Pioneering work in the field of luminescence by lanthanidomesogens in the liquid-crystal state has been done by Bu¨nzli and coworkers. In a seminal paper, they monitored the luminescence intensity and the excited state
mesogenic pendant arms (ligandL46, seeFig. 59) as a function of the temper- ature, with the aim to detect phase transitions (Suarez et al., 2003). The integrated intensities of the5D0!7F2transitionIobsand the Eu(5D0) lifetime tobswere found to decrease with increasing temperatures due to more efficient nonradiative relaxation of the excited state at higher temperatures. The ln(tobs/ t295K) versus 1/Tand ln(Iobs/I295K) versus 1/Tcurves showed upon heating a sigmoidal shape, with a marked variation at the melting point. Therefore, the luminescence measurements allowed to accurately detect the transition of the crystalline state to the hexagonal columnar phase during the first heat- ing process. The corresponding ln(Iobs/I295K) versus 1/T curve cooling curve was quite monotonic, whereas the ln(tobs/t295K) versus 1/T cooling curve closely followed the variation observed during heating. By monitoring the integrated intensity of the 5D4!7F5 transition of Tb3þ, it was possible to determine the crystal-to-mesophase transition in a terbium-containing metal- lomesogen (Suarez et al., 2004; Terazzi et al., 2005). Further luminescence studies on R(NO3)3 complexes of 2,6-bis(1-ethylbenzimidazol-2-yl)pyridine derivatives synthesized by Piguet and coworkers showed that the transition of a crystalline phase to a cubic mesophase could be detected by measurement of the luminescence properties during the first heating process (Terazzi et al., 2006a). However, the vitrification of the cubic mesophase upon cooling could not be observed. This shows that this luminescence technique cannot be applied to the detection of glass transitions.
Yang et al.(2006)reported on low-melting lanthanidomesogens consisting of Lewis-base adducts of a nonmesomorphic salicylaldimine Schiff’s base ligand L and tris(2-2-thenoyltrifluoroacetonato)lanthanide(III) or tris(benzoyl- trifluoroacetonate) lanthanide(III) complexes, [R(tta)3L2] or [R(bta)3L2].
These compounds form a smectic A phase at room temperature. The lumines- cence spectra of the Eu(III), Sm(III), Nd(III) and Er(III) complexes have been measured in the mesophase (Fig. 62). Of interest is the fact that near-infrared emission by Nd(III) and Er(III) could be observed. Plots of the luminescence decay time of the5D0level of [Eu(tta)3L2] or the4G5/2level of [Sm(tta)3L2] as a function of the temperature show a gradual decrease of the luminescence lifetime with increasing temperatures. A distinct lowering could be observed in these curves at the SmA!I transition (Fig. 63). The intrinsic quantum yield of the complexes [Eu(tta)3L2] and [Eu(bta)3L2] have been determined by photoacoustic methods (Li et al., 2008b; Yang et al., 2008a,b). Although glass transitions of the vitrified mesophase could not be detected by lumines- cence measurements, the photoacoustic spectra clearly showed this transition.
Yin et al. (2009)observed luminescence in the mesophase for benzoic-acid- terminated surfactant complexes of the europium-containing polyoxo- metalates [Eu(W10O36)]9, [Eu(PW11O39)2]11, and [Eu(SiW11O39)2]13. A review of the luminescence of metallomesogens (including lanthanidome- sogens) in the liquid-crystal state was published byBinnemans (2009).
675 700 725 bta
tta
650 Wavelength (nm)
Intensity (a.u.)
625 600 575
FIGURE 62 Luminescence spectra of [Eu(tta)3L2] (gray curve) and [Eu(bta)3L2] (black curve) as a thin film in the mesophase at 25C. The excitation wavelength was 370 nm. All transitions start from the5D0level and end at the differentJlevels of the7F term (J¼0–4 in this spectrum).
Adapted fromYang et al. (2006).
10 0.1 0.2 0.3 0.4 0.5
20 30 40
Temperature (°C)
Luminescence lifetime (ms)
50 60 70
FIGURE 63 Luminescence decay time of the5D0level of the [Eu(tta)3L2]complex as a function of the temperature. The luminescence was monitored at 613 nm (5D0!7F2line) and the excita- tion wavelength was 370 nm. The measurements were made during cooling of the sample. The transition SmA$I (clearing point) can be observed as a jump in the curve at about 60C.
