Rare-earth-containing magnetic liquid crystals

Article abstract of DOI:10.1021/ja993351q

Rare-earth-containing metallomesogens with 4-alkoxy-N-alkyl-2- hydroxybenzaldimine ligands are reported. The stoichiometry of the complexes is [Ln(LH)₃(NO₃)₃], where Ln is the trivalent rare-earth ion (Y, La, and Pr to Lu, except Pm) and LH is the Schiff base. The Schiff base ligands are in the zwitterionic form and coordinate through the phenolic oxygen only. The three nitrate groups coordinate in a bidentate fashion. The X-ray single- crystal structures of the nonmesogenic homologous complexes [Ln(LH)₃(NO₃)₃], where Ln = Nd(III), Tb(III), and Dy(III) and LH = CH₃OC₆H₃(2-OH)CH=NC₄H₉, are described. Although the Schiff base ligands do not exhibit a mesophase, the metal complexes do (SmA phase). The mesogenic rare-earth complexes were studied by NMR, IR, EPR, magnetic susceptibility measurements, X-ray diffraction, and molecular modeling. The metal complexes in the mesophase have a very large magnetic anisotropy, so that these magnetic liquid crystals can easily be aligned by an external magnetic field.

Full text of DOI:10.1021/ja993351q

J. Am. Chem. Soc. 2000, 122, 4335-4344
4335
Rare-Earth-Containing Magnetic Liquid Crystals
Koen Binnemans,*,† Yury G. Galyametdinov,‡ Rik Van Deun,† Duncan W. Bruce,§
Simon R. Collinson,§ Arkadiy P. Polishchuk,| Ildar Bikchantaev,‡ Wolfgang Haase,
Andrey V. Prosvirin,‡ Larisa Tinchurina,‡ Igor Litvinov,# Ajdar Gubajdullin,#
Ajdar Rakhmatullin,¶ Koen Uytterhoeven,† and Luc Van Meervelt†
Contribution from the Department of Chemistry, Katholieke UniVersiteit LeuVen, Celestijnenlaan 200F,
B-3001 LeuVen, Belgium, Physical-Technical Institute, Kazan Branch RAS, Sibirsky Tract 10/7,
420029 Kazan, Russia, School of Chemistry, UniVersity of Exeter, Stocker Road,
Exeter EX4 4QD, England, National Technical UniVersity of Ukraine, pr. Pobedi 37,
252056 KieV, Ukraine, Institut fu¨r Physikalische Chemie, Technische UniVersita¨t Darmstadt, Condensed
Matter Group, Petersenstrasse 20, D-64287 Darmstadt, Germany, and A. E. ArbuzoV Institute of Organic
and Physical Chemistry, ArbuzoV Str. 8, 420088 Kazan, Russia
ReceiVed September 15, 1999. ReVised Manuscript ReceiVed February 23, 2000
Abstract: Rare-earth-containing metallomesogens with 4-alkoxy-N-alkyl-2-hydroxybenzaldimine ligands are
reported. The stoichiometry of the complexes is [Ln(LH)3(NO3)3], where Ln is the trivalent rare-earth ion (Y,
La, and Pr to Lu, except Pm) and LH is the Schiff base. The Schiff base ligands are in the zwitterionic form
and coordinate through the phenolic oxygen only. The three nitrate groups coordinate in a bidentate fashion.
The X-ray single-crystal structures of the nonmesogenic homologous complexes [Ln(LH)3(NO3)3], where Ln
) Nd(III), Tb(III), and Dy(III) and LH ) CH3OC6H3(2-OH)CHdNC4H9, are described. Although the Schiff
base ligands do not exhibit a mesophase, the metal complexes do (SmA phase). The mesogenic rare-earth
complexes were studied by NMR, IR, EPR, magnetic susceptibility measurements, X-ray diffraction, and
molecular modeling. The metal complexes in the mesophase have a very large magnetic anisotropy, so that
these magnetic liquid crystals can easily be aligned by an external magnetic field.
Introduction
a large magnetic anisotropy ∆ø. The |∆ø| values of diamagnetic
liquid crystals range between 0 and 60 × 10-6 cm3 mol-1 2
.
Liquid crystals are fascinating materials, as they exhibit
properties of both solids and liquids. Classic rodlike organic
liquid crystals are diamagnetic and can be easily oriented by
an electric field. In theory, it is possible to align liquid crystals
by a magnetic field, in a manner analogous to alignment in an
electric field. Relatively strong magnetic fields are necessary
for the alignment of diamagnetic liquid crystals, and this is not
compatible with small devices such as displays. In recent years,
the design of metal-containing liquid crystals (metallomesogens)
has attracted a lot of attention, and this research field is described
in several reviews.1 By incorporating metal ions with unpaired
electrons into liquid crystals, it is possible to create paramagnetic
liquid crystals. The required magnetic field strength to align
paramagnetic liquid crystals is much smaller than the field
strength necessary to align diamagnetic liquid crystals. For
alignment by weak magnetic fields, the compounds need to have
This value can be increased by introducing a paramagnetic
center in the liquid-crystalline compounds.3,4
Our search for magnetic liquid crystals brought us to low
molecular weight metallomesogens containing rare-earth (or
lanthanide) ions, because these ions often have a large magnetic
anisotropy. Only a few examples of rare-earth containing
metallomesogens have been described in the literature. The first
calamitic () rodlike) lanthanide-containing metallomesogens
were reported by Galyametdinov and co-workers, being a series
of mesomorphic lanthanide complexes of liquid-crystalline
Schiff base ligands with two aromatic rings.5 All the complexes
formed a highly viscous smectic mesophase, which was
afterward identified as a smectic A phase. Later, Galyametdinov
et al.6,7 used simpler one-ring Schiff bases LH as the ligand.
(2) Muller, H.; Haase, W. J. Phys. (France) 1983, 44, 1209.
(3) (a) Marcos, A.; Romero, P.; Serrano, J. L.; Barbera´, J.; Levelut, A.
M. Liq. Cryst. 1990, 7, 251. (b) Haase, W.; Gehring, S.; Borchers, B.
Multifunctional Materials. In MRS Symp. Proc. 1990, 175, 249; Buckley,
A. J., Gallagher-Daggitt, G., Karasz, F. E., Ulrich, D. R., Eds. (c) Barbera´,
J.; Levelut, A. M.; Marcos, M.; Romero, P.; Serrano, J. L. Liq. Cryst. 1991,
10, 119. (d) Campillos, E.; Marcos, M.; Serrano, J. L.; Alonso, P. J. J.
Mater. Chem. 1991, 1, 197. (e) Haase, W.; Borchers, B. Magnetic Molecular
Materials. In NATO ASI Ser. E: Appl. Sci. 1991, 198, 245. (f) Borchers,
B.; Haase, W. Mol. Cryst. Liq. Cryst. 1991, 209, 319. (g) Alonso, P. J.;
Mart´ınez, J. I. Liq. Cryst. 1996, 21, 597. (h) Deschenaux, R.; Schweissguth,
M.; Vilches, M. T.; Levelut, A. M.; Hautot, D.; Long, G. J.; Luneau, D.
Organometallics 1999, 18, 5553.
* Address correspondence to this author.
† Katholieke Universiteit Leuven.
‡ Physical-Technical Institute.
§ University of Exeter.
| National Technical University of Ukraine.
Technische Universita¨t Darmstadt.
# A. E. Arbuzov Institute of Organic and Physical Chemistry.
¶ Kazan State University.
(1) (a) Giroud-Godquin, A. M.; Maitlis, P. M. Angew. Chem., Int. Ed.
Engl. 1991, 30, 375. (b) Hudson, S. A.; Maitlis, P. M. Chem. ReV. 1993,
93, 861. (c) Polishchuk, A. P.; Timofeeva, T. V. Russ. Chem. ReV. 1993,
62, 291.(d) Metallomesogens, Synthesis, Properties and Applications;
Serrano, J. L., Ed.; VCH: Weinheim, 1996. (e) Bruce, D. W. In Inorganic
Materials, 2nd ed.; Bruce, D. W., O’Hare, D., Eds.; Wiley: Chichester,
1996; Chapter 8, p 429. (f) Donnio, B.; Bruce, D. W. Struct. Bond. 1999,
95, 193.
(4) Bikchantaev, I.; Galyametdinov, Yu.; Prosvirin, A.; Griesar, K.; Soto-
Bustamante, E. A.; Haase, W. Liq. Cryst. 1995, 18, 231.
(5) Galyametdinov, Yu. G.; Ivanova, G. I.; Ovchinnikov, I. V. Bull. Acad.
Sci. USSR, DiV. Chem. Sci. 1991, 40, 1109.
10.1021/ja993351q CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/21/2000

4336 J. Am. Chem. Soc., Vol. 122, No. 18, 2000
Binnemans et al.
