The rheological properties of polyimide film surfaces have an important influence on contrast reduction during the in-plane switching mode of liquid crystal displays. To clarify these properties, the slight difference of deviation angles of liquid crystal directors from the rubbing direction were measured during prolonged exposure to alternating electric fields. The results indicate that the data can be well described using the Kelvin-Voigt model. The relation between the in-plane shear modulus G and the strain at the polyimide surface was also investigated based on the torque balance between the energy density of the electric field and the elastic energy density of the polyimide surface. It was found that much smaller G than bulk polyimide materials existed on the polyimide surface in liquid crystal display.
© 2014 Optical Society of America
Recent progress in the performance of liquid crystal displays (LCDs) are remarkable. An example of techniques to enhance LCD performance is the in-plane switching (IPS) mode proposed by Oh-e et al [1, 2]. The IPS allows a wider viewing angle and yields superior color shift properties, and has become familiar in television, personal computer monitors and high resolution displays [3–5].
In the IPS mode, liquid crystal (LC) molecules are initially aligned by rubbing method. They are reoriented using an alternating current (AC) driving voltage allowing the polarized light from a backlight unit to pass through. The twist direction during reorientation is uniquely determined by the rubbing direction and the electrode configuration. There are two main types of IPS electrode designs for interdigitated electrode: single domain type and bent dual domain type. The use of the dual domains with different reorientation directions allows achievement of wider viewing angle. The dual domain type is extensively used in recent IPS mode devices [4–6].
One of the most serious performance issues affecting current LCDs is image sticking, in which a previous pattern is temporarily retained on the screen. It is important to work out solutions for this problem.
Figure 1 shows an example of image sticking. In Fig. 1(a), a black and white checker pattern is displayed on the screen for about a few days. In Fig. 1(b), an attempt is made to display a uniform gray image. However, a slight memory of the original checker image remains leading to a faint ghost pattern appearing.
Image sticking is caused by application of both direct current (DC) and AC driving. It has been reported that DC-type image sticking is mainly the result of a residual DC voltage induced by impurity ions in the LC layer [7–9]. Such impurities are due to contamination either from other surrounding materials in LCD (e.g. sealant and so on) or introduced during the manufacturing process. When an external DC offset voltage is applied, impurity ions can be adsorbed on the surface of the alignment films, and they do not immediately become desorbed even after the DC voltage is removed. This phenomenon can be detected using flicker minimization and other methods [10, 11]. Clarification of the mechanism involved in DC-type image sticking has improved the quality of LCDs by purification of the LC and polyimide alignment material and by using improved driving methods [12, 13].
The mechanism involved in the image sticking of AC field driving is not still well understood. Particularly in the case of IPS mode in which an AC voltage is applied about one month, the continuous force field leads to director deviation from the initial alignment direction on the surface polyimide. This deviation leads to a reduction in the display contrast. For the present image sticking of IPS mode with AC field driving, we can not detect remarkable DC offset by using flicker minimization method. We cannot explain the phenomena by a DC bias due to space charges.
There have been many studies on alignment deviations of LC molecules on polyvinyl-alcohol alignment layers and inorganic material surfaces [14–23]. They discussed comparatively large deviation angle (more than 10°) considering weak anchoring and using desorption re-adsorption model. On the other hand we have focused and discussed the slight deviation (less than 1°) on rubbed polyimide alignment film considering strongly anchored LC molecules in LCD. We have also previously investigated the mechanism involved in the AC memory effect .
We proposed that the deviation is due to a small rheological deformation of the surface molecular conformation of the polyimide film in LCD. To achieve a significant reduction of AC image sticking, the effect of the surface rheology of the rubbed polyimide alignment film must be considered.
