Synthesis and magnetic properties of new mono- and binuclear iron complexes with salicyloylhydrazono dithiolane ligand

Article abstract of DOI:10.1021/ic800522h

We report here experimental evidence for the formation in the solid state of a new binuclear Fe^III₂-OMe)₂(HL)₄ complex (H₂L is 2-salicyloylhydrazono-1,3-dithiolane). The isostructural Mn^III₂(μ-OMe)₂(HL)₄ complex has provided the strongest ferromagnetic interaction value (J ≈ 20 cm^-1) between Mn^III ions to date. The new iron binuclear compound presented in this study shows antiferromagnetic intramolecular coupling, which agrees with the theoretical study that we previously proposed. During our synthetic work, we also observed an unexpected spontaneous reduction of the new Fe^III(HL)₂Cl,S complex to the new Fe ^IIH₂L)₂Cl₂ high-spin mononuclear complex. This process has been checked by cyclo-voltammetry as well as pseudosteady voltammetry.

Full text of DOI:10.1021/ic800522h

Inorg. Chem. 2008, 47, 7623-7630
Synthesis and Magnetic Properties of New Mono- and Binuclear Iron
Complexes with Salicyloylhydrazono Dithiolane Ligand
Nouri Bouslimani,† Nicolas Cle´ment,† Guillaume Rogez,‡ Philippe Turek,§ Maxime Bernard,§
Samuel Dagorne,† David Martel,| Hoan Nguyen Cong,| and Richard Welter*,†
Laboratoire DECOMET, UMR 7177 CNRS, UniVersite´ Louis Pasteur, 4, rue Blaise Pascal,
67070 Strasbourg Cedex, France, I.P.C.M.S., UMR 7504 CNRS-ULP, Groupe des Materiaux
Inorganiques, 23 Rue du Loess, BP 43, F-67034 Strasbourg Cedex 2, France, Laboratoire
POMAM, UMR 7177 CNRS, UniVersite´ Louis Pasteur, 1, rue Blaise Pascal, BP 296 R8,
67008 Strasbourg Cedex, France, and Laboratoire d’Electrochimie et de Chimie Physique du
Corps Solide, UMR 7177 CNRS-ULP, UniVersite´ Louis Pasteur Strasbourg I,
F-67070 Strasbourg cedex, France
Received March 22, 2008
We report here experimental evidence for the formation in the solid state of a new binuclear FeIII (µ-OMe)2(HL)4
2
complex (H2L is 2-salicyloylhydrazono-1,3-dithiolane). The isostructural MnIII (µ-OMe)2(HL)4 complex has provided
2
the strongest ferromagnetic interaction value (J ≈ 20 cm-1) between MnIII ions to date. The new iron binuclear
compound presented in this study shows antiferromagnetic intramolecular coupling, which agrees with the theoretical
study that we previously proposed. During our synthetic work, we also observed an unexpected spontaneous
reduction of the new FeIII(HL)2Cl,S complex to the new FeII(H2L)2Cl2 high-spin mononuclear complex. This process
has been checked by cyclo-voltammetry as well as pseudosteady voltammetry.
Introduction
[Mn12O12(CH3COO)16(H2O)4] ([Mn12(OAc)])5,6 that still re-
mains a golden standard due to combination of a large spin
state (S ) 10) and negative zero-field splitting axial
parameter (D ) -0.5 cm-1). A record value for the energy
barrier to magnetization reversal of SMM was reported
recently for a hexanuclear MnIII compound (S ) 12, D )
-0.43 cm-1, Ueff ) 86.4 K).7 Although the feasibility of
SMM for technological application in information storage
or quantum computing8 is still quite questionable (especially
in view of the recent findings9), studies of SMM contribute
to an understanding of magnetic phenomena such as, for
instance, quantum tunnelling of the magnetization. On the
other hand, ECC with lower nuclearity are versatile building
blocks for cooperatively associated magnetic systems.10 They
can also be incorporated in crystal lattices together with
The control of exchange magnetic coupling at the molec-
ular scale is one of the most challenging themes in the field
of molecular magnetism. Among a large variety of magneti-
cally interesting molecules,1 special attention has been paid
to exchange coupled clusters (ECC). An ECC molecule
typically consists of a small number (∼4 to 30) of paramag-
netic transition-metal ions, which are linked together by
simple bridges such as O2-, HO-, CH3O-, F-, Cl-, RCOO-.
The bridges provide a superexchange pathway, giving rise
to an isotropic coupling on the order of 1 to 100 cm-1. Some
ECC show slow relaxation of magnetization of purely
molecular origin and they are usually called single molecule
magnets (SMM).2–5 Probably, the most famous example is
* Author to whom correspondence should be addressed. E-mail: welter@
chimie.u-strasbg.fr, Tel: +33 90 24 15 93, Fax: +33 90 24 12 29.
