Dinuclear lanthanide Schiff-base complexes forming a rectangular columnar mesophase

Article abstract of DOI:10.1002/ejic.200500702

The Schiff base LH₃ obtained by condensation of 3-formyl-4-hydroxyphenyl-3,4,5-tris(tetradecyloxy)benzoate with 1,3-diamino-2-propanol reacts with lanthanide(III) acetate salts to form dinuclear complexes of the type Ln₂L₂

Full text of DOI:10.1002/ejic.200500702

FULL PAPER
DOI: 10.1002/ejic.200500702
Dinuclear Lanthanide Schiff-Base Complexes Forming a Rectangular
Columnar Mesophase
Koen Binnemans,*[a] Katleen Lodewyckx,[a] Thomas Cardinaels,[a] Tatjana N. Parac-Vogt,[a]
Cyril Bourgogne,[b] Daniel Guillon,[b] and Bertrand Donnio*[b]
Keywords: Columnar mesophases / Lanthanides / Liquid crystals / Metallomesogens / Rare earths
The Schiff base LH3 obtained by condensation of 3-formyl-
Nd, Sm, Gd). These stable and neutral metallomesogens ex-
4-hydroxyphenyl-3,4,5-tris(tetradecyloxy)benzoate with 1,3- hibit a rectangular columnar mesophase over a broad tem-
diamino-2-propanol reacts with lanthanide(III) acetate salts
to form dinuclear complexes of the type Ln2L2 with the tri-
valent ions from the first half of the lanthanide series (Ln =
perature range.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
lumnar phase over a broad temperature range. Depending
on the radius of the lanthanide() ion, either dinuclear f-
d[7] or trinuclear f-d-f complexes were formed.[7,8] Note that
all these complexes can be considered as adducts of a me-
sogenic Cu(salen) or Ni(salen) complex of a trivalent lan-
thanide ion. Because the Cu(salen) and Ni(salen) complexes
are electrically neutral, three nitrate groups are necessary to
compensate for the positive charge of the lanthanide ion in
the resulting mixed f-d complexes. Additionally, if such a
doubly negatively charged ligand is to be used to prepare
purely lanthanide complexes, a further small inorganic
anion (e.g. X = Cl, NO3) has to be present to obtain electric
neutrality in the corresponding LnXL complexes, otherwise
neutral complexes with a different stoichiometry such as
Ln2L3 type may be formed instead. By modifying the salen-
type Schiff-base ligand, it should be possible to obtain an
organic species that contains three deprotonable OH groups
and which can be used as triply negatively charged anionic
ligand to form neutral complexes with the trivalent lantha-
nide cations. Such a ligand can be obtained by condensing
the aldehyde that is used to prepare the salen-type Schiff
base, with 1,3-diamino-2-propanol instead of 1,2-diaminoe-
thane.
In this paper, we show that the chemical modification of
the salen-type Schiff base by using 1,3-diamino-2-propanol
instead of 1,3-diaminopropane in the reaction with the ap-
propriate 3-formyl-4-hydroxyphenyl-3,4,5-tris(tetradecyl-
oxy)benzoate appears to be a judicious approach to prepare
neutral dinuclear complexes of the type Ln2L2 with tri-
valent lanthanide ions. Moreover, these complexes are
mesomorphic and exhibit a rectangular columnar meso-
phase over a large temperature range, as shown by small-
angle X-ray diffraction. With the help of molecular me-
chanics, a model for the organization of these compounds
in the columns is proposed.
Introduction
N,NЈ-Bis(salicylidene)ethylenediamine (salen) and similar
Schiff-base ligands derived from diamines other than ethyl-
enediamine have often been used as building blocks for me-
tal-containing liquid crystals.[1] This work was focused
mainly on copper(), nickel() or oxovanadium() (vana-
dyl) complexes that exhibit smectic mesophases,[2,3] essen-
tially smectic A and smectic C phases, but also the more
ordered smectic E phase.[2e] Similar mesomorphic proper-
ties were also observed in calamitic cobalt() complexes,[2g]
and the enhancement of the molecular anisotropy of such
NiII, CuII and VOIV salen complexes by the use of alkyl-
phenylazo groups results in new mesomorphic systems
showing a high temperature SmC phase, and in some cases,
the nematic phase.[4] Binnemans et al. reported the first
mesomorphic lanthanide complexes of salen-type Schiff
base, but owing to the elevated transition temperatures, un-
equivocal identification of the mesophase formed was not
possible.[5] Swager and co-workers studied vanadyl com-
plexes derived from such Schiff bases, possessing a hexaca-
tenar structure, that exhibit a rich columnar mesomor-
phism,[6] whereas Binnemans et al. synthesized mixed f-d
metallomesogens, derived form the related hexacatenar li-
gand, that contain both a transition metal ion and a lantha-
nide() ion.[7] These compounds showed a hexagonal co-
[a] Katholieke Universiteit Leuven, Department of Chemistry,
Celestijnenlaan 200F, 3001 Leuven, Belgium
E-mail: Koen.Binnemans@chem.kuleuven.be
[b] Institut de Physique et Chimie des Matériaux de Strasbourg –
Groupe Matériaux Organiques, UMR 7504 CNRS-Université
Louis Pasteur,
B. P. 43, 23 rue du Loess, 67034 Strasbourg Cedex 2, France
E-mail: bdonnio@ipcms.u-strasbg.fr
150
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 150–157

Dinuclear Lanthanide Schiff-Base Complexes
FULL PAPER
the Schiff base (0.5 equiv.) in chloroform at refluxing tem-
perature. The synthesis of ligand and complexes 1–3 is out-
Results and Discussion
The Schiff-base ligand LH3 was obtained by condensa- lined in Scheme 1.