In the present study, in order to clarify the rheological properties of the polyimide surface, the deviation angles of LC molecules from the rubbing direction were investigated during prolonged exposure to alternating electric fields of various strengths. The orientation of LC molecules was determined from the measurement of the darkest brightness for a simple test cell with in-plane bent interdigitated electrodes using the method proposed by Suzuki et al . A deviation angle was used to characterize the memory effect. Finally, the in-plane shear modulus and viscosity of the polyimide surface was estimated based on the torque balance between the LC twist elastic energy from the electric field and the polyimide surface.
A simple 10 cm × 10 cm glass substrate test cell was fabricated with bent interdigitated transparent indium tin oxide (ITO) electrodes. The dimensions of each electrode were 1 cm × 1 cm and a total of 9 electrodes were arranged in a 3 × 3 grid in the test cell. The in-plane bent interdigitated electrode width of ITO was 4 μm and the spacing between electrodes was 5, 10 or 15 μm. The angle between the lines of electrodes and the rubbing direction used for LC alignment was 20°. Two glass substrates, one containing the electrodes, were stacked using 3-μm-height column spacers to achieve a uniform cell gap above each electrode.
A polyimide film (Nissan Chemicals; SE7492, approximately 100 nm thick) was spin coated onto each of the glass substrates and baked at 170 °C for 30 min in a heat chamber. Weak rubbing was carried out using cotton cloth at a pile impression of 0.1 mm. The pile impression corresponds to the distance of the rubbing roller and glass substrate from the rubbing zero point, and is an indication of the rubbing strength. An epoxy resin heat sealant (Mitsui Chemicals; XS-21S) was dispensed around the edges of the stacked substrates, leaving an about 10 mm width hall for LC injection. The epoxy resin was then cured under a mechanical pressure of 30 kgf in a heat chamber for 3 h at 150 °C. A positive type LC mixture (Merck Japan; ZLI4792) was injected into the assembled test cell under vacuum. The injection hall was then sealed using UV sealant (Threebond; 3052) and the test cell was heated at 110 °C in the heat chamber for 10 min to realign the LC molecules along the rubbing direction. The cell gap above each ITO electrode region was measured to confirm the uniformity achieved by the column spacers.
Figure 2 schematically shows the measurement method used in this work. Figure 2(a) shows the initial state immediately after sample preparation. The rubbing direction is from the bottom to the top, and the LC molecules are aligned along this direction. Figure 2(b) shows the situation during AC driving. Due to the use of the bent interdigitated electrodes, the direction of LC twist under the influence of the electric field is always opposite in the upper and lower electrode regions. Figure 2(c) shows an example of LC molecules with slight deviations from their initial alignments after prolonged AC driving. Since the deviation angles are opposite in the upper and lower electrode regions, the precise deviations can be easily determined by adding the two angles.
The orientations of the LC molecules were determined using a cross-nicols polarizer arrangement in which the LC sample cell was placed. The polarizer was rotated keeping the cross-nicoles condition until we obtain the darkest image. This rotation angle of the darkest image reflects the LC director. This angle was measured using an LCD analyzer (Meiryo Technica; LCA-LU4A10) before and after driving the cell using a 60 Hz AC square electric wave from a function generator (Tektronix; AFG3022). The amplitude of the AC electric field and the size of the electrode gap are summarized in Table 1. The LC director deviation angle Δθ is defined asFig. 2(c). Further details of the method for measuring Δθ are given in a previous paper .
In the initial state, Δθ was first determined for each ITO electrode region of the sample, and was found to be less than 0.02° in all cases. The change in Δθ was monitored during about one month of AC driving in room temperature. For each measurement, the sample cell was temporarily removed from the AC driving circuit, and it was placed in the LCD analyzer with its electrodes shunted. After the measurement, the sample cell was again placed in the AC driving circuit to continue the experiment.
