1. Introduction

The reactive mesogen (RM) materials have been widely used in the display applications [1,2]. The chemical structure of the general RM materials consists of a mesogenic moiety and a reactive one [3,4]. The former results in the anisotropic physical properties like birefringence, while the latter enables solidification keeping the anisotropic properties. The RM materials have been used for various applications since 1990’s. For example, the stabilization of the liquid crystal (LC) orientation using the RM molecules were actively studied [5–7]. The LC pretilt angle was controlled using the RM materials for the construction of a multi-domain structure in the vertically-aligned (VA) liquid crystal display (LCD) [8]. In these days, the retarder film made from the RM molecules draw much attention for the compensation film of display devices [1,9,10].

The RM-based retarder films have several advantages compared to the traditional stretched-polymer films. First, the thickness of the RM film can be reduced about one tenth compared to the stretched-film keeping the same retardation. Second, the RM film can be fabricated by a coating process without the complicated stretching process. In addition, the orientation of the optic axis of the RM film can be easily controlled by photoalignment technique, while that is possible by a dual axis stretching for the stretched-film retarder. Based on the merits above, the stretched-film retarders are being gradually replaced by the RM retarders.

Various kinds of the RM materials have been synthesized and developed for the display compensation films [1,9,11–15]. The positive a-plate [11,12], negative a-plate [13], positive c-plate [14], and negative c-plate [15] materials have been reported. In addition, the hybrid-aligned discotic RM material has been widely used for the compensation of the twisted-nematic (TN) LCDs [9]. On the other hand, the biaxial retarder using the RM materials has been rarely reported [16]. For the efficiency of the optical compensation scheme, the biaxial retarder is much useful and has a great potential for various optical applications [17–19]. In this paper, we synthesized new kinds of T-shaped reactive materials with a biaxial shape. We mixed the T-shaped materials with rodlike RM molecules and investigated their optical properties. The chemical structures of the T-shaped molecules were varied and its effect on the in-plane and the out-of-plane optical anisotropy was examined.

2. Experimental

Figure 1 shows the chemical structures of the rodlike RM and the T-shaped reactive molecules which were used in this study. The rodlike RM is a commercial produced named as LC242 (Daken) [Fig. 1(a)]. The T-shaped reactive molecules were newly synthesized and called as T1, T2, T3, T4, and T5, respectively, hereinafter. The T-shaped molecules have a different central core X [Fig. 1(b)]. The T-shaped molecules were mixed with LC242 at various concentrations in a toluene solvent. A photoinitiator (Irgacure369, Ciba Chem) was mixed in the solution and its concentration was 2.0 wt% with respect to the whole solid contents in the solution. The total concentration of the solid contents in the solution was 20 wt%. The mixtures were stirred at 70 °C for 10 min.

 

Fig. 1. Molecular structure of the reactive molecules used in this study (a) LC242, (b) T1, T2, T3, T4, and T5.

Download Full Size | PDF

For fabrication of the retarder film, a commercial polyimide (PIA-X189-KU1, JNC) was spin-coated on a glass substrate at 1300 and 2000 rpm for 10 and 20 sec, respectively. The polyimide-coated substrate was soft baked at 120 °C for 3 min, then hard baked at 230 °C for 1 h. The baked substrate was unidirectionally rubbed with a commercial cotton cloth which was developed for liquid crystal display (HY7018, Hyperflex). The RM solution was spin-coated at 1300 rpm for 25 sec and the coated substrate was soft baked at 100 °C to evaporate the solvent. For checking the phase transition temperature of the sample, the polarizing optical microscopy (POM) texture was observed. After checking the nematic-isotropic phase transition temperature (T_NI) of each mixture, the sample was first heated up to isotropic state and then cooled down to the nematic phase for getting better alignment state. Then, the UV light was exposed to the sample for 20 min in the nematic phase with nitrogen gas purged. A monochromatic UV lamp (LF215LS, Uvitec) was used with an emission wavelength of 365 nm and intensity of 1.5 mW/cm². The temperature was maintained at 35 °C using a heating stage (Linkam) during the UV polymerization process.