Adapted fromYang et al. (2006).
smectic A and nematic phases (Galyametdinov et al., 2008). The ligand was a rather unusual b-diketone with a cyclohexyl ring (Fig. 44, see also Sec- tion 6). When a thin film of the corresponding europium(III) complex in the nematic phase was supercooled to a glass phase in a liquid-crystal cell with alignment layers for planar alignment, polarized luminescence could be observed for the aligned samples (Fig. 64). The lanthanide complexes were also well soluble in nematic liquid-crystal mixtures. The polarization effects were observable as differences in the intensities of crystal-field transitions for the two mutually perpendicular polarization directions. The energy trans- fer in mesogenic europium(III) b-diketonate complexes with 5,50-diheptade- cyl-2,20-bipyridine as coligand was studied byLapaev et al. (2008, 2009).
Another approach to obtain luminescent lanthanide-containing LCs is by dissolving a luminescent lanthanide complex in a suitable liquid-crystal host matrix (guest–host approach). The advantage of this method is that the lumi- nescence and mesomorphic properties can be optimized independently. This allows easier access to nematic lanthanide-containing LC mixtures. However, one is often facing the poor solubility of the molecular lanthanide complexes in the liquid-crystalline host matrix. The guest–host concept involving lumi- nescent lanthanide complexes was first applied by Yu and Labes (1977) who doped the nematic LC 4-n-pentyl-40-cyanobiphenyl (5CB) with euro- pium(III) 2-thenoyltrifluoroacetonate trihydrate, Eu(tta)33H2O. The LC
580 590 600 610 620 630 640
Intensity (a.u.)
Wavelength (nm)
FIGURE 64 Polarization effects in the room-temperature luminescence spectra of an aligned supercooled thin film of the europium(III) complex of the type shown inFig. 44. The polarizer is either parallel (gray line) or perpendicular (black line) to the alignment layers in the liquid- crystal cell.Adapted fromGalyametdinov et al. (2008).
mixture gave an intense red photoluminescence with an emission maximum at 612 nm, when excited with UV light. Doping the LC mixture with different chiral additives such as cholesteryl nonanoate or cholesteryl chloride, allowed obtaining luminescent chiral LC mixtures. The chiral LC mixtures in a cell of indium-tin-oxide coated glasses and mylar spacers experienced a cholesteric to nematic phase transition upon application of an electric potential. The homeotropically aligned nematic phase scatters much less light than the cho- lesteric phase and can absorb the excitation light less efficiently. As a result the luminescence intensity of the europium(III) complex in the cholesteric LC phase is much higher than in the nematic phase. Contrast ratios as high as 9:1 could be observed for the cholesteric phase (off state) in comparison with the nematic phase (on state). The off-on response time was about 100 ms (Labes, 1979). The actual values of the contrast ratios were dependent on the temperature, the cell thickness, the pitch of the cholesteric phase and the concentration of the europium complex. However, the polarization of the luminescence was found to be very small. Tb(tta)33H2O did not show luminescence in the same LC mixture, but the corresponding complex of ben- zoylmethane did exhibit a weak green luminescence (Labes, 1979).Larrabee (1976)reported luminescence for a europium tetrakis b-diketonate complex dissolved in a commercial nematic LC mixture with a nematic phase range between 26 and 88C.Zolin et al. (1980)studied the luminescence of differ- ent types of europium(III) complexes dissolved in mixtures of 5CB and cho- lesteryl nonanoate, in pure cholesteryl nonanoate, as well as in commercial nematic LC mixtures. They authors also investigated the luminescence prop- erties of suspensions of the europium phosphor Y2OS2:Eu3þin these LC mix- tures. The differences in the luminescence intensities in the different liquid mixtures were attributed to changes in the light scattering: the stronger the light scattering by the LC, the more intense the luminescence. Strong light scattering by the LC can be considered as an increase of the effective optical path length for the excitation radiation. This leads to a stronger light absorp- tion and thus to a brighter luminescence. A 50–100% decrease of the lumines- cence decay times of the europium chelate complexes was observed for the solid-to-nematic transition and for the nematic-to-isotropic transition. The decrease of the decay times were explained by a more efficient quenching of the excited states in solvents with a lower viscosity. Small changes were observed for the relative intensities of crystal-field transitions for the same complex dissolved in different LC host matrices. These differences were explained by orientational effects of the chelate molecules in the LC solvent.