(4) Gatteschi, D.; Sessoli, R. Angew. Chem., Int. Ed. 2003, 42, 268.
(5) Caneschi, A.; Gatteschi, D.; Sessoli, R. J. Am. Chem. Soc. 1991, 113,
5873.
†
Laboratoire DECOMET, Universite´ Louis Pasteur.
I.P.C.M.S., Groupe des Materiaux Inorganiques.
Laboratoire POMAM, Universite´ Louis Pasteur.
Laboratoire d’Electrochimie et de Chimie Physique du Corps Solide,
‡
§
(6) Lis, T. Acta Crystallogr. 1980, B36, 2042.
(7) Milios, C. J.; Vinslava, A.; Wernsdorfer, W.; Moggach, S.; Parsons,
S.; Perlepes, S. P.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc.
2007, 129, 2754–2755.
|
Universite´ Louis Pasteur Strasbourg I.
(1) Kahn, O., Molecular Magnetism; Wiley-VCH: New York, 1993.
(2) Single-Molecule Magnets and Related Phenomena. In Structure
Bonding; Winpenny, R., Ed.; Berlin, 2006; 122.
(3) Sessoli, R. Mol. Cryst. Liq. Cryst. 1995, 274, 145.
(8) Leuenberger, M. N.; Loss, D. Nature, 2001, 410, 789.
(9) Ruiz, E.; Cirera, J.; Cano, J.; Alvarez, S.; Loose, C.; Kortus, J. Chem.
Commun. 2008, 52.
10.1021/ic800522h CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 17, 2008 7623
Published on Web 07/29/2008

Bouslimani et al.
Scheme 1. Synthesis of C1, C2, and C3 from H2L
Figure 1. Hypothetical situation between paramagnetic TIII mononuclear
and binuclear complexes.
conducting molecules, leading to innovative magnetic/
conducting bifunctional materials.11 Because of the depen-
dence of the coupling pathways on structure and symmetry
of the organic ligands, dinuclear ECC offer interesting
possibilities to tune metal-to-metal interactions by small
structural changes of the organic periphery. Following this
approach, we have been working on small modifications of
the organic periphery as well as the metallic centers of a
very exciting dinuclear MnIII complex [Mn2(HL)4(µ-OCH3)2]
(where H2L is 2-salicyloylhydrazono-1,3-dithiolane), which
present the largest J value (J ) +19.7 cm-1) ever reported
so far for a MnIII-MnIII interaction.12 The peculiarity of the
[Mn2(HL)4(µ-OCH3)2] structure (in the solid state) is an
unsymmetrical arrangement of the ligands that is responsible
for ferromagnetic ground state.12 This unexpected situation
has to be related to intramolecular nonclassical hydrogen
bonds occurring between hydrogen atoms of the dithiolane
ring and the centroid of the phenol groups (Figure 1). In the
present article, we specifically focus on iron compounds
obtained with the same H2L ligand to investigate in solution
(with the use of EPR studies) the equilibrium between mono-
and binuclear complexes.
isolated as air-sensitive single crystals after a slow diffusion
of diethyl ether into the DMF solution (Figure 2).
The solvation of the single crystals of C2 by DMF solvent
(under air) gives back, instantly, a blue solution of C1 (step
F, Scheme 1). In addition, these yellow crystals of C2, as
they are air-sensitive, were found to slowly convert (within
a few days) to blue solid under air exposure. A similar
observation has been done with a DMF solution of C2 (under
stirring), which can be quickly (within few seconds) oxidized
to C1 under air exposure.
From electrochemical studies as well as EPR, magnetic
measurements, and crystal-structures determination, we can
assume that the iron atoms in C1 are in a (+III) oxidation
state, both in solution and in crystals, whereas those in C2
are in a (+II) oxidation state. A preliminary study of this
unusual spontaneous reduction of FeIII to FeII was carried
out by electrochemical investigations (Figures 11 and 12).
Results and Discussion
Ligand H2L was prepared as previously reported where
H2L is 2-salicyloylhydrazono-1,3-dithiolane).13,14 The first
step in the present work was to examine the possibility to
obtain well-defined molecular complexes when the ligand
H2L is reacted with FeIII salts. Considering the results
described hereafter, commercial crystalline FeCl3 and/or
Fe(acac)3 have been selected as main-metal sources.
The reaction of the ligand H2L (2 equiv) with iron(III)
chloride in DMF as solvent affords C1 (step A, Scheme 1)
in a good yield. Dark-blue single crystals have been obtained
by diffusion of diethyl ether into the DMF solution of the
complex under air. When the recrystallization process was
performed in sealed glass tubes, we observed the formation
of a yellow complex C2 (step F′, Scheme 1), which was
(10) Ferbinteanu, M.; Miyasaka, H.; Wernsdorfer, W.; Nakata, K.; Sugiura,
K. I.; Yamashita, M.; Coulon, C.; Cle´rac, R. J. Am. Chem. Soc. 2005,
127, 3090.