For measurement of the in-plane retardation R_in ≡ (n_x - n_y)d, the light source from a halogen lamp (KLS-150H-RC-15E, Kwangwoo) consecutively passed through a bandpass filter, a linear polarizer, a pinhole, the sample, and a polarimeter (PAX-1000VIS, Thorlab). The diameter of the pinhole was 1.5 mm and the hole shape was circular to avoid a polarization state change. Because the molecules were not uniformly oriented near the edge, we measured 5 data points near the sample and the standard deviation was less than 4.1 nm through each sample. We measured the retardation data at 5 points near the center of the sample at The slow axis of the sample was parallel to the x-axis and the transmission axis of the linear polarizer was at 45° from the x-axis. Stokes parameters S₁, S₂, and S₃ after passing through the sample was recorded and converted to ${R_{in}}$ of the sample as below.

For measurement of the out-of-plane retardation R_th ≡ [(n_x + n_y)/2 - n_z]d, the retardation was measured at various incident angles θ. The relative differences between each component of the refractive indices ${n_x}$, ${n_y}$, and ${n_z}$ were approximated by fitting the sets of the retardation data. The NZ coefficient ≡ R_th /R_in + 0.5 = (n_x - n_z)/(n_x - n_y) can be obtained from R_th. The details of ${R_{in}}$ and ${R_{th}}$ measurements are identical with previous paper [20].

The thickness of the sample was measured by checking the interference pattern of the reflected light using a UV-Vis. spectrometer (UV2600, Shimadzu). The thickness of the pure LC242 was 1.01 µm, while that of the mixture with 20 wt% T1, T2, T3, T4, and T4 was 1.05, 1.00, 1.07, 1.10, and 1.07, respectively. Thus, the difference of thickness between the pure LC242 and the LC242 + T-shaped molecule mixture was less than 5%.

3. Results and discussion

Figure 2 shows the averaged ${R_{in}}$ (${\boldsymbol{\lambda }}$) data and the polarizing optical microscopy (POM) textures of the samples made from the pure LC242 and the LC242 + T-shaped molecules mixtures. For better comparison of the data from various samples, the vertical scales in the ${\boldsymbol{R_{in}}}$ (${\boldsymbol{\lambda }}$) graphs were fixed same. The concentration of the T-shaped molecules in the POM images was 20 wt%.

 

Fig. 2. Averaged ${R_{in}}$ (${\rm{\lambda }}$) of the samples made from the (a) pure LC242 and the LC242 + (b) T1, (c) T3, (d) T4, and (e) T5-mixed samples. The T2 mixture showed heterogenous orientation and the data was not added. The pictures are the POM image of the corresponding samples between crossed polarizers. The concentration of the T-shaped molecules in the POM images were 20 wt%. The bright and the dark images were obtained when the rubbing direction of the sample was at 45° and 0°, respectively.

Download Full Size | PDF

Figure 2(a) shows the averaged ${R_{in}}$ (${\rm{\lambda }}$) data of the pure LC242 sample. ${R_{in}}$ (550 nm) was 115.9 nm and the standard deviation from 5 measurement points was 2.8 nm. The images are the polarizing optical microscopy (POM) textures of the sample between crossed polarizers. The bright and the dark textures were obtained when the rubbing direction of the sample was at 45° and 0°, respectively. The vivid contrast between two state indicates uniform orientation of the RM molecules.

Figure 2(b)–2(e) correspond to the data of the samples with T1, T3, T4, and T4 molecules were mixed, respectively. Because the T2 mixture showed heterogenous orientation and its data was not added. Averaged ${R_{in}}$ (550 nm) of the 40 wt% T1 mixture sample was 117.3 nm [Fig. 2(b)]. Thus, averaged ${R_{in}}$ (${\rm{\lambda }}$) of the LC242 + T1 sample was nearly same. ${R_{in}}$ (${\rm{\lambda }}$) was decreased after further doping of T1 due to the poor orientation of the molecules. The POM texture of the 20 wt% T1 mixed sample still showed a similar contrast compared with the pure LC242 sample, but the uniformity was lower than the pure LC242 sample.