3. Results and discussion
The variation in the deviation angle Δθ during AC driving period is shown in Fig. 3 for the applied electric fields and electrode spacings listed in Table 1. It is interesting that Δθ slowly increased with AC driving period, and the size of the increase depended on the electric field strength. For example, for the strongest electric field of 2 V/μm (No .1 in Table 1), its value after 700 h of AC driving was about 0.7°. Figure 4 shows a photograph of the electrode region of the test cell subjected to the strongest electric field. This photograph was taken through the crossed-nicols polarizers. Even for a small deviation angle of 0.7°, a clear difference is observed between the brightness of the upper and lower ITO electrode regions. Hence it was confirmed that AC-type image sticking was induced.
From previous experiment, we concluded that such deviation in LC alignment was mainly influenced by the rheology of the polyimide surface , and the well-known Kelvin-Voigt (KV) model was used to describe results. The KV model is a simple model involving a parallel arrangement of a dashpot representing the viscosity (η) and an elastic spring with a modulus (G). This model is used to explain creep phenomenon, and describes the change in the strain γ with time t under a constant stress σ. With this model, the strain γ is expressed as
From Eq. (1) Δθ corresponds to twice the deviation angle in each electrode region. In the present study, we assume26], and the length of liquid crystal molecules is about 2nm, we estimate the order of the coefficient c. When the unit of Δθ is radian, Eq. (1).
Figure 3 shows a comparison of the experimental data and the fitted curves obtained using the KV model for the experimental conditions in Table 1. The KV model well agrees the experimental data except for the high electric field condition. In the high electric field region, a continuing plastic deformation which could not be expressed by the simple KV model was induced.
The parameter τ of Eq. (2) expresses relaxation time. It is calculated by fitting curves of Eq. (2) for plots in Fig. 3. Figure 5 shows obtained retardation time τ of Eq. (2) depending on electric field. These data show that τ ranges from 40 to 140 h. This rheological phenomenon of AC-type image sticking progressed so slowly.
To estimate the stress σ. We examine expression of torque due to electric field. The electrical energy density fe, which is expressed in terms of the electric field E and the electric displacement D is provided byEq. (7), ε⊥ and ε‖ are dielectric constants in the directions perpendicular and parallel to the directer of the nematic liquid crystal, respectively. The torque density associated with this electrostatic energy is given by
For the present nematic material, we have ε‖ − ε⊥ = 5.2ε0, where ε0 is the the vacuum permitivity. The angle between the electric field and the nematic director for the present configuration of electrodes and rubbing direction of Fig. 2 is given asEqs. (3), (5), (6) and (8) we estimated G and η. We summarize used parameters and obtained viscoelastic parameters in Table 2 which also lists R-squared values to show agreements of data fittings by Eq. (2). We estimate G of polyimide surface as 104 Pa for the present study. This magnitude corresponds to an order of elastic modulus of silicone gel substrates . The G value of polyimide film was reported to be of the order of 109 Pa [28, 29], which is about 105 times larger than the present results. Viscosity η given by Eq. (4) for polyimide surface is estimated in the order of 109–1010 Pa·sec.
The rubbing depth on the surface is less than about 10 nm  so that the present G is not that of the bulk polyimide film. The results of the present study reflected the surface properties. The modulus G of the present study was much smaller than that of bulk solid of polyimide materials.
Figure 6 shows a plot of the electric density vs. strain γ(∞). It indicates a relationEq. (8) provide a quadratic dependence of stress σ on strain γ: Eq. (11) becomes
We expect following mechanisms are related to modification of the surface of polyimide substrates. The general LC alignment polyimide used in the present study had a main-chain structure without linkings of network structure . After the rubbing process, the molecules of polyimide surface aligned roughly and axially with an excess free volume around rubbed polymer interface. This excess free volume causes the softening of the surface. In addition, the test cell used in the present study can be considered to be a ‘wet’ environment, in which the LC molecules can enter the polyimide surface as a solvent and cause surface swelling that drastically reduces the interaction between polyimide molecules.