tion of the aldehyde 3-formyl-4-hydroxyphenyl-3,4,5-tris(te-
A schematic representation of the dinuclear lantha-
tradecyloxy)benzoate (1 equiv.) and 1,3-diamino-2-propa- nide() complexes is given in Scheme 1 (compounds 1, 2,
nol (0.5 equiv.) in toluene, with azeotropic removal of the 3). Because each ligand carries a triple negative charge, the
water. The three dinuclear lanthanide() complexes (Ln = positive charge of the two trivalent lanthanide ions is coun-
Nd, Sm, Gd) were synthesized by reaction between the lan- terbalanced and the dinuclear complexes are neutral; there-
thanide acetate salt Ln(CH3COO)3·xH2O (1.1 equiv.) and fore no counterions are necessary. This is in contrast with
Scheme 1. Synthesis of the Schiff-base ligand and the corresponding dinuclear Ln–Ln complexes 1–3. Reagents and conditions: (a) RBr
(3 equiv.), K2CO3 (6 equiv.), DMF, KI (catalytic amount), reflux overnight; (b) NaOH, ethanol, reflux for 4 hours; acidification by dilute
HCl solution; (c) DCC, DMAP, dichloromethane, 24 hours at room temp.; (d) 1,3-diamino-2-propanol (0.5 equiv.), glacial acetic acid
(catalyst), toluene, Dean–Stark trap, 3 hours at reflux; (e) Ln(CH3COO)3·xH2O (1.1 equiv.), chloroform, reflux overnight.
Eur. J. Inorg. Chem. 2006, 150–157
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
151

K. Binnemans, B. Donnio et al.
FULL PAPER
the f-d complexes with salen-type ligands,[7,8] the lanthanide ronment for the manganese() ions. The authors showed
complexes with salicylaldimine Schiff bases which have that the ligands adopt a hairpin conformation and that two
counterions such as nitrate groups.[1f,9,10] or lipophilic of them wrap the two metallic ions in an antiparallel mode
bis(benzimidazole)pyridine-based ligands.[11] The reasons to form the neutral mesocate Mn2L2 complex as schemati-
why a dinuclear complex is formed instead of a mononu- cally drawn in Figure 1, b. This particular dimeric structure
clear species are due to the tendency of the lanthanide() could be representative of the chelating mode of this type
ions to achieve a high coordination number and to the na- of ligand with two metallic nuclei and generalized to other
ture of the ligand itself, auspicious to such a coordination dinuclear species. We have thus extrapolated this mode of
mode (fused chelating parts). Typically, the coordination chelation to the presently studied dinuclear lanthanide()
number of the trivalent lanthanide ions in coordination complexes, the close proximity of the two sets of donor
compounds is eight or nine. Also six-coordinate lantha- atoms of the bidentate bridging ligand, and the presence of
nide() complexes are known, for example the tris(β-dike- the central hydroxy coordinating group slightly disfavoring
tonato)lanthanide() complexes. When the ligand does not the formation of the helicate structure at the expense to
offer a sufficient number of donor atoms to the lantha- the mesocate one; this assumption is comforted by some
nide() ion to achieve a high coordination number, some molecular calculations, too (vide infra).
of the donor atoms can coordinate in a bridging mode. Al-
ternatively, adducts can be formed with Lewis bases, such
as water molecules. Unfortunately, we were unable to obtain
single crystals, so that the structure of the complex is still
unknown. It was also not possible to grow single crystals
of similar complexes with shorter alkyl chains and due to
1
the presence of paramagnetic ions, H NMR spectra of the
complexes were not resolved. It is however unlikely to con-
Figure 1. Schematic representation of the two possible structures
sider the structure of the dinuclear lanthanide() complex
as totally flattened and planar (as drawn in Scheme 1) by
simply considering the coordination chemistry of d and f
elements and entropic phenomena. The geometric prefer-
ences of lanthanides to adopt coordination numbers higher
than six support that such extended planarity is absent, and
given by the Ln2L2 complexes. The two bischelate bent-like ligands,
may entwine the two metallic species to form a double-stranded
helicate (a, D2 symmetry) or sandwich the metal ions to form a
non-helical meso structure or mesocate instead (b, C2h symmetry).