Interestingly, the averaged ${R_{in}}$ (${\rm{\lambda }}$) was increased when the T3 molecules were mixed [Fig. 2(c)]. These mixtures also retained liquid crystalline state up to 60 wt% of T3. Averaged ${R_{in}}$ (550 nm) of the 40 wt% T3 mixed sample was 196.6 nm. The thickness of this sample was 3% larger than the pure LC242 sample. Hence, the increase of ${R_{in}}$ (${\rm{\lambda }}$) is not due to the thickness effect, but due to the chemical structure of the T3. The 40 wt% T3-mixed sample showed a bright and uniform texture when its rubbing direction was at 45° from the transmission axis of the polarizer. With further mixing of T3 over 40 wt%, ${R_{in}}$ (${\rm{\lambda }}$) was decreased due to the poor orientation of the molecules.

The averaged ${R_{in}}$ (550 nm) of the 20 wt% T4 mixture was 136.4 nm, but the thickness of the sample was about 10% larger than the pure LC242. Thus, the increment of ${R_{in}}$ was not very large [Fig. 2(d)]. The LC242 + T5 mixture also showed a similar ${R_{in}}$ (${\rm{\lambda }}$) compared to the pure LC242 [Fig. 2(e)]. The T4- and T5-mixed sample showed a lower contrast between crossed polarizers than the other samples.

Thus, the averaged ${R_{in}}$ (${\rm{\lambda }}$) was particularly increased when the T3 molecules were added. The LC242-T3 mixture also showed a wide liquid crystalline phase up to 60 wt% of T3. The T3 molecule has a longer core length with 5 benzene rings which presumably contributed to the greater magnitude of ${R_{in}}$ (${\rm{\lambda }}$). The T2 molecule has a diphenyl fluorene group in the core and this seems to hinder the ordering of the molecules, resulting in narrow liquid crystalline phase and nonuniform ordering. The increase of the side-chains like T1, T3 and T4 showed no contribution to the averaged ${R_{in}}$ (${\rm{\lambda }}$).

As one of the main results of this paper, we investigated the biaxiality of the LC242 and T-shaped molecule mixtures by measuring the out-of-plane retardation ${R_{th}}$ and the NZ coefficients of the retarder films. In order to compare the effect of the T-shaped molecules chemical structure, we fixed the concentration of the T-shaped molecules at 20 wt%. All of the mixtures except for T2 showed homogenous orientation at that concentration. As described in the experimental section, ${R_{th}}$ is given by [(n_x + n_y)/2 - n_z]d, while the NZ coefficient is given by R_th/R_in + 0.5. The difference between each refractive index n_x - n_y, n_y – n_z, and n_z – n_x can be obtained by approximating retardation data of the sample measured at various incident angles $\theta $ [ Fig. 3(a)]. ${R_{th}}$ and the NZ coefficient can be obtained by the sets of the refractive index difference data [Fig. 3(b)]. The optical property of the biaxial retarders vs. the NZ coefficient is summarized in [Fig. 3(c)]. When the NZ coefficient is negative, the retarder is a positive b-plate. When the NZ coefficient is between 0 and 1, it corresponds to a z-plate. When the NZ coefficient is over 1, it becomes a negative b-plate.

 

Fig. 3. (a) Averaged retardation of the LC242 and T-shaped molecules mixtures vs. incident angle $\theta $. The wavelength of the probe beam was 550 nm. The weight fraction of the T-shaped molecules was 20 wt%. (b) ${R_{th}}$ (${\boldsymbol{\lambda }}$) and NZ coefficient of the samples. (c) Classification of the optical retarders. (d) Schematic illustrations of molecule orientations for being a negative b-plate.

