In our experiments mentioned above, we used two different dimensions of nano-particles (mean diameter: 50 & 90 nm) to create protrusions in proposed NE-Pi-cells. The distribution of the nano-particles on the ITO glass was measured by AFM. The diameters of the protrusions were found in the region of 200-300 nm, with
the height of 150-270 nm after PI coating with respect to 0.2 wt% 50 nm and 90 nm nano-particles. The AFM photos of 0.2 wt% nano-particle (90nm) distributions, fabricated by Method 1 & 2 are shown in Figs. 4-4 and 4-5 respectively. As the results, the aggregations were improved greatly and the particle density was also increased from 6.78 to 11.6μm-2 by using Method 2, shown in Fig. 4-6. Some aggregations of SiO2 particles were still found when the concentration was higher than 0.2 wt%; not only 90 nm but also 50nm nano-particle distribution shown in Fig. 4-7. Thus, the concentration of nano-particles must be confined to no more than 0.2wt% in the process of NE-Pi-cells. The other AFM photos of 50 and 90 nm nano-particles fabricated in different concentrations are shown in Figs. A-1, A-2 ~ A-8 (Appendix A).
Fig. 4-4 0.2wt% 90nm nano-particles are coated twice by Method 1, the particle density is about 6.78μm-2. The aggregations are serious.
Fig. 4-5 0.2wt% 90nm nano-particles are coated by Method 2, the particle density is about 11.6μm-2. The aggregation issue is improved greatly.
90nm nano‐particle
0 2 4 6 8 10 12 14 16
0 0.05 0.1 0.15 0.2 0.25
Concentration (wt%)
Density
‐2 (μm)
method 1 method 2
Fig.4-6 Concentration (wt%) v.s. Density(μm-2) of 90nm nano-particles fabricated by Method 1 and 2, respectively. The density can be increased from 6.78 to 11.6 μm-2.
Fig. 4-7 0.2wt% 50nm nano-particles are coated by Method 2. Some aggregations were still found when the concentration is over 0.2wt%.
In order to investigate densities of different dimensions of nano-particles, we proposed a specific metric “protrusion volume ratio,” which was the total volume of protrusions covered with PI layer in a maximum affected volume. The maximum affected volume was defined by maximum nano-particle size coated on the surface, to unify the densities of different nano-particle sizes (Maximum size of 90nm was used in the experiment). The protrusion volume ratio whose physical meaning was similar to protrusion density was defined as following Eq. (4-1). The illustration is shown in Fig. 4-8. In the experiments, the maximum nano-particle size and highest protrusion covered with PI are 90nm and 103.5nm, respectively.
Protrusion volume ratio (%)
[ ]
Fig. 4-8 The illustration of calculated volume of a nano-particle covered with PI on the surface.
By using the novel metric, the unified parameter for different nano-particle sizes can be achieved. It’s useful for the investigation between transition rate and nano-particle enhanced effect. The relationship between concentration (wt%) and protrusion volume ratio of the nano-particle was investigated in Fig. 4-9. The average protrusion volume ratios of the nano-particles possess linear relationship with the concentration of SiO2 solution.
0 5 10 15 20 25
0 0.05 0.1 0.15 0.2 0.25
concentration (wt%)
protrusion volume ratio (%) 50nm 90nm
Fig. 4-9 Concentration (wt%) v.s. protrusion volume ratios of the nano-particles.
h
r
θr‐h
a
PI
Nano‐particleh
On surface volume Protrusion with PI
4.4 Transition Time v.s. Nano-particle Density
In order to observe the transition time by video camera, the 6V AC square wave was applied on test cells. As shown in Fig. 4-10, the nucleation process of 5μm NE-Pi-cell can be clearly identified. In Fig. 4-10 (b) and (d), there are two nucleation processes; one is elevated splay stay, also known as asymmetry splay state, transits to bend state (Vapp > Vcr) and another one is twist state relaxes to ground splay state (Vapp = 0), respectively.
Fig. 4-10 Nucleation process of a 5 μm NE-Pi-cell. (a) Splay state, (b) Ha-to-Bend state, (c) Bend state, and (d) Twist-to-Splay state.
