FeII Spin crossover materials based on dissymmetrical N4 Schiff bases including 2-pyridyl and 2R-imidazol-4-yl rings: Synthesis, crystal structure and magnetic and M?ssbauer properties

Article abstract of DOI:10.1016/j.poly.2006.12.017

The synthesis and characterization of new symmetrical Fe^II complexes, [FeL^A(NCS)₂] (1), and [FeL^Bx(NCS)₂] (2-4), are reported (L^A is the tetradentate Schiff base N,N′-bis(1-pyridin-2-ylethy

Full text of DOI:10.1016/j.poly.2006.12.017

Polyhedron 26 (2007) 1745–1757
www.elsevier.com/locate/poly
FeII Spin crossover materials based on dissymmetrical N4 Schiff
bases including 2-pyridyl and 2R-imidazol-4-yl rings: Synthesis,
crystal structure and magnetic and Mossbauer properties
¨
a
a
a
a
Nicolas Brefuel , Isabelle Vang , Sergiu Shova a,b, Franc¸oise Dahan , Jean-Pierre Costes ,
´
a,
*
Jean-Pierre Tuchagues
a
Laboratoire de Chimie de Coordination du CNRS, UPR 8241, 205 route de Narbonne, 31077 Toulouse Cedex, France
Institute of Applied Physics, Academy of Sciences of Moldova, Academiei Street 3, 2028 Chisinau, Republic of Moldova
b
Received 12 October 2006; accepted 8 December 2006
Available online 21 December 2006
Abstract
The synthesis and characterization of new symmetrical FeII complexes, [FeLA(NCS)2] (1), and [FeLBx(NCS)2] (2–4), are reported (LA
is the tetradentate Schiff base N,N0-bis(1-pyridin-2-ylethylidene)-2,2-dimethylpropane-1,3-diamine, and LBx stands for the family of
tetradentate Schiff bases N,N0-bis[(2-R-1H-imidazol-4-yl)methylene]-2,2-dimethylpropane-1,3-diamine, with: R = H for LB1 in 2,
R = Me for LB2 in 3, and R = Ph for LB3 in 4). Single-crystal X-ray structures have been determined for 1 (low-spin state at 293 K),
2 (high-spin (HS) state at 200 K), and 3 (HS state at 180 K). These complexes remain in the same spin-state over the whole temperature
range [80–400 K]. The dissymmetrical tetradentate Schiff base ligands LCx, N-[(2-R2-1H-imidazol-4-yl)methylene]-N0-(1-pyridin-2-ylethy-
lidene)-2,2-R1-propane-1,3-diamine (R1 = H, Me; R2 = H, Me, Ph), containing both pyridine and imidazole rings were obtained as their
[FeLCx(NCS)2] complexes, 5–10, through reaction of the isolated aminal type ligands 2-methyl-2-pyridin-2-ylhexahydropyrimidine
(R1 = H, 5–7) or 2,5,5-trimethyl-2-pyridin-2-ylhexahydropyrimidine (R1 = Me, 8–10) with imidazole-4-carboxaldehyde (R2 = H: 5, 8),
2-methylimidazole-4-carboxaldehyde (R2 = Me: 6, 9), and 2-phenyl-imidazole-4-carboxaldehyde (R2 = Ph: 7, 10) in the presence of
iron(II) thiocyanate. Together with the single-crystal X-ray structures of 7 and 9, variable-temperature magnetic susceptibility and Mo¨ss-
bauer studies of 5–10 showed that it is possible to tune the spin crossover properties in the [FeLCx(NCS)2] series by changing the
2-imidazole and/or C2-propylene susbtituent of LCx
.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Keywords: Tetradentate schiff bases; Dissymmetrical ligands; FeII Complexes; Spin crossover materials; Single-crystal X-ray structures; Magnetic and
Mo¨ssbauer studies
1. Introduction
(LS, S = 0, 1A1 M HS, S = 2, 5T2g) is easily triggered ther-
mally, magnetically, by pressure or by light irradiation [4].
N6 but also N4O2 [5], N3O2 [6], P2N2X [7], P4X [8] or P4X2
[9] (X = Cl, Br) ligand environments around FeII may yield
SC behavior, but there are no defined conditions for
designing SC materials. Slight modifications of the
ligand(s) can provide conditions favoring SC, depending
on the balance between the ligand field and the mean
spin-pairing energy.
Numerous 3d4–3d7 octahedral transition-metal com-
plexes exhibit a thermally induced crossover between the
low-spin (LS) and high-spin (HS) states [1]. Spin crossover
(SC) materials are increasingly investigated due to their
potential technological applications in molecular electronics
[2], and memory devices [3]. Among them, most interest has
been focused on FeII complexes because their bi-stability
FeII SC complexes characterized by an [FeN6] core may
be classified into two main families. One family includes the
cationic complexes [Fe(L)x]2+ where x = 1, 2, 3 or 6 when
*
Corresponding author. Tel.: +33 561 333154; fax: +33 561 553003.
E-mail address: tuchague@lcc-toulouse.fr (J.-P. Tuchagues).
0277-5387/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2006.12.017

1746
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
L is hexadentate [10], tridentate [11], bidentate [12] or
monodentate [13], respectively. These cationic species show
a large variety of SC behaviors, depending on the non-
coordinated counter anions and solvent of crystallization.
The other family includes neutral compounds [Fe(L)y(A)2]
where y = 1 or 2 when L is tetradentate [14] or bidentate
[15], respectively. The monodentate ligands A are most
often NCSꢀ or NCSeꢀ anions because halides or pseud-
R1 = H in 5–7, and Me in 8–10; R2 = H for LC1 in 5 and
8, R2 = Me for LC2 in 6 and 9, and R2 = Ph for LC3 in 7
and 10). Together with the single-crystal X-ray structures
of 7 and 9, variable-temperature magnetic susceptibility
and Mo¨ssbauer studies of 5–10, showed that it is possible
to tune the spin-crossover properties in the [FeLCx(NCS)2]
series by changing the 2-imidazole and/or C2-propylene
substituent of LCx
2. Experimental
2.1. Materials
.
ohalides like N3 or NCOꢀ yield HS compounds (low
ꢀ
ligand-field) while LS complexes can be obtained with
CNꢀ (strong ligand field). To our knowledge, very few
SC complexes have been prepared with tetradentate ligands
[14,16] and an unique example of dissymmetrical ligand is
depicted for an FeII complex [17]. The synthesis of new
SC compounds is of utmost importance in this research
area, particularly when the synthesis is performed in a con-
trolled manner with the aim of obtaining compounds
exhibiting a predictable SC behavior.
All reagents and solvents were commercially available
(Aldrich, Wako) and were used without further purification.
2.2. Syntheses
In the present contribution, we report an original syn-
thetic route for the synthesis of new SC materials. We first
synthesized new FeII complexes from symmetrical tetrad-
entate Schiff base ligands: [FeLA(NCS)2] (1) (LA stands
for N,N0-bis(1-pyridin-2-ylethylidene)-2,2-dimethylpropane-
1,3-diamine), and [FeLBx(NCS)2] (2–4) (LBx stands for
N,N0-bis[(2-R-1H-imidazol-4-yl)methylene]-2,2-dimethyl-
propane-1,3-diamine, with: R = H for LB1 in 2, R = Me
for LB2 in 3, and R = Ph for LB3 in 4 (Fig. 1)). Complexes
1–4 remaining in the same spin-state over the 80–400 K
range, dissymmetrical tetradentate Schiff bases LCx con-
taining both pyridine and imidazole rings (Fig. 1) were then
prepared as their [FeLCx(NCS)2] complexes, 5–10
(LCx stands for N-[(2-R2-1H-imidazol-4-yl)methylene]-N0-
(1-pyridin-2-ylethylidene)-2,2-R1-propane-1,3-diamine, with:
2.2.1. Ligands
The ligands were characterized by 1H and in some cases
by 13C NMR in CDCl3 (d, ppm; s, singlet; d, doublet;
t, triplet; dt = doublet of triplets; m, multiplet; b, broad).
2.2.1.1. Symmetrical ligands LA and LBx. 2,2-Dimethyl-1,3-
diaminopropane (0.102 g, 10ꢀ3 mol) and 2-acetyl-pyridine
(0.242 g, 2 · 10ꢀ3 mol) for LA, imidazole-4-carboxaldehyde
(0.192 g, 2 · 10ꢀ3 mol) for LB1
, 2-methylimidazole-4-
carboxaldehyde (0.220 g, 2 · 10ꢀ3 mol) for LB2, 2-pheny-
limidazole-4-carboxaldehyde (0.344 g, 2 · 10ꢀ3 mol) for
LB3, were stirred for 2 h in 20 mL of methanol at room
temperature. Slow evaporation of the solvent yielded
brown-yellow oils (yield: 0.280 g (90%); 0.210 g (80%);
0.243 g (85%); 0.330 (80%) for LA, LB1, LB2, and LB3
,
Fig. 1. Schematic representation of ligands LA, LBx, LCx and of the half units L and L0 (R1 = H, Me; R2 = H, Me, Ph).

