Vertical alignment of high birefringence and negative dielectric anisotropic liquid crystals for projection displays

Transcription

1 Invited Paper Vertical alignment of high birefringence and negative dielectric anisotropic liquid crystals for projection displays Chien-Hui Wen a, Benjamin Wu b, Sebastian Gauza a, and Shin-Tson Wu a a College of Optics and Photonics, University of Central lorida, Orlando, L b Duke University, Durham, NC ABSTRACT Physical properties and alignment performance of biphenyl and terphenyl negative dielectric anisotropic liquid crystal (LC) compounds are investigated. Results show biphenyl compounds align well in homeotropic LC cells and the alignment of terphenyls are relatively poor. We have developed a new method to align these high birefringence LC compounds. Adding a few percent of positive dielectric anisotropic or nonpolar LC material not only enhances the contrast ratio but also improves the overall figure-of-merit. Molecular modeling and experimentation are demonstrated to support this concept. Keywords: Negative Dielectric Anisotropic, Liquid Crystal, Alignment, Vertical Aligned LC cell 1. INTRODUCTION Homeotropic cell (also called vertical alignment, VA) [1-2] exhibits an excellent contrast ratio and has been used extensively for direct-view and projection displays. This contrast ratio is insensitive to the incident light wavelength, liquid crystal (LC) cell gap (d), and operating temperature [3]. In a VA cell, negative dielectric anisotropic ( ε= ε // ε < 0) LCs are needed. To get flicker free images [4], a high resistivity LC is another crucial requirement for obtaining a high voltage-holding-ratio. In addition, thin LC cell is highly attractive to achieve fast response time. The response time is quadratic proportional to the cell gap. To realize the advantages of the thin cell approach, the LC birefringence ( n) must be increased in order to retain sufficient phase retardation. or example, the required minimum d n is 275 nm for TV applications. If d= 1 µm, then the required n value is ~ 0.3 [5]. or infrared applications [6], the role of high birefringence and low viscosity LC materials become even more important because of the resultant longer wavelength and reduced birefringence [7]. Therefore, it is essential to develop high birefringence, large negative ε, and low viscosity LC mixtures [8-9]. The common feature of dielectrically negative liquid crystals is that lateral polar substituents induce a dipole moment perpendicular to the principal molecular axis [10-11]. Laterally (2,3) difluorinated biphenyl, terphenyl [12] and tolane [13], are good candidates in this approach. However, we found that some laterally fluorinated LC compounds are difficult to align [14-15]. A poor alignment leads to light leakage in the dark state and degrades the contrast ratio. Without good alignment, the VA cell is ineffective. In this paper, we report a simple method for aligning the high birefringence ( n~ ) laterally difluoro-biphenyl, -terphenyl and -tolane LC compounds. irst, we investigated these compounds individually by filling them into a buffed polyimide VA cell. Since these compounds have different nematic ranges, we compared their electro-optic properties at the same reduced temperature (T r = 0.96), which is defined as T/T c. We found that the biphenyls are much easier to align than either the terphenyls or the tolanes. To align these highly conjugated negative ε LC compounds, we formulated an LC host mixture for this study. By doping some positive ε or nonpolar LC materials into the host LC, we obtain a very good dark state and sharp threshold voltage. To understand the molecular alignment mechanisms, we evaporated a positive LC compound (5CB) onto a glass substrate, made a cell, and then filled it with the negative LC compounds. We found that these agents form a buffer layer to neutralize the negative electrostatic potential and help align the negative LCs. This doping method is not limited by the molecular structure of polyimide and the process of rubbing technology. Moreover, we also applied this doping method to commercial LC mixture, MLC-6608 (Merck), with proper dopants, and found the mixture s figure-of-merit is improved significantly. Liquid Crystal Materials, Devices, and Applications XI, edited by Liang-Chy Chien, Proc. of SPIE Vol. 6135, 61350U, (2006) X/06/$15 doi: / Proc. of SPIE Vol U-1

2 2. EXPERIMENT We selected several LC compounds with different birefringence for the alignment study. Table I shows the molecular structure, clearing temperature, birefringence, and dielectric anisotropy of these LC compounds. We measured the phase transition temperature of these LC compounds by using a differential scanning calorimeter (DSC; TA-100). Since the compounds we studied are solid at room temperature, we mixed 10 wt% of the compounds into an LC host ZLI-4330 and extrapolated their birefringence at T~ 23 C. To monitor the LC birefringence, we filled a 5-µm homeotropic cell (pretilt angle ~ 87 o ) and measured its phase retardation (δ) [16]. At a given temperature, the phase retardation is related to cell gap d, birefringence n, and wavelength λ as δ= 2πd n/λ. All the measurements were performed using a He-Ne laser (λ= 633 nm) at T= 23 C. The dielectric and elastic constants of these three compounds were measured by the capacitance method [17-18] of a single homeotropic cell, using a computer controlled Displaytech APT III instrument. In the molecular alignment study, we prepared VA cells with a thin polyimide (PI) alignment layer on each surface. To prepare VA cells, we spin-coated a commercial polyimide SE-1211 (from Nissan Chemicals) onto the indium-tin-oxide (ITO) glass substrates, baked the substrates at 80 C for 5 minutes and then 180 C for 1 hour. We then gently rubbed the ITO-glass with cloth in anti-parallel directions. The rubbing induced pretilt angle was about 87. We also compared our results with those non-rubbed LC cells. To study the dopant-induced alignment enhancement effect, we selected MLC-6608 (Merck) as the host mixture, and doped some biphenyl compounds. The new mixture is named as UC-A. We measured the temperature dependent birefringence and figure-of-merit (om) of MLC-6608 and UC-A, and compared their performance. Table I. Molecular structure, clearing temperature (T c ), extrapolated birefringence ( n), and dielectric anisotropy ( ε) of studied compounds. n and ε are extrapolated from ~ 10% guest-host systems in ZLI-4330 host. T= 23 o C and λ= 633 nm. No. Molecular structure T c ( o C) n ε N1 C 3 H 7 CO OC 2 H N2 N3 C 5 H 11 OC 2 H * C 5 H 11 OMe N4 C 4 H 9 C 2 H N5 N6 C 4 H 9 OC 2 H C 3 H 7 C 3 H * This comes from super cooling effect. 3. RESULTS AND DISSCUSSIONS 3.1 Physical properties Among the compounds we studied, compound N5 exhibits a smectic phase, and compound N2 exhibits a monotropic phase. rom compound N1 to compound N6, the number of phenyl rings increases from one to three. Because of the Proc. of SPIE Vol U-2

3 increase of unsaturated elements, birefringence increases from for compound N1 to ~ 0.25 for compounds N5 and N6. Concerning the terminal substituents, compounds N4 and N6 have an alkyl chain rather than an alkoxy group. The oxygen not only increases the conjugation length, but also increases the dipole moment perpendicular to the long axis of the molecule. This results in a large dielectric anisotropy ( ε~-5.1) of difluorinated alkoxy compounds, such as N2, N3 and N Alignment of single compounds in VA cell To investigate the alignment performance of a single compound in rubbed VA cells, we injected each of the compounds we listed into VA cells. Because the compounds have their nematic phases at different ranges, we measured the voltage dependent transmittance (V-T) curve between crossed polarizers at the same reduced temperature ( T r ~ 0.96) except N2. Compound N2 exhibits a monotropic phase and its nematic range is quite narrow. Thus, we measured its V-T curve at T = 3 o C. All the measurements were performed at λ= 633 nm. Results are plotted in igs igure 1 plots the measured V-T curves of compounds N1, N2, and N3, while igs. 2, 3 and 4 are for compounds N4, N5 and N6, respectively. rom these curves, we found that compounds N1, N2 and N3 have a fairly good voltage-on and -off states, but compounds N4, N5 and N6 have a very poor on state, which means the molecular alignment of terphenyls (N4, N5, N6,) in VA cells is imperfect. 0.8 T (%) N1 N2 N Voltage ( V rms ) igure 1. Voltage dependent transmittance of N1, N2, and N3 homeotropic cells between crossed polarizers. N1 and N3 were measured at T r = 0.96 and N2 at T= 3 o C. λ= 633 nm. Table II. Compositions of mixtures A to. Host Wt% of Mixture Dopant Compound dopant A MLC N4 B ZLI C MLC N5 D ZLI E MLC N6 ZLI To improve the alignment capabilities of these terphenyl compounds, we doped some positive or some neutral LC mixtures into compounds N4, N5, and N6. Results are shown in igs. 2, 3, and 4, respectively. The positive and neutral LC mixtures we selected are MLC ( ε= 4) and ZLI-3086 ( ε ~ 0), respectively. Both are Merck s mixtures. The required weight percentage of dopants to have the best performance depends on the host compounds. Table II shows the required weight percentages of MLC and ZLI-3086 for compounds N4, N5 and N6. By doping these positive ε or neutral LC mixtures, a V-T curve with a very sharp threshold voltage and improved on-state transmittance is obtained. or compound N6, the dopants not only improve the on-state transmittance, but also make the V-T curve Proc. of SPIE Vol U-3

4 smoother in the high voltage regime. Besides LC mixtures, we also doped some positive or neutral single compounds into the host compounds, and observed the same results. The required percentage of dopant is ~ 10 wt%. 0.8 N4 A B T (%) Voltage( V rms ) igure 2. Voltage dependent transmittance of homeotropic cells between crossed polarizers. The dotted, dashed, and solid lines represent compound N4, mixture A, and mixture B, respectively. T r = 0.96, λ= 633 nm. 0.8 N5 C D T (%) Voltage( V rms ) igure 3. Voltage dependent transmittance of homeotropic cells between crossed polarizers. The dotted, dashed, and solid represent compound N5, mixtures C, and mixture D, respectively. T r = 0.96, λ= 633 nm T (%) N6 E Voltage ( V rms ) igure 4. Voltage dependent transmittance of homeotropic cells between crossed polarizers. The dotted, dashed and solid lines represent compound N6, mixtures E, and mixture, respectively. T r = 0.96, λ= 633 nm. Proc. of SPIE Vol U-4

5 3.3 Physical mechanisms The compounds we selected are mainly two and three ring systems, either saturated cyclohexane or unsaturated phenyl rings. We found that the alignment capability is related to the molecular conjugation length. or example, the molecules with two or fewer phenyl rings have better alignment than those with three phenyl rings. Moreover, compounds N4 and N5, except for the different difluoro positions and terminal alkyl/alkoxy groups, are both lateral difluoro terphenyls. The two fluoro groups of N4 are in the middle ring and its terminal group is alkyl rather than alkoxy. Both N4 and N5 have poor V-T curves and the on-state transmittance of compound N5 is worse than that of N4. Therefore, both the conjugation length and lateral positions of the difluoro groups jointly determine the alignment capability. In order to understand the physical mechanisms of the LC alignment, molecular modeling was used to investigate the electrostatic potential distribution of these single compounds. Molecular modeling (HyperChem- molecular modeling software) was performed by running Austin Model I (AM1) and Modified Neglect of Diatomic Overlap (MNDO) calculations in order to estimate the molecular conformations, electrostatic potentials, and Huckel charges. igure 5 shows the space filling models of compounds we studied by using color codes. Here, blue stands for a strong negative electrostatic potential; red, a strong positive one; and green, a moderately negative one. or example, the fluoro group has a strong negative electrostatic potential and, thus, it is presented by the blue color. To balance the charges, the remaining phenyl ring has a strong positive electrostatic potential. The electrostatic potential distributions from ig. 5 show that compounds N1, N2, and N3 (collectively called group I) have a strong negative electrostatic potential only in the vicinities of the fluoro groups. These negative charges spread to the nearby phenyl ring because of electron conjugation. Although compounds N4 and N5 have similar terphenyl structures, their electrostatic potentials are quite different. In N4, the relatively strong negative electrostatic potential distributes over the entire molecule. By contrast, the strong negative electrostatic potential of compound N5 is more localized; only a moderate negative charge spreads to the other two phenyl rings. N1 N2 N3 N4 N5 igure 5. Molecular models of compounds N1-N5. The color is used for visualizing electrostatic potentials. Here, blue stands for a strong negative electrostatic potential; red, a strong positive one; and green, a moderately negative one. We have determined from our experimental results that the group I compounds are easy to align while the compounds of group II (N4, N5, and N6) are not. Therefore, the charge distribution seems to shed some light on the understanding of the alignment mechanism. rom ig. 5, the negative electrostatic potentials of the group I compounds spread out a smaller distance because of their shorter electron conjugation. Due to the shorter electron conjugation, the repulsive force Proc. of SPIE Vol U-5

6 from the polyimide is weaker. As a result, a good homeotropic alignment is achieved. On the other hand, the negative electrostatic potentials in the group II compounds spread over to the entire compounds. The repulsive force from the polyimide is stronger, resulting in a larger pretilt angle. A large pretilt angle causes light leakage in the dark state. Although we do not know the exact molecular structure of the PI we employed, molecules in group II have strong interactions with the alignment layer along the long side of the molecules and result in a large pretilt angle. Therefore, we believe that the added positive or neutral LC materials help neutralize the strong negative charge of the group II molecules and serve as a buffer layer for aligning the difluoro terphenyl and tolane compounds and mixtures. The required amount of positive or neutral LC material depends on how the negative electrostatic potential is distributed over the entire molecule. In order to investigate this dopant effect at room temperature, we formulated a binary eutectic mixture (called G) by mixing two N4-type terphenyl homologues. The terminal groups of these two homologues are R 1 = C 2 H 5, R 2 = C 4 H 9, and R 1 = C 3 H 7, R 2 = C 5 H 11. Mixture G consists of 35% and 65% of the -24 and -35 homologues, respectively. We selected 5CB as the dopant for this part of the study because of its low clearing temperature (35.3 o C). We used two different methods to coat 5CB onto the glass substrate. or the first, we vaporized 5CB onto the glass substrate. or the second, we dissolved 5CB into a cyclohexane solution and then dripped this solution onto a glass substrate. In the first method, we filled a glass vial with 5CB and placed a glass substrate, which has the PI alignment layer, over the opening. We heated the 5CB to ~ 120 o C for 20 minutes, evaporating the 5CB onto the substrate, and then removed the glass substrate. We then repeated the process for another substrate and then joined the two substrates coated with 5CB, creating a cell. The sealed cell was injected with mixture G. We observed this LC cell under a polarizing optical microscope. Results are shown in ig. 6. Between crossed polarizers, the area coated with 5CB vapor appears darker than the adjacent uncoated areas, indicating that the coated 5CB indeed helps align mixture G. igure 6. A microscope photo of a VA cell filled with 5CB under a crossed polarizing microscope. The PI alignment layer is coated with vaporized 5CB molecules. The cell was then filled with mixture G. The dark area (bottom right corner) is the area coated with the 5CB vapor. A 20X magnification objective lens was used. In the second method, we prepared a 5CB solution in cyclohexane. The molar concentration of the 5CB solution was 5.3x10-3 mol/l. We dripped the solution onto the glass substrate inside an oven (~ 90 o C) in order to vaporize the cyclohexane, leaving just the 5CB on the substrate. We placed 10 drops, each time waiting for the cyclohexane to evaporate before placing another drop, in order to deposit enough 5CB molecules on the glass substrate. After being coated with the 5CB solution, the cell was sealed and injected with mixture G. Results are shown in ig. 7. In ig. 7, the left upper corner is the area without 5CB, where the light leaks severely because of the poor LC alignment. The lower right corner is the area with the 5CB solution. The slightly diagonal lines are the rubbing direction. The 5CB solution, indeed, enhances the contrast ratio of a VA cell. We found that, compared to ig. 6, the solvent method forms a larger and more uniform area. This is because the vaporized 5CB molecules could aggregate on the PI-coated glass substrate and cause non-uniformity. Proc. of SPIE Vol U-6

7 igure 7. A microscope photo of VA cell under crossed polarizers. The surfaces of the VA cell were coated with 5CB dissolved in cyclohexane. The cell was then filled with mixture G. A 20X magnification objective lens was used. The bottom right corner is the area with 5CB solvent. In our experiments, we used different PI alignment layers and also found similar results. In addition, we tested our concepts using non-rubbed PI substrates. Results are shown in ig. 8. ig. 8(a) and 8(b) represents the non-rub VA cell filled with Mixture G when the driving voltage is V= 0V rms and V= 9V rms, respectively. To improve the alignment performance, we doped 10 % of MLC into mixture G. The new mixture is named as mixture H. Results are shown in ig. 8(c) and 8(d) with driving voltage V= 0V rms and V= 9V rms, respectively. Results show mixture H has better dark state than mixture G, and the dopant helps to have larger domain size. This larger domain size results in a higher transmittance when the LC cell is placed between two crossed polarizers. The non-buffed PI substrates have been used in multi-domain structures [19-20] for widening the viewing angle. Therefore, the dopant-enhanced contrast ratio of VA cells can be applied to varieties of liquid crystal display devices. (c) loojim r igure 8. A microscope photo of an non-rubbed polyimide VA cell filled with (a) Mixture G, V= 0V rms, (b) Mixture G, V= 9V rms, (c) Mixture H, V= 0V rms, and (d) Mixture H, V= 9V rms under a crossed polarizing microscope. Proc. of SPIE Vol U-7

8 To see the dopant effect on birefringence and viscosity, we demonstrated an experiment by using commercial LC mixture, MLC-6608 (Merck), as host mixture, and doped some biphenyl compounds properly. This mixture is named as UC-A. We compared the temperature dependent birefringence and figure-of-merit (om = K 33 n/γ 1 ) of these two mixtures. Results are shown in ig. 9 and ig. 10, respectively. At T= 50 C, the om of UC-A is 75% higher than that of MLC MLC-6608 UC-A Birefringence Temperature ( o C) igure 9. Temperature-dependent n of MLC-6608 (Squares) and UC-A (Triangles). Solids lines are fitting results. λ= 633 nm MLC-6608 UC-A 2.0 om Temperature ( o C) igure 10. Temperature-dependent om of MLC-6608 (Squares) and UC-A (Triangles). Solid lines are fitting results. λ= 633 nm. CONCLUSION We have developed a simple method for aligning the high birefringence and negative ε LCs. By doping ~ 10% of some positive ε or non-polar materials into a negative LC host, the buffed PI cells exhibited an excellent homeotropic alignment. We also conducted two experiments (vapor and solvent) to investigate the alignment mechanisms. The dopant effect indeed helps align the laterally difluoro LCs well in a homeotropic cell, no matter if the PI is rubbed or not. A wide-spread application of this alignment method to different display devices is foreseeable. REERENCES [1] M.. Schiekel and K. ahrenschon, Deformation of nematic liquid crystals with vertical orientation in electrical fields, Appl. Phys. Lett., 19, , [2]. J. Kahn, Orientation of liquid crystals by surface coupling agents, Appl. Phys. Lett., 22, , Proc. of SPIE Vol U-8

9 [3] S. T. Wu and D. K. Yang, Reflective Liquid Crystal Displays, Wiley, New York, [4] P. Kirsch and M. Bremer, Nematic liquid crystals for active matrix displays: Molecular design and synthesis, Angew. Chem. Int. Ed., 39, , [5] D. Pauluth and K. Tarumi, Advanced liquid crystals for television, J. Mater. Chem., 14, , [6] S.T. Wu, U. Efron, and L. D. Hess, Infrared birefringence of liquid crystals, Appl. Phys. Lett., 44, , 1984 [7] S. T. Wu, Birefringence dispersions of liquid crystals, Phys. Rev. A, 33, , [8] V. Reiffenrath, J. Krause, H. J. Plach and G. Weber. New liquid crystalline compounds with negative dielectric anisotropy, Liq. Cryst., 5, , [9] K. Tarumi, M. Bremer, and B. Schuler, Development of New Liquid Crystal Materials for TT LCDs, IEICE Trans. Electron., E79-C, , [10] M. Klasen, M. Bremer and K. Tarumi, New liquid-crystal materials for active matrix displays with negative dielectric anisotropy and low rotational viscosity, Jpn. J. Appl. Phys., 39, L1180-L1182, [11] P. Kirsch and K. Tarumi, A novel type of liquid crystals based on axially fluorinated cyclohexane units, Angew. Chem. Int. Ed., 37, , [12] G. W. Gray, M. Hird, and K. J. Toyne, The synthesis of several lateral difluoro-substituted 4,4 -diakyl- and 4,4 -alkoxyalkyo-terphenyls and a rationalization of the effect of such substitution on mesophase type and transition temperatures, Mol. Cryst. Liq. Cryst., 204, 43-64, [13] S.T. Wu, C.S. Hsu, and J. M. Chen, Room temperature difluoro-tolane and diphenyl-diacetylene liquid crystals with negative dielectric anisotropy, Mol. Cryst. Liq. Cryst., 304, , [14] C. H. Wen, S. Gauza, J. Li, H. Wang, and S. T. Wu, High contrast homeotropic alignment of difluoro-tolane liquid crystals, Liq. Cryst., 32, , [15] C. H. Wen, S. Gauza, and S. T. Wu, High-contrast vertical alignment of lateral difluoro-terphenyl liquid crystals, Appl. Phys. Lett., 87, , 2005 [16] S. T. Wu, U. Efron, and L. D. Hess, Birefringence Measurements of Liquid Crystals, Appl. Opt., 23, , [17] I. C. Khoo and S. T. Wu, Optics and Nonlinear Optics of Liquid Crystals, World Scientific, Singapore, [18] M. G. Clark, E. P. Raynes, R. A. Smith, and R. J. A. Tough, Measurement of the permittivity of nematic liquid crystals in magnetic and electric fields using extrapolation procedures, J. Phys. D: Appl. Phys., 13, , [19] K. Kim and S. Kim, Advance of PVA technology for multi-medium applications, Soc. Inf. Display Tech. Digest, 34, , [20] K. Okamoto, Recent development in MVA-LCDs, Proc. Int l Display Manufacturing Conference, , Proc. of SPIE Vol U-9