In this model, the two lanthanide() ions are in a six-
that the ligands are wrapped around the metallic centers. coordinate environment, with the coordination polyhedron
Thus, a more energetically favorable structure for these di- being a (distorted) octahedron. The two coordinating octa-
nuclear species would imply the ligand to be bent or V- hedral cages of the two lanthanide ions share a common
shaped (due to the flexible spacer), and that they are either edge, and the alkoxide group bridges the two lanthanide()
entwined around the two metal ions as in helicates[12–14] centers (Figure 2). For a further investigation of the confor-
(Figure 1a, D2 symmetry) or coordinated to the two cations mation of this dinuclear complex and to support the mesoc-
in such a way that a sandwich complex with a mesocate ate structure, molecular simulations were performed on
structure is formed (part b in Figure 1, C2h symmetry).[13] both mesocate and helicate conformers (Gaussian 03,
Whatever the exact structure is, it has been shown recently rb3lyp/3-21g**). For compatibility with the calculation
for some M2L3 complexes, that both species exist in equilib- bases, the lanthanide() cations were replaced in these
rium in solution, and the helicate to mesocate isomerization models by yttrium(), chosen because of its great proximity
only requires inversion at one metal center,[15] and that this to the lanthanide series (located just above the lanthanum
low-energy cost equilibrium may be generalized to other in the periodic table of elements) and for the similarity of
systems. The driving forces for the formation of helicates the ionic radius between YIII and LnIII. The minimization
and/or mesocates result from the minimization of energy of yielded two stable structures with an energetic gap of
the ultimate self-assembly. Such a formation depends on the 44.9 kJ·mol–1 in favor of the mesocate conformation. For
metal-imposed (number and coordination geometry) and li- both conformers in these theoretical structures (taken as
gand-imposed constraints (type, nature of the spacer and isolated molecule in the gas phase), the angle formed by the
distance between bridging units), as well as to some degree phenyl groups of the ligands is quite wide in comparison
of pre-organization of the ligand. In the case of lanthanides, with the manganese complex just mentioned,[16] but a se-
helicates mainly possess a three-stranded helicate Ln2L3 arch in the Cambridge Crystallographic Database shows
structure (in such case, the two lanthanide ions are wrapped that the very flexible salen-type ligand can be very distorted
by three ligands), while mesocate complexes of the type by packing constraints in condensed phase. It is then rea-
M2L2 – schematically shown in Figure 1, b – are common sonable to propose that the present dinuclear lanthanide()
for both d and f elements.[13] For instance, Gelasco and Pe- complexes adopt preferentially a mesocate form, with the
coraro described the crystal structure of a closely related ligands in the hairpin conformation. Furthermore, with this
dimeric manganese() complex of the Schiff base 1,3-bis(5- assumption, the pinched ligand structure could be even
nitrosalicylidenimino)-2-propanol:[16] this dimeric manga- more stabilized by intramolecular π-stacking of the benzo-
nese complex has a µ2-alkoxy core with a symmetric envi- ate rings (vide infra).
152
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 150–157

Dinuclear Lanthanide Schiff-Base Complexes
FULL PAPER
Figure 2. Minimized structure of the dinuclear diligand species (see text; purple: lanthanide cation; blue: nitrogen atom; red: oxygen
atom, green: carbon atom. Hydrogen atoms have been omitted for clarity.).
The Schiff-base ligand LH3 is itself mesomorphic, not
totally unexpected, because of its polycatenar-like molecu-
lar structure, a priori compatible with mesophase forma-
tion.[17] Observation by polarizing optical microscopy
(POM) indicated a monotropic liquid-crystalline phase be-
havior. On heating the sample, an isotropic liquid is ob-
tained at 82 °C, which was confirmed by an intense endo-
thermic peak transition (82 kJ·mol–1) in the first heating
run of the DSC curve. Heating was continued above 100 °C
without detecting additional thermal events, and the sample
was then cooled down. At 62 °C, a phase transformation is
detected by the growth of birefringent dendritic features
from the isotropic liquid, remembering one of the typical
texture of the hexagonal columnar phase (Figure 3). A peak
Figure 3. Optical texture of hexagonal columnar phase formed by
of –2.50 kJ·mol–1 in the DSC curve corresponds to this
the Schiff-base ligand LH3 at 58 °C.