(11) Hiraga, H.; Miysaka, H.; Nakata, K.; Kajiwara, T.; Takaishi, T.;
Oshima, Y.; Nojiri, H.; Yamashita, M. Inorg. Chem. 2007, 46, 9661.
(12) Beghidja, C.; Rogez, G.; Kortus, J.; Wesolek, M.; Welter, R. J. Am.
Chem. Soc. 2006, 128, 3140.
Figure 2. Left, starting point of the diffusion process of diethyl ether
(colorless liquid) in C3 solution (blue, in DMF); middle, color change
(blue to yellow) when the diethyl ether diffuses in the DMF solution;
right, end of the diffusion process with the yellow crystals on the glass
tube bottom.
(13) Beghidja, C.; Wesolek, M.; Welter, R. Inorg. Chim. Acta 2005, 358,
3881.
(14) Chen, H. C.; Gong, X. Q.; Luo, J. X.; Huang, F.; Fang, J. Y.; Pan,
Z. J.; Liu, W. J. Chin. J. Org. Chem. 2000, 5, 833.
7624 Inorganic Chemistry, Vol. 47, No. 17, 2008

Mono- and Binuclear Iron Complexes
Nevertheless, the actual mechanism of this reduction process
is not yet well understood as of today.
Single crystals of C2 were obtained in numerous solvent/
nonsolvent conditions (Experimental Section). These three
refined crystal structures are given as Supporting Information
(CIF files, hydrogen bonds, tables), and their compared
description is given hereafter.
The binuclear FeIII C3 could be prepared by two methods
depending on the FeIII salt used for the preparation. The direct
reaction of the ligand H2L with FeCl3 in methanol in the
presence of sodium acetate (AcONa) (step B, Scheme 1)
yields a dark-red polycrystalline powder of C3, which can
be recrystallized from a THF solution by slow diffusion of
methanol to afford well-formed dark-red crystals of C3.
Likewise, the reaction of the ligand H2L with Fe(acac)3 in
DMF or THF (step C, Scheme 1) affords the hypothetical
(nonisolated) mononuclear species C4, which can be con-
sidered as an intermediate in the formation of C3 as
evidenced by EPR spectroscopy (L, L′ are the remaining
acetylacetonate ligand from step C, or the methoxy group
and a solvent ligand resulting from the solvation of C3 in
the step E′). Slow diffusion of methanol into a THF solution
of C4 (step E, Scheme 1) gave dark-red crystals of C3. An
alternative method can be used for the preparation of C3
starting from the mononuclear C1 (which can also be
considered as an intermediate in the formation of C3) by
addition of sodium acetate to a methanol solution of C1 (step
G, Scheme 1). Finally, in light of what was explained above,
the hypothetical situation between C1 and C3 (step G′
Scheme 1) can be rationalized. Moreover, we succeeded in
synthesizing the complex C2 from direct reaction of H2L
and FeCl2 in DMF. This procedure led to the rapid formation
of a yellow solution of C2, which was isolated and
characterized as yellow single crystals (obtained by diethyl
ether diffusion in DMF).
Figure 3. ORTEP view of C1 with partial labeling scheme. The ellipsoids
enclose 50% of the electronic density. The dashed lines show intramolecular
hydrogen bonds. Selected bonds length (angstroms) and angles (degrees):
Fe-O2, 1.973(4); Fe-O4, 1.977(4); Fe-O5, 2.035(4); Fe-N4, 2.145(5);
Fe-N2, 2.213(5); Fe-Cl, 2.3318(14); O2-Fe-O4, 159.84(17); O5-Fe-N4,
166.56(17).
Figure 4. ORTEP view of C2(Et2O) with partial labeling scheme. The
ellipsoids enclose 50% of the electronic density. The Et2O solvent molecule
is omitted for clarity. The dashed lines show intramolecular hydrogen bonds.
Selected bonds length (angstroms) and angles (degrees): Fe1-O2, 2.115(1);
Fe1-O4, 2.122(1); Fe1-N4, 2.199(1); Fe1-N2, 2.229(1), Fe1-Cl1
2.4396(7), Fe1-Cl2 2.4613(7); O2-Fe1-O4, 175.41(4); O4-Fe1-N4,
74.55(4).
These three iron paramagnetic complexes (C1, C2 and C3)
have been completely characterized by single-crystal dif-
fraction.
Description of the Crystal Structures. NB: Crystal-
lographic data (as well as full hydrogen bonding scheme)
are gathered in the Supporting Information, and the selected
bond lengths and angles are given in the corresponding figure
captions (Figures 3, 4 and 5).