Binnemans and Moors (2002)showed by high-resolution luminescence spec- troscopy that well-resolved crystal-field fine structures could be observed for [Eu(tta)3(phen)] doped in the liquid-crystal hosts N-(4-methyloxybenzyli- dene)-4-butylaniline (MBBA) and 5CB (Fig. 65). The spectra are more remi- niscent to what is observed for europium(III) ions doped in crystalline hosts rather than europium(III) complexes dissolved in organic solvents (Fig. 66).
Binnemans and coworkers were the first to observe near-infrared lumines- cence from lanthanide-doped LC mixtures (Van Deun et al., 2003a). They studied the spectroscopic properties of the lanthanide(III) b-diketonate com- plexes [R(dbm)3(phen)], where R¼Nd, Er, Yb, and dbm is dibenzoyl- methane, in the LC MBBA. By incorporation of an erbium(III)-doped nematic LC (ErCl3dissolved in E7) in the pores of microporous silicon, nar- rowing of the erbium(III) emission band in the near-infrared was observed
N
N
O CF3
O Eu
S 3 [Eu(tta)3(phen)]
MBBA 5CB
C N C5H11
H3CO
N C4H9
FIGURE 65 Structures of the nematic liquid-crystal host matrices MBBA and 5CB, and struc- ture of the europium(III) complex [Eu(tta)3(phen)].
650 Wavelength (nm)
Intensity (a.u)
675 700 725 625
600 575
FIGURE 66 Room temperature luminescence spectrum of [Eu(TTA)3(phen)] in the nematic liquid-crystal host MBBA (doping concentration: 4 wt.%.lexc¼396 nm).Adapted fromBinne- mans and Moors (2002).
(Weiss et al., 2007). Luminescent optically active LCs were obtained by doping Eu(tta)33H2O into a mixture of cholesteryl nonanoate, cholesteryl tet- radecanoate and the ternary LC mixture ZLI 1083 from Merck (Boyaval et al., 1999, 2001). To improve the solubility of lanthanide complexes in chiral LC mixtures, chiral dopants with coordinating groups have been developed (Hapiot, 2006). Up to 6 mol.% of R(tta)3(R¼Nd, Sm, Eu) could be dissolved in the LC matrix. Long alkyl chains on the dopant led to a higher dopant sol- ubility in the LC solvent.Driesen et al.(2007)constructed a switchable near- infrared emitting LC cell. The authors doped the lanthanide complexes [Nd (tta)3(phen)] and [Yb(tta)3(phen)] into the chiral nematic liquid-crystal mix- ture of E7 and cholesteryl nonanoate. However, the contrast ratios were low. The photoluminescence intensity of europium(III) and terbium(III) com- plexes dissolved in nematic LC 4-(isothiocyanotophenyl)-1-(trans-4-hexyl) cyclohexane (6CHBT) strongly depends on the strength of the applied electric field (Palewska et al., 2007). This effect was tentatively assigned to a complex dipolar orientational mechanism in the LC matrix.Driesen et al. (2009)inves- tigated the luminescence spectra of different europium(III) complexes dis- solved in the LC 40-pentyl-4-cyanobiphenyl (5CB). Upon alignment of the europium(III)-doped nematic liquid-crystal host in a liquid-crystal cell with alignment layers, polarization effects were observed in the emission spectra.
These polarization effects were visible as differences in the relative intensities of the crystal-field components of the transitions. Although the europium(III) complexes do not need to be liquid-crystalline themselves, some structural anisotropy (rodlike shape) seems to required for good alignment in the liquid-crystal host and for generation of linearly polarized light. No polariza- tion effects were observed for unaligned bulk samples.
Driesen and Binnemans (2004) prepared glass-dispersed LC films doped with a europium(III) b-diketonate complex [Eu(dbm)3(gly)], where dbm is dibenzoylmethanate and gly is monoglyme (1,2-dimethoxyethane). The films consisted of tiny droplets of the LC 5CB in a silica-titania glass matrix, and were prepared via a sol-gel process. Below the clearing point, the films are strongly scattering light. When the film is heated above the clearing point, the refractive index of the isotropic liquid becomes similar to that of the glass and hardly any light scattering occurs. The luminescence intensity of the film was measured as a function of temperature. A sharp decrease in luminescence intensity was observed for the transition from the nematic phase to the isotro- pic phase (Fig. 67). This decrease in luminescence intensity at the clearing point was attributed to weaker light scattering and thus less efficient use of the excitation light in the isotropic state. Upon cooling of the sample an increase in light intensity was observed at the clearing point.