monotropic phase transition. The phase remained stable
down to room temperature. On second heating, the com-
pound cleared straight into the isotropic liquid at 62 °C, rized in Table 1. At room temperature, the complexes are in
that is 20 °C lower than during the first heating. Subsequent a semi-crystalline or condis (conformationally disordered)
cooling and heating repeated the behavior of the second phase as revealed by XRD: only broad and diffuse scat-
cycle so crystallization from a solvent must result in a crys- terings were observed in the angular range of study (0 Ͻ 2θ
talline form that needs to rearrange on heating before it Ͻ 35°). The halo in the wide angle part at 4.5 Å corre-
clears into the isotropic liquid. The X-ray diffractogram of sponds to the liquid-like order of the molten chains,
the mesophase registered on cooling the sample from the whereas those in the small angle part, obtained within the
isotropic liquid does not contradict the mesophase assign- ratio 1:2:3 (at ca. 30.5, 15.3, and 10.2 Å) indicate some
ment made by POM, nor gives conclusive information short-range layering correlations of the system. This lam-
about the mesophase symmetry, and so the mesophase re- ellar structure is formed again on subsequent cooling/heat-
mains assigned from its optical texture alone. The pattern ing runs, proving the high reversibility of the phase trans-
recorded at 50 °C on cooling showed only one single, but formation and the good thermodynamic stability of this
sharp and intense small-angle diffraction peak at ca. 38 Å, phase. Above 100 °C, and for the three dinuclear complexes,
which was indexed as the [10] reflection of a 2D hexagonal a change in the optical texture concomitant to a phase
lattice. The liquid-like order of the aliphatic chains was con- transformation was detected. However, the optical textures
firmed by the presence of the broad scattering at ca. 4.5 Å. observed by microscopy were not sufficiently characteristic
To conclude, the optical texture shows enough evidence that to permit the unambiguous identification of the mesophase.
a hexagonal columnar phase was formed (Figure 3), and With regards to the molecular structure of the complexes,
the thermal properties of the ligand can be summarized as which resembles to that of related dinuclear complexes con-
follows: Cr 82 (Colh 62) I.
structed from hexacatenar ligands, the mesophase is likely
The lanthanide() complexes 1–3 are also mesomorphic, columnar.[18] The definitive mesophase identification was
and exhibit a mesophase over a temperature range of more achieved by temperature-dependent X-ray diffraction. The
than 100 °C. The clearing temperatures of the complexes XRD pattern of the neodymium() complex 1 at 200 °C,
could not be determined due to extensive decomposition chosen as representative example for the three complexes, is
well above 200 °C. The transition temperatures are summa- shown in Figure 4 and the relevant data are summarized in
Eur. J. Inorg. Chem. 2006, 150–157
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
153

K. Binnemans, B. Donnio et al.
FULL PAPER
Table 2. The X-ray pattern recorded at 200 °C was chosen Table 2. XRD data of the neodymium() complex 1 at 200 °C.[a]
instead of those recorded at lower temperatures as it pro-
[c]
[c]
Reflection no.[b]
dexp
[hk]
I[d]
dtheo
Cell parameters[c,e]
vided the most suitable information. Indeed, the reflections
appear to sharpen and to become more intense as the tem-
perature was raised, a result of the temperature-induced
micro-segregation. The pattern reveals two diffuse scat-
tering halos (5 and 6) observed in the mid and wide-angle
region. The halo at 4.5 Å (6), corresponds to the liquid-
like order of the aliphatic chains and to the loose lateral
interactions between rigid parts of the complex, whereas the
other diffuse halo (5), corresponding to an average distance
of ca 10 Å, very likely conveys to the existence of dimeric
species (vide infra). In the small angle region, four large
reflections were observed and indexed into a rectangular
2D lattice with the pairs of Miller’s indices [hk] = [11], [20],
[22] and [40] (Table 2). This pattern is typical for a centered
rectangular lattice with a c2mm plane group. Since two in-
dexations of the reflections are possible (only harmonics of
the fundamental reflections were measured), we have cho-
sen the most likely one (Table 2). The first option consists
of chosing [11] = 26.0 Å and [20] = 38.25 Å, but this yields
to a highly anisometric rectangular lattice, the parameters
of which (a = 76.5 Å and b = 27.45 Å) are highly incompat-
ible with the molecular dimensions (length, width and
thickness) of the dinuclear complexes. In contrast, the other
1
2
3
4
5
6
38.25
26.0
19.1
13.15
10.0
4.5
[11]
[20]
[22]
[40]
–
vs
vs
m
m
broad
broad
38.25
26.0
19.1
13.0
–
Colr – c2mm
a = 52.0 Å
b = 56.45 Å
S = 2.935 Å2
Vcell = 14.675 Å3
Ncol = 2
–
[a] For complexes 2 and 3, similar data were obtained: Sm complex
2: a = 52.2 Å, b = 56.3 Å, Vmol = 7360 Å3, Ncol = 2; Gd complex
3: a = 52.4 Å, b = 56.1 Å, Vmol = 7380 Å3, Ncol = 2. [b] The num-
bering of X-ray reflections corresponds to the one in Figure 4.