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1747
respectively).1H (250 MHz): LB1, d 7.91 (s, H5, 2H), d 7.67–
7.65 (sb, H1 and H3, 4H), d 3.59 (s, H6, 4H), d 1.06 (s, H8,
6H); LB2, d 7.98 (s, H5, 2H), d 7.40 (s, H3, 2H), d 3.37
(s, H6, 4H), d 2.40 (s, H1, 6H), d 0.93 (s, H8, 6H); LB3, d
7.97 (s, H5, 2H), d 7.900–7.30 (m, H1, 10H), d 7.69 (s,
Caution: Although no such behavior was observed dur-
ing the present work, perchlorate salts are potentially
explosive and should be handled in small quantities and
with much care.
1
H3,2H), 3.42 (s, H6, 4H), d 1.06 (s, H8, 6H). H NMR
2.2.2.2. General procedure (1–4). To a methanolic solution
(10ꢀ3 mol in 10 mL) of the ligand was added slowly a meth-
anolic solution (10ꢀ3 mol in 10 mL) of [Fe(NCS)2] Æ
xMeOH. The solution immediately changed from pale yel-
low to dark blue (1) or orange–yellow (2–4) and was left
under stirring for 30 min.
spectra of LA could not be assigned due to the presence
of a mixture including the aminal resulting from single con-
densation followed by cyclization, and the Schiff base
resulting from the usual bis-condensation [18]. Neverthe-
less, complexation with Fe(NCS)2 Æ xsolvent led to quanti-
tative formation of the desired symmetrical complex.
[FeLA(NCS)2] (1): Slow crystallization (over 1 month)
yielded dark blue crystals suitable for X-ray analysis.
Yield: 0.34 g, 68%. Anal. Calc. for FeC21H24N6S2
(480.4 g molꢀ1): C, 52.50; H, 5.04; N, 17.49; S, 13.35; Fe,
11.62. Found: C, 52.25; H, 5.05; N, 17.27; S, 13.29; Fe,
11.57%. IR (cmꢀ1): 2110–2090 (NCS); 1600 (imine).
[FeLB1(NCS)2] (2): Orange crystals suitable for X-ray
analysis were obtained after addition of Et2O to the meth-
anolic solution of 2. Yield: 0.32 g, 75%. Anal. Calc. for
FeC15H18N8S2 (430.3 g molꢀ1): C, 41.87; H, 4.22; N,
26.04; S, 14.90; Fe, 12.98. Found: C, 41.78; H, 3.91; N,
25.88; S, 14.65; Fe, 12.89%. IR (cmꢀ1): 2110–2090 (NCS);
1603 (imine).
2.2.1.2. Synthesis of the half-units L and L0. 2-Acetyl-pyri-
dine (0.121 g, 10ꢀ3 mol) was added to a solution of
1,3-diaminopropane (0.074 g, 10ꢀ3 mol) for L or of 2,2-
dimethyl-1,3-diaminopropane (0.102 g, 10ꢀ3 mol) for L0
in 10 mL of CH2Cl2. The resulting pale yellow solution
was stirred for 30 min and kept undisturbed at room tem-
perature until complete evaporation of the solvent (yield:
0.17 g, 85% for L; 0.15 g, 83% for L0). Colorless crystals
and an oil were obtained for L and L0, respectively. Both
ligands were characterized by elemental analyses, IR (L),
and 2D 1H–13C NMR. Ligand L: Anal. Calc. for
C10H15N3 (177 g molꢀ1): C, 67.76; H, 8.53; N, 23.71.
Found: C, 66.63; H, 8.25; N, 23.90%. IR (cmꢀ1): 1589,
1558, 1463, 1429, 786. 1H NMR (400 MHz): d 8.59
(d, H1, J = 5.5 Hz, 1H), d 7.10 (t, H2, J = 5 Hz, 1H), d
7.73 (t, H3, J = 1.5 Hz, 1H), d 7.76 (d, H4, J = 6 Hz,
1H), d 1.35 (s, H7, 3H), d 2.62 and 2.93 (s, H8,
[FeLB2(NCS)2] (3): Slow crystallization (over one week)
yielded orange crystals suitable for X-ray analysis. Yield:
0.32 g, 70%. Anal. Calc. for FeC17H22N8S2 (458.4 g molꢀ1):
C, 44.55; H, 4.84; N, 24.45; S, 14.00; Fe, 12.20. Found: C,
44.47; H, 4.55; N, 24.44; S, 13.93; Fe, 11.61. IR (cmꢀ1):
2110–2090 (NCS); 1611 (imine).
2H + 2H),
d
1.63 (m, H9, 2H). 13C {1H}NMR
[FeLB3(NCS)2] (4): Yellow microcrystalline powder.
Yield: 0.38 g, 65%. Anal. Calc./found for FeC27H27N8S2
(583.5 g molꢀ1): C, 55.67; H, 4.50; N, 19.24; S, 11.00; Fe,
9.58. Found: C, 55.60; H, 4.27; N, 19.05; S, 10.29; Fe,
9.20%. IR (cmꢀ1): 2110–2090 (NCS); 1603 (imine).
(100.6 MHz): d 164.3 (C5), d 158.2 (C1), d 137.2 (C3), d
121.3 (C2), d 121.2 (C4), d 71.7 (C6), d 41.8 (C8), d 27.0
(C9). Ligand L0. Anal. Calc. for C12H19N3 (205 g molꢀ1):
C, 70.20; H, 9.33; N, 20.47. Found: C, 69.23; H, 9.26; N,
1
20.54%. H (400 MHz): d 8.61 (d, H1, J = 5.5 Hz, 1H), d
7.77 (d, H4, J = 7 Hz, 1H), d 7.71 (td, H3, J = 2 Hz and
J = 7 Hz, 1H), d 7.19 (td, H2, J = 5.5 Hz and J = 2 Hz,
1H), d 1.41 (s, H7, 3H), d 2.38 and 2.99 (s, H8,
2H + 2H), d 0.69 and 1.14 (s, H10, 6H). 13C {1H} NMR
(100.6 MHz):d 164.1 (C5), d 149.4 (C1), d 137.1 (C3), d
122.2 (C2), d 122.0 (C4), d 71.3 (C6), d 53.7 (C8), d 28.7
(C9), d 23.3 and 23.88 (C10).
2.2.2.3. General procedure (5–10). To a pale yellow solu-
tion (10ꢀ3 mol, 10 mL methanol) of the aminal L (5–7) or
L0 (8–10) the freshly prepared Fe(NCS)2 solution
(10ꢀ3 mol, 10 mL methanol) was slowly added. The color
of the reaction mixture immediately turned to dark blue
and was stirred for 5 min. The appropriate 2R-imidazole-
4-carboxaldehyde (10ꢀ3 mol, 10 mL methanol) was then
added (imidazole-4-carboxaldehyde for 5 and 8, 2-methy-
limidazole-4-carboxaldehyde for 6 and 9, and 2-phenyl-
imidazole-4-carboxaldehyde for 7 and 10). No further
color change was observed; the solution was stirred for
30 more minutes and then left undisturbed until a micro-
crystalline powder or crystals quantitatively formed.
[FeLC1(NCS)2] (5): Dark blue powder. Yield: 0.32 g,
75%. Anal. Calc. for FeC16H17N7S2 (427.3 g molꢀ1): C,
44.97; H, 4.01; N, 22.94; S, 15.01; Fe, 13.07. Found: C,
45.15; H, 4.00; N, 22.85; S, 14.67; Fe, 12.72%. IR (cmꢀ1):
2112–2060 (NCS); 1620 (imine).
2.2.2. Complexes
All complexation reactions and sample preparations for
physical measurements were carried out in a purified nitro-
gen atmosphere within a glovebox (Vacuum Atmospheres
H.E.43.2) equipped with a dry train (Jahan EVAC 7).
2.2.2.1. Fe(NCS)2 Æ xMeOH. Fe(ClO4)2 Æ xH2O (10ꢀ3 mol)
was reacted with potassium thiocyanate (0.195 g,
2 · 10ꢀ3 mol) in 10 mL of MeOH following a procedure
adapted from the preparation of Fe(py)4(NCS)2 [19]. After
removal of potassium perchlorate through filtration, the
freshly prepared Fe(NCS)2 Æ xMeOH was used immediately
for the preparation of complexes 1–10.
[FeLC2(NCS)2] (6): Dark blue powder. Yield: 0.32 g,
73%. Anal. Calc. for FeC17H19N7S2 (441.4 g molꢀ1): C,
46.26; H, 4.34; N, 22.22; S, 14.53; Fe, 12.65. Found: C,