Although these ligands were not mesomorphic, the correspond-
ing lanthanide complexes were mesomorphic (SmA phase), and
the viscosity and transition temperatures of these complexes
were much lower than those of earlier described complexes.
These complexes were supposed to have the stoichiometry [LnL-
(LH)2X2], showing that one ligand coordinates in its deproto-
nated form, whereas the other two coordinate as the neutral
species. Binnemans et al.8 showed that the choice of the
counterion has an important influence on the transition tem-
peratures of these complexes with Schiff base ligands. The
change of the transition temperatures of Schiff base complexes
with the nitrate anion reflects the lanthanide contraction, in the
sense that a gradual decrease of the mesophase stability range
over the lanthanide series was observed.9,10 Magnetic studies
on this type of lanthanide-containing metallomesogens were
published by Galyametdinov and co-workers.6,11-14 A huge
value for the magnetic birefringence was found for the Tb
complex [TbL(LH)2(NO3)2], where LH is the Schiff base C12H25-
OC6H3(2-OH)CHdNC18H37.11
In this paper, we describe the structure and magnetic
properties of mesogenic rare-earth complexes with 4-(alkoxy)-
N-alkyl-2-hydroxybenzaldimine Schiff base ligands. The crystal
structures of nonmesogenic Nd(III), Tb(III), and Dy(III) com-
plexes with Schiff base ligands were obtained. On the basis of
NMR, EPR, and IR data, in combination with molecular
modeling, these results are extrapolated to liquid-crystalline rare-
earth-containing Schiff base complexes. The magnetic alignment
of the metal complexes is discussed.
Scheme 1. Synthesis of the Schiff Base Ligands and Their
Rare-Earth Complexesa
a Reaction conditions: (i) RBr (1 equiv), KHCO3 (1 equiv), DMF,
reflux (3 h); (ii) R′NH2 (1 equiv), acetic acid (catalyst, 5 drops), absolute
ethanol, reflux (3 h); (iii) absolute ethanol, room temperature, 2 h.
Modifications of step (i) for ligands with short alkoxy chains are
described in the Experimental Section
the atomic positions are available as Supporting Information.
The Dy atom lies in a general position and the metal atom is
coordinated by three bidentate nitrate groups, while the three
neutral Schiff base ligands coordinate in a monodentate fashion
via the phenolic oxygen. The nitrogen atom of the Schiff base
remains noncoordinated. The nearest environment of the Dy
atom consists of three oxygen atoms of the neutral Schiff base
ligands and six oxygen atoms of the nitrate groups (Figure 1).
In this way, the Dy atom is in a nine-coordinate environment.
The coordination polyhedron is a distorted monocapped square
antiprism. The dihedral angle between planes of any two nitrate
groups in complex 1 is approximately 90° (79-87°). A special
feature of the crystal structure is that the Schiff bases coordinate
as neutral ligands to the Dy atom. However, the oxygen atoms
of the OH group are deprotonated and the hydrogen atoms
have moved to the nitrogen atom. In this way, a zwitterionic
structure is formed (negative charge on the oxygen atom,
positive charge on the imine nitrogen atom). Intramolecular
H-bonds N-H‚‚‚O exist between the protonated nitrogen and
the deprotonated oxygen atoms. Two of the three hydrogen
atoms form a second H-bond with an oxygen atom of a nitrate
group. The six-membered rings closed by intraligand H-bonds
are planar (root-mean-square deviations 0.005-0.024 Å) and
virtually coplanar with the benzene rings (deviation from
planarity is 1.5-2.3°).
Results
Synthesis of Rare-Earth Complexes 1-42. The synthesis
of the 4-(alkoxy)-N-alkyl-2-hydroxybenzaldimine Schiff bases
and the rare-earth complexes 1-42 is summarized in Scheme
1. A summary of the rare-earth complexes can be found in Table
1. The stoichiometry of the complexes is consistent with the
formula [Ln(LH)3(NO3)3]. The CHN analysis data of the ligands
and complexes, together with the melting points of the ligands,
are available as Supporting Information.
The crystal structure of the Nd(III) complex [Nd(L1H)3-
(NO3)3] (2) is very similar to the crystal structure of the
corresponding Dy(III) complex 1. Because the Nd(III) ion has
a slightly larger ionic radius than the Dy(III), the unit cell
volume of the Nd(III) complex 2 is larger than the unit cell
volume of the Dy(III) complex 1. The same trend is of course
observed for the cell parameters a, b, and c (Table 2). The
analogous Tb(III) compound 3 proved to be isostructural with
the Nd(III) and Dy(III) compounds. The cell parameters of [Tb-
(L1H)3(NO3)3] are a ) 11.372(2) Å, b ) 13.777(1) Å, c )
14.035(1) Å, R ) 98.647(8)°, â ) 98.29(1)°, γ ) 92.68(1)°,
and V ) 2145.9(5) Å3. From these results, we can conclude
that the rare-earth complexes of the type [Ln(L1H)3(NO3)3] are
isostructural, at least for the elements of the middle of the
lanthanide series. Calculations have shown that the packing of
the atoms in the crystal structures of compounds 1, 2, and 3 is
very dense.
Thermal Behavior of the Rare-Earth Complexes. The
4-alkoxy-N-octadecyl-2-hydroxybenzaldimine Schiff bases them-
selves are not mesomorphic. All the rare-earth complexes, except
complexes 1-3, exhibit a mesophase, the mesophase being
identified as a smectic A phase by hot-stage polarized optical
microscopy (Table 2). The presence of this mesophase type was
also proved by high-temperature X-ray diffraction on an
analogous compound.13 The mesophase is highly viscous, but
more fluid than the mesophase of the previously described
Crystal Structures of Complexes 1-3. The X-ray single-
crystal structure was determined for the Dy(III) 1 and the Nd-
(III) complex 2. For the Tb(III) complex 3 the unit cell
parameters were determined. These complexes have short alkyl
chains, but it is supposed that the coordination sphere is
unaltered by increasing the chain lengths.
The Dy(III) complex 1 crystallizes in the triclinic system
(space group P1h, No. 2), with two formula units in the unit
cell. The crystallographic data are summarized in Table 2, and
(6) Galyametdinov, Yu. G.; Ivanova, G. I.; Prosvirin, A. V.; Kadkin, O.
Russ. Chem. Bull. 1994, 43, 938.
(7) Galyametdinov, Yu. G.; Athanassopoulou, M. A.; Griesar, K.;
Kharitonova, O.; Soto Bustamante, E. A.; Tinchurina, L.; Ovchinnikov, I.;
Haase, W. Chem. Mater. 1996, 8, 922.
(8) Binnemans, K.; Galyametdinov, Yu. G.; Collinson, S. R.; Bruce, D.
W. J. Mater. Chem. 1998, 8, 1551.
(9) Binnemans, K.; Van Deun, R.; Galyametdinov, Yu. G.; Bruce, D.
W. Chem. Phys. Lett. 1999, 300, 509.
(10) Binnemans, K.; Bruce D. W.; Collinson, S. R.; Van Deun, R.;
Galyametdinov, Yu. G.; Martin, F. Philos. Trans. R. Soc. London A 1999,
357, 3063.
(11) Galyametdinov, Yu. G.; Athanassopoulo, M.; Haase, W.; Ovchin-
nikov, I. V. Russ. J. Coord. Chem. 1995, 21, 718.
(12) Ovchinnikov, I. V.; Galyametdinov, Yu. G.; Prosvirin, A. V. Russ.
Chem. Bull. 1995, 44, 768.
(13) Galyametdinov, Yu. G.; Ivanova, G.; Ovchinnikov, I.; Prosvirin,
A.; Guillon, D.; Heinrich, B.; Dunmur, D. A.; Bruce, D. W. Liq. Cryst.
1996, 20, 831.
(14) Turanov, A. N.; Ovchinnikov, I. V.; Galyametdinov, Yu. G.;
Ivanova, G. I.; Goncharov, V. A. Russ. Chem. Bull. 1999, 48, 690.

Rare-Earth-Containing Magnetic Liquid Crystals
J. Am. Chem. Soc., Vol. 122, No. 18, 2000 4337
Table 1. Overview of the Rare-Earth Complexes [Ln(LH)3(NO3)3] and Their Thermal Behaviora
Cr f SmAc
SmA f Ic
∆H (kJ mol-1
)
complex
R
R′
metal ion
T (°C)
∆H (kJ mol-1
)
T (°C)
1b
2b
CH3
CH3
CH3
C4H9
C4H9
C4H9
Dy(III)
Nd(III)
Tb(III)
Y(III)
La(III)
Pr(III)
n.d.d
n.d.
n.d.
22.7
2.5
8.7
15.0
12.4
15.5
16.9
17.6
21.4
22.2
n.d.
n.d.
n.d.
n.d.
11.0
11.8
13.2
11.8
12.4
13.3
12.7
10.6
11.8
10.8
n.d.
n.d.
n.d.
n.d.