[c] dexp and dtheo are the experimentally measured and theoretical
diffraction spacings. Distances are given in Å. a, b are the param-
eters of the Colr phase, S is the lattice area and Vcell the volume of
the rectangular cellule 10 Å thick. [d] Intensity of the reflection. vs:
very strong, m: medium. [e] a, b are deduced from the measured
spacings d11 and d20: a = 2d20; b = A, with
;
dtheor is deduced from the following mathematical expression and
. S = a×b, and Vcell
the lattice parameters a and b:
= ½×S×10. Ncol is the aggregation number or the number of mol-
ecular equivalents per stratum of column 10 Å thick, Ncol = Vcell
Vmol
/
.
option ([11] = 38.25 Å and [20] = 26.0 Å), leads to lattice understand the molecular packing of the complexes in the
parameters that are more relevant here (a = 52.0 Å and b columns and to propose a model, we first compared the
= 56.45 Å) and in agreement with the size of the complexes. molecular dimensions and volume with the columnar cross-
section and rectangular cell volume of the Colr phase, and
Table 1. Transition temperatures of the dinuclear lanthanide com-
followed the methodology explained elsewhere[19] to evalu-
plexes.
ate the number of molecules per a given length of column,
Compound
Ln
Transition temperatures[a] [°C]
h. Of course, h can be chosen arbitrarily, but in the follow-
ing calculation a value of 10 Å has been assumed, corre-
sponding to the broad signal 5 seen in the X-ray pattern
(vide supra), which suggests that there is a periodicity at
this distance and the breadth of the signal that there is a
correlation which persists over several molecules. The mini-
mization of the complex also showed a thickness of about
10 Å, concordant with this assumption. Thus, from the
molecular volumes Vmol of the Ln2L2 complexes (for Ln =
Nd, Vmol = 7340 Å3 at 200 °C), it was found that two com-
plexes molecular equivalents occupy the rectangular cell,
10 Å thick. Similar data were obtained for the correspond-
ing dinuclear gadolinium and samarium complexes (see
Table 2).
1
2
3
Nd
Sm
Gd
X · 116 · Colr-c2mm · 257 (dec.)
X · 108 · Colr-c2mm · 227 (dec.)
X · 106 · Colr-c2mm · 236 (dec.)
[a] X: semi-crystalline lamellar solid or condis (conformationally
disordered) phase; Colr-c2mm: rectangular columnar phase with
the centered c2mm symmetry; dec.: decomposition.
On the basis of the molecular minimization shown above
(Figure 2), it is possible to consider the formation of a di-
meric species which takes an overall “bow-tie” conforma-
tion, with a thickness of approximately 10 Å, and be stabi-
lized by lateral intermolecular interactions between the
phenyl rings of the ligand strands as shown in the Figure 5.
This interaction, which depends on the molecular structure,
is short range and is limited in the plane to a dimeric species
Figure 4. X-ray diffractogram of neodymium() complex 1 at
200 °C.
The similarity of the lattice parameters a and b in this only because of the important number of irradiating ali-
case expresses that the rectangular lattice is in fact nearly phatic chains (‘bow-tie’-like dimer, Figure 5), which results
square-like, that is that the cross-section of the columnar in the strong curvature of the hard-flexible interface and
hard core (i.e. constituted of the rigid molecular parts) is disfavors molecular associations higher than in pairs. The
not so elliptic, but rather more circular instead. In order to formation of columnar stacking is mainly driven by the
154
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 150–157

Dinuclear Lanthanide Schiff-Base Complexes
FULL PAPER
micro-segregation between the hard and flexible parts (Fig- during the simulation. This model led to a consistent result,
ure 6). The hard part of the dimer forms the rigid spine of with a density of the packing very close to unity, similar to
the column, and the voids between these “hard” columns is that of the mesophase.
filled by the homogeneous matrix of aliphatic chains. The
stability of the edifice is mainly ensured by weak van der
Waals interactions and by inter-dimeric interactions along
the columnar axis (Figure 6). These columns then self-orga-
nize into the 2D rectangular lattice (Figure 7). A molecular
dynamics model has been built from two dimers disposed
respectively on the edge and at the center of a rectangular
cell based on the parameters determined by XRD. Due to
the incompatibility of the lanthanide ions with common
forcefields used for organic compounds, and according to
our previous discussion about the shape of these salen-type
dimetallic complexes, the geometry of the central part of
the complex was tethered to a template structure obtained
from the crystallographic data of the dinuclear manganese
complex mentioned above.[15] Nevertheless, each molecule
was allowed to move freely from each other within the cell
Figure 7. Arrangement of the ‘bow-tie’-like dimers within the 2D
rectangular lattice.
Conclusions
In this paper, we report the first examples of neutral,
liquid-crystalline dinuclear f-f complexes, i.e. complexes
that contain two lanthanide ions. The ligand was designed
in such a way that the resulting complexes are electrically
neutral, so that no additional inorganic or organic counteri-
ons are needed to counterbalance the electric charges of
the complex. Due to the synthetic methodology specifically
Figure 5. The ‘bow-tie’-like dimeric species viewed along the co-
lumnar axis (i.e. cross-section of column).
elaborated here, only homodinuclear f-f complexes were
prepared. The complexes exhibit a rectangular columnar
mesophase, and a model, based on X-ray diffraction data
and molecular modelling, is proposed for the molecular
self-organization.