The transition times of splay-to-bend and twist-to-splay states can be calculated from captured video clips and the results of 3 and 5μm cells with different protrusion volume ratios were investigated in Fig. 4-11 (a) and (b) respectively. In ~5μm NE-Pi-cell, the both transition time of splay-to-bend and the twist-to-splay states were improved from 141 and 228sec to almost 0 and could not be measured, when the protrusion volume ratio was over 3%. The transition time was also related with the cell gap, so the ~3μm cell transited faster than the 5μm cell. As the results, the higher protrusion volume ratios of the nano-particles, the less transition times of the
(a) splay (b) Ha to bend
(c) bend (d) twist to splay
splay-to-bend and twist-to-splay states. The fitting curves of transition time were calculated by exponential functions (correlation coefficients R2 were about 90%), which were matched with the nucleation theory.
Gap = 4.67μm
Fig. 4-11 The relationship between protrusion volume ratios and transition times of (a) 5μm cells (gapave. = 4.67μm), and (b) 3μm cells (gapave. = 2.8μm).
Furthermore, as the protrusion volume ratio was over ~3% and (concentration ≧ 0.1wt %), the response time of the NE-Pi-cells were raised from ~3 to ~13ms and
~1.5 to ~6ms with respect to 5 and 3μm cells, no matter nano-particles of 50nm or 90nm. The concentration limitation implied the protrusion quantities affect the potential nucleation sites. According to the detail investigation, we found the protrusion density was also limited under 4.2μm-2. The protrusion density raise meant the potential nucleation sites were increased. If the protrusion density was over the limitation, the morphology of alignment layers were affected, and the LC directors could not be arranged quickly toward the same direction when applied voltage was released. In other words, the raises of protrusion volume ratio and protrusion density were limited by the response time. Anyway, the levels of the response time of NE-Pi-cells were not related with nano-particle sizes but related with their cell gaps significantly. The detail data of response time and protrusion densities for 50 nm and 90nm nano-particle are summarized in Tabs. 4-3 (a), (b), Tabs. 4-4 (a) and (b) individually. Then, the response time of ~5μm NE-Pi-cells coated by nano-particles of 50 and 90nm with different nano-particle concentrations and protrusion densities are shown in Figs. 4-12 and 4-13 respectively.
Tab. 4-3(a) The response time of ~5μm cells coated by 50nm nano-particles.
Concentration (wt%)
no nano‐particle 0 0 4.89 3.22 0.2102 3.43
0.004 0.35 1.9 4.72 3.2 3.35E‐04 3.20
0.0125 0.49 2.8 4.54 3.25 2.69E‐04 3.25
0.05 1.24 4.2 4.68 2.98 3.03E‐04 2.98
0.1 2.30 11.2 4.35 13.8 3.45E‐04 13.80
0.2 4.11 17.7 4.41 14 2.57E‐04 14.00
Tab. 4-3(b) The response time of ~3μm cells coated by 50nm nano-particles.
no nano‐particle 0 0 3.21 1.677 0.09378 1.77
0.004 0.35 1.9 2.8 1.30E+00 2.34E‐04 1.30
0.0125 0.49 2.8 3.03 1.5 1.64E‐04 1.50
0.05 1.24 4.2 2.44 1.2 2.03E‐04 1.20
0.1 2.30 11.2 2.76 5.3 2.31E‐04 5.30
0.2 4.11 17.7 2.88 5.4 2.30E‐04 5.40
Tab. 4-4(a) The response time of ~5μm cells coated by 90nm nano-particles.
Concentration (wt%)
no nano‐particle 0 0 4.89 3.22 0.2102 3.43
0.004 0.47 0.3 4.71 3.17 0.2451 3.42
0.0125 0.88 0.6 4.81 3.54 0.1781 3.72
0.05 3.09 2.3 4.74 3.25 0.256 3.51
0.1 9.75 6.2 4.89 12.86 0.3643 13.22
0.2 21.5 11.6 4.62 12.32 0.2914 12.61
Tab. 4-4(b) The response time of ~3μm cells coated by 90nm nano-particles.