1748
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
46.33; H, 4.49; N, 20.32; S, 14.39; Fe, 12.50%. IR (cmꢀ1):
2060 (NCS); 1638 (imine).
2.4. Crystallographic data collection and structure
determination for 1, 2, 3, 7 and 9
[FeLC3(NCS)2] (7): Dark blue single crystals. Yield:
0.30 g, 60%. Anal. Calc. for FeC22H21N7S2 (503.4 g molꢀ1):
C, 52.49; H, 4.21; N, 19.48; S, 12.74; Fe, 11.09. Found: C,
52.19; H, 3.84; N, 19.09; S, 12.70; Fe, 10.90%. IR (cmꢀ1):
2068 (NCS); 1641 (imine).
The X-ray data for compounds 1, 2, 7 and 9 were col-
lected on a Stoe Imaging Plate Diffractometer System
(IPDS) equipped with an Oxford Cryosystems cooler
˚
device using a graphite monochromator (k = 0.71073 A).
[FeLC4(NCS)2] (8): Dark blue powder. Yield: m =
0.32 g, 72%. Anal. Calc. for FeC18H21N7S2 (455.4 g molꢀ1):
C, 47.48; H, 4.65; N, 21.53; S, 14.08; Fe, 12.26. Found: C,
47.82; H, 4.47; N, 20.74; S, 13.81; Fe, 11.80%. IR (cmꢀ1):
2108–2058 (NCS); 1618 (imine).
Data were collected [21] using u rotation movement with
the crystal-to-detector distance equal to 70 mm (u = 0.0-
270ꢁ, Du = 1.5ꢁ for 1, u = 0.0–200ꢁ, Du = 1.5ꢁ for 2 and
9 and u = 0.0–250ꢁ, Du = 1.4ꢁ for 7). Diffraction data for
3 were collected using an Oxford-Diffraction XCALIBUR
CCD diffractometer equipped with a graphite-monochro-
mated Mo Ka radiation source. The crystal was placed
50 mm from the CCD detector. More than the hemisphere
of reciprocal space was covered by combination of four sets
of exposures; each set had a different u-angle (0ꢁ, 90ꢁ, 180ꢁ,
270ꢁ) and each exposure of 20 s covered 0.75ꢁ in x. Cover-
age of the unique set is 99.5% complete up to 2h = 63.22ꢁ.
The unit cell determination and data integration were car-
ried out using the CrysAlis package of Oxford Diffraction
[22]. All structures were solved by direct methods using
SHELXS-97 [23] and refined by full-matrix least-squares on
F 2o with SHELXL-97 [24] with anisotropic displacement
parameters for non-hydrogen atoms. The hydrogen atoms
[FeLC5(NCS)2] Æ H2O (9): Dark blue single crystals.
Yield: 0.35 g, 72%. Anal. Calc. for FeC19H23N7S2 Æ H2O
(487.4 g molꢀ1): C, 46.82; H, 5.17; N, 20.12; S, 13.16; Fe,
11.46. Found: C, 47.10; H, 4.52; N, 20.29; S, 12.73; Fe,
11.30%. IR (cmꢀ1): 2058 (NCS); 1634 (imine).
[FeLC6(NCS)2] (10): Dark blue microcrystals. Yield:
0.35 g, 66%. Anal. Calc. for FeC24H25N7S2 (531.5 g molꢀ1):
C, 54.24; H, 4.74; N, 18.45; S, 12.06; Fe, 10.51. Found: C,
54.04; H, 4.31; N, 18.16; S, 11.64; Fe, 10.51, 10.37%. IR
(cmꢀ1): 2072 (NCS); 1643 (imine).
2.3. Physical measurements
˚
Elemental analyses were carried out on a Perkin–
Elmer 2400 series II device (C, H, N) at the Microanalyt-
ical Laboratory of the Laboratoire de Chimie de
Coordination in Toulouse, and on a Maxim Thermoelec-
tron device (S, Fe) at the Service Central de Microanaly-
were placed in calculated positions (C–H 0.96 A) and they
were included in the refinement in the riding-model approx-
imation with Uiso = 1.1 U (atom of attachment). Positional
parameters of the H(water) atoms in compound 9 were
obtained from difference Fourier syntheses and verified
by the geometric parameters of the corresponding hydro-
gen bonds. Scattering factors were taken from the standard
compilation [25]. The molecular plots were obtained by
using the ZORTEP program [26]. Numerical absorption cor-
rections [27] were applied for 3, while for compounds 1, 2, 7
and 9 the absorption corrections were introduced by semi-
empirical methods based on equivalent reflections, using
the program MULTISCAN [28]. The final full-matrix least-
1
ses du CNRS in Vernaison, France. 1D H and 1D 13C
NMR spectra were performed on an AC 250 FT spec-
trometer (Bruker) and 2D 1H13C spectra on an AMX
400 spectrometer (Bruker). Chemical shifts are given in
ppm versus TMS (1H and 13C) using CDCl3 as solvent.
IR spectra were measured in the 400–4000 cmꢀ1 range
on a 9800 FTIR spectrometer (Perkin–Elmer). Samples
were run as KBr pellets. Mo¨ssbauer spectra were
recorded on a constant-acceleration conventional spec-
trometer with a 50 mCi source of 57Co (Rh matrix).
The absorber was a powdered sample enclosed in a
20 mm diameter cylindrical plastic sample holder, the size
of which has been determined to optimize the absorption.
Spectra were obtained in the 80–400 K range using a
MD306 Oxford cryostat (4–293 K) or an home-made vac-
uum oven (300–500 K), the thermal scanning being mon-
itored by an Oxford ITC4 servo control device ( 0.1 K
accuracy). A least square computer program [20] was
used to fit the Mo¨ssbauer parameters and to determine
their standard deviations of statistical origin (given in
parentheses). Isomer shift values (d) are relative to iron
foil at 293 K. Variable temperature magnetic susceptibil-
ity data were collected on powdered samples on a
MPMS-55 Quantum Design SQUID magnetometer or a
Faraday type susceptometer under applied magnetic fields
lower than 0.1 T. Diamagnetic corrections were applied
by using Pascal’s constants.
h
i
2
ꢀ
ꢁ
P
2
squares refinement, minimizing
w jF 2oj ꢀ jF 2cj
, con-
verged at the values of R and wR listed in Table 1, together
with the main crystallographic parameters. The relatively
high R-factors for complex 7 are due to the low quality
of the crystals: among several sets of diffraction data
obtained from different crystals which all diffracted poorly,
the best one was used to solve this structure.
3. Results
3.1. Syntheses
While the reaction of 2 equiv of 2R-imidazole-4-carbox-
aldehyde with 1 equiv of 2,2-dimethyl-1,3-diaminopropane
yielded pure samples of the corresponding acyclic tetraden-
tate bis-Schiff bases LBx, the corresponding condensation
reaction carried out with 2-acetyl-pyridine yielded a mix-
ture including the cyclic aminal L0 resulting from single

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1749
condensation, and the Schiff base resulting from the usual
bis-condensation. This lower reactivity of ketone compared
to aldehyde carbonyl function is well documented [29] but
has been scarcely exploited to prepare so-called ‘half-unit’
synthons through 1:1 condensation with the aim to gener-
ate dissymmetrical ligands through opening of the aminal
ring and condensation of the amine function with a differ-
ent carbonyl [30]. One of the reasons is the occurrence of
redistribution reactions that yield mixtures of the respec-
tive symmetrical bis-Schiff bases corresponding to each
carbonyl synthon, including or not a small percentage of
the targeted dissymmetrical bis-Schiff base. Using a metal
ion in order to favor the acyclic Schiff base versus the cyclic
aminal intermediate through complexation, and to subse-
quently favor formation of the tetracoordinated dissym-
metric bis-Schiff base has occasionally allowed avoiding
redistribution reactions, and has thus allowed preparing
dissymmetric bis-Schiff base cleanly and quantitatively in
a very few cases [31].
Based on these results, we explored the 1:1 condensation
reactions between 2-acetyl-pyridine and 1,3-diaminopro-
pane or 2,2-dimethyl-1,3-diaminopropane. These reactions
yielded the isolated aminal type ligands L: 2-(2-methyl-
hexahydro-2-pyrimidinyl)pyridine and L0: 2-(2-methyl-
5-dimethylhexahydro-2-pyrimidinyl)pyridine, respectively
(first reaction sketched in Fig. 2).
The NMR spectra clearly showed that, in the case of L
and L0, in the absence of metal ion the acyclic single con-
densation intermediate is not stable and the equilibrium is
totally shifted toward the cyclic aminal form. Addition of
2R-imidazole-4-carboxaldehyde to L (L0) yielded mixtures
of the respective symmetrical bis-Schiff bases and possibly
a small amount of the targeted dissymmetrical bis-Schiff
base, as suggested by the NMR spectra that showed a large
number of intricated absorption resonances.
Clean and quantitative syntheses of the dissymmetrical
ligands LCx could only be performed in the presence of
Fe(NCS)2 Æ xMeOH yielding their [FeLCx(NCS)2] (x =
1–6) complexes (Fig. 2). These results suggest that the
metal ion displaces totally the equilibrium from the cyclic
aminal toward the acyclic Schiff base intermediate through
complexation as sketched in the second reaction of Fig. 2.
The presence of the pyridine aromatic ring conjugated with
FeII results in a strong MLCT band [32] responsible for the
intense blue color of compounds 5–10 which are crystals (7,
9), microcrystals (10), or more or less microcrystalline
powders (5, 6, 8). The homogeneity (in color and texture)
of all samples has been checked by microscopy. Moreover,
Mo¨ssbauer spectroscopy has confirmed the presence of
both HS and LS spin isomers in temperature dependent
ratios for all compounds in this series, except for 6 which
is 100% HS (see Mo¨ssbauer section and Table SM.2).
Compounds 5–10 are oxidized in aerobic solutions. How-
ever, in the solid state, they are reasonably stable in aerobic
conditions provided the samples are not exposed to mois-
ture and/or heat. The absence of oxidation for all samples
of compounds 1–10 in the solid state has been checked by