3.8
3b
4
5
6
7
8
9
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
C8H17
CH3
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C18H37
C16H33
C16H33
C16H33
C16H33
C16H33
132
83
90
141
165
163
159
154
151
150
148
146
144
142
142
141
139
127
147
149
158
155
156
155
158
158
157
158
158
155
154
152
150
150
146
144
138
148
147
145
147
147
Nd(III)
Sm(III)
Eu(III)
Gd(III)
Tb(III)
Dy(III)
Ho(III)
Er(III)
Tm(III)
Yb(III)
Lu(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
Nd(III)
La(III)
Pr(III)
98
105
108
121
128
131
134
135
139
138
135
119
132
112
119
115
110
104
100
101
103
100
100
102
103
103
104
101
101
100
81
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
n.d.
n.d.
n.d.
37.9
21.9
15.6
14.0
15.7
13.7
13.7
15.4
16.8
16.9
15.5
11.0
15.5
15.3
16.4
19.0
15.1
19.7
19.6
3.7
C2H5
C3H7
C4H9
7.3
8.9
10.4
10.9
11.5
11.3
11.4
11.0
10.8
11.7
11.7
10.9
10.8
10.7
10.4
10.1
10.3
10.1
10.3
n.d.
n.d.
n.d.
n.d.
n.d.
C5H11
C6H13
C7H15
C9H19
C10H21
C11H23
C12H25
C13H27
C14H29
C15H31
C16H33
C17H35
C18H37
C19H39
C20H41
C12H25
C12H25
C12H25
C12H25
C12H25
C12H25
105
105
107
107
107
n.d.
n.d.
n.d.
n.d.
Nd(III)
Gd(III)
Tb(III)
Er(III)
n.d.
a LH is the Schiff base ligand with alkoxy chain R and N-alkyl chain R′ (see Scheme 1). b Not mesomorphic. c Cr: crystalline phase. SmA:
smectic A phase. I: isotropic liquid. d n.d. not determined.
complexes of the two-ring Schiff base ligands.5 The viscosity
gradually decreases at higher temperatures, and a strong increase
in fluidity is observed close to the clearing point. When cooling
the mesophase, no crystallization is observed in the microscope
and the texture of the mesophase is frozen into the solid state,
forming a glassy mesophase. The small values of the melting
enthalpies of the complexes in the second heating run of the
DSC-thermogram indicate that the compounds only crystallize
partially (Table 2).
The rare-earth complexes of ligand L9H show very clearly
that a correlation exists between the ionic radius and the
transition temperatures. The ionic size decreases over the
lanthanide series (lanthanide contraction), and simultaneously,
the mesophase stability range decreases. This correlation is also
illustrated by the fact that the yttrium complex 4 and the
holmium complex 13 have very comparable transition temper-
atures, in agreement with the nearly identical ionic size. Since
the ion gets smaller over the lanthanide series and the charge
remains, the electrostatic attraction between the lanthanide ion
and the coordinating groups becomes stronger and therefore the
melting temperature increases. It is more difficult to explain
the reduction of the clearing point over the lanthanide series.
Probably the reduction in ion size leads to steric problems in
the arrangement of the ligands around the complexes and thus
more structural distortions occur, leading to a less anisotropic
complex and reduction of the clearing point. Because of the
overlap between melting and clearing peaks, the melting and
clearing enthalpy could not be determined separately for the
complexes at the end of the lanthanide series (Er-Lu).
For investigating the influence of the alkoxy chain length on
the transition temperatures of the mesogenic Schiff base
complexes, we prepared the Nd(III) complexes of the Schiff
base ligands L2H to L21H, giving all the alkyl chains between
CH3O and C20H41O. The N-alkyl chain was fixed to C18H37.
All the Nd(III) complexes were found to be liquid crystals, even
the complexes with the shortest chain lengths. However, the
alkoxy chain length of the ligand has less influence on the
transition temperatures than the rare-earth ion has. The transition
temperatures of the Nd(III) complexes of the different ligands
are rather constant for the different alkoxy chain lengths (Figure
2); for the shortest alkoxy chain lengths a smooth variation is
absent. The octadecyl chain on the imine nitrogen provides
enough structural anisotropy to the complexes (i.e., a high
length/width ratio) so that the complexes behave as calamitic

4338 J. Am. Chem. Soc., Vol. 122, No. 18, 2000
Binnemans et al.
Table 2. Summary of Crystal Data, Intensity Measurements, and
Structure Refinement for [Dy(L1H)3(NO3)3] (1) and
[Nd(L1H)3(NO3)3] (2)
[Dy(L1H)3(NO3)3] (1) [Nd(L1H)3(NO3)3] (2)
formula
mol wt
cryst syst
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
C36H51N6O15Dy
970.31
C36H51N6O15Nd
952.05
triclinic
P1h (No. 2)
11.455(2)
13.859(3)
14.101(4)
98.50(2)
98.87(1)
92.52(2)
2182.4(9)
2
triclinic
P1h (No. 2)
11.380(1)
13.800(5)
14.027(5)
98.70(3)
98.15(2)
92.67(2)
2150(1)
2
1.498
990
1.81
13477
Z
D
Figure 2. Evolution of the transition temperatures of the [Nd(LH)3-
(NO3)3] complexes as a function of the alkoxy chain length.
calc, g‚cm-3
1.449
978
1.85
7708
F(000)
µ
Mo KR, cm-1
reflns measd
which the deprotonated Schiff base coordinates via the phenolic
oxygen and via the azomethine nitrogen to the copper(II) ion.15
In these copper(II) complexes, the CdN stretching vibration is
shifted to lower wavenumbers, indicating less double-bond
character in the CdN bond. The broad weak O-H stretching
vibration of the ligand at ca. 2850 cm-1 (on which the CH
modes are superimposed) is replaced in the metal complexes
by a band at ca. 3300 cm-1, due to the N-H vibration in Cd
N+-H. The broadness of the band shows that this hydrogen
atom is still involved in intramolecular H-bonding with the
phenolic oxygen. The phenolic C-O stretching vibration, which
θ range (deg)
2 < θ < 26
2 < θ < 25
no. of unique reflcns 10305
7328
no. of reflcns obsda
no. of variables
R
9711
728
0.0250
6418
566
0.0376
a I > 3σ(I) for the Dy(III) complex, I > 2σ(I) for the Nd(III)
complex.
is observed in the Schiff base ligands around 1280 cm-1
,
overlaps in the metal complexes with a vibration due to the
nitrate groups. Four bands in the IR spectra of the complexes
near 1475, 1283, 1120, and 850 cm-1 can be assigned to
vibrations of the coordinated nitrate group (respectively the
vibrations ν4, ν1, ν2, and ν6). The difference between the ν4
and the ν1 peak positions is ca. 190 cm-1, which is typical for
bidentate nitrate groups (monodentate nitrate groups display a
much smaller splitting).
1H and 13C NMR investigations were carried out both in
solution and in the solid state on the ligand L13H and on the
diamagnetic La(III) complex [La(L13H)3(NO3)3] (37). The
1
analysis of the H NMR data of these compounds in solution
(CDCl3) is consistent with the composition [La(L13H)3(NO3)3]
of the lanthanum complex. The signal corresponding to the
imine proton (CHdN) in the complex (δ ) 7.63 ppm) is much
broader than that in the ligand (δ ) 7.18 ppm). For other La-
(III) complexes, we sometimes even observed a splitting of the
imine proton signal. In the ligand L13H, the signal of the
phenolic proton (OH) is found at δ ) 13.80 ppm. In the La-
(III) complex 37, the signal seems to be shifted to δ ) 12.29
ppm. The relative integrated intensity of this signal corresponds
to three protons. However, we found that selective irradiation
of the signal at δ ) 12.29 ppm removed the broadening of the
imine signal. This proves that the protons are coupled. Thus,
the signal at δ ) 12.29 ppm does not correspond to the proton
of the OH group, but to the proton of the dN+H group. In the
1H NMR spectrum at 293 K, the value of the coupling constant
3J ) 10.3 Hz points to a trans-disposition of the -CHd and
dN+H- protons.16 An increase in the temperature leads to a
broadening of the signal of the CHdN proton in the ligand from
4.1 Hz at 293 K to 13.7 Hz at 333 K. The 13C NMR spectra of
the ligand L13H and the corresponding La(III) complex in
solution (acetone-d6) were compared with the 13C CP/MAS
NMR spectra of these compounds in the solid state at room
Figure 1. Crystal structure of complex [Dy(L1H)3(NO3)3] (1).
mesogens and are able to form a smectic A phase. The thermal
stability of the metal complexes was studied by thermogravi-
metry. All the samples which were examined indicate a high
thermal stability of the complexes: the complexes only start to
decompose above 300 °C (more than 100 °C above the clearing
point).