Experimental Section
Techniques: The nuclear magnetic resonance (NMR) spectra were
recorded on a Bruker Avance 300 spectrometer (operating at
300 MHz) and on a Bruker AMX-400 (operating at 400 MHz),
using CDCl3 as solvent and tetramethylsilane (TMS) as internal
standard. The NMR shifts were assigned with the help of 13C
DEPT spectroscopy, two dimensional correlation spectroscopy
(COSY) and proton-carbon heteronuclear multiple-quantum co-
herence (HMQC) spectroscopy. FTIR spectra were recorded on a
Bruker IFS-66 spectrometer, using the KBr pellet method. Elemen-
tal analyses were obtained on a CE-Instrument EA-1110 elemental
analyzer. MALDI-TOF mass spectra were measured on a VG Tof-
spec SE (Micromass, UK) equipped with a N2-laser (337 nm). Op-
tical textures of the mesophases were observed with an Olympus
BX60 polarizing microscope equipped with a LINKAM THMS600
hot stage and a LINKAM TMS93 programmable temperature-con-
troller. DSC traces were recorded with a Mettler- Toledo DSC821e
module. The XRD patterns were obtained with two different exper-
Figure 6. View of the stacking along the columnar axis, with the
stacking periodicity, h, also representing the molecular thickness.
Eur. J. Inorg. Chem. 2006, 150–157
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
155

K. Binnemans, B. Donnio et al.
FULL PAPER
imental set-ups, and in all cases, the powdered sample was filled in
Lindemann capillaries of 1 mm diameter. A linear monochromatic
Cu-Kα1 beam (λ = 1.5405 Å) obtained with a sealed-tube generator
(900 W) and a bent quartz monochromator were used (both gener-
ator and monochromator were manufactured by Inel). One set of
diffraction patterns was registered with a curved counter Inel CPS
120, for which the sample temperature is controlled within
0.05 °C; periodicities up to 60 Å can be measured. The other set
of diffraction patterns was registered on Image Plate. The cell pa-
rameters are calculated from the position of the reflections at the
smallest Bragg angle, which are in all cases the most intense. Peri-
odicities up to 90 Å can be measured, and the sample temperature
is controlled within 0.3 °C. The molecular modeling calculations
were performed on a SGI Origin 2800 20 CPU computer and on
an SGI Octane² calculators using Insight II and Discover 3 soft-
ware from Accelrys (www.accelrys.com) with the cvff forcefield. For
all models, prior to the dynamics, the systems were minimized to a
gradient of 0.05 kcalmol–1. The simulation then consisted of a 200
ps isotherm at 473 K in the NVT-PBC ensemble and with a 1 fs
time step.
= 3920.71 m/e; M (found) = 3921.65 m/e. IR (KBr): ν = 1633
˜
w
(C=N) cm–1.
Gd Complex 3: C230H382Gd2N4O26 (3934.01): C 70.22, H 9.79, N
1.42; found C 69.67, H 9.96, N 1.34. IR (KBr): ν = 1633 (C=N)
˜
cm–1.
Acknowledgments
K. B. and T. N. P.-V. thank the F.W.O.-Flanders (Belgium) for a
postdoctoral fellowship. T. C. is research assistant of the F.W.O.-
Flanders. Financial support by the F.W.O.-Flanders (G.0117.03)
and by the K.U. Leuven (GOA, 03/03) is gratefully acknowledged.
Funding for travel was obtained via a Tournesol project (Project
T2004.10). C. B., D. G. and B. D. thank CNRS and ULP for sup-
port and fundings.
[1] a) Metallomesogens, Synthesis Properties and Applications (Ed.:
J. L. Serrano), VCH, Weinheim (Germany), 1996; b) D. W.
Bruce, in: Inorganic Materials, 2nd ed., chapter 8, p. 429 (Eds.:
D. W. Bruce, D. O’Hare), Wiley, Chichester, 1996; c) K. Binne-
mans, C. Görller-Walrand, Chem. Rev. 2002, 102, 2303; d) B.
Donnio, D. Guillon, R. Deschenaux, D. W. Bruce, in: Compre-
hensive Coordination Chemistry II, J. A. McCleverty, T. J.