Download Full Size | PDF

The NZ coefficient of the pure LC242 sample was 0.95. Because the chemical structure of the LC242 is certainly uniaxial, the NZ coefficient of the pure LC242 should be 1 provided that all of the molecules were perfectly planar aligned. The slightly smaller NZ coefficient than unity is probably due to some gradient of molecular orientation. It is generally known that the RM molecules are often vertically oriented at the interface of air [17,21,22]. Such a polar gradient of the molecules results in a greater ${n_z}$ and a smaller NZ coefficient. On the other hand, the NZ coefficients of the samples which were made from the mixture of LC242 and T-shaped molecules were greater than 1 [Fig. 3(b)]. Thus, the samples doped with T-shaped molecules became negative b-plates whose in-plane refractive indices ${n_x}$ and ${n_y}$ were greater than ${n_z}$. This is possible provided that more fraction of the T-shaped molecules is oriented parallel to the surface plane [Fig. 3(d)].

The asymmetric angular dependence of the retardation is also due to the orientational gradient of the molecules. In order to quantify the materials intrinsic property, R_th and the NZ coefficient are determined in the rotated coordinates where the slow axis of the sample is parallel to the one of the principal axes. The detailed conditions and descriptions are explained in our previous paper [20].

Because the LC242-T3 mixture showed the widest liquid crystalline phase among the various mixtures, we investigated the dependence of the ${R_{in}}$, ${R_{th}}$, and the NZ coefficient on the T3 concentration [Fig. 4]. The wavelength of the probe beam was 550 nm. In order to measure those data, we similarly measured the retardation at various incident angles $\theta $ and calculated the NZ coefficient. ${R_{in}}$ was increased up to 40 wt% of T3, but decreased with further mixing due to poor orientation. ${R_{th}}$ value also showed a similar dependence on the T3 concentration. It is observed that the NZ coefficient was gradually increased with a greater concentration of T3 up to 30 wt%. The NZ coefficient of 30 wt% T3-mixed LC242 was 1.25, whereas that of the pure LC242 was 0.95. The physical reason for the NZ coefficient decreases from 40 wt% of T3 can be interpreted as following. The samples showed homogeneous texture in POM and ${R_{in}}$ was not significantly decreased even with 40 wt% T3. From the experimental data, the T3 molecules prefer planar orientation whose central linking groups are parallel on the surface at low concentration. When the T3 concentration is over 40 wt%, the long molecular axis of the T3 molecules are oriented parallel to the LC242, but the side biphenyl groups are randomly oriented with respect to the long molecular axis.

 

Fig. 4. Averaged ${R_{in}}$, ${R_{th}}$ and NZ coefficient of the LC242 and T3 mixtures vs. weight fraction of the T3 molecules. The wavelength of the probe beam was 550 nm.

Download Full Size | PDF

We measured the transmittance (TR) spectra of the samples in Fig. 5. TR was measured using a UV-Vis spectrometer and the polyimide-coated substrate was inserted in the path of the reference beam. Thus, the 100% TR in Fig. 5 means TR of the polyimide-coated sample. The MgF₂ substrate was used as the substrate to avoid the absorption at the short wavelength region. TR of the pure LC242 sample was 87% at 550 nm, while TR of the 20 wt% T1, T3, T4, and T5 samples was 83, 86, 84, 85%, respectively, at the corresponding wavelength. TR of the films made from the mixture of the LC242 and the T-shaped molecules was not reduced to much which is good for the application as the compensation films for the display devices.

 

Fig. 5. Transmittance data of the films made from the pure LC242 and the T-shaped molecules. The weight fraction of the T-shaped molecules was 20 wt%. The data was measured using a UV-Vis. spectrometer where a polyimide-coated film was used as a reference.

Download Full Size | PDF

4. Conclusion

In this paper, 5 kinds of T-shaped reactive molecules were newly synthesized. The optical anisotropy of the retarder films made from the rodlike RM and the T-shaped molecules mixtures were investigated. The mixtures showed a greater magnitude of in-plane retardation and the NZ coefficient over 1. Thus, the retarder films represented a negative b plate, which can be used for the compensation film of the VA-LCD. The biaxial retarders made from the RM materials have been rarely reported and the suggested results can help the development of the materials for various optical applications.