The slight deviation angles of LC molecules from the rubbing direction in LCD were investigated during prolonged exposure to alternating electric fields with different strengths. From the relation of the electric field density and the strain determined using the test cell and by fitting with the KV model, the dependence of the elastic modulus on the strain was obtained. The results indicated that for strain-free conditions, the elastic modulus approaches zero.
Three reasons were suggested for much smaller G value than bulk polyimide film determined for the rubbed polyimide surface: (1) The polyimide thin film for LCD had a long main-chain structure without networking; (2) The rubbed polyimide molecules align roughly and axially; (3) The polyimide surface in the present study can be considered to be a ‘wet’ environment with LC molecules. This wet condition dramatically decreases the interaction of polyimide molecules so that G decreases on the surface within 10 nm depth in LCD.
We specially thank Mr. Hayato Ishiguro who assisted with the present experiments for many long hours. We would like to dedicate this paper to Dr. Woo-yeol Kim who created the basis for collaboration between Yamagata University and LG Display. We would like to thank Mr. Kaoru Furuta, Mr. Chang-Ryong Seo, and Dr. Sun-Dong Min for financial and documentation assistance in Japan. This work was financially supported by JSPS KAKENHI Grant Number 23350108.
References and links
1. M. Oh-e and K. Kondo, “Electro-optical characteristics and switching behavior of the in-plane switching mode,” Appl. Phys. Lett. 67, 3895–3897 (1995). [CrossRef]
2. M. Oh-e, M. Ohta, S. Aratani, and K. Kondo, “Principles and characteristics of electro-optical behavior with in-plane switching mode,” The 15th IDRC (Asia Display ’95), 577–580 (1995).
3. Z. Tajima, “IPS-TFT-LCDs technology trends,” Asia Display/IMID ’04 Dig., 15–17 (2004).
4. S. L. Wright, “IBM 9.2-Megapixel flat-panel display: technology and infrastructure,” IBM Research Report, RC22363(W0203-023), March 8 (2002).
5. T. Ueki, “Requirements for large-sized high-resolution TFT-LCDs,” J. Soc. Inf. Display , 9, 151–154 (2001). [CrossRef]
6. Y. Nagae, “Recent trends in wide-viewing angle color TFT-LCDs,”, Proc. SPIE , 4079, 152–159 (2000). [CrossRef]
7. M. Mizusaki, T. Miyashita, and T. Uchida, “Behavior of ion affecting image sticking on liquid crystal displays under application of direct current voltage,” J. Appl. Phys. 108,104903 (2010). [CrossRef]
8. M. Mizusaki, T. Miyashita, and T. Uchida, “Kinetic analysis of image sticking with adsorption and desorption of ions to a surface of an alignment layer,” J. Appl. Phys. 112,044510 (2012). [CrossRef]
9. M. Mizusaki, T. Miyashita, and T. Uchida, “The mechanism of image sticking on LCD and its evaluation parameters related to LC and alignment materials,” SID Symp. Dig. Tech. Pap. 37, 673–676 (2006). [CrossRef]
10. S. -C. Park, K. -H. Lim, S. -H. Choi, and H. -S. Soh, “Quantitative analysis of image Sticking in LCDs,” SID Symp. Dig. Tech. Pap. 38, 1042–1045 (2007). [CrossRef]
11. H. J. Park, L. Lai, S. H. Lin, and K. H. Yang, “Analysis of IPS mura, image-sticking and flicker caused by internal DC effects,” SID Symp. Dig. Tech. Pap. 34, 204–207 (2003). [CrossRef]