4, the compound consists of a mononuclear complex of
iron(+II). The presence of hydrogen atoms on the N1 and
N3 nitrogen atoms (clearly pointed out by Fourier differences
and by the special conformation of the hydroxyl groups)
indicates that both bidentate ligands are neutral. Considering
both chloride atoms (-I) to complete the octahedral sur-
rounding of the iron cation, it is in +II oxidation state. This
is confirmed by solid-state EPR as well as by magnetometric
measurements (Curie constant C ) 3.68 emu · K· mol-1, in
perfect agreement with a high-spin FeII complex, Figure S4
in the Supporting Information). In C2, both bidendate H2L
ligands are in a trans configuration, as observed in the
analogous Mn+II(HL)2(Py)2 compounds.13
Twelve hydrogen bonds have been detected in the crystal
structure of C2(Et2O); these most likely account for the great
stability of this FeII compound in the solid state. The large
Fe-N, Fe-O and Fe-Cl bond distances deduced from the
diffraction data (Figures 4, 5 and 6) are in accordance with
a high-spin electronic configuration (S ) 2) for this complex.
Nevertheless, the formation of C2 from C1 remains unclear
to us: it constitutes, to the best of our knowledge, the first
example of spontaneous reduction in solution of a high spin
C1 crystallizes in the monoclinic space group P21/c. As
clearly shown in Figure 3, both HL- ligands around the FeIII
cation remain planar and are in a cis configuration, as already
13
observed for the analogous MnII or MnIII12 complexes. The
octahedral coordination environment of the FeIII ion in C1
is classical and includes a chloride atom and one DMF
molecule. Three intramolecular hydrogen bonds are de-
tected15 and are located between O1 and N1 via H1O, O3,
and N3 via H3O and at least between O5 and C22 via H22C.
From the crystal packing point of view, a 3D arrangement
is observed, the packing being related to numerous van der
Waals contacts.
Single crystals of C2 have been obtained as described
above (also Figure 2). This compound crystallizes in the
j
triclinic centrosymmetric space group P1. As shown in Figure
Inorganic Chemistry, Vol. 47, No. 17, 2008 7625

Bouslimani et al.
Figure 5. Partial view17 of the crystal packing in complex C2(Et2O). The hydrogen atoms are omitted for clarity. The dashed lines represent the Cl1-Fe1-Cl2
bonds.
Figure 6. Structural comparison of the three refined crystal structures of the C2 complexes. Only the direct iron coordination sphere is represented with
bond labels. The circles around the FeII atoms (C2_(MeOH)2 and C2_(CH2Cl2)2) indicate symmetry centers. Symmetry operator for equivalent positions: -x
+ 2, -y, -z + 1.
FeIII complex to a high spin FeII complex (one related case
can be found in ref 16), well characterized in the solid state.
A packing view of C1 in the solid state is given in Figure 5.
The structure can be described as a pseudo-bidimensionnal
arrangement of complex C2(Et2O) (all Cl1-Fe1-Cl2 axes
are aligned), the molecules being connected to each other
by hydrogen bonds (Table S2 in the Supporting Information)
and/or van der Waals contacts.
Crystals of C2 can also be obtained from THF solution
of FeCl3 and 2 equiv of the ligand H2L by slow diffusion
of methanol and from ethanol solution of the same mixture
by slow diffusion of CH2Cl2. In both cases, well-formed
yellow crystals of C2 (named C2_methanol and C2_dichlo)
are obtained and characterized. The three different crystal
structures of C2 can be compared regarding the following
points:
• The lattice symmetry is monoclinic for C2_MeOH and
C2_CH2Cl2, whereas this symmetry is triclinic for C2(Et2O).
Moreover, in this last case, there is no symmetry center on
the iron atom.
• By changing the crystallization conditions, different
solvent molecules are incorporated in the crystal lattice: one
Et2O molecule for C2(Et2O) per iron complex, two MeOH
and two CH2Cl2 per iron complex for C2_methanol and
C2_dichlo, respectively.
can be described as a neutral asymmetric complex, where
each iron ion is chelated by two HL- bidentate ligands and
bridged by two methoxy anions.
As already observed for the analogous MnIII compound,
C3 features a nonsymmetrical situation for the two metal
ions. In fact, Fe1 and Fe2 are in two different distorted
octahedra. Around Fe1, both nitrogen atoms N2-N4 are in
a trans configuration, and around Fe2 both nitrogen atoms
N6-N8 are in a cis configuration. From another point of
view, the four oxygen atoms are quite square-planar around
Fe1, whereas the four oxygen atoms are in tetrahedral
configuration around Fe2. This special situation is related
to both nonclassical intramolecular hydrogen bonds found
between C10 and the centroid of the C31_C36 phenol group
and C20 and the centroid of the C21_C26 phenol group
(Figure 5). It is important to note that this µ-methoxo
binuclear iron compound adopts the same molecular structure
and the same crystal symmetry as that previously described
for the analoguous MnIII complex.12 From a crystal packing
point of view, a 3D arrangement is observed, the packing
being related to numerous hydrogen bonds as well as van
der Waals contacts (Table S2 in the Supporting Information).