The LC host matrix is not restricted to nematic LCs. Bu¨nzli and coworkers dissolved different europium(III) salts in the smectogenic ionic LC 1-dodecyl- 3-methylimidazolium chloride [C12mim]Cl, which exhibits a smectic A phase between 2.8 and 104.4C (Guillet et al., 2004). It was observed that
concentrations of europium(III) salts as high as 10 mol.% do not appreciable alter the liquid-crystalline behavior of the host matrix. Interestingly, the emis- sion color of the europium(III)-containing LC mixture could be tuned from blue (emission color of the host matrix) to red (emission color of Eu3þ), depending on the excitation wavelength and the counter ion of europium (III). Intense near-infrared emission has been observed by doping the b-diketonate complexes [R(tta)3(phen)] (R¼Nd, Er, Yb) in this ionic LC host (Puntus et al., 2005). By doping the lanthanide bromides EuBr2, SmBr3, TbBr3, and DyBr3 in the ionic LC 1-dodecyl-3-methylimidazolium bromide [C12mim]Br, luminescent ionic LCs could be obtained (Getsis and Mudring, 2010). [C12mim]Br forms a smectic A phase between 42 and 100C. The doping concentrations of about 1 mol.% did not strongly influence the meso- phase behavior of [C12mim]Br. However, it was found that the mesophase of [C12mim]Br is stabilized by addition of larger amounts of the lanthanide bro- mides to such an extent that the samples stayed liquid-crystalline to tempera- tures below room temperature. [C12mim]Br itself showed bluish-white emission upon excitation wit UV light, due to the imidazolium cation. The samples doped with EuBr2 showed blue emission due to 4f–5d transitions, SmBr3-doped samples a red emission, TbBr3-doped samples green emission and DyBr3-doped samples an orange emission. Emission by SmBr3, TbBr3, and DyBr3is due to 4f–4f transitions. An interesting phenomenon is that the emission of the TbBr3-doped samples was tunable from bluish-white to green, depending on the excitation wavelength. The bluish-white emission originates from the imidazolium cation and the green emission is due to TbBr3. In a
0 25 50 75 100
20 25 30 35 40 45 50 50 45 40 35 30 25 20 0 25 50 75 100
Relative intensity (%)
Relative intensity (%)
Temperature (°C) Heating Cooling
FIGURE 67 Relative intensity change of the5D0!7F2transition upon heating and cooling of a glass-dispersed liquid crystal film doped with the europium(III) complex [Eu(dbm)3(gly)].
Adapted fromDriesen and Binnemans (2004).
similar way, the emission of the DyBr3-doped samples could be tuned from bluish-white to orange-yellow. If the 1-dodecyl-3-methylimidazolium cation was replaced by the N-dodecyl-N-methylpyrrolidinium cation, the emission became less intense, but the luminescence decay times became somewhat lon- ger (Getsis and Mudring, 2011). The possibility to tune the luminescence color from bluish-white to green by a change of the excitation wavelength is lost by doping TbBr3inN-dodecyl-N-methylpyrrolidinium bromide.
14.2. Magnetism
A LC can be macroscopically aligned by applying an external magnetic field.