1750
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
Fig. 2. Synthetic pathway for the family of complexes [FeLCx(NCS)2] based on the dissymmetrical tetradentate Schiff base ligands LCx (R1 = H, Me;
R2 = H, Me, Ph).
EPR spectroscopy. Formation of the intermediate com-
plexes FeL(NCS)2(MeOH) and FeL0(NCS)2(MeOH)
sketched in Fig. 2 as resulting from addition of the ferrous
salt to L (L0) is assessed on the basis of the color change
observed upon addition of Fe(NCS)2 Æ xMeOH and on
the rationale that complexation with the acyclic Schiff base
intermediate is more likely due to the entropy gain upon
conversion of the bidentate cyclic aminal into the tridentate
Schiff base. However, these dark blue intermediate com-
plexes are very soluble and we have not attempted to iso-
late them.
radentate Schiff base ligand (LA (1), LBx (2 and 3), and LCx
(7 and 9)) in the equatorial positions, while the apical posi-
tions are occupied by nitrogen atoms from two NCSꢀ
ligands. Within the estimated standard deviations, the iron
atom is in the mean plane of the four equatorial nitrogen
donors. However, the tetradentate Schiff base ligands, as
a whole, exhibit more or less significant deviations from
planarity. This can be due to the steric constraints imposed
by the pyridine and/or imidazole moieties upon coordina-
tion to the metal center, and results in tilts of the aromatic
rings from the mean plane of the tetradentate ligand. The
larger tilt is observed for the py–py dihedral angle of 1
(32.6ꢁ), and the smaller one for the py–imidazole dihedral
angle of 9 (4.8ꢁ). The Fe–N bond distances are characteris-
3.2. Description of the structures
II
˚
ORTEP plots of complexes 1, 3, and 9 are shown in
Figs. 3–5, respectively. Selected inter-atomic distances
and angles for complexes 1, 2, 3, 7 and 9 are collated in
Table 2.
The overall molecular structure of the five complexes is
similar: the distorted octahedral [FeN6] coordination envi-
ronment includes four nitrogen atoms belonging to the tet-
tic of the spin-state of Fe : the average 1.949 A value for 1
(293 K) is in good agreement with those observed for sim-
ilar LS iron(II) complexes, while the average Fe–N bond
˚
˚
distances of 2.185 A (2, 200 K), 2.189 A (3, 180 K),
˚
˚
2.196 A (7, 180 K) and 2.150 A (9, 293 K) indicate that
the FeII center is in the HS state at the temperatures of
the X-ray data collections for the four remaining com-
pounds that have been structurally characterized [4,18].
In average, in the HS structures the N–Fe–N angles depart
more from those expected for a regular octahedron
(Table 2). It may thus be concluded that, except for the
metric differences resulting from the difference in FeII
spin-state, the complexes based on the LA, LBx and LCx
ligands have similar molecular structures.
On the other hand, these compounds provide a diver-
sity of crystal packing modes, with a large variety of net-
work topologies and these peculiarities deserve detailed
consideration.
The crystal packing of 1 results from an extended 3D
network of C–Hꢁ ꢁ ꢁS contacts. A fragment of the crystal
structure showing five H-bonded molecules is displayed
in Fig. 6. The Cꢁ ꢁ ꢁS distances between C–Hꢁ ꢁ ꢁS bonded
˚
molecules are in the 3.59–3.69 A range, in keeping with pre-
vious reports [18,33]. In addition, p–p stacking interactions
involving the pyridine rings of molecules related by a crys-
tallographic inversion center participate in the crystal
Fig. 3. ORTEP plot of complex 1 at the 40% probability level. H-atoms
have been omitted for clarity.

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1751
Fig. 4. ORTEP plot of complex 2 at the 50% probability level. H-atoms have been omitted for clarity.
Fig. 6. Intermolecular interactions responsible for the supramolecular
structure of [FeLA(NCS)2] (1). p–p stacked pairs of pyridine rings are
linked by broken double lines.
Fig. 5. ORTEP plot of complex 9 at the 30% probability level. H-atoms
have been omitted for clarity.
packing, as evidenced by the short interplanar (3.64 and
3.93 A) and centroid to centroid (3.88 and 4.47 A) dis-
tances between interacting pyridine rings.
The asymmetrical unit of 2 contains two crystallograph-
ically independent [FeLB1(NCS)2] complex molecules
denoted A and B (Fig. 7). Molecules A and B are linked
˚
˚
Table 2
˚
Selected bond lengths (A) and angles (ꢁ) for complexes 1, 2, 3, 7 and 9
Compound
[FeLA(NCS)2] (1)
[FeLB1(NCS)2] (2)
[FeLB2(NCS)2] (3)
[FeLC3(NCS)2] (7)
[FeLC5(NCS)2] Æ H2O (9)
Molecule A
Molecule B
Fe–N1
Fe–N2
Fe–N3
Fe–N4
Fe–N5
Fe–N6
1.964(5)
1.953(6)
1.939(5)
1.961(6)
1.938(6)
1.939(6)
2.216(2)
2.188(2)
2.180(2)
2.216(2)
2.189(3)
2.121(3)
2.241(2)
2.191(2)
2.194(2)
2.245(2)
2.115(3)
2.132(3)
2.2838(11)
2.1603(11)
2.2170(10)
2.2178(10)
2.1124(12)
2.1409(12)
2.28(2)
2.11(2)
2.284(15)
2.31(2)
2.109(15)
2.082(16)
2.188(17)
2.158(15)
2.184(13)
2.125(15)
2.135(10)
2.112(12)
N1–Fe–N2
N1–Fe–N4
N2–Fe–N3
N3–Fe–N4
N5–Fe–N6
81.3(3)
102.1(2)
96.8(2)
80.6(2)
173.3(2)
77.35(9)
118.05(9)
87.10(8)
77.46(8)
172.36(10)
77.03(9)
120.42(9)
86.05(9)
76.43(8)
168.90(11)
76.09(4)
120.10(4)
86.58(4)
77.13(4)
166.04(5)
79.2(9)
117.3(8)
90.4(9)
73.2(8)
168.8(6)
76.3(8)
119.5(8)
88.5(7)
75.6(6)
164.2(5)