Meyer (Eds.), Elsevier, Oxford, 2003, vol. 7 (Eds.: M. Fujita,
A. Powell, C. Creutz), chapter 7.9, pp. 357–627.
Synthesis of the Schiff-Base Ligand LH3: The aldehyde 3-formyl-
4-hydroxyphenyl-3,4,5-tris(tetradecyloxy)benzoate was synthesized
as described elsewhere.[7] To
a solution of 3-formyl-4-hy-
droxyphenyl-3,4,5-tris(tetradecyloxy)benzoate (1.33 g, 0.002 mol)
in 200 mL of toluene was added 1,3-diamino-2-propanol (90 mg,
0.001 mol) and 5 drops of glacial acetic acid (as the catalyst). The
mixture was heated for 3 hours at reflux and water formed by the
reaction was removed azeotropically (Dean–Stark trap). After al-
lowing the solution to cool down to room temperature, the solvent
was removed at reduced pressure. The crude product was purified
by crystallization from absolute ethanol. Yield: 80% (1.11 g). 1H
NMR (CDCl3): δ = 0.89 (m, 18 H, CH3), 1.20–1.40 (m, 120 H,
CH2), 1.49 (m, 12 H, CH2–CH2–CH2–O), 1.70–1.90 (m, 12 H,
CH2–CH2–O), 3.70–3.95 (m, 4 H, N–CH2), 4.05 (m, 12 H, CH2–
O), 4.30 (m, 1 H, CH), 7.01 (d, 2 H, Harom), 7.14 (s, 2 H, Harom),
7.16 (d, 2 H, Harom), 7.39 (s, 4 Harom), 8.40 (s, 2 H, CH=N), 13.0
(br., OH) ppm. 13C NMR (CDCl3): δ = 14.1 (CH3), 22.7 (CH2),
26.1 (CH2), 29.3–30.3 (CH2), 31.9 (CH2), 63.3 (CH2–N), 69.3
(CH2–O), 70.5 (CH), 73.6 (CH2–O), 108.7 (Carom), 117.9 (Carom),
118.5 (Carom), 123.7 (Carom), 123.9 (Carom), 126.2 (Carom), 158.8
[2] a) R. Paschke, H. Zaschke, A. Mädicke, J. R. Chipperfield,
A. B. Blake, P. G. Nelson, G. W. Gray, Mol. Cryst. Liq. Cryst.
Lett. Sect. 1988, 6, 81; b) T. D. Shaffer, K. A. Sheth, Mol.
Cryst. Liq. Cryst. 1989, 172, 27; c) K. Miyamura, K. Sato, Y.
Gohshi, Bull. Chem. Soc. Jpn. 1989, 62, 45; d) R. Paschke, D.
Balkow, U. Baumeister, H. Hartung, J. R. Chipperfield, A. B.
Blake, P. G. Nelson, G. W. Gray, Mol. Cryst. Liq. Cryst. 1990,
188, 105; e) K. Ohta, Y. Morizumi, T. Fujimoto, I. Yamamoto,
K. Miyamura, Y. Gohshi, Mol. Cryst. Liq. Cryst. 1992, 214,
161; f) A. Serrette, P. J. Carroll, T. M. Swager, J. Am. Chem.
Soc. 1992, 114, 1887 (Corrigendum: J. Am. Chem. Soc. 1993,
115, 11656); g) R. Paschke, S. Diele, I. Letko, A. Wiegeleben,
G. Pelzl, K. Griesar, M. Athanassopoulou, W. Haase, Liq.
Cryst 1995, 18, 451; h) A. B. Blake, J. R. Chipperfield, W. Hus-
sain, R. Paschke, E. Sinn, Inorg. Chem. 1995, 34, 1125; i) R.
Paschke, D. Balkow, E. Sinn, Inorg. Chem. 2002, 41, 1949; j) I.
Sakata, K. Miyamura, Chem. Commun. 2003, 156.
(C=O), 166.7 (C=N) ppm. IR (KBr): ν = 1641 (C=N), 1201 (C–
˜
[3] N. Hoshino, Coord. Chem. Rev. 1998, 174, 77.
O) cm–1. C115H194N2O13 (1812.78): calcd. C 76.19, H 10.79, N 1.55;
found C 76.18, H 10.99, N 1.42.
[4] a) I. Aiello, M. Ghedini, F. Neve, D. Pucci, Chem. Mater. 1997,
9, 2107; b) I. Aiello, M. Ghedini, M. La Deda, D. Pucci, O.
Francescangeli, Eur. J. Inorg. Chem. 1999, 1367.
Synthesis of the Lanthanide(III) Complexes: The dinuclear lantha-
nide complexes (Ln = Nd, Sm, Gd) of the Schiff-base ligand LH3
were synthesized by reaction between the lanthanide acetate salt
Ln(CH3COO)3·xH2O (1.1 equiv.) and the Schiff base (1 equiv.) in
chloroform at refluxing temperature overnight. The acetate ions act
as the base to deprotonate the ligand. The complexes were purified
by crystallization from a water/ethyl acetate mixture, and were ob-
tained in moderate yields: 74% for the Nd complex 1, 62% for
the Sm complex 2 and 49% for the Gd complex 3. MALDI-TOF
measurements show a molecular peak that corresponds to the dinu-
clear species with the stoichiometry [Ln2L2] for the neodymium()
and samarium() complexes.