With two additional electrons for this complex versus its
manganese analogue (MnIII f FeIII), it is of course highly
relevant to analyze the magnetic properties of this new
complex.
• Finally, the molecular cores of the three compounds are
very similar (distances and angles are the same), as clearly
shown in Figure 6, which confirm the +II oxidation and high-
spin state of the iron ion in these C2 complexes.
C3, obtained as well-formed dark-red single crystals,
crystallizes in the monoclinic centrosymmetric space group
P21/n. This compound consists of a µ-methoxo binuclear FeIII
complex, as expected. The molecular structure (Figure 7)
Magnetic Measurements of the µ-Methoxo Binuclear
Iron C3. The magnetic properties of the C3 µ-methoxo
binuclear iron(III) complex were investigated in the solid
state in the 300-1.8 K temperature range with an applied
field of 5 kOe (Figure 8). The ꢀT value at room temperature
(5.74 emu · K · mol-1) is well below the expected value for
two high-spin FeIII (8.75 emu · K · mol-1 assuming g ) 2).
7626 Inorganic Chemistry, Vol. 47, No. 17, 2008

Mono- and Binuclear Iron Complexes
Upon cooling, the ꢀT product decreases continuously, down
to almost 0 (0.05 emu · K · mol-1 at 1.8 K). This indicates
the occurrence of a relatively strong antiferromagnetic
intramolecular interaction between the two FeIII ions, which
is in perfect accordance with the orbital model previously
described for the analogous MnIII dimer.12 In the present case,
all orbitals are half-filled, and thus the ferromagnetic
pathways resulting from the interaction between one half-
filled and one empty orbitals are no longer available.
The data were fitted using the following spin Hamiltonian
where all parameters have their usual meaning and the spin
operator S is defined as S ) SFe1 + SFe2
:
H ) -JSFe1SFe2 + gꢁHS
To reproduce the data satisfactorily we had to consider a
5
certain amount F of paramagnetic impurity (Simpur ) /2).
The fit leads the following values: J ) -27.4(1) cm-1, g )
2.05(1), and F ) 0.79(1) % with an excellent agreement
factor R ) 7.6 × 10-5 (R is defined as R ) ∑(ꢀexptl - ꢀcalcd)/∑
ꢀ2exptl).
The value of exchange coupling constant determined for
C3(MeOH)2 is perfectly in line with the values reported for
other symmetric bis-µ-alkoxo bridged diiron(III) complexes
with similar structural features (Fe-Fe distances and
Fe-O-Fe angles).18
Despite many attempts, it is very difficult to give a general
tendency for an accurate magneto-structural correlation for
weakly coupled FeIII binuclear complexes. Although it is clear
that some parameters (the Fe-Fe distance, the Fe-O-Fe
angle) play a role, their relative contribution to the coupling
constant J is still questioned. As an example, the very nice
correlation established by Le Gall et al. between J and the
Fe-O-Fe angle for a family of binuclear FeIII complexes
with ꢁ-diketonate-alkoxide peripheral ligands19 does not
apply to all the other complexes reviewed for instance by
R. Werner et al.18 The semiempirical relation established by
Gorun and Lippard20 is probably the best compromise
between simplicity and accuracy. It gives J as a function of
a coupling distance P, defined as half the shortest superex-
change pathway between the two FeIII ions:
Figure 7. ORTEP view of C3(MeOH)2 with partial labeling scheme. The
ellipsoids enclose 50% of the electronic density. Both MeOH solvent
molecules are omitted for clarity. The dashed lines show representative
intramolecular hydrogen bonds and the arrows represent the CH-π
interactions. Selected bonds length (angstroms) and angles (degrees):
Fe1-O2, 1.974(2); Fe1-O4, 1.966(2); Fe2-O1, 1.967(2); Fe2-O8,
1.964(2); Fe1-N2, 2.173(3); Fe1-N4, 2.173(3); Fe2-N6, 2.150(3);
Fe2-N8, 2.171(3); Fe1-Fe2,, 3.085(4).
EPR Studies of the µ-Methoxo Binuclear Iron Com-
plex C3. The EPR spectra of the compound C3 have been
recorded at different temperatures both in frozen solution
and in the solid state for polycrystalline powders. The EPR
signal of powders is temperature dependent, both in shape
and in intensity (Figure 9). It corresponds to a high-spin
hexacoordinated FeIII ion at 4 K with a rhombic g-tensor: g1
) 8.28, g2 ) 4.38, g3 ) 2.07.21 As the temperature increases,
extra lines superimpose to this base signal, which intensity
decreases. Considering these spectra parallel to the static
magnetic behavior, the low-temperature features should be
attributed to isolated monomer impurities as evidenced by
the Curie tail in the SQUID susceptibility (Figure 8).
Starting from the nonmagnetic ground state (S ) 0) where
only Curie impurities are observed, higher-spin multiplets
are thermally populated and the additional features should
be ascribed to these multiplets. A detailed study of these
fine structure lines has not been performed in the explored
X-band frequency range because a higher frequency is more
relevant for such investigation.