1752
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
Fig. 7. Intermolecular interactions responsible for the supramolecular
structure of [FeLB1(NCS)2] (2). Metric parameters (d, A; \, ꢁ):
˚
N7Aꢁ ꢁ ꢁS1B0, 3.296(3); Hꢁ ꢁ ꢁS1B0, 2.48; \N7A–Hꢁ ꢁ ꢁS1B0, 160.1;
N8Aꢁ ꢁ ꢁS2B, 3.353(3); Hꢁ ꢁ ꢁS2B, 2.62; \N8A–Hꢁ ꢁ ꢁS2B, 144.4; N8Bꢁ ꢁ ꢁ
N5A, 2.887(4); Hꢁ ꢁ ꢁN5A, 2.03; \N8B–Hꢁ ꢁ ꢁN5A, 172.0; N7Bꢁ ꢁ ꢁS2A00,
3.339(2); Hꢁ ꢁ ꢁS2A00, 2.70; \N7B–Hꢁ ꢁ ꢁS2A00, 132.4; symmetry codes: (0) x,
y ꢀ 1, z; (00) ꢀx + 1, ꢀy + 1, ꢀz.
Fig. 8. Intermolecular interactions responsible for the supramolecular
structure of [FeLB2(NCS)2] (3). Metric parameters (d, A; \, ꢁ): N7ꢁ ꢁ ꢁS2 ,
3.320(1); Hꢁ ꢁ ꢁS20, 2.457; \N7–Hꢁ ꢁ ꢁS20, 146.7ꢁ; N8ꢁ ꢁ ꢁS100, 3.340(1);
Hꢁ ꢁ ꢁS100, 2.569; \N8–Hꢁ ꢁ ꢁS100, 166.7ꢁ; symmetry codes: (0) 1 ꢀ x, 1 ꢀ y,
2 ꢀ z; (00) 1.5 ꢀ x, 0.5 + y, 1.5 ꢀ z.
0
˚
into a 3D supramolecular network via NHꢁ ꢁ ꢁS and
NHꢁ ꢁ ꢁN hydrogen bonds, responsible for the crystal pack-
ing. A view of this supramolecular network showing four
H-bonded molecules is displayed in Fig. 7, along with the
hydrogen-bonding parameters. The four non-coordinated
nitrogen atoms of molecules A and B act as proton donors,
thus fulfilling all H-bonds opportunities in the crystal. The
complex molecules inside each [A, B] pair are linked
through N8B–Hꢁ ꢁ ꢁN5A and N8A–Hꢁ ꢁ ꢁS2B hydrogen
bonds. It is noteworthy, that one of these hydrogen bonds
involves as acceptor the nitrogen atom N7A coordinated to
FeA, while the other one involves as acceptor the non-
coordinated atom S2B. Further association of the complex
molecules results from the N7B–Hꢁ ꢁ ꢁS2A short contact
which leads to the formation of tetra-molecular centrosym-
metrical aggregates and finally to a 3D network through
the N7A–Hꢁ ꢁ ꢁS1B0 hydrogen bond (Fig. 7).
The crystal packing of 3 results from an extended 3D
network of intermolecular interactions. Each [FeLB2
-
Fig. 9. Projection of the crystal structure of [FeLC3(NCS)2] (7) perpen-
dicularly to an ac plane. Centroid to centroid distances between
overlapping rings are materialized by broken double lines.
(NCS)2] complex molecule is involved in NHꢁ ꢁ ꢁS hydrogen
bonds with its three nearest neighbors. A view of the supra-
molecular network showing four H-bonded molecules is
displayed in Fig. 8, along with the hydrogen-bonding
parameters, illustrating how this 3D supramolecular net-
work is generated by two short intermolecular contacts
involving the two non-coordinated imidazole nitrogen
atoms and both sulfur atoms of the coordinated NCSꢀ
anions. As for compound 2, all H-bond opportunities are
realized.
(i) the N7–Hꢁ ꢁ ꢁS20 (x, 0.5 ꢀ y, ꢀ0.5 + z) hydrogen bond
0
0
0
˚
˚
(N7ꢁꢁꢁS2 , 3.37(2) A; HꢁꢁꢁS2 , 2.52 A; \N7–HꢁꢁꢁS2 , 170.3ꢁ);
(ii) the p-stacking interaction between the pyridine and
imidazole rings of adjacent molecules (centroid to centroid
˚
distances of 3.87 A).
A view of the crystal packing of complex 9 is shown in
Fig. 10. The supramolecular network is built from aggre-
gates of complex molecules connected via two intermolec-
ular H-bonds through the solvated water molecules. The
water molecule acts as proton acceptor towards the non-
coordinated imidazole nitrogen atom (N7ꢁ ꢁ ꢁO1w,
The molecular structure and crystal packing of 7 are
shown in Fig. 9. The latter consists of alternate two-dimen-
sional layers oriented along an ac crystallographic plane.
Each layer is built from infinite chains connected through
˚
˚
˚
short intermolecular contacts (3.797(6) A) between the S1
2.83(2) A; Hꢁ ꢁ ꢁO1w, 1.97 A; \N7–Hꢁ ꢁ ꢁO1w, 173.8ꢁ). The
water molecule also acts as proton donor towards the sul-
fur atom S2 of the symmetry related (0.5 ꢀ x, y ꢀ 0.5,
and S2 atoms related by translation along a. Two types
of interactions are responsible for the association of the
neutral [FeLC3(NCS)2] molecules into a chainlike structure:
0
0
˚
z ꢀ 0.5) complex molecule (O1wꢁ ꢁ ꢁS2 , 3.59(2) A; Hꢁ ꢁ ꢁS2 ,

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1753
Fig. 12. Thermal variation of the vMT product for the [FeLCx(NCS)2]
complexes 5–10.
Fig. 10. View of the crystal packing of FeLC5(NCS)2] ꢁ H2O (9). Centroid
to centroid distances between overlapping rings are materialized by
broken double lines.
and HS (2, 200 K; 3, 180 K) states, respectively. However,
the vMT product at room temperature for 2–4, as well as
that for 7 and 10 that will be discussed next, is higher than
the spin-only value of ꢂ3 cm3 molꢀ1 K expected for S = 2
(g = 2), as often observed for HS FeII compounds. The
presence of small amounts of HS FeIII (possibly resulting
from oxidation of the prepared FeII materials) being dis-
carded by their Mo¨ssbauer spectra (see following subsec-
tion), these high vMT values are attributed to operation
0
˚
2.74 A; \O1w–Hꢁ ꢁ ꢁS2 , 152.8ꢁ). The chains of H-bonded
molecules extend into a 3D network through p-stacking
interactions between pyridine and imidazole rings belong-
ing to adjacent chains, as evidenced by their centroid to
˚
centroid distances of 3.77 A (Fig. 10). The dihedral angles
between p-stacked rings equals 5.3ꢁ.
5
3.3. Magnetic studies
of spin-orbit coupling (SOC). The T2 ground state is split
by SOC yielding levels separated by energies on the order
of kT (thermally populated in the high temperature range)
which results in slightly temperature dependent vMT values
above ca. 60 K [34].
The thermal variation of the magnetic susceptibility vM
has been explored for complexes 1–10 and the vMT data for
selected temperatures (T) are collated in Table SM.1. The
complete sets of vMT versus T data are shown in Figs. 11
and 12 (1–4) (5–10). The vMT product for 1 remains lower
than 0.12 cm3 molꢀ1 K between 110 and 400 K, indicating
that the iron(II) center in this material is LS over the whole
With the exception of compound 6, and at variance with
the two previous series, the vMT product for the
[FeLCx(NCS)2] complexes 5–10 is not constant over the
explored temperature range (Fig. 12). The vMT product
for complexes 7 and 10 (R2(imidazole) = Ph) is practically
constant from 300 to 160 K and characteristic of a large
predominance of HS FeII. Then it decreases between 150
and 50 K: for 7, vMT decreases from 3.03 to 2.25 cm3
molꢀ1 K, indicating a partial HS ! LS transition centered
around 100 K of ca. 20% FeII centers. The transition is
much more gradual for 10 (Table SM.1). The thermal var-
iation of the vMT product for complex 9 (R1 = Me,
R2(imidazole) = Me) showing a very gradual and partial
SC between 250 and 50 K (ca. 25%) is similar to that
described above for 7 and 10 (R2(imidazole) = Ph). All
three compounds exhibit a sharp vMT decrease below
30 K resulting from zero field splitting (ZFS) of HS FeII.
The vMT product for complex 6 (R1 = H, R2(imidazole) =
Me) is practically constant over the 80–300 K temperature
range indicating the absence of SC for FeII in this material
above 80 K.
temperature range. The vMT product for the [FeLBx
-
(NCS)2] complexes 2–4 is comprised between 4.12 and
3.70 cm3 molꢀ1 K in the 80–300 K temperature range, indi-
cating that the iron(II) center in this series of materials is
HS over the whole temperature range. These data agree
with the structural studies of complexes 1–3 that indicate
Fe–N bond lengths characteristic of the LS (1, 293 K)
On the other hand, the magnetic behavior of complexes
5 and 8 (R2(imidazole) = H) is strikingly different: their
vMT product, practically constant and lower than 0.10
(5) and 0.13 (8) cm3 molꢀ1 K from 80 to 300 K, is charac-
teristic of LS FeII. Then, above 330 K, vMT increases up
to 1.29 and 1.37 cm3 molꢀ1 K at 400 K for 5 and 8, respec-
tively, indicating a LS ! HS transition centered above
Fig. 11. Thermal variation of the vMT product for [FeLA(NCS)2] (1), and
the [FeLBx(NCS)2] complexes 2–4.