The magnetic alignment arises from the anisotropic magnetic energy of the LCs (Yamaguchi and Tanimoto, 2006). Because of this anisotropy, the mag- netic field and the director of the mesophase are coupled. As a consequence, the LC experiences a torque in the magnetic field and orients itself into a direction where its magnetic energy is minimum. Isolated small molecules or ions do not undergo alignment in a magnetic field, because the coupling energy between a molecule (or ion) and the applied magnetic field is small compared to the thermal energy kBT at room temperature, and thermal agita- tion will destroy alignment effects. Aggregates or ordered assemblies (e.g., liquid-crystal phases) be magnetically aligned, since their anisotropic mag- netic energy exceeds the thermal agitation at room temperature. The molecu- lar orientation in a mesophase by a magnetic field is a cooperative effect involving more than 108molecules in a liquid-crystal microdomain (Alonso, 1996; de Gennes, 1974). The molecules in a mesophase microdomain have orientational order as a consequence of the anisotropic intramolecular interac- tions between the molecules. Alignment of mesophases by a magnetic field is of importance for the study of mesophases by XRD, because the information content of an X-ray diffractogram of an aligned mesophase is much higher than that of a corresponding diagram of an unaligned mesophase. Also for the measurements of anisotropy of physical properties (e.g., the dielectric per- mittivity, electrical conductivity or viscosity) aligned mesophases are required. An advantage of magnetic alignment over mechanical alignment by rubbing or by alignment layers is that alignment of the bulk sample rather than of the surface layers can be achieved, because the magnetic field lines can penetrate the material. The direction along which the magnetic field is applied can easily be controlled, so that alignment inclined to a surface is pos- sible. With alignment layers or with alignment by means of an electric field, only parallel or perpendicular alignment can be achieved.
For easy alignment of mesogens in a magnetic field a high magnetic anisotropy is required. Themagnetic anisotropyDwis the difference between the magnetic susceptibility parallel (w||) and perpendicular (w?) to the director n:~ Dw¼w||w?. The free energy for alignment of LCs in a magnetic field can be expressed as:
whereHis the applied magnetic field and n~is the director. The molar mag- netic anisotropy of most organic LCs is less than 100106cm3mol1. This value can be increased by incorporation of paramagnetic metal ions in the LC. Lanthanidomesogens are very attractive for this purpose as they can exhibit a very high magnetic anisotropy, up to one or two orders of magnitude larger than that of organic LCs. The orientation of LCs in a magnetic field depends on the sign and magnitude of the magnetic anisotropyDw. Depending on the sign ofDw, the directorn~of the LC is oriented parallel or perpendicular to the external magnetic field. If Dw is positive, the molecules are oriented with their molecular long axis parallel to the magnetic field. IfDwis negative, the molecules are aligned with their molecular long axis perpendicular to the magnetic field. A low viscous mesophase is desirable for getting a well aligned sample in a magnetic field. The overall magnetic anisotropy Dw is the sum of a diamagnetic (Dwdia) and a paramagnetic contribution (Dwpara).
For calamitic molecules, Dwdia is nearly always positive, but Dwpara can be either positive or negative. The diamagnetic contribution to the magnetic anisotropy of most paramagnetic molecules is large in comparison with the paramagnetic contribution. The diamagnetic contribution to the magnetic anisotropy can be calculated by an additive scheme (Pascal’s scheme) (Bou- dreaux and Mulay, 1976) and the paramagnetic contribution can be obtained by subtracting the calculated Dwdiavalue from the experimentally measured Dwvalue. In the case whereDwparaDwdiaand with the two contributions hav- ing an opposite sign, the resultingDwvalue will be very small, and an aligned monodomain is difficult to obtain by means of a magnetic field.
The situation is quite different in the case of lanthanidomesogens, because the trivalent lanthanide ions (especially Tb3þ, Dy3þ, Ho3þ, Er3þ, Tm3þ) have a very large magnetic anisotropy in comparison with other paramagnetic ions (e.g., Cu2þor vanadyl). In the case of the paramagnetic lanthanide ions (all rare-earth ions except La3þ, Lu3þ, Y3þ,and Sc3þ), the diamagnetic con- tributions toDwcan be neglected in comparison to the paramagnetic contribu- tion. An exception is the Gd3þion, which is magnetically isotropic due to the S ground state 8S7/2. The large magnetic anisotropy of lanthanidomesogens was discovered soon after the first calamitic lanthanide complexes had been prepared. Because the sign of the magnetic anisotropy depends on the lantha- nide ion, it is possible to obtain with the same ligand and with the proper choice of the lanthanide ion compounds which can either be aligned perpen- dicular or parallel to the magnetic field (Binnemans et al., 2001c; Galyamet- dinov et al., 2007). Analysis of the experimental magnetic susceptibility data shows that these compounds can be classified into two distinct groups, depending on the sign of Dw. The first group contains Ce3þ, Pr3þ, Nd3þ, Sm3þ, Tb3þ, Dy3þ,and Ho3þcompounds, while the second group contains Eu3þ, Er3þ, Tm3þ, and Yb3þ compounds (Galyametdinov et al., 2007;