The compound is EPR silent in fluid solution at room
temperature. The EPR signal in frozen solution exhibits
hyperfine structure superimposed to the signal of an isolated
high-spin hexacoordinated FeIII ion throughout the studied
temperature range (4-100 K). This high-spin FeIII spectrum
has a similar g tensor than the one observed at 4 K for the
powder (Figure 10). The hyperfine pattern is ascribed to the
four nonequivalent 14N nuclei directly bound to the iron ions.
The broadening of the hyperfine lines makes them disappear
as the temperature increases. Unlike the behavior of the
powder material, the line shape as a whole does not
J ) A exp(B × P) (referring to the following Hamiltonian:
H ) -JSFe1SFe2
)
The parameters reported by Gorun and Lippard20 have
been refined by R. Werner et al.18 who have reported A )
-2.16 × 1013 cm-1 and B ) -13.9 Å-1.
For the present-studied complex C3(MeOH)2, P ) 1.977
Å, this semiempirical formula gives J ) -25.1 cm-1, in very
good agreement with the experimental value.
(15) Spek, A. L. PLATON, J. Appl. Crystallogr. 2003, 36, 7.
(16) Patra, A. K.; Olmstead, M. M.; Mascharak, P. K. Inorg. Chem. 2002,
41, 5403.
(17) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;
Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004,
25, 1605.
(18) Werner, R.; Ostrovsky, S.; Griesar, K.; Haase, W. Inorg. Chim. Acta
2001, 326, 78.
(19) Le Gall, F.; Biani, F. F.; Caneschi, A.; Cinelli, P.; Cornia, A.; Fabretti,
A. C.; Gatteschi, D. Inorg. Chim. Acta 1997, 262, 123.
(20) Gorun, S. M.; Lippard, S. J. Inorg. Chem. 1991, 30, 1625.
Inorganic Chemistry, Vol. 47, No. 17, 2008 7627

Bouslimani et al.
Figure 8. ꢀ (open circles) and ꢀT (open squares) as a function of
temperature for C3(MeOH)2. Full line corresponds to the fit of the ꢀ )
f(T) curve (text) and to the corresponding ꢀT product.
Figure 11. Cyclic voltammogram of C1 (1 mM) in DMF at a platinum
electrode, supporting electrolyte 0.2 M nBU4PF6, scan rate 100 mV ·s-1
,
ferrocene as internal standard. Insert, in the same conditions, cyclic
voltammogram of the H2L ligand.
Figure 9. Experimental EPR powder spectra of C3(MeOH)2 at 4 and 196
K.
Figure 12. Quasi-steady-state current-potential curves measured on both
C1 (1 mM) and C2 (1 mM) complexes in DMF at glassy carbon electrode,
supporting electrolyte 0.2 M nBU4NPF6, scan rate 2 mV ·s-1, rotation rate:
1000 rpm.
Figure 10. Comparison of some selected experimental EPR spectra of
complex C3 in polycrystalline form and in frozen solution, at 4 K.
significantly change in solution between 4 and 95 K. This
signal must be attributed to isolated high-spin hexacoordi-
nated FeIII ions resulting from the decoupling of the Fe-Fe
dimer, as observed in the solid state, that is, it corresponds
to the separation of the dimer complex in solution. This is
evidenced by the actual similarity of the spectra recorded
for the dimer in solution and for a monomer prepared with
a closely related ligand (Figure S6 in the Supporting
Information).
As a summary, the EPR spectra in the solid state cor-
respond to a magnetic behavior parallel to the one observed
by SQUID. Moreover, the studies in solution point to the
decomplexation of the dimer units, resulting in monomer
units. Such unreacted monomers are probably responsible
for the Curie tail in the SQUID susceptibility and of the EPR
signal of the powder at 4 K.
in the mixture indicates the viability of these electrochemical
measurements. The C1 electrochemical signature is in
agreement with a reversible C1/C1- system, whereas the H2L
ligand only exhibits the oxidation part of the electro-
chemical redox cycle. Considering the quasi-steady vol-
tammograms (Figure 12) measured on both C1 and C2
complexes, we assume that the C1/C1- electrochemical
signal observed in the cyclic voltammogram corresponds
to the redox system C1/C2. In fact, Figure 12 clearly
shows both oxidation and reduction waves of C2 (yellow
solution in DMF) and C1 (dark blue solution in DMF)
complexes, respectively.
In summary, the electrochemical study is consistent with
the fact that the color change from blue to yellow during
the crystallization process corresponds to the reduction of
FeIII to FeII, that is, a reduction of C1 to C2 in solution.