[5] K. Binnemans, D. W. Bruce, S. R. Collinson, R. Van Deun,
Yu. G. Galyametdinov, F. Martin, Phil. Trans. Royal Soc., A
1999, 357, 3063.
[6] a) A. G. Serrette, T. M. Swager, J. Am. Chem. Soc. 1993, 115,
8879; b) D. Killian, D. Knawby, M. A. Athanassopoulou, S. T.
Trzaska, T. M. Swager, S. Wróbel, W. Haase, Liq. Cryst. 2000,
27, 509; c) W. Haase, D. Killian, M. A. Athanassopoulou, D.
Knawby, T. M. Swager, S. Wróbel, Liq. Cryst. 2002, 29, 133.
[7] a) K. Binnemans, K. Lodewyckx, B. Donnio, D. Guillon,
Chem. Eur. J. 2002, 8, 1101; b) K. Binnemans, K. Lodewyckx,
Supramol. Chem. 2003, 15, 485.
[8] K. Binnemans, K. Lodewyckx, B. Donnio, D. Guillon, Eur. J.
Inorg. Chem. 2005, 1506.
[9] Yu. Galyametdinov, M. A. Athanassopoulou, K. Griesar, O.
Kharitonova, E. A. Soto Bustamante, L. Tinchurina, I. Ovch-
innikov, W. Haase, Chem. Mater. 1996, 8, 922.
[10] K. Binnemans, Yu. G. Galyametdinov, R. Van Deun, D. W.
Bruce, S. R. Collinson, A. P. Polishchuk, I. Bikchantaev, W.
Haase, A. V. Prosvirin, L. Tinchurina, I. Litvinov, A. Gu-
bajdullin, A. Rakhmatullin, K. Uytterhoeven, L. Van Meerv-
elt, J. Am. Chem. Soc. 2000, 122, 4335.
Nd Complex 1: C230H382N4Nd2O26 (3907.99): calcd. C 70.69, H
9.85, N 1.43; found C 70.79, H 9.84, N 1.44. MALDI-TOF:
Mw(isotopic) = 3907.68 m/e; Mw(found) = 3909.30 m/e. IR (KBr):
ν = 1633 (C=N) cm–1.
˜
Sm Complex 2: C230H382N4O26Sm2 (3920.23): C 70.47, H 9.82, N
1.43; found C 69.97, H 9.69, N 1.41. MALDI-TOF: Mw(isotopic)
156
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 150–157

Dinuclear Lanthanide Schiff-Base Complexes
FULL PAPER
[11] E. Terazzi, S. Torelli, G. Bernardinelli, J.-P. Rivera, J.-M.
Bénech, C. Bourgogne, B. Donnio, D. Guillon, D. Imbert, J.-
C. G. Bünzli, A. Pinto, D. Jeannerat, C. Piguet, J. Am. Chem.
Soc. 2005, 127, 888.
[12] a) Supramolecular Chemistry: Concepts and Perspectives, J.-M.
Lehn, VCH, Weinheim (Germany), 1995; b) Supramolecular
Chemistry, J. W. Steed, J. L. Atwood, Wiley, Chichester, 2000.
[13] a) C. Piguet, G. Bernardinelli, G. Hopfgartner, Chem. Rev.
1997, 97, 2005; b) G. F. Swiegers, T. Malefetse, J. Chem. Rev.
2000, 100, 3483; c) M. Albrecht, Chem. Rev. 2001, 101, 3457.
[14] R. Ziessel, Coord. Chem. Rev. 2001, 216–217, 195.
[15] J. Xu, T. N. Parac, K. N. Raymond, Angew. Chem. Int. Ed.
1999, 38, 2878.
[18] a) A. El-ghayoury, L. Douce, A. Skoulios, R. Ziessel, Angew.
Chem. Int. Ed. 1998, 37, 2205; b) R. Ziessel, L. Douce, A. El-
ghayoury, A. Harriman, A. Skoulios, Angew. Chem. Int. Ed.
2000, 39, 1489; c) L. Douce, R. Ziessel, Mol. Cryst. Liq. Cryst.
2001, 362, 133.
[19] a) F. Morale, R. W. Date, D. Guillon, D. W. Bruce, R. L. Finn,
C. Wilson, A. J. Blake, M. Schröder, B. Donnio, Chem. Eur. J.
2003, 9, 2484; b) B. Donnio, B. Heinrich, H. Allouchi, J. Kain,
S. Diele, D. Guillon, D. W. Bruce, J. Am. Chem. Soc. 2004, 126,
15258.
Received: August 4, 2005
Published Online: November 11, 2005
[16] A. Gelasco, V. L. Pecoraro, J. Am. Chem. Soc. 1993, 115, 7928.
[17] a) J. Malthête, H.-T. Nguyen, C. Destrade, Liq. Cryst. 1993,
13, 171; b) H.-T. Nguyen, C. Destrade, J. Malthête, Adv. Mater.
1997, 9, 375.
Eur. J. Inorg. Chem. 2006, 150–157
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
157