1754
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
400 K (ca. 30% FeII centers are involved between 330 and
400 K).
3.4. Mo¨ssbauer spectroscopy
Mo¨ssbauer spectra have been recorded at selected tem-
peratures in the 20–400 K range for complexes 1–10. Val-
ues of the Mo¨ssbauer parameters (isomer shift, d,
quadrupole splitting, DEQ, and line width, C) resulting
from fitting the 80 and 293 K spectra are collated in Table
SM.2. The isomer shift values slightly decrease with
increasing temperature, indicating operation of second-
order Doppler effect [35].
3.4.1. [FeLA(NCS)2] and [FeLBx(NCS)2] complexes
As expected from the structural and magnetic data, the
Mo¨ssbauer spectrum of 1 at 293 K consists of a unique
doublet characteristic of LS FeII (d, 0.328(2) mm sꢀ1
;
DEQ, 0.865(3) mm sꢀ1). In agreement with the structural
and magnetic data, Mo¨ssbauer spectra of complexes 3
and 4, recorded between 80 and 293 K, show a unique
quadrupole split doublet characteristic of HS FeII (Table
SM.2). At variance with those of 3 and 4, the Mo¨ssbauer
spectra of 2 evidence two HS doublets in agreement with
the presence of two distinct FeII sites, FeII(A) and FeII(B),
in the crystal structure. These doublets present similar iso-
mer shifts (ꢂ1.15 mm sꢀ1 at 80 K, Table SM.2) as expected
for identical N6 donor sets around the FeII(A) and FeII(B)
sites and quite different DEQ values (ꢂ1.51 and
ꢂ2.58 mm sꢀ1 at 80 K for FeII(A) and FeII(B), respec-
tively), in agreement with different low symmetry environ-
ments [36]. The differences in Fe–N bond lengths and
N–Fe–N angles associated to the different hydrogen con-
tacts lead to significant differences in symmetry of the
ligand environment for the FeII centers of molecules A
and B, see Section 3.2.
Fig. 13. Selected Mo¨ssbauer spectra of [FeLC3(NCS)2] (7).
illustrates how important it is to use a local probe like
Mo¨ssbauer spectroscopy in addition to bulk susceptibility
measurements in order to assess the presence or absence
of SC, and the relative percentages of spin isomers. Among
compounds 6–10 for example, 7 shows the second highest
vMT product in the high temperature range, suggesting
the absence of LS fraction at high temperature. However,
Mo¨ssbauer spectroscopy evidences a partial SC with
ꢂ20% of residual LS fraction at 293 K, the second highest
among materials 6–10. On the other hand, 6 shows the low-
est vMT product at room temperature while Mo¨ssbauer
spectroscopy evidences the absence of LS fraction at
293 K.
3.5. IR spectroscopy
3.4.2. [FeLCx(NCS)2] complexes
The IR study has been focussed on the NCS stretching
absorptions (2000–2100 cmꢀ1 range) which constitute a
good sensor for the cis or trans conformation of coordi-
nated thiocyanate anions. Usually, trans isomers give a sin-
gle absorption whereas cis isomers are characterized by a
splitting of the mCN absorption [15b,37]. However, this crite-
rion must be considered with much caution because the
splitting of this band is very sensitive to slight deformations
of the coordination sphere (in particular the N–M–N angle
between the two thiocyanate ligands coordinated to the
metal center). The structural study of complex 3 has evi-
denced a distorted octahedral geometry around FeII with
a trans conformation of the thiocyanate ligands. Two well
distinct mCN absorption bands were observed in the IR spec-
tra at 2093 and 2065 cmꢀ1 in compliance with the strong
deviation from an ideal trans conformation evidenced by
the 158.8ꢁ (C17–N8–Fe) and 137.5ꢁ (C16–N7–Fe) angles.
The same conclusion can be drawn from the IR spectra of
complexes 2 and 4, as well as for complexes 5–10
The Mo¨ssbauer spectra of complexes 6–10 confirm the
spin-states deduced from the magnetic data for these mate-
rials; however, due to poor counting statistics, the high
temperature spectra (350–400 K) of complexes 5 and 8
could not be fitted. As an example, in the following we
describe the behavior of complex 7, [FeLC3(NCS)2], for
which representative Mo¨ssbauer spectra recorded in the
cooling mode are shown in Fig. 13. These spectra, mea-
sured in the 20–293 K range, confirm the partial SC
deduced from the magnetic measurements.
The relative area of the HS doublet (d = 1.113(2) and
DEQ = 2.563(4) mm sꢀ1 at 80 K) slightly decreases from
80% at 293 K to 60% at 20 K, confirming the gradual
and incomplete character of the SC in this material. The
Mo¨ssbauer parameters of the LS doublet (d = 0.439(3)
and DEQ = 0.732(5) mm sꢀ1 at 80 K) are in agreement with
previously reported values [4,18,35].
Comparison of Table SM.1 and Figs. 11 and 12 on the
one hand to Table SM.2 and Fig. 13 on the other hand
([FeLCx(NCS)2] series). The larger splitting (64 cmꢀ1
)

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1755
observed for 4 suggests a larger deformation of the ligand
environment symmetry around FeII, in agreement with
the large quadrupole splitting evidenced by Mo¨ssbauer
spectroscopy. The third band (low intensity) observed for
2 can be explained by the presence of two types of NCS
ligands, depending on the atom through which they partic-
ipate into hydrogen bonds (NHꢁ ꢁ ꢁSCN or NHꢁ ꢁ ꢁNCS that
leads to two FeII sites distinguishable crystallographically
and by Mo¨ssbauer spectroscopy, see the corresponding sub-
sections above). In line with our previous report on parent
iron(II) compounds with related dissymmetric Schiff bases
[32], the mC@N(imine) absorptions observed for compounds
1–10 do not seem to depend on the spin-state of iron(II).
This is possibly because they are no longer pure mC@N
(imine) stretches, but correspond to the simultaneous
involvement of mC@N(imine), mC@N(py) and mC@N(imidazole)
vibrators.
In this study, the 2:1 condensation of acetylpyridine and
2,2-dimethyl-1,3-diaminopropane followed by reaction
with ferrous thiocyanate yielded [FeLA(NCS)2] (1) where
the tetradentate ligand is the symmetrical bis-schiff base
including two N(pyridine) and two N(imine) donors. This
new ferrous material, although based on a tetradentate
ligand closely related to bptn [14b] (bptn results from
the 2:1 condensation of pyridine carboxaldehyde with
1,3-diaminopropane), is significantly different from
[Fe(bptn)(NCS)2]: while the NCS ligands are cis in the
latter, they are trans in compound 1, and while [Fe-
(bptn)(NCS)2] is a SC material, 1 is LS in the 80–400 K
temperature range. The differences between
1 and
[Fe(bptn)(NCS)2] are the substituents (Me, H, respectively)
on the central carbon atom of the alkyl chain, the substit-
uents (Me, H, respectively) on the imine carbon atoms, and
the trans versus cis arrangement of the NCS ligands. The
possible respective roles of these three factors in the differ-
ence of electronic behavior between the two materials can-
not be untangled yet: it is one of the aims of [32], this
report, and a paper in preparation, [40], to independently
assess the role of each of these factors.
Condensation of 2R-imidazole-4-carboxaldehyde with
the same diamine in the 2:1 ratio followed by reaction with
ferrous thiocyanate was expected to yield ferrous com-
plexes with symmetrical tetradentate ligands similar to that
in 1, but better suited to stabilize HS FeII or to yield FeII
SC materials, depending on the decrease in ligand field
on going from LA to LBx, and on the cis or trans arrange-
ment of the thiocyanates ligands. Indeed, compounds 2–4
include the expected symmetrical tetradentate ligands LBx
and two trans-coordinated thiocyanates. As shown by the
results described in the previous section, the ligand field
generated by the LBx ligands associated with two trans-
NCS stabilizes the HS FeII state, regardless of the imidaz-
ole substituent R2 (H, Me, Ph). This is in line with the
significant difference between LA and LBx that results from
the replacement of two pyridyl (strong ligand field N
donor) by two imidazolyl rings (weaker ligand field N
donor).
4. Discussion
Tetradentate ligands including exclusively nitrogen
donors have been scarcely used in association with coor-
dinating pseudohalide anions to prepare FeII SC com-
plexes with a N6 ligand environment. The first examples
were reported by Toftlund et al. [14a]. and included tripo-
dal ligands of the tris-pyridylamine (tpa) type associated
with two cis-coordinated pseudohalide anions: varying
the tpa substituents or the pseudohalide yielded different
SC behaviors or mixtures of spin isomers. A related type
of FeII SC material, also including a tripodal N4 ligand
(dppa) and two cis-coordinated NCSꢀ, was reported later
by Matouzenko et al. [14c]. Finally, Toftlund et al. [14b]
also reported a series of FeII compounds including acyclic
Schiff bases resulting from condensation of R-pyridine-
carboxaldehydes (R = H, Me) or acetylpyridine with
ethylenediamine or propylenediamine (bpn ligand type)
associated with two pseudohalide anions: again, depend-
ing on the ligand substituents and alkyl bridge, they
observed different magnetic behaviors. While the cis-
coordination mode of NCS anions was imposed by the
tpa or the dppa topology, no such constraint was imposed
by the bpn ligands; however the NCS anions were also
cis-coordinated in the bpn series of FeII compounds
[14b]. The diversity of reported SC behaviors in contrast
to the scarcity of SC materials built from tetradentate
ligands including exclusively nitrogen donors associated
with coordinating pseudohalide anions led us to explore
the use of such ligand environments with the aim to tune
the SC behavior through minute changes in the ligand.
An additional incentive to this study was the possibility
to benefit from the versatile arrangement of the two
pseudohalide ligands: indeed, while the pseudohalide
anions are cis-coordinated in many SC materials [1,38],
a recent work has evidenced that the trans-NCS isomer
of a FeII compound including a bidentate triaryltriazole
ligand exhibits SC while its cis-NCS counterpart is a HS
material [15b].
Considering that the NCSꢀ ligands are trans-coordi-
nated in compounds 1–4, regardless of the tetradentate
ligand (LA or LBx), it was anticipated that combination
of pyridyl and imidazolyl rings to yield a dissymmetrical
LCx bis-Schiff base chelated to FeII would not affect the
trans-coordination mode of the NCSꢀ anions. Then, the
only difference in ligand field would result from the change
in the tetradentate ligand. Indeed, the results described in
the previous section confirm that the NCSꢀ ligands are
trans-coordinated in compounds 5–10, allowing us to con-
sider the differences between the tetradentate ligands LA,
LBx and LCx as the sole origin of the different electronic
behaviors among all materials reported in this study.
Then, comparison of the electronic behavior of 1–4 on
the one hand to that of 5–10 on the other hand evidences
that combination of pyridyl and imidazolyl rings to yield
a dissymmetrical tetradentate LCx ligand has indeed