Electrochemical Studies of the C1 and C2 Iron
Complexes. C1 has been investigated by cyclic voltammetry
as well as pseudosteady state voltammetry. In Figure 11 are
gathered cyclic voltammograms measured on C1 as well as
H2L ligand (in insert). In both cases, the presence of ferrocene
(21) Walker, F-A. Chem. ReV. 2004, 104, 589.
7628 Inorganic Chemistry, Vol. 47, No. 17, 2008

Mono- and Binuclear Iron Complexes
Calcd for C23H25N5O5S4ClFe: C, 41.17; H, 3.76; N, 10.44. Found:
C, 40.78; H, 4.08; N, 10.56.
Conclusions
We have reported the synthesis, structures, and magnetic
properties of some new FeIII and FeII complexes, which
illustrate the very rich coordination chemistry of the multi-
dentate hydrazido salicyl derivative ligands bearing sulfur
groups. From a synthetic point of view, we have demon-
strated that the reaction of 2-salicylichydrazono-1,3-dithi-
olane ligand (H2L) with iron (+III) chloride and/or acety-
lacetonate salts readily affords new monomeric complexes
as well as µ-methoxo binuclear complexes in good yields.
All new complexes have been fully characterized by single-
crystal X-ray diffraction.
3. Fe(H2L)2(Cl)2. C2. A yellow solution of FeCl3 (19 mg, 0.12
mmol) in DMF (1 mL) is added to a well stirred colorless solution
of the ligand (60 mg, 0.24 mmol) in DMF (2 mL). The color
changes immediately from yellow to dark blue. The homogeneous
mixture was left under stirring overnight at room temperature. A
yellow solution and, when the diffusion process is ended, crystalline
powder of the titled complex were obtained after a slow diffusion
(2 to 3 days) of diethyl ether into the crude reaction mixture in a
sealed glass tube (Yield: 52 mg, 70%). Anal. Calcd for
C20H20N4O4S4Cl2Fe (0.8 Et2O): C, 40.09; H, 4.05; N, 8.06. Found:
C 39.71; H, 4.04; N, 7.67.
4. Fe2(HL)4(µ-OCH3)2. C3. This complex can be synthesized
by two ways. Starting from Fe(acac)3, a solution of this salt (0.1
mmol 35.5 mg in 2 mL of DMF) was added to a well stirred
solution of H2L (0.2 mmol, 50.8 mg in 2 mL of DMF), and the
resulting mixture color changes quickly from colorless to dark
brown. The product of this step (typically the mononuclear species)
was extracted by THF to yield a dark solution; slow diffusion of
methanol into this solution (during 6 to 12 h) affords well-formed
dark-red crystals of the titled complex with good yield (45 mg,
73%). Anal. Calcd for C42H42N8O10S8Fe2 (1·MeOH): C, 41.30; H,
3.45; N, 9.19. Found: C, 40.98; H, 3.75, N, 9.32. IR (KBr, cm-1):
3513, 2814, 2921, 1619, 1588, 1520, 1487, 1364, 1252. C3 can
also be prepared starting from FeCl3 salt in one step, by reacting
(0.2 mmol) of H2L with (0.1 mmol) of FeCl3 in methanol in the
presence of sodium acetate (NaOAc). The recrystallization from
THF by slow diffusion of methanol yields well-formed crystals of
the complex.
5. Magnetic Measurements. Magnetic measurements were
performed at the Institut de Physique et Chimie des Mate´riaux de
Strasbourg (UMR CNRS-ULP 7504) using a Quantum Design
MPMS-XL SQUID magnetometer. The susceptibility measurement
was performed in the 300-1.8 K temperature range with an applied
field of 5 kOe. Magnetization measurements at different fields at
room temperature confirm the absence of ferromagnetic impurities.
Data were corrected for the sample holder and diamagnetism was
estimated from Pascal constants.
6. EPR Measurements. The EPR spectra have been recorded
at X-band (ca. 9.8 GHz) with an ESP-300E spectrometer (Bruker)
equipped with a rectangular TE 102 cavity and an ESR 900
continuous flow cryostat for liquid helium circulation (Oxford). The
static field was controlled with a Hall probe, whereas the microwave
frequency was simultaneously recorded with a frequency counter
(HP-5350 B). Dilute solutions (ca. 10-3 mol) were degassed with
argon before recording the spectra.
7. Electrochemistry. Dimethylformamide (DMF) was used as
solvent. Before used, DMF was in contact with molecular sieves
for 12 h. Tetrabuthylammonium hexafluorophosphate; nBU4NPF6
(Fluka, electrochemical grade) was used as supporting electrolyte.
All electrochemical measurements were carried out at ambient
temperature (20 ( 2 °C) in a classical one compartment three
electrodes cell. The working electrodes were platinum electrode
(2 mm diameter) or a rotating-disk glassy-carbon electrode (3 mm
diameter), the counter electrode was a platinum wire, and a second
platinum wire was used as pseudo reference. The cell was connected
to a PGSTAT 20 potentiostat (Eco Chemie, Holland). The solution
was bubbled with argon before measurements.