Concentration (wt%)
no nano‐particle 0 0 3.21 1.677 0.09378 1.77
0.004 0.47 0.3 2.94 1.41 0.08812 1.50
0.0125 0.88 0.6 2.43 1.41 0.1194 1.53
0.05 3.09 2.3 2.83 1.085 0.1652 1.25
0.1 9.75 6.2 2.76 6.4 0.1291 6.53
0.2 21.5 11.6 2.77 6.56 0.07999 6.64
Fig. 4-12 The response time of ~5μm NE-Pi-cells coated by nano-particles of 50 and 90nm with different nano-particle concentrations.
Fig. 4-13 The response time of ~5μm NE-Pi-cells coated by nano-particles of 50 and 90nm with different protrusion densities.
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
0 0.05 0.1 0.15 0.2 0.25
Concentration (%)
Response time (ms)
90nm 50nm
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
0 5 10 15 20
Protrusion Density (μm‐2)
Response time (ms)
90nm 50nm
4.2
However, under the limitation of the protrusion density, the response time and voltage versus transmittance curves (V-T curve) of the NE-Pi-cells were almost the same as the conventional Pi-cell driven over the critical voltage (Vcr). And the critical voltages of NE-Pi-cells are only related with cell gaps and independent of nano-particle sizes. The V-T curves and critical voltages of the NE-Pi-cells with different nano-particle concentrations are shown in Fig. 4-14 and Fig. 4-15 with respect to 50nm and 90nm nano-particle treatments. Moreover, both transition times of splay-to-bend and twist-to-splay states had 99.9% reductions without high voltage pulse driving.
To further investigate splay-to-bend transition, there was strong negative correlation between the transition time and driving voltage. Due to the high reduction rate of transition time in a NE-Pi-cell, it’s difficult to recognize the varieties of transition times with different protrusion volume ratios. In order to observe easily, the transition time needs to be measured under lower driving voltage which is just higher than the critical voltages to slow the transition rates. The results are shown in Figs.
4-16 (a) and (b) with respect to ~5μm and ~3μm cells. As the Fig. 4-16 (a), in ~5μm cells, the transition time is strong correlated with protrusion volume ratios by negative exponential function and the R2 (the square of correlation coefficient), means coefficient of determination in linear regression, is over 90%. However, the R2 is lower in ~3μm cells shown in Fig. 4-16 (b), because of larger measuring deviations of transition times and worse cell uniformities for the cells with low cell gaps.
50nm‐‐gap=3μm
Fig. 4-14 The V-T curves of different nano-particle (50nm) concentrations with respect to (a) 3μm cell gap, and (b) 5μm cell gap.
Vcr ~ 1.7 V
Vcr ~ 2.0 V
90nm‐‐gap=3μm
Fig. 4-15 The V-T curves of different nano-particle (90nm) concentrations with respect to (a) 3μm cell gap, and (b) 5μm cell gap.
Vcr ~ 2.0 V Vcr ~ 1.7 V
Gap = 4.67 μm Splay‐to‐Bend (V=0→3V)
y = 868.6e‐0.0964x R2 = 0.9166
0 200 400 600 800 1000 1200
0 5 10 15 20 25
protrusion volume ratio (%) (a)
transition time (sec)
Gap = 2.8 μm Splay‐to‐Bend (V=0→2.5V)
y = 508.15e‐0.1201x R2 = 0.8333
0 100 200 300 400 500 600 700
0 5 10 15 20 25
protrusion volume ratio (%) (b)
transition time (sec)
Fig. 4-16 The relationship between protrusion volume ratios and transition time of (a)
~5μm cells driven at 3V (Vcr ~ 2V), and (b) 3μm cells driven at 2.5V (Vcr ~ 1.7V).