Structure of the Mesogenic Rare-Earth Complexes. Be-
cause single crystals of Schiff bases with long alkyl chains and
single crystals of the corresponding complexes could not be
obtained, we were forced to probe the structure of these
compounds by means of spectroscopic methods (IR, NMR, and
EPR), and by molecular modeling. We made the assumption
that extension of the chain lengths of the Schiff base ligands
has only a minor effect on the first coordination sphere of the
trivalent rare-earth ion. Of course, the chain length will have a
significant effect on the packing of the complexes in the crystal
structure.
In the infrared spectra of the rare-earth complexes, the CdN
stretching vibration is shifted to higher wavenumbers (by 30-
40 cm-1 to about 1662 cm-1) in comparison to the same
transition in the Schiff base ligands. This behavior is opposite
to what is observed for copper(II) Schiff base complexes in
(15) Kovacic, J. E. Spectrochim. Acta 1967, 23A, 183.
(16) Bystrov, V. F. Prog. NMR Spectrosc. 1976, 10, 41.

Rare-Earth-Containing Magnetic Liquid Crystals
J. Am. Chem. Soc., Vol. 122, No. 18, 2000 4339
temperature. For both ligand L13H and the corresponding La-
(III) derivative 37 the chemical shifts in the powder spectra are
very similar to the chemical shifts in the solution spectra,
indicating that the magnetic fields surrounding the structure in
solution and in the solid state are comparable.
The structure of the SmA phase of a Dy(III)-containing
metallomesogen was simulated by molecular modeling. The
atomic positions of the Dy(III) complex 1 were taken as a
starting point. Small-angle X-ray diffraction studies11,13 have
shown that the thickness of the SmA layers in compound [Tb-
(L13H)3(NO3)3] (31.1 Å) is substantially less than the overall
length of this complex, being ca. 49 Å. The simulation of a
probable form of the mesogenic complex [Dy(L13H)3(NO3)3]
was carried out with the assumption that the coordination of
the Dy(III) ion in the solid state is retained in the liquid-
crystalline (LC) state. Therefore in the complex [Dy(L13H)3-
(NO3)3] the optimization of the geometry of the metal unit was
not made, i.e., the mutual arrangement of the (Dy,O,N) atoms
in the metal unit remained unaltered. The geometry of complex
1 was optimized, and then we added the missing links of alkyl
tails to obtain the complex [Dy(L13H)3(NO3)3]. The main
difference in the optimized structure of the central fragments
of complexes 1 and [Dy(L13H)3(NO3)3] is the various orienta-
tions of the phenyl rings; the phenyl rings in the mesogenic
complex are nearly parallel to each other, in contradistinction
to the phenyl rings in 1. The overall shape of the complex is
rodlike, with a bulge due to the nitrate groups coordinated to
the dysprosium atom. A slightly steplike curvature in the rod
explains why the overall calculated length of the complex with
the alkyl chains in the all-trans conformation is longer than the
experimental layer distance. All the alkoxy chains are situated
at the same side of the complex (and the same holds for the
N-alkyl chains). The thickness of the layers is determined by
the long alkyl tails of the ligands and exactly equal to the
distance between the Dy atoms of the adjacent layers. Taking
into account the close electronic nature of the Tb and Dy atoms,
it is possible to expect that the central fragment in mesogenic
complex [Tb(L13H)3(NO3)3] is similar to the fragment in
complex [Dy(L13H)3(NO3)3]. The layer periodicity of the SmA
phase of [Tb(L13H)3(NO3)3] was calculated for the optimized
structure to be 30.3 Å. This magnitude correlates well with the
experimentally determined periodicity of the layer in the
mesophase of compound [Tb(L13H)3(NO3)3], being 31.1 Å.7
Figure 3. Inverse magnetic susceptibility of the liquid-crystalline Nd-
(III) complex [Nd(L22H)3(NO3)3] (39) in the temperature range of
4-300 K. The solid line is obtained via a Curie-Weiss fit.
The magnetic susceptibilities of powdered samples were mea-
sured between 4.2 and 300 K on the mesomorphic complexes
38-42. The complexes follow the usual Curie-Weiss behavior
of the magnetic susceptibility. The typical temperature depen-
dence of the inverse magnetic susceptibility of the Nd(III)
complex 39 at low temperatures is shown in Figure 3. The molar
susceptibilities and the room-temperature effective magnetic
moments of the rare-earth complexes are displayed in the
Supporting Information. The experimental high-temperature
limits of µeff of the two metal complexes are in good agreement
with the expected theoretical values for the free-ion approxima-
tion.17
The magnetic birefringence of the rare-earth complexes 38-
42 in CCl4 solution was measured to study the behavior of these
metal complexes in magnetic fields. Especially the magnetic
birefringence can be used to determine the sign of the magnetic
anisotropy ∆ø. In the experiment, the dependence of ∆R on
the squared magnetic field strength is determined. ∆R is the
difference between the molar refractions of the ordinary and
extraordinary rays:
6n(n0 - ne)
C(n2 + 2)2
∆R )
(3)
n0, ne, and n being the refraction indices of the ordinary and
extraordinary ray and their mean value, respectively. C is the
concentration (in mol L-1). The molar constant of the magnetic
birefringence mS is defined as ∆R/H2. The limiting value of mS
for the magnetic field strength approaching zero is proportional
to the anisotropy of the molecular electronic polarizability ∆R
and to the anisotropy of the magnetic susceptibility ∆ø:
Magnetic and Magnetooptical Properties
To obtain additional information on the structure of the
metallomesogenic complexes and to investigate the magnetic
anisotropy, we carried out temperature-dependent magnetic
susceptibility and magnetic birefringence measurements. The
molar magnetic susceptibility øM (hereafter abbreviated to ø)
of a free metal ion, with total angular momentum J, in the
ground state is
4πN
∆R
H2
mS )
≈
A‚∆ø‚∆R (H f 0)
(4)
45kT
The electronic polarizability anisotropy is determined by the
ligand part of the metal complex only. The anisotropy of the
magnetic susceptibility depends on the central metal ion. The
values of ∆R and ∆ø calculated by the additive method are ∆R
2
NAgJ2 µB
øM )
J(J + 1)
(1)
3k(T - Θ)
) 21 × 10-24 cm-3 mol-1 and ∆ø ) 17 × 10-29 cm-3 mol-1 18
.
where NA is Avogadro’s number, µB is the Bohr magneton, k is
the Boltzmann constant, T is the absolute temperature (in
Kelvin), Θ is the Curie-Weiss constant (in Kelvin), and gJ is
the Lande´ factor. The effective magnetic moment µeff (in Bohr
magneton) can be calculated as follows:
The magnetic anisotropy of the rare-earth ions is about 1-2
orders of magnitude larger than the magnetic anisotropy of the
ligands. The molar constants of the magnetic birefringence,
(17) Siddall, J. H., III In Theory and Applications of Molecular
Paramagnetism; Boudreaux, E. A., Mulay, L. N., Eds.; Interscience: New
York, 1976; p 257.
(18) (a) Vul’fson, S. G.; Chevela, V. V.; Salnikov, Yu. I.; Vereschgin,
A. N. Bull. Acad. Sci. USSR, DiV. Chem. Sci. 1990, 39, 38. (b) Vul’fson,
S. G. Molecular Magnetochemistry; Gordon and Breach: Amsterdam, 1998.
3k
µeff
)
1/2(ø‚T)1/2 ) 2.828(ø‚T)1/2
(2)
2
(
)
NAµB

4340 J. Am. Chem. Soc., Vol. 122, No. 18, 2000
Binnemans et al.
Table 3. Magnetic Properties of Ligand L22H and of the Rare-Earth Complexes [Ln(L22H)3(NO3)3] (38-42)
mSd (cm3 mol-1
)
a
µeff
b
c
compd
exp
theor
Θ (K)
ø iso
ø or
øor - øiso
∆øexp
exp
theor
L22H
-400
3850
3750
18200
27980
27270
-400
4160
4020
18200
29830
28345
∼0
∼0
-930
-810
∼0
5.1 × 10-15
-4.5 × 10-9
-4.1 × 10-9
1.3 × 10-10
-3.1 × 10-8
9.4 × 10-9
4.9 × 10-15
-5.5 × 10-13
-3.8 × 10-13
1.4 × 10-14
38 (Pr)
39 (Nd)
40 (Gd)
41 (Tb)
42 (Er)
3.60
3.55
7.8
9.65
9.55
3.58
3.62
7.9
9.7
9.6
-78.2
-54.7
∼0
310
270
∼0
-3.5
-2.0
1850
1075
-5550
1610
-3.4 × 10-12
1.3 × 10-12
a The effective magnetic moments µeff are expressed in Bohr magnetons (µB) and were measured at room temperature. b All magnetic susceptibility
values ø are expressed in 10-6 cm3 mol-1 at the melting point. c The calculation of the ∆øexp values from the (øor - øiso) values is discussed in the
text. d The molar constants of the magnetic birefringence mS )∆R/H2 are approximated to zero field.