1756
N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
allowed to bring the ligand field from the high (1) and low-
field (2–4) ranges to the intermediate SC ligand field range
(5–10).
033; or e-mail: deposit@ccdc.cam.ac.uk. Supplementary
data associated with this article can be found, in the online
version, at doi:10.1016/j.poly.2006.12.017.
Within the series, the differences originate from: (i) the
substituents on the central carbon atom of the alkyl chain
(R1 = H (5–7), Me (8–10)), and (ii) the 2-substituent on the
imidazolyl ring (R2 = H (5, 8), Me (6, 9), Ph (7, 10)). The
most important difference in the SC properties within this
series is the SC temperature T1/2: it is higher than 400 K
for 5 and 8, while it is around 100 K for the remaining com-
pounds. In other words, substitution of the phenyl or
methyl group by an hydrogen atom at the C2 imidazolyl
shifts the transition temperature T1/2 by ꢂ300 K towards
higher temperatures.
References
[1] P. Gutlich, H.A. Goodwin, Topics Curr. Chem. 233 (2004) 1–47, and
¨
references cited therein.
[2] O. Kahn, J.-P. Launay, Chemtronics 3 (1988) 140.
[3] (a) O. Kahn, J. Krober, C. Jay, Adv. Mater. 4 (1992) 718;
(b) O. Kahn, C. Jay-Martinez, Science 279 (1998) 44.
[4] P. Gutlich, H.A. Goodwin, Topics Curr. Chem. 233–235 (2004).
¨
[5] (a) V. Petrouleas, J.-P. Tuchagues, Chem. Phys. Lett. 137 (1987) 21;
(b) D. Boinnard, A. Bousseksou, A. Dworkin, J.-M. Savariault, F.
Varret, J.-P. Tuchagues, Inorg. Chem. 33 (1994) 271;
´
(c) G. Psomas, N. Brefuel, F. Dahan, J.-P. Tuchagues, Inorg. Chem.
This result evidences that a prominent electronic effect
is exerted by R2 but not R1: in other words, the tuning of
the ligand field is essentially obtained through modifica-
tions of the heterocyclic aromatic rings of the tetradentate
ligand. In this report we concentrate on the imidazolyl
ring; the effect of pyridyl ring substituents has been
described in a previous report [32]. Considering that the
ligand fields exerted by LC1 (5) and LC4 (8) (R2 = H)
are weaker (higher T1/2) than those exerted by the remain-
ing LCx ligands (R2 = Me or Ph), it may be suggested that
the steric effect of the R2 substituent overpasses its induc-
tive (+I, Me) or mesomeric (+M, Ph) effect. As estimated
through roughly approximated T1/2 values, the role of R1
in tuning the ligand field is much more limited. However,
although subtle, the trend is the same within the three
pairs (5, 8), (6, 9) and (7, 10): the lower T1/2 is associated
with R1 = H (5–7). This seems to indicate that the induc-
tive effect of R1 (Me for 8–10) slightly lowers the ligand
field with regard to its counterpart (5–7) in each pair. This
trend is further substanciated by the results described in
[32].
43 (2004) 4590;
(d) L. Salmon, A. Bousseksou, B. Donnadieu, J.-P. Tuchagues, Inorg.
Chem. 44 (2005) 1763.
[6] S.M. Nelson, P.D.A. McLlroy, C.S. Stevenson, E. Ko¨nig, G. Ritter,
J. Waigel, J. Chem. Soc., Dalton Trans. (1986) 991.
[7] W.S. Kelly, G.H. Ford, S.M. Nelson, J. Chem. Soc. A (1971) 388.
[8] M. Bacci, C.A. Ghilardi, A. Orlandini, Inorg. Chem. 23 (1984) 388.
[9] (a) W. Levason, C.A. McAuliffe, M.M. Khan, S.M. Nelson, J. Chem.
Soc., Dalton. Trans. (1975) 1778;
(b) F. Cecconi, M. Di Varia, S. Midollini, A. Orlandini, L. Sacconi,
Inorg. Chem. 20 (1981) 3423;
(c) E. Ko¨nig, G. Ritter, S.K. Kulshreshtha, J. Waigel, L. Sacconi,
Inorg. Chem. 23 (1984) 1241;
(d) C.-C. Wu, J. Jung, P.K. Gantzel, P. Gutlich, D.N. Hendrickson,
¨
Inorg. Chem. 36 (1997) 5339.
[10] (a) L.J. Wilson, D. Georges, M.A. Hoselton, Inorg. Chem. 14 (1975)
2968;
(b) H. Toftlund, S. Yde-Andersen, Acta. Chem. Scand. A35 (1981) 575;
(c) L.L. Martin, K.S. Hagen, A. Hauser, R.L. Martin, A.M. Sargeson,
J. Chem. Soc., Chem. Commun. (1988) 1313;
(d) A.H.R. Obaidi, J.J. McGarvey, K.P. Taylor, S.E.J. Bell, K.B.
Jensen, H. Toftlund, J. Chem. Soc., Chem. Commun. (1993) 536;
(e) A.H.R. Al-Obaidi, K.B. Jensen, J.J. McGarvey, H. Toftlund, B.
Jensen, S.E.J. Bell, J.G. Carroll, Inorg. Chem. 35 (1996) 5055;
(f) Y. Sunatsuki, Y. Ikuta, N. Matsumoto, H. Ohta, M. Kojima, S.
Iijima, S. Hayami, Y. Maeda, S. Kaizaki, F. Dahan, J.-P. Tuchagues,
Angew. Chem., Int. Ed. 42 (2003) 1614;
(g) Y. Ikuta, M. Ooidemizu, Y. Yamahata, M. Yamada, S. Osa, N.
Matsumoto, S. Iijima, Y. Sunatsuki, M. Kojima, F. Dahan, J.-P.
Tuchagues, Inorg. Chem. 42 (2003) 7001;
(h) M. Yamada, M. Ooidemizu, Y. Ikuta, S. Osa, N. Matsumoto, S.
Iijima, M. Kojima, F. Dahan, J.-P. Tuchagues, Inorg. Chem. 42 (2003)
8406;
For clarity, only intramolecular effects have been taken
into account in the above considerations: this is of course a
first approximation because intermolecular effects may also
indirectly but significantly contribute to the ligand field
onto the FeII center [39].
Acknowledgements
(i) Y. Sunatsuki, H. Ohta, M. Kojima, Y. Ikuta, Y. Goto, N.
Matsumoto, S. Iijima, H. Akashi, S. Kaizaki, F. Dahan, J.-P.
Tuchagues, Inorg. Chem. 43 (2004) 4154;
(j) S. Arata, H. Torigoe, T. Iihoshi, N. Matsumoto, F. Dahan, J.-P.
Tuchagues, Inorg. Chem. 44 (2005) 9288.
Financial support by the French Ministry for Edu-
cation, Research, and Technology, through a doctoral
grant to N.B., and by the CNRS-DFG collaborative pro-
gramme, through a postdoctoral grant to S.S. are gratefully
acknowledged.
[11] (a) R.N. Sylva, H.A. Goodwin, Aust. J. Chem. 20 (1967) 479;
(b) R.N. Sylva, H.A. Goodwin, Aust. J. Chem. 21 (1968) 1081;
(c) E. Ko¨nig, G. Ritter, H.A. Goodwin, Chem. Phys. 1 (1973) 17;
(d) H.A. Goodwin, R.N. Sylva, Aust. J. Chem. 21 (1968) 2881;
(e) H.A. Goodwin, D.W. Mather, Aust. J. Chem. 25 (1972) 715;
(f) K.H. Sugiyarto, D.C. Craig, A.D. Rae, H.A. Goodwin, Aust.
J. Chem. 46 (1993) 1269;
Appendix A. Supplementary material
CCDC 613879, 613880, 613881, 613882 and 613883 con-
tain the supplementary crystallographic data for 1, 2, 3, 7
and 9. These data can be obtained free of charge via
htpp://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-
(g) B.J. Sugiyarto, D.C. Craig, A.D. Rae, H.A. Goodwin, Aust.
J. Chem. 47 (1994) 869;
(h) K.H. Childs, D.C. Craig, K.A. Ross, M.L. Scudder,
H.A. Goodwin, Aust. J. Chem. 47 (1994) 891;
(i) G. Lemercier, A. Bousseksou, M. Verelst, J.-P. Tuchagues,
F. Varret, J. Magn. Magn. Mater. 150 (1995) 227;

