. Worch, A.C. Mayer, C. Bolm, in: T. Toru, C. Bolm (Eds.), Organosulfur
Chemistry in Asymmetric Synthesis, Wiley-VCH, Weinheim, (2008). P. 209.
[25] N. Guihéry, V. Robert, F. Neese, J. Phys. Chem. A 112 (2008) 11979.
[26] J. Bonner, M. Fisher, Phys. Rev. 135 (1964) 640.
[10] (a) M.H. Torre, G. Facchin, E. Kremer, E.E. Castellano, O.E. Piro, E.J. Baran, J.
Inorg. Biochem. 94 (2003) 200;
[27] S. Shova, G. Novitchi, M. Gdaniec, A. Caneschi, D. Gatteschi, L. Korobchenko,
V.K. Voronkova, Y.A. Simonov, C. Turta, Eur. J. Inorg. Chem. (2002) 3313.
[28] Y. Deng, J. Liu, B. Wu, C. Ambrus, T. Keene, O. Waldmann, S. Liu, S. Decurtins, X.
Yang, Eur. J. Inorg. Chem. (2008) 1712.
(b) G.M. Golzar Hossain, A.J. Amoroso, A. Banu, K.M.A. Malik, Polyhedron 26
(2007) 967.
[11] J.B. Tommasino, G. Pilet, F. Renaud, D. Luneau, Polyhedron 30 (2011) 1663.
[12] (a) B. Bleaney, K.D. Bowers, Proc. R. Soc. Lond. A 214 (1952) 451;
(b) G.M. Brown, R. Chidambaram, Acta Crystallogr., Sect. B 29 (1973) 2393.
[13] CRYSALISPRO, Oxford Diffraction Ltd., Version 1.171.33.46 (release 27-08-2009
CrysAlis171 .NET).
[29] R. Feyerherm, S. Abens, D. Gunther, T. Ishida, M. Meissner, M. Meschke, T.
Nogami, M. Steiner, J. Phys.: Condens. Matter 12 (2000) 8495.
[30] S. Eggert, I. Affleck, M. Takahashi, Phys. Rev. Lett. 73 (1994) 332.
[31] M. Julve, M. Verdaguer, J. Faus, F. Tinti, J. Moratal, A. Monge, E. Gutierrez-
Puebla, Inorg. Chem. 26 (1987) 3520.
[14] CRYSALISPRO, Oxford Diffraction Ltd., Version 1.171.33.46 (release 27-08-2009
CrysAlis171 .NET); Empirical absorption correction using spherical harmonics,
implemented in SCALE3 ABSPACK scaling algorithm.
[15] J. De Meulenaar, H. Tompa, Acta Crystallogr., Sect. A 19 (1965) 1014.
[16] G. Cascarano, A. Altomare, C. Giacovazzo, A. Guagliardi, A.G.G. Moliterni, D.
Siliqi, M.C. Burla, G. Polidori, M. Camalli, Acta Crystallogr., Sect. A 52 (1996)
C-79.
[32] S.T. Ochsenbein, F. Tuna, M. Rancan, R.S.G. Davies, C.A. Muryn, O. Waldmann,
R. Bircher, A. Sieber, G. Carver, H. Mutka, F. Fernandez-Alonso, A. Podlesnyak,
L.P. Engelhardt, G.A. Timco, H.U. Gudel, R.E.P. Winpenny, Chem. Eur. J. 14
(2008) 5144.
[33] S. Cabaleiro, R. Calvo, J. Castro, A. Garcia-Vazquez, L.M.B. Napolitano, O.R.
Nascimento, P. Perez-Lourido, J. Romero, A. Sousa, J. Chem. Crystallogr. 38
(2008) 71.
[17] D.J. Watkin, C.K. Prout, J.R. Carruthers, P.W. Betteridge, CRISTAL Issue 11,
Chemical Crystallography Laboratory, Oxford, UK, 1999.
[18] P. Pascal, Ann. Chim. Phys. 19 (1910) 5.
[19] Matlab, The MathWorks Inc., 2000.
[20] Optimiseur-Visialiser, EPFL, Lausanne, 2001.
[34] R. Cejudo-Marin, G. Alzuet, S. Ferrer, J. Borras, A. Castineiras, E. Monzani, L.
Casella, Inorg. Chem. 43 (2004) 6805.
[35] J. Casanova, G. Alzuet, J. Latorre, J. Borras, Inorg. Chem. 36 (1997) 2052.
[36] J. Ellena, E. Kremer, G. Facchin, E.J. Baran, O.R. Nascimento, A.J. Costa-Filho,
M.H. Torre, Polyhedron 26 (2007) 3277.
[21] (a) S. Messaoudi, V. Robert, N. Guihéry, D. Maynau, Inorg. Chem. 45 (2006)
3212;
[37] A. Bodoki, G. Alzuet, A. Hangan, L. Oprean, F. Estevan, A. Castineiras, J. Borras,
Inorg. Chim. Acta 363 (2010) 3139.
(b) J.-B. Rota, L. Norel, C. Train, N. Ben Amor, D. Maynau, V. Robert, J. Am.
Chem. Soc. 130 (2008) 10380.
[38] L. Gutierrez, G. Alzuet, J. Borras, A. Castineiras, A. Rodriguez-Fortea, E. Ruiz,
Inorg. Chem. 40 (2001) 3089.
[22] MOLCAS7: The Next Generation, J. Comp. Chem. 31 (2010) 224. Available from:
<dx.doi.org/10.1002/jcc.21318/>.
[39] C.J. Calzado, J. Cabrero, J.-P. Malrieu, R. Caballol, J. Chem. Phys. 116 (2002)
3985.
[23] B.O. Roos, R. Lindh, P.-A. Malmqvist, V. Veryazov, P.-O. Widmark, J. Phys. Chem.
A 109 (2005) 6575.
[40] J.B. Rota, B. Le Guennic, V. Robert, Inorg. Chem. 49 (2010) 1230.
[41] (a) G.M. Brown, Adv. Biochem. 35 (1971) 35;
[24] (a) J. Miralles, O. Castell, R. Caballol, J.-P. Malrieu, Chem. Phys. 172 (1993) 33;
(b) N. BenAmor, D. Maynau, Chem. Phys. Lett. 286 (1998) 211.
(b) G.H. Miller, P.H. Doukas, J.K. Seydal, J. Med. Chem. 15 (1972) 700.