Figure 5. Effective magnetic moments µeff of the liquid-crystalline
Pr(III) complex [Pr(L22H)3(NO3)3] (38) in the temperature range of
300-442 K.
Figure 4. Dependence of the magnetic birefringence (∆R) on the
squared magnetic field strength (H2) of the liquid-crystalline Pr(III)
complex [Pr(L22H)3(NO3)3] (38). The dotted line is the theoretical curve
calculated by eq 4.
the maximum component of the magnetic susceptibility tensor
is perpendicular to the long molecular axis. For the Gd(III)
complex, the magnetic birefringence has a very small positive
value.
calculated theoretically by eq 4, and the experimental values
extrapolated to zero magnetic field are presented in Table 3.
The dependence of the magnetic birefringence (∆R) on the
magnetic field strength (H2) obtained for the Pr(III) complex
38 is presented in Figure 4, the theoretical curve being given
by a dotted line. This theoretical curve represents the behavior
of a single molecule in a magnetic field, and the curve is a
straight line. Figure 4 shows the huge increase in the Cotton-
Mouton effect (3 to 4 orders of magnitude higher than the
calculated value), even in weak fields (<100 G). A deviation
toward the X-axis for strong magnetic field strengths is observed
for the experimental curve, indicating a saturation effect. This
was previously observed for metal-free lyotropic liquid crystals,
for colloids, and for biological macromolecules, and was related
to particle association (Majorane effect).19 According to the
mStheor to mSexp ratio, the metal complexes are associated forming
oriented domains, and each of these domains contains more than
104 molecules of the metal complexes. The strong Cotton-
Mouton effect is caused by the high values of the magnetic
anisotropy of the metal ion, the electronic polarization of the
ligand, and the association of the metal complexes in solution.
The magnetic birefringence of the Er(III) complex has a positive
sign, in agreement with model calculations (∆R > 0, ∆ø > 0).
In this case, the maximum component of the electronic
polarizability tensor coincides with the direction of the long
molecular axis. For the Pr(III), Nd(III), and Tb(III) complexes,
the magnetic birefringence is negative (as in the case of Dy-
(III) complexes). This indicates a negative magnetic anisotropy
(∆R > 0, ∆ø < 0). In other words, in an axial approximation,
To obtain quantitative information on the degree of magnetic
anisotropy of the mesogenic complexes 38-42, we carried out
magnetic susceptibility measurements in the mesophase. Align-
ment of the samples in a static magnetic field (field strength
12 000 G) was observed. The samples exhibited a Curie-Weiss
behavior for the temperature dependence of their magnetic
properties, when heated from the solid state through the SmA
phase until the isotropic phase was reached. When the samples
were cooled from the isotropic phase to the mesophase (in the
presence of a magnetic field), in the vicinity of the clearing
point a drastic increase in the magnetic moment µeff was
observed in comparison to the initial values recorded by heating
of the samples. The increase takes place over a narrow
temperature range, and upon further cooling, the magnetic
properties vary according to the Curie-Weiss law, but with a
different higher µeff than in the heating cycle. This behavior
can be referred to as a magnetic-field-induced orientation in
the liquid-crystalline phase of a magnetically anisotropic sample
with its axis of maximum magnetic susceptibility parallel to
the magnetic field. The orientation of these molecular magnetic
materials in the mesophase can be observed by cooling from
the isotropic phase rather than by heating the solid sample,
because of the high viscosity of the SmA phase. The oriented
mesophase can be frozen into a glassy state. During the second
heating run, the same values for ø are measured as in the first
cooling run. The temperature dependence of the magnetic
moments of the mesogenic Pr(III) complex 38 in the temperature
range from 300 to 410 K (heating and cooling) is shown in
Figure 5.
(19) Maret, G.; Dransfeld, K. In Strong and Ultrastrong Magnetic Fields
and Their Applications; Herlach, F., Ed.; Springer-Verlag: Berlin, 1985; p
80.
In the case of a powdered solid sample (i.e. beneath the
melting temperature of the complex), the measured magnetic

Rare-Earth-Containing Magnetic Liquid Crystals
J. Am. Chem. Soc., Vol. 122, No. 18, 2000 4341
susceptibility is equal to øiso (isotropic magnetic susceptibility).
Also in the isotropic phase, the isotropic magnetic susceptibility
øiso is measured. øiso is increasing with increasing temperature.
The magnetic susceptibility øor measured in the mesophase is
equal to (or less than, if alignment is not complete) the greater
of the two components ø or ø|. øor corresponds to the maximum
of the ø-tensor. The actual value of the experimental magnetic
anisotropy ∆øexp depends on the relative values of ø or ø|, or
to say it in other words, on the sign of ∆øexp
:
3
2
∆øexp ) (øor - øiso) if ø| > ø
(5)
(6)
∆øexp ) -3(øor - øiso) if ø| < ø
The sign of ∆øexp can be determined from EPR and from
magnetic birefringence measurements. If ∆ø ) ø| - ø < 0, or
ø| < ø , the molecular long axis is oriented perpendicular to
the direction of the magnetic field, and in the oriented
mesophase, ø will be measured. If ∆ø ) ø| - ø > 0, or ø| >
ø , the molecular long axis is oriented parallel to the direction
of the magnetic field, and in the oriented mesophase, ø| will be
measured. Therefore, we can calculate the magnitude of ∆øexp
in the Pr(III), Nd(III), and Tb(III) compounds, including the
orientational behavior, using information of the EPR and
magnetic birefringence measurements about the negative sign
of ∆ø ) ø| - ø . For the Er(III) compound, we can propose a
positive sign of ∆ø ) ø| - ø on the basis of EPR and magnetic
birefringence measurements. The experimental data for øor and
Figure 6. The experimental EPR spectra of the Gd(III) complex in
the smectic A mesophase of the Tb(III)-containing host matrix. γ is
the angle between the magnetic field and the direction of initial
alignment of the sample.
øiso at the phase transition temperatures and the estimated values
for ∆øexp can be found in Table 3. As a result after a heating
and a cooling cycle in a magnetic field, we obtained a
magnetically anisotropic molecular material. This anisotropy is
obvious from the cosine-like angle dependence of the magnetic
susceptibility, measured on aligned rare-earth compounds in the
glassy state as a function of the R angle between the directions
of the predominant orientation and the magnetic field.
Figure 7. The experimental EPR spectra of the Gd(III) complex in
the smectic A mesophase of the Er(III)-containing host matrix. γ is
the angle between the magnetic field and the direction of initial
alignment of the sample.
EPR Measurements. The mesogenic Tb(III) and Er(III)
complexes 41 and 42 were studied by electron paramagnetic
resonance (EPR). In the past, only EPR spectra of Gd(III)-
containing metallomesogens were studied.20 The spin lattice
relaxation time of Tb(III) and Er(III) is much shorter than that
of Gd(III), so that EPR spectra of Tb(III) and Er(III) compounds
can only be detected at liquid-helium temperature and not in
the mesophase (above +100 °C). Gd(III) is the only trivalent
lanthanide ion with an observable EPR signal at room temper-
ature (or at higher temperatures).21 To avoid this obstacle, we
used Gd(III) complex 40 as a spin probe to study the orienta-
tional behavior of the mesomorphic Tb(III) and Er(III) com-
plexes in a magnetic field. The Gd(III) complex itself cannot
be oriented very easily in a magnetic field, because of the 8S7/2
ground state. The Q-band EPR spectrum of Gd(III) reflects this
orientational behavior. The EPR spectra of Gd(III) mixed in
small amount (5 wt %) with the Tb(III) or Er(III) compound
are given in Figures 6 and 7, and alignment of the samples is
obvious from the different features in the EPR spectra at
different orientations. As a control experiment, the samples were
cooled in the absence of an external magnetic field: no
alignment occurred and the EPR spectra at different orientations
proved to be identical.
The positions of the EPR lines of the Gd(III) complex can
be calculated by use of the spin-Hamiltonian:22
H ) gâH‚S +
B qO q
(7)
∑
k
k
k,g
where â is the Bohr magneton, H the magnetic field strength,
S the electron spin, and g the g-factor. The Okq are spin-operators
with coefficients Bkq. The number of (k,q) combinations depends
on the symmetry of the crystal field around the Gd(III) ion,
and k ) 2, 4, or 6 and 0 e q e k. The contribution of the
Zeeman term in the spin-Hamiltonian is much greater than the
contribution of the ligand-field part. In the smectic mesophase
of both Tb(III)- and Er(III)-containing liquid crystals, we
observed a typical “powder-type” EPR spectrum with a fine
structure, consisting of two sets of EPR peaks with “parallel”
and “perpendicular” features. The former may be assigned to
the symmetry axis of the crystal field, the latter to the plane
perpendicular to this axis. The positions of the EPR lines of
the Gd(III) complex were interpreted by a spin-Hamiltonian with
(20) Bikchantaev, I.; Galyametdinov, Yu.; Kharitonova, O.; Ovchinnikov,
IV; Bruce, D. W.; Dunmur, D. A.; Guillon, D.; Heinrich, B. Liq. Cryst.