N. Bre´fuel et al. / Polyhedron 26 (2007) 1745–1757
1757
(j) A. Bousseksou, M. Verelst, H. Constant-Machado, G. Lemercier,
J.-P. Tuchagues, F. Varret, Inorg. Chem. 35 (1996) 110;
(k) G. Lemercier, M. Verelst, A. Bousseksou, F. Varret, J.-P.
Tuchagues, Magnetism: A supramolecular function, in: O. Kahn
(Ed.), NATO ASI Series C, Vol. 484, Kluwer Academic Publishers,
Dordrecht, 1996, p. 335;
[20] K. Lagarec, Recoil, Mo¨ssbauer Analysis Software for windows.
<http://www.physics.uottawa.ca/~recoil>.
[21] STOE, IPDS Manual. Version 2.93. Stoe
Germany, 1997.
& Cie, Darmstadt,
[22] CrysAlis RED, Version 1.170.32, Oxford Diffraction Ltd., Oxford,
UK, 2003.
(l) A. Thiel, A. Bousseksou, M. Verelst, F. Varret, J.-P. Tuchagues,
Chem. Phys. Lett. 302 (1999) 549.
[12] (a) G.A. Renovitch, W.A. Baker, J. Am. Chem. Soc. 89 (1967) 6377;
[23] G.M. Sheldrick, SHELXS-97. Program for Crystal Structure Solution.
University of Go¨ttingen, Go¨ttingen, Germany, 1990.
[24] G.M. Sheldrick, SHELXL-97. Program for the Refinement of Crystal
Structures from Diffraction Data. University of Go¨ttingen, Go¨ttin-
gen, Germany, 1997.
(b) M. Sorai, J. Ensling, K.M. Hasselbach, P. Gutlich, Chem. Phys.
¨
20 (1977) 197;
(c) A.M. Greenaway, E. Sinn, J. Am. Chem. Soc. 100 (1978) 8080;
(d) A.M. Greenaway, C.J. O’Connor, A. Schrock, E. Sinn, Inorg.
Chem. 18 (1979) 2692;
(e) B.A. Katz, C.E. Strouse, J. Am. Chem. Soc. 101 (1979) 6214;
(f) M. Mikami, M. Konno, Y. Saito, Chem. Phys. Lett. 63 (1979) 566;
(g) H.A. Goodwin, R.N. Sylva, Aust. J. Chem. 21 (1968) 83;
(h) R.J. Dosser, W.J. Eilbeck, A.E. Underhill, J. Chem. Soc. A (1969)
810;
[25] International Tables for Crystallography, Vol. C, Kluwer Academic
Publishers, Dordrecht, The Netherlands, 1992.
[26] L. Zsolnai, H. Pritzkow, G. Huttner, ZORTEP. Ortep for PC, Program
for Molecular Graphics, University of Heidelberg, Heidelberg,
Germany, 1996.
[27] Stoe, X-SHAPE. Crystal Optimisation for Numerical Absorption
Corrections, Revision 1.01. Stoe & Cie, Darmstadt, Germany, 1996.
[28] R.H. Blessing, Acta Crystallogr. A 51 (1995) 33.
[29] J.P. Costes, M.I. Fernadez-Garcia, Inorg. Chim. Acta 237 (1995) 57,
and references cited therein.
(i) D.M.L. Goodgame, A.A.S.C. Machado, Inorg. Chem. 8 (1969)
2031;
(j) Y. Sasaki, T. Shigematsu, Bull. Chem. Soc. Jpn. 46 (1973) 3438;
(k) Y. Sasaki, T. Shigematsu, Bull. Chem. Soc. Jpn. 47 (1974) 109;
(l) D.M.L. Goodgame, Bull. Soc. Chim. Fr. (1972) 3;
(m) J.R. Sams, J.C. Scott, T.B. Tsin, Chem. Phys. Lett. 18 (1973) 451;
(n) J.R. Sams, T.B. Tsin, J. Chem. Soc., Dalton Trans. (1976) 488;
(o) J.R. Sams, T.B. Tsin, Inorg. Chem. 15 (1976) 1544;
(p) A.T. Baker, H.A. Goodwin, Aust. J. Chem. 30 (1977) 771.
[30] (a) D.E. Fenton, R. Moody, U. Casellato, P.A. Vigato, R.J. Graziani,
Heterocycl. Chem. 22 (1985) 1639;
(b) J.P. Costes, F. Dahan, A. Dupuis, J.P. Laurent, New J. Chem. 21
(1997) 1211.
[31] (a) J.P. Costes, F. Dahan, A. Dupuis, J.P. Laurent, Inorg. Chem. 39
(2000) 169;
(b) J.P. Costes, F. Dahan, A. Dupuis, Inorg. Chem. 39 (2000) 5994;
(c) A. Bencini, J.P. Costes, F. Dahan, A. Dupuis, J. Garcia-Tojal, D.
Gatteschi, F. Totti, Inorg. Chim. Acta 357 (2004) 2150.
[13] (a) E.W. Muller, J. Ensling, H. Spiering, P. Gutlich, Inorg. Chem. 22
¨
(1983) 2074;
´
(b) O. Roubeau, A.F. Stassen, I. Ferrero Gramage, E. Codjovi, J.
Linares, F. Varret, J.G. Haasnoot, J. Reedjik, Polyhedron 20 (2001)
1709.
[32] N. Brefuel, S. Shova, J. Lipkowski, J.-P. Tuchagues, Chem. Mater. 18
(2006) 5467.
[33] (a) G.R. Desiraju, V. Nalini, J. Mater. Chem. 1 (1991) 201;
(b) J.-A. van Derberg, K.R. Seddon, Crystal Growth Design 3 (2003)
643;
[14] (a) F. Hojland, H. Toftlund, S. Yde-Andersen, Acta Chem. Scand.
A37 (1983) 251;
(b) H. Toftlund, E. Pedersen, S. Yde-Andersen, Acta Chem. Scand.
A38 (1984) 693;
(c) G.S. Matouzenko, A. Bousseksou, S. Lecocq, P.J. van Konings-
bruggen, M. Perrin, O. Kahn, A. Collet, Inorg. Chem. 36 (1997) 5869.
(c) G.R. Desiraju, Chem. Commun. (2005) 2995.
[34] F.E. Mabbs, D.J. Machin, Magnetic and Transition Metal Com-
plexes, Chapman and Hall, London, UK, 1973.
[35] N.N. Greenwood, T.C. Gibbs, Mo¨ssbauer Spectroscopy, Chapman
and Hall, New York, 1971, pp. 9–11, 50–53.
´
[15] (a) J.-F. Letard, P. Guionneau, E. Codjovi, D. Lavastre, G. Bravic,
D. Chasseau, O. Kahn, J. Am. Chem. Soc. 119 (1997) 10861;
(b) D. Zhu, Y. Xu, Z. Yu, Z. Guo, H. Sang, T. Liu, X. You, Chem.
Mater. 14 (2002) 838, and references cited therein.
[16] (a) A.L. Nivororozhkin, H. Toftlund, M. Nielsen, J. Chem. Soc.,
Dalton. Trans. (1994) 361;
[36] A.K. Boudalis, J.-M. Clemente-Juan, F. Dahan, J.-P. Tuchagues,
Inorg. Chem. 43 (2004) 1574, and references cited therein.
[37] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, fourh ed., John Wiley & Sons, New
York, 1986.
(b) G. Brewer, J. Jasinski, W. Mahany, L. May, S. Prytkov, Inorg.
Chim. Acta 232 (1995) 183.
[38] (a) J.A. Real, J. Zarembovitch, O. Kahn, X. Solans, Inorg. Chem. 26
(1987) 2939;
[17] M.A. Hoselton, L.J. Wilson, R.S. Drago, J. Am. Chem. Soc. 97
(1975) 1722.
(b) H. Toftlund, Coord. Chem. Rev. 94 (1989) 67.
´
[39] J.-P. Tuchagues, A. Bousseksou, G. Molnar, J.J. McGarvey, F.
´
[18] N. Brefuel, C. Lepetit, S. Shova, F. Dahan, J.-P. Tuchagues, Inorg.
Varret, in: P. Gutlich, H.A. Goodwin (Eds.), Topics Curr. Chem. 235
¨
Chem. 44 (2005) 8916.
(2004), pp. 85–103, and references cited therein.
´
[40] N. Brefuel, S. Shova, C. Duhayon, J.-P. Tuchagues, in preparation.
[19] N.E. Erikson, N. Sutin, Inorg. Chem. 5 (1966) 1834.