4.5 Discussions
In the novel NE-Pi-cell, the nanostructures are coated on the ITO glass substrates before PI coating, which differ from prior art by mixing the nano-particles with PI solution. In nanostructure fabrications, the nano-particles need to be uniformly distributed on the glass substrate without aggregation. Therefore, the nano-particles can be covered with PI layer completely, and the anchoring energy between the side chains of alignment layer (i.e. PI layer) and LC molecules will not be changed under the protrusion density limitation. Based on the concept, the critical voltage of a NE-Pi-cell was unchanged as a conventional Pi-cell. The mechanism of nanostructure enhanced transition rate can be confined in heterogeneous nucleation, but not anchoring effect related. Since the cell structure is purified, the discussion will be focus on the effects and limitations caused by densities, distribution and dimensions of nano-particles.
In the study, to find an appropriate solvent to disperse nano-particles uniformly is very important. In the beginning, the nano-particles were dispersed in DI water whose surface tension was very large with respect to nano-particles. Thus, the high spin speed had to be implemented to avoid the nano-particle aggregations that resulted in the limitation of the nano-particle density on the surface. However, to raise the nano-particle density is the key point in proposed NE-Pi-cells; the higher densities of the nano-particles, the faster transition rates under the response time limitations.
Finally, the solvent composed of 10% PGMEA and 90% EG was found to resolve the fabrication issue of high nano-particle density treatment.
Since high protrusion volume ratio and protrusion density (μm-2) would seriously affect the morphology of alignment layers, the LC directors could not be
arranged perfectly toward the same direction. As the results, the rise time of a NE-Pi-cell became longer when the LC molecules were relaxed, which investigated in Tab. 4-3 and 4-4. The phenomena explained that the imperfect alignment resulted in the disordered LC flow to increase the relaxation time. Although, the protrusion volume ratio was limited; the transition time was still reduced to about 99.9% (i.e.
splay-to-bend transition is invisible under 6V driving.) when the protrusion volume ratio was around 3%. The response time of NE-Pi-cells could be remained as the same level of conventional Pi-cells.
The photographs of nucleation processes of a conventional Pi-cell and a NE-Pi-cell (2.5x2cm2, cell gap~4.8μm) are shown in Fig. 4-17. The transition time was counted by the sequential video images captured by video camera under 6V driving. The frame rate was 25 fps (frames per second). It’s convenient to observe the whole transition process and nucleation from spherical cluster appearing.
Fig. 4-17 (a)~(d) and (e)~(h) are the transition processes of a conventional Pi-cell and a NE-Pi-cell, respectively.
(e) Splay (f) Ha→Bend (g) (h)
Bend
Bend Twist→Splay
Splay
Twist Bend
Splay Ha→Bend Twist→Splay
4.6 Summary
In this study, the transition time of splay-to-bend and twist-to-splay states have about 99.9% reduction with protrusion volume ratio over 3%. The nucleation process could be reduced to nearly invisible with the novel cell configuration and fabrication.
Moreover, the limitation of the protrusion density was clarified by 4.2μm-2. However, the electro-optical properties of the NE-Pi-cell were remained and almost independent of the nanostructures. In other words, the proposed NE-Pi-cell could not lead the critical voltage reduced or eliminated. It only enhanced the transition rate of the cells.
Therefore, we suggested another method — RMM-Pi-cell to resolve the transition issue.
Chapter 5
Splay-to-Bend Transition-Free Reactive Monomer Modified Pi-cell
A reactive monomer modified Pi-cell (RMM-Pi-cell) comprising a layer of liquid crystal reactive monomer on one surface was prepared to control the surface pretilt angle. The simulation results suggested that a transition-free Pi-cell can be prepared by asymmetrical cell with one 8° pretilt angle and the other surface greater than 47°
when the cell gap smaller than 4μm. The nematic reactive monomer (RM) layer has molecular average tilt angle over 80o which allowed the liquid crystal molecules arranged in favored bend state in the asymmetrical cell. The critical voltages cannot be found in all 3μm RMM-Pi-cells. The cell retardation data confirmed the initial bend orientation with zero voltage applied. Moreover, the light leakage of dark state was reduced. The contrast ratio (CR) of RMM-Pi-cell was improved by a factor of 11 compared with an original Pi-cell without using compensation film.