1996, 20, 489.
7
effective electron spin S ) /2 and assuming a crystal-field
(21) Stephens, E. M. Gadolinium as an EPR probe. In Lanthanide Probes
in Life, Chemical and Earth Sciences, Theory and Practice; Bu¨nzli, J. C.
G., Choppin, G. R., Eds.; Elsevier: Amsterdam, 1989; Chapter 6, p 181.
(22) Abragam, A.; Bleaney, B. Electron Paramagnetic Resonance of
Transition Ions; Clarendon Press: Oxford, 1970.

4342 J. Am. Chem. Soc., Vol. 122, No. 18, 2000
Binnemans et al.
symmetry C3h, as in Gd(III) ethyl sulfate.23 The spin-Hamilto-
An interesting feature of the liquid-crystalline rare-earth
complexes (SmA phase) is their ability to be oriented in the
mesophase by an applied magnetic field, due to their large
magnetic anisotropy. The sign of magnetic anisotropy ∆ø
depends on the choice of the rare-earth ion and this sign
determines whether the metal complexes orient with the director
parallel to the magnetic field (∆ø > 0) or perpendicular to the
magnetic field (∆ø < 0). Although the value of ∆ø depends on
the crystal-field perturbation of the rare-earth ion, the sign of
∆ø is a function of the electronic angular momenta of the rare-
earth ion, and is constant for a given rare-earth ion.25 A positive
magnetic anisotropy (∆ø > 0) is expected for Eu(III), Er(III),
Tm(III), and Yb(III), and a negative magnetic anisotropy (∆ø
< 0) for Pr(III), Nd(III), Tb(III), Ho(III), and Dy(III). No
measurable magnetic anisotropy is expected for La(III), Gd(III),
and Lu(III) ions (except from diamagnetic contributions of the
ligands), and a small positive anisotropy for Sm(III). These
predictions were experimentally proved in this study for Pr-
(III), Nd(III), Gd(III), Tb(III), and Er(III) complexes. The
orientational behavior of the rare-earth ions is in contrast to
what is observed for other paramagnetic metallomesogens. All
vanadyl-containing metallomesogens orient with the director
parallel to the magnetic field (∆ø > 0). The orientational
behavior of the copper(II) complexes depends on the number
of aromatic rings present in the complex (4 or 6): ∆ø > 0 for
6-ring systems. For 4-ring systems the situation is more
complicated: ∆ø > 0 or ∆ø < 0 depending on the molecular
structure. For the rare-earth complexes the orientational behavior
is determined by the paramagnetic contribution of the metal
ion to ∆ø only, because this contribution is much larger than
the diamagnetic contributions of the aromatic rings in the
complex.
0
0
3
nian parameters are g ≈ 1.98, B2 ≈ 60 × 10-4 cm-1. B4 , B4 ,
0
3
B6 , and B6 are too small to be determined from such
insufficiently resolved EPR data. This set of parameters implies
that the crystal field is practically axially symmetric, the mean
orientation of this axis corresponding to the mesophase direc-
tor.20 The intensity of each (“parallel” or “perpendicular”) peak
may be roughly considered as proportional to the number of
molecules directed along the external magnetic field with the
molecular axis assigned to this peak. From the relative intensities
of the different kinds of peaks, one can conclude which
molecular axis (long or short axis) is predominantly directed
along the magnetic field at a given orientation of the sample.
The most obvious and convenient way to examine the orientation
in a magnetic field is to compare the “ ” -3/2 f -1/2 peak
with the 1/2 f -1/2 one (marked in Figures 6 and 7). From a
theoretical point of view, the intensity of the central 1/2 f -1/2
peak must be constant for different orientations of the sample,
because in a first approximation this transition is isotropic in
nature. This ensures that any changes in the intensity of the
“ ” -3/2 f -1/2 peak relatiVe to the 1/2 f -1/2 peak can
be considered as reflecting the orientational behavior of the
molecules in a magnetic field. When γ varies from 0 to 90°,
the intensity of the “ ” -3/2 f -1/2 peak appreciably falls
down relative to the 1/2 f -1/2 peak for the Tb(III) compound,
whereas the intensity increases for the Er(III) compound, proving
unambiguously that the molecular alignments of the Tb(III) and
Er(III) compounds are opposite. The “ ” -3/2 f -1/2 peak is
associated with a plane perpendicular to the C3-axis of the crystal
field (long molecular axis). The intensity of this transition may
be considered as proportional to the number of molecules
aligned with the long axis perpendicular to the external magnetic
field at any particular orientation of the sample. Since the
intensity of “ ” -3/2 f -1/2 peak at the initial orientation of
the sample (the sample was aligned upon cooling in a magnetic
field; γ ) 0°) is maximal for the Tb(III) compound, we can
conclude that the Tb(III) complexes tend to orient perpendicu-
larly to the magnetic field. The opposite conclusion follows for
the Er(III) complex. The magnetic anisotropy of mesophase is
positive for Er(III) and negative for Tb(III)-containing liquid
crystals. The angular dependence of the other peaks in the EPR
spectra confirms these conclusions.
Conclusions
In this paper, we described rare-earth-containing metallome-
sogens with 4-alkoxy-N-alkyl-2-hydroxybenzaldimine Schiff
bases as ligand. The ligands are not liquid crystals themselves,
but complex formation with rare-earth nitrates induced meso-
morphism. All the complexes (except those with very short alkyl
chains) display a viscous smectic A phase. It was shown the
rare-earth complexes have the composition [Ln(LH)3(NO3)3],
with the Schiff base ligands in a zwitterionic form. Due to their
large magnetic anisotropy, the paramagnetic liquid crystals can
be aligned in an external magnetic field. Depending on the
choice of the rare-earth ion, the complexes will align with their
long molecular axis either parallel or perpendicular to the
magnetic field. Therefore, the direction of orientation can be
changed simply by replacing the rare-earth ion by another one,
without the need for modifying the ligands themselves. Although
we demonstrated the possibility of magnetic alignment, the
actual switching rate is slow, because of the high viscosity of
the complexes. A next step is to try to reduce the viscosity and
to lower the transition temperatures without altering the thermal
stability too much. The high viscosity is not only a disadvantage,
because the mesophase can be supercooled to a glassy me-
sophase, forming an anisotropic glass. When the glass formation
is carried out in a magnetic field, anisotropic magnetic materials
are obtained.
Discussion
Previously, the stoichiometry of the metal complexes was
claimed as [LnL(LH)2(NO3)2]7 with a metal:ligand ratio of 1:3,
but with only one of the Schiff base ligands deprotonated. In
this study, we show that the stoichiometry should be reformu-
lated as [Ln(LH)3(NO3)3], with the Schiff base ligands in a
zwitterionic form and coordinating to the metal ion via the
negatively charged phenolic oxygen only. The three nitrate
groups coordinate in a bidentate fashion. The coordination
number of the lanthanide ion is nine, and the coordination
polyhedron can be described as a distorted monocapped square
antiprism. Spectroscopic and molecular modeling data indicate
that the first coordination sphere around the lanthanide ion is
not changed by elongation of the alkyl chains. The Schiff base
complexes have the advantage of being much easier to
synthesize than other rare-earth-containing compounds which
could be the starting point of mesomorphic materials, such as
complexes of 2,6-bis(benzimidazol-2′-yl)pyridines.24
Experimental Section
1
Techniques. H NMR spectra were obtained on a Bruker AMX-
400 spectrometer (400 MHz), a Bruker WM-250 spectrometer (250
MHz), or on a Varian UNITY-300 spectrometer (300 MHz) using
(23) Low, W. Paramagnetic Resonance in Solids; Academic Press: New
York, 1960.
(24) Nozary, H.; Piguet, C.; Tissot, P.; Bernardinelli, G.; Bu¨nzli, J. C.
G.; Deschenaux, R.; Guillon, D. J. Am. Chem. 1998, 120, 12274.
(25) Bleaney, B. J. Magn. Reson. 1972, 8, 91.

Rare-Earth-Containing Magnetic Liquid Crystals
J. Am. Chem. Soc., Vol. 122, No. 18, 2000 4343
CDCl3 as solvent and tetramethylsilane (TMS) as internal standard.
13C NMR solution spectra were measured on Varian UNITY-300
spectrometer (75.43 MHz). 13C CP/MAS NMR spectra (75.43 MHz)
in the solid phase were obtained on a Varian UNITY-300 spectrometer,
with a Doty Scientific probe. The samples were contained in 7 mm
diameter zirconia or silicium nitride rotors with Kel-F end caps.