Three important results should be highlighted:
• Thanks to EPR studies, this work demonstrates that the
FeIII µ-methoxo binuclear complex exists in the crystalline
solid state but not in THF solution. This observation seems
to indicate that the CH-π hydrogen bonds, as clearly
detected by single-crystal diffraction, is a key parameter for
the stability of the binuclear asymmetric complex. It is is
very likely that these hydrogen bonds are broken in THF
and/or DMF solvents. This knowledge is crucial for further
investigations with such potential magnetic species. Our aim
is now to stabilize such µ-methoxo binuclear complexes in
solution by specific functionalization.
• The second important result concerns the predicted
antiferromagnetic behavior observed for the FeIII2(HL)4(µ-
OCH3)2 complex in the crystalline solid state. This behavior
is completely in line with the orbital model proposed for
the analogous ferromagnetic MnIII2(HL)4(µ-OCH3)2 com-
plex.12
• The third highly relevant observation is the spontaneous
reduction of C1 (FeIII) to complex C2 (FeII). The reduction
process remains unclear to us but both the electrochemical
studies and crystal structure determinations of many C2
complexes clearly show that the reduction (FeIII f FeII)
occurs during the crystallization process. It constitutes, to
the best of our knowledge, the first example of spontaneous
reduction of a high-spin Fe(III) complex in solution to a high-
spin Fe(II) complex in solution as well as in the crystalline
solid state. Great attention will be focused in this very
exciting phenomenon, which opens a very broad spectrum
of new opportunities for further investigations. We presently
work on CrIII, CoIII, and CoII chemistry as well as on specific
modifications of the ligands to tune and better understand
the intramolecular hydrogen bonds in binuclear compounds.
Experimental Section
All manipulations were performed under aerobic conditions using
commercial reagents and solvents.
1. Ligand H2L. This ligand was prepared as previously de-
scribed.13,14
2. Fe(HL)2(DMF)(Cl). C1. To a well-stirred solution of H2L in
DMF (64.5 mg, 0.3 mmol in 2 mL) was added FeCl3 (13.7 mg,
0.1 mmol), after 5 h of stirring at room temperature, 8 mL of diethyl
ether were added carefully on the surface of the dark solution which
was then left undisturbed for slow precipitation (2 to 3 days). The
resulting dark crystals of the titled complex were collected by
filtration and washed with diethyl ether. (Yield: 35 mg, 62%). Anal.
8. Crystal Structure Determinations. Single crystals of C1,
C2(Et2O), C2(MeOH)2, C2(CH2Cl2)2, and C3(MeOH)2 were mounted
on a Nonius Kappa-CCD area detector diffractometer (Mo KR λ
Inorganic Chemistry, Vol. 47, No. 17, 2008 7629

Bouslimani et al.
) 0.71073 Å). The complete conditions of data collection (Denzo
software22) and structure refinements are given in Table S1 in the
Supporting Information. The cell parameters were determined from
reflections taken from one set of 10 frames (1.0° steps in phi angle),
each at 20 s exposure. The structures were solved using direct
methods (SHELXS97) and refined against F2 using the SHELXL97
and CRYSTALBUILDER softwares.23,24 The absorption was not
corrected. All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were generated according to stereochemistry and
refined using a riding model in SHELXL97. Crystallographic data
(excluding structure factors) have been deposited in the Cambridge
Crystallographic Data Centre as Supplementary publication no.
CCDC 679931-679933, 682062, 682063. Copies of the data can
be obtained free of charge on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (fax: (+44)1223-336-033; e-mail:
deposit@ccdc.cam.ac.uk).
Acknowledgment. We thank the CNRS and the Ministe`re
de la Recherche (Paris) for the Ph.D. grant of Nicolas
Cle´ment and the Agence Nationale de la Recherche (ANR
contract no. ANR-06-JCJC-0008) for funding. We also thank
Dr. Pierre Braunstein and Dr. Dominique Mandon for fruitful
discussions.
Supporting Information Available: Fully labeled ORTEP
views; figures of 1/ꢀ ) f(T) plot for complex C2, Et2O, and
additional experimental EPR spectra; CIFs; and tables of crystal
data and X-ray structure refinement parameters at 173 K for all
complexes. Hydrogen bonds detected in the crystal structures of
C1, [C2, Et2O], [C2,(MeOH)2], [C2,(CH2Cl2)2], and [C3,(MeOH)2].
This material is available free of charge via the Internet at
http://pubs.acs.org.
IC800522H
(22) Kappa CCD Operation Manual, Nonius B. V., Ed.; Delft: The
Netherlands, 1997.
(23) Sheldrick, G-M., SHELXL97, Program for the Refinement of Crystal
Structures; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(24) Welter, R. Acta Crystallogr. 2006, A62, s252, The Crystalbuilder
Project.
7630 Inorganic Chemistry, Vol. 47, No. 17, 2008