5.1 Introduction
Pi-Cell possesses the advantages of wide viewing angles and fast response time for liquid crystal display (LCD) application. To be useful, the conventional Pi-cells are challenged by two major issues: the splay-to-bend transition, for one, is inevitably present. Without applying a higher voltage, the active bend configuration relaxes into splay configuration [13][25][58]. The other drawback is the low contrast ratio (CR) caused by the light leakage in the dark state in which LC molecules can not be
vertically arranged under strong boundary condition. Furthermore, the optical axis of the LC molecules diverges by 45 degrees from the absorption axes of the polarizers.
Therefore, a Pi-cell can cause a high degree of optical retardation, which in turn results in increased light leakage in dark state. Many research efforts have been intensively involved in solving the splay-to-bend transitional issue by using various methods: multi-domain alignment [24][63-64], an electrical twist fieldin Pi-cell [66], polymer-stabilized walls [26][67-68], and increasing surface pre-tilted angle [13]
[24]…, etc.
Fig. 5-1 The scheme of RMM-Pi-cell under (a) 0V and (b) 6V.
(a)
Polarizer Glass ITO
Alignment layer
RMM‐19B Alignment layer ITO
Glass Analyzer LC
(b)
Polarizer Glass ITO
Alignment layer
RMM‐19B Alignment layer ITO
Glass Analyzer LC
In this study, we report a transition-free reactive monomer modified Pi-Cell (RMM-Pi-cell),assembled with RM at one side and buffed PI surface at the other side, configured as Fig. 5-1, in which the splay-to-bend transition is suppressed and contrast ratio is enhanced by a factor of 11 than a conventional Pi-cell [69-70].
5.2 Experiments and Simulations
5.2.1 RMM-Pi-cell Fabrication
The polyimide alignment film (PI: PIA-5580-01, Chisso Co.), compounded into the solvent ( PI: solvent = 3:1), was prepared by spin-coated on a 2 cm × 2.5 cm indium-tin-oxide (ITO) glass and hard baked at 220 oC for 1 hr. The 8°~12° pre-tilt angles were obtained after buffing the cured PI film. The propylene glycol mono-methyl ether acetate (PGMEA) solutions, comprising various concentrations of reactive monomer (RMM-19B, Merck Chemicals Ltd., see the Appendix C), were spin-coated on a PI buffed ITO glass. Then, the reactive monomer (RM) film sample was exposed by UV lamp and baked at ~45°C for 5 minutes at the same time. The thickness of RM film was controlled by solution’s concentration and spin coater’s speed. The process conditions of RM films are listed in Tab. 5-1. The film thickness was measured by the height of the cross section, using atomic force microscope (AFM). The relationships of film thicknesses and coater’s rotation speeds with respect to different concentrations of the RM solution are shown in Fig. 5-2.
Tab. 5-1 The spin-coating conditions of RM films.
Step 1: RM film coating
Spin (rpm) Time(sec)
1st spin 500 30
6000 4000 2nd spin
(for different thickness)
2000
60
Step2: (Baking45 °C + UV Exposure)x 5min
RM solution’s concentration: 4wt%, 2wt% and 1wt%
50 70 90 110 130 150 170 190
0 2000 4000 6000 8000
coater's spin speed (rpm)
4wt%
2wt%
1wt%‐
RM FilmThickness(nm)
Fig. 5-2 The relationships of film thicknesses and spin speeds of coater with respect to different concentrations of the RM solution.
RMM-Pi-cell was assembled with RM on one side and buffed PI surface on the other side, as shown in Fig. 5-1. The cell gap was controlled at either ~3 or 5 μm. The liquid crystal (LC: ZCE-5096XX, Chisso Co., see the Appendix B.) was filled by
capillary force into the cell for electro-optical characterizations. The retardations of RMM-Pi-cells and RM thin films were measured by Soleil-Babinet Compensator. The fabrication processes are illustrated in Fig. 5-3.
Fig. 5-3 RMM-Pi-cell fabrication processes in laboratory.
5.2.2 Critical Pretilt Angle Estimation
In order to verify the initial bend orientation without splay configuration in RMM-Pi-cell, we calculated the lowest pretilt angle requirements and probed the RM’s surface pretilt angle, in addition to cell retardation’s measurement. In the calculation, the symmetrical