Spinning rates were between 2 and 3 kHz. Chemical shifts were
measured relative to adamantane as external standard. IR spectra were
recorded on a Perkin-Elmer 580B spectrometer with the sample pressed
in a KBr-pellet technique, mixed with vaseline to a mull, or dissolved
in CCl4. Elemental analyses (CHN) were obtained on a CE-Instrument
EA-1110 elemental analyzer. Differential scanning calorimetry (DSC)
measurements were carried out on a Mettler-Toledo DSC821e module
(scan rate of 10 °C min-1 under a nitrogen flow). Optical textures of
the mesophases were observed with an Olympus BX60 polarizing
microscope equipped with a LINKAM TMS600 hot stage and a
LINKAM TMS93 programmable temperature-controller. Thermogravi-
metric analysis was done with a Polymer Laboratories STA1000H TG-
DTA apparatus, using a static air atmosphere. The temperature
dependence of the magnetic susceptibilities in the range 4.2-300 K of
the powdered metallomesogens were recorded using a Faraday type
magnetometer. Data collection was done via a PC-controlled Cahn
D-200 microbalance and a Bruker B-MN 200/60 power supply.26 The
applied magnetic field had a strength of 12 000 G. The magnetic
susceptibilities of the compounds in the temperature range 300-500
K were obtained using a home-built apparatus.27 The molar susceptibili-
ties were corrected for the underlying diamagnetism applying Pascal’s
scheme.28 The ø(T) fits were done with computer programs developed
in Darmstadt. The magnetic birefringence was measured on a previously
described setup.29 The magnetic field strength ranged between 0 and
20 000 G, and the magnetic field was applied perpendicularly to the
laser beam. A He-Ne laser (λ ) 632.8 nm) was used as the light source.
The EPR experiments were performed on a Q-band EPR spectrometer
RE-1308 (34 GHz), with an external magnetic field strength of 14 000
G, and with a temperature controller capable of stabilizing the
temperature within less than 0.2 K. Alignment was achieved by slow
cooling of the sample from the isotropic to smectic A phase in the
magnetic field. A copper-constantan thermocouple inserted into a
sample tube and positioned just above the sample surface was used
for temperature measurements. Organic reagents were obtained from
ACROS, Aldrich, Sigma, or Fluka. Hydrated rare earth nitrates were
purchased from Aldrich. The molecular modeling calculations were
performed using HyperChem 5.02 for Windows 95/NT (force field
MM+).
mL) and a few drops of glacial acetic acid were added as the catalyst.
Immediately after mixing the reagents, the solution turned yellow. The
mixture was heated during 3 h at reflux. The solution was left to cool
to room temperature and the Schiff base precipitated. The crude Schiff
base was recrystallized twice from absolute ethanol. The crystallization
was completed in an ice bath. The yellow crystals were filtered on a
Bu¨chner funnel, washed with cold absolute ethanol, and dried in vacuo.
1
Yield: 4.21 g (43%). H NMR (δ, CDCl3, 250 MHz): 0.88 (t, 3H,
CH3), 1.10-1.35 (m, 30H, CH2), 1.38 (t, 3H, CH3), 1.67 (m, 2H,
NCH2CH2), 3.50 (t, 2H, NCH2), 4.04 (t, 2H, OCH2), 6.30 (d, 1H, H-aryl,
Jm ) 2.4 Hz), 6.36 (dd, 1H, H-aryl, Jo ) 8.5 Hz, Jm ) 2.4 Hz), 7.05
(d, 1H, H-aryl, Jo ) 8.5 Hz), 8.09 (s, 1H, CHdN), 14.19 (s, 1H, OH).
(b) Preparation of 2-[(Octadecylimino)methyl]-5-(octyloxy)phe-
nol, L9H. 4-Octyloxy-2-hydroxybenzaldehyde was prepared by reflux-
ing for 3 h 2,4-dihydroxybenzaldehyde (50 mmol, 6.91 g) with
1-bromooctane (50 mmol, 9.66 g) in DMF, with KHCO3 (50 mmol,
5.01 g) as the base. After the reaction mixture was cooled to room
temperature it was poured into an aqueous HCl solution (6 N). The
organic layer was separated and the aqueous layer was extracted with
diethyl ether. The combined organic layers were dried by anhydrous
MgSO4 and the solvent was removed under reduced pressure. The crude
aldehyde was purified on a silica column with n-heptane/ethyl acetate
(90:10) as the eluent. Yield: 73% (9.14 g). 1H NMR (400 MHz, CDCl3,
δ ppm): 0.88 (t, 3H, CH3), 1.00-1.55 (m, 10H, CH2), 1.79 (quintet,
2H, -CH2-CH2-O-), 4.00 (t, 2H, -CH2-O-), 6.40 (d, 1H, H-aryl),
6.52 (dd, 1H, H-aryl), 7.41 (d, 1H, H-aryl), 9.70 (s, 1H, CHdO, Jo )
8.5 Hz, Jm ) 2.5 Hz).
The Schiff base L9H was prepared by refluxing 9.14 g (36.5 mmol)
of 4-octyloxy-2-hydroxybenzaldehyde with 9.84 g (36.5 mmol) of
n-octadecylamine in absolute ethanol for 3 h, with a few drops of glacial
acetic acid as the catalyst. The crude product was crystallized several
1
times from hot absolute ethanol. Yield 79% (14.5 g). H NMR (400
MHz, CDCl3, δ ppm): 0.88 (t, 6H, CH3), 1.25-1.44 (m, 40H, CH2),
1.67 (quintet, 2H, -CH2-CH2-N), 1.77 (quintet, 2H, -CH2-CH2-
O-), 3.50 (t, 2H, -CH2-N-), 3.95 (t, 2H, -CH2-O-), 6.30 (dd,
1H, H-aryl, Jm ) 2.5 Hz), 6.34 (d, 1H, H-aryl, Jo ) 8.5 Hz, Jm ) 2.5
Hz), 7.04 (d, 1H, H-aryl, Jo ) 8.5 Hz), 8.08 (s, 1H, CHdN), 14.08 (s,
1H, OH).
(c) Preparation of 2-[(Octadecylimino)methyl]-5-(tetradecyloxy)-
phenol (34), L15H. 4-Tetradecyloxy-2-hydroxybenzaldehyde was pre-
pared by refluxing for 3 h 2,4-dihydroxybenzaldehyde (20.73 g, 0.15
mol) with 1-bromotetradecane (41.59 g, 0.15 mol) in DMF, with
KHCO3 (15.04 g, 0.15 mol) as the base. After the reaction mixture
was cooled to room temperature it was poured into an aqueous HCl
solution (6 N). The organic layer was separated and the aqueous layer
was extracted with diethyl ether. The combined organic layers were
dried by anhydrous MgSO4 and the solvent was removed under reduced
pressure. The crude aldehyde was purified by crystallization from hot
acetonitrile. Yield: 54% (27 g). 1H NMR (250 MHz, CDCl3, δ ppm):
0.90 (t, 3H, CH3), 1.0-1.6 (m, 22H, CH2), 1.80 (quintet, 2H, -CH2-
CH2-O-), 4.00 (t, 2H, -CH2-O-), 6.40 (d, 1H, H-aryl), 6.50 (dd,
1H, H-aryl), 7.40 (d, 1H, H-aryl), 9.70 (s, 1H, CHdO).
Synthesis. (a) Preparation of 5-Ethoxy-2-[(octadecylimino)meth-
yl]phenol, L3H. 2,4-Dihydroxybenzaldehyde (50 mmol, 6.91 g) was
dissolved in ethanol (100 mL). To the solution was added KOH (50
mmol, 2.81 g) and iodoethane (50 mmol, 7.80 g). The mixture was
heated for 15 h at 60 °C (oil bath) with stirring, in a nitrogen
atmosphere. After the reaction the dark brown reaction mixture was
left to cool to room temperature, water (200 mL) was added, and
4-ethoxy-2-benzaldehyde was removed from the reaction mixture by
steam distillation. The milky emulsion of the aldehyde in water (2 L)
was extracted by diethyl ether (4 × 200 mL). The combined organic
layers were dried by anhydrous MgSO4 and the solvent was removed
on a water bath using a rotavap at ambient pressure. 4-Ethoxy-2-
hydroxybenzaldehyde was obtained as a yellow oil, which crystallized
to a beige solid by leaving the oil overnight in a freezer. Yield: 3.87
g (47%). 1H NMR (δ ppm, CDCl3, 250 MHz): 1.43 (t, 3H, CH3), 4.08
(quart, 2H, CH2O), 6.40 (d, 1H, H-aryl, Jm ) 2.3 Hz), 6.52 (dd, 1H,
H-aryl, Jo ) 8.6 Hz, Jm ) 2.3 Hz), 7.42 (d, 1H, H-aryl, Jo ) 8.5 Hz),
9.70 (s, 1H, CHdO), 11.45 (s, 1H, OH).
The Schiff base L15H was prepared by refluxing 16.73 g (50 mmol)
of 4-tetradecyloxy-2-hydroxybenzaldehyde with 13.50 g (50 mmol) of
n-octadecylamine in n-heptane for 3 h, with a few drops of glacial
acetic acid as the catalyst. The crude product was crystallized several
1
times from hot n-heptane. Yield 40% (11.6 g). H NMR (400 MHz,
CDCl3, δ ppm): 0.88 (t, 6H, CH3), 1.10-1.50 (m, 52H, CH2), 1.67
(quintet, 2H, -CH2-CH2-N), 1.77 (quintet, 2H, -CH2-CH2-O-),
3.50 (t, 2H, -CH2-N-), 3.95 (t, 2H, -CH2-O-), 6.32 (dd, 1H,
H-aryl, Jm ) 2.5 Hz), 6.36 (d, 1H, H-aryl, Jo ) 8.5 Hz, Jm ) 2.5 Hz),
7.05 (d, 1H, H-aryl, Jo ) 8.5 Hz), 8.08 (s, 1H, CHdN), 14.1 (s, 1H,
OH). Calcd for C39H71NO2: C 79.94, H 12.21, N 2.39. Found: C 80.23,
H 12.16, N 2.30.
4-Ethoxy-2-hydroxybenzaldehyde (3.85 g, 23.2 mmol) and octade-
cylamine (6.24 g, 23.2 mmol) were dissolved in absolute ethanol (200
(d) Synthesis of La(III) Complex [La(L9H)3(NO3)3] (5). Compound
5 was prepared by adding a solution of 0.87 g (2.00 mmol) of La-
(NO3)3‚6H2O in absolute ethanol dropwise to a stirred solution of 1.00
g (2.00 mmol) of the ligand L9H in absolute ethanol, at a temperature
not higher than 50 °C. After ca. 15 min, the solution turned cloudy,
indicating that the product started to precipitate. Stirring was continued
for at least 3 h. The pale yellow precipitate was filtered on a crucible,
(26) Gehring, S.; Fleischhauer, P.; Paulus, H.; Haase, W. Inorg. Chem.
1993, 32, 54.
(27) Merz, L.; Haase, W. J. Chem. Soc., Dalton Trans. 1980, 875.
(28) Weiss, A.; Witte, H. Magnetochemie; Verlag Chemie: Weinheim,
1973.
(29) Abdullin, I. R.; Vul’sfon, S. G.; Kazakova, E. Kh Bull. Acad. Sci.
USSR, DiV. Chem. Sci. 1987, 36, 2551.

4344 J. Am. Chem. Soc., Vol. 122, No. 18, 2000
Binnemans et al.
washed with ice-cold absolute ethanol, and dried in vacuo. Yield 84%
(1.00 g). H NMR (250 MHz, CDCl3, δ ppm): 0.88 (m, 6H, CH3),
The crystals had a greenish yellow color. The data were measured at
16 °C on a Siemens P4-Pc diffractometer (Mo KR radiation) using a
crystal of size 0.20 × 0.25 × 0.35 mm3. The unit cell parameters were
obtained by a least-squares analysis of a diffractometer angular setting
of 35 well-centered reflections in the range 7° < θ < 13 °. Intensities
of 7708 reflections with 2° < θ < 25° were measured by the ω-scan
method and corrected for absorption (empirical correction based on
ψ-scans for 12 reflections). The structure was solved by direct methods
and refined by full-matrix least-squares on F2 using the SHELXTL
program package.32 Non-hydrogen atoms were anisotropically refined
and the hydrogen atoms in the riding mode allowed the C-H distances
to be refined (except the imino hydrogens, which were localized in the
difference Fourier synthesis). The isotropic temperature factors were
fixed at 1.2 times U(eq) of the parent atoms. The final R-values were
R1 ) 0.0376 for the observed reflections and wR2 ) 0.0971 for all
1
1.10-1.40 (m, 40H, CH2), 1.62 (m, 4H, N-CH2-CH2 and CH2-CH2-
O), 3.46 (t, 2H, N-CH2), 3.64 (t, 2H, CH2-O), 6.15 (dd, 1H, H-aryl),
6.33 (d, 1H, H-aryl), 6.90 (d, 1H, H-aryl), 7.60 (d, 1H, CHdN), 12.2
(s, 1H, NH, Jo ) 8.5 Hz, Jm ) 2.5 Hz). IR (KBr-pellet, cm-1): 3200
(m, ν(NH)), 2920, 2860 (s, aliphatic C-H stretch), 1662 (s, ν(CdN)),
1530 (s, ν(CdC)), 1475 (s, ν4(NO3)), 1283 (s, ν(CPh-O) and/or ν1-
(NO3)), 1230 (s, νas(C-O-C)), 1030 (s, νs(C-O-C)), 590 (m, ν (La-
O)). Calcd for C99H177N6O15La: C 64.96, H 9.75, N 4.59. Found: C
64.94, H 9.87, N 4.41.
All the other rare-earth complexes were synthesized in an analogous
way.
X-ray Crystal Structure Determination of [Dy(L1H)3(NO3)3] (1).
Single crystals were obtained by slow isothermal evaporation of a
solution of the complex in an ethanol/chloroform (1:1) mixture. The
crystals had a yellow color. The data were measured at room
temperature (20 °C) on an Enraf-Nonius CAD4 diffractometer (Mo
KR radiation) using a crystal of size 0.2 × 0.2 × 0.3 mm3. The unit
cell parameters were obtained by the least-squares analysis of a
diffractometer angular setting of 25 well-centered reflections in the
range 14° < θ < 15°. Intensities of 13 477 reflections with 2° < θ <
26° were measured by the ω/θ-scan method, 10 305 of them with I >
3σ. An empirical absorption correction was applied (7 reflections with
ø g 80° were measured by rotation around the ψ-vector with a step of
10°). The structure was solved by a direct method using the SIR
program.30 The difference Fourier synthesis revealed the positions of
all hydrogen atoms. The structure was refined by the full-matrix least-
squares method with anisotropic temperature factors for non-hydrogen
atoms and isotropic temperature factors for hydrogen atoms. The final
difference electron density map showed minimum and maximum values
data (weighting scheme w-1 ) σ2(Fo ) + (0.0564P)2 + 0.67P, where
2
2
P ) [Max(Fo ,0) + 2Fc2]/3). A summary of the crystal data, intensity
measurements, and structure refinements is reported in Table 2.
Acknowledgment. K. Binnemans is Postdoctoral Fellow and
L. Van Meervelt is Research Director of the F.W.O.-Flanders
(Belgium). K. Binnemans thanks the F.W.O. for a “Krediet aan
Navorsers” and Prof. C. Go¨rller-Walrand for laboratory facilities
(GOA 98/03 and F.W.O. research grant G.0243.99). R. Van
Deun (specialisatiebursaal IWT) is indebted to the “Flemish
Institute for the Encouragement of Scientific and Technological
Research in the Industry (IWT)” for financial support. K.
Binnemans, R. Van Deun and D.W. Bruce wish to thank the
British Council and the F.W.O.-Flanders for travel grants
(Academic Research Collaboration Program). W. Haase is
grateful to the Deutsche Forschungsgemeinschaft for support.
D.W. Bruce, Yu. G. Galyametdinov, A. Polishchuk, A.V.
Prosvirin and W. Haase are receiving support from the INTAS-
project (contract INTAS 96-1198). K. Binnemans and Yu. G
Galyametdinov wish to thank NATO for a collaborative linkage
grant (PST.CLG.974907). Ms. L. Jongen (K.U.Leuven) is
acknowledged for TG-DTA measurements and Ms. P. Bloemen
(K.U.Leuven) for CHN microanalyses.
2
of 0.07(3) and 0.32(3) eÅ-3. The weighting scheme w-1 ) 4Fo /[σ(I)2
2
+ (0.04Fo )]2 was used in the later stages of the refinement. The final
values of the discrepancy indices were R ) 0.025 and Rω ) 0.030. All
calculations were made on the DEC Alpha Station 200 computer by
means of the program package MolEN.31 A summary of the crystal
data, intensity measurements, and crystal structure refinements is
reported in Table 2.
X-ray Crystal Structure Determination of [Nd(L1H)3(NO3)3] (2).
Single crystals were obtained by slow isothermal evaporation of a
solution of the complex in a hexane/dichloromethane (1:1) mixture.
Supporting Information Available: CHN elemental analy-
sis data and melting points of the ligands (Sup-1, PDF) and the
CHN data of the complexes (Sup-2) (PDF). An X-ray crystal-
lographic file (CIF). This material is available free of charge
via the Internet at http://pubs.acs.org.
(30) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Viterbo, D. Acta
Crystallogr. 1991, A47, 744.
(31) Straver, L. H.; Schierbeek, A. J. MolEN. Structure Determination
System. Nonius B.V.; 1994, Vols. 1 and 2.
(32) SHELXTL-PC, Manual Version 5.1, Bruker Analytical X-ray
Systems Inc.: Madison, Wisconsin, 1997.
JA993351Q