12. S. Naemura, “Liquid-crystal-material technologies for advanced display applications,” J. Soc. Inf. Display 8, 5–9 (2000). [CrossRef]
13. Y. Park, S. Kim, and E. Lee, “A study on reducing image-sticking artifacts in wide-screen TFT-LCD monitors,” J. Soc. Inf. Display , 15, 969–973 (2007). [CrossRef]
14. V. P. Vorflusev, H.-S. Kitzerow, and V. G. Chigrinov, “Azimuthal surface gliding of a nematic liquid crystal,” Appl. Phys. Lett. 70, 3359–3361 (1997). [CrossRef]
15. R. Yamaguchi and S. Sato, “Tortional torque effects of twisted nematic bulk on the polymer surface alignment,” Mol. Cryst. and Liq. Cryst. 367, 379–386 (2001). [CrossRef]
18. P. Vetter, Y. Ohmura, and T. Uchida, “Study of memory alignment of nematic liquid crystals on polyvinyl alcohol coatings,” Jpn. J. App. Phys. 32, L1239–L1241 (1993). [CrossRef]
19. R. Barberi, I. Dozov, M. Giocondo, M. Iovane, Ph. Martinot-Lagarde, D. Stoenescu, S. Tonchev, and L. V. Tsonev, “Azimuthal anchoring of nematic on undulated substrate: elasticity versus memory,” Eur. Phys. J. B6,83–91 (1998). [CrossRef]
20. S. Faetti, M. Nobili, and I. Raggi, “Surface reorientation dynamics of nematic liquid crystals,” Eur. Phys. J. B11,445–453 (1999). [CrossRef]
21. I. Dozov, D. N. Stoenescu, S. Lamarque-Forget, Ph. Martinot-Lagarde, and E. Polossat, “Planar degenerated anchoring of liquid crystals obtained by surface memory passivation,” Appl. Phys. Lett. , 77, 4124–4126 (2000). [CrossRef]
22. D. N. Stoenescu, I. Dozov, and Ph. Martinot-Lagarde, “Long-time behavior of the azimuthal anchoring strength and easy axis gliding of nematic liquid crystal,” Mol. Cryst. Liq. Cryst. , 351, 427–434 (2000). [CrossRef]
23. A. Romanenko, V. Reshetnyak, I. Pinkevich, I. Dozov, and S. Faetti, “Magnetic field induced director reorientation in the nematic cell with time-dependent anchoring due to adsorption/desorption of LC molecules,” Mol. Cryst. Liq. Crystl , 439, 1867–1888, (2005).
24. Y. Momoi, K. Tamai, K. Furuta, T. -R. Lee, K. -J. Kim, C. -H. Oh, and T. Koda, “Mechanism of image sticking after long-term AC field driving of IPS mode,” J. Soc. Inf. Display 18, 134–140 (2010). [CrossRef]
25. T. Suzuki, J. Matsushima, Y. Sakaki, M. Sugimoto, H. Tanaka, C. Mizoguchi, S. Onda, K. Mimura, and K. Sumiyoshi, “High contrast ratio in-plane-switching TFT-LCD with ion beam irradiated polyimide film as LC alignment layer,” IDW/ASIA DISPLAY Dig., 57–60 (2005).
26. J. Stöhr, M.G. Samant, A. Cossy-Favre, J. Diaz, Y. Momoi, S. Odahara, and T. Nagata, “Microscopic origin of liquid crystal alignment on rubbed polymer surfaces,” Macromolecules 31, 1942–1946 (1998). [CrossRef]
28. M. Fukuhara and A. Sampei, “Temperature dependence of elastic moduli and internal dilational and shear frictions of polyimide,” J. Polym. Sci., Part B, Polym. Phys. , 34, 1579–1582 (1996). [CrossRef]
29. S. H. Cho, G. Kim, T. J. McCarthy, and R. J. Farris, “Orthotropic elastic constants for polyimide film,” Polym. Eng. Sci. , 41, 301–307 (2001). [CrossRef]
30. M. Kwak, D. Han, H. Kwon, S. Choi, Y. Choi, D. Koo, K. Kim, and B. Kim, “Studies of the directional property on rubbed alignment films by rubbing condition,” Mol. Cryst. Liq. Cryst. 546, 1481–1486 (2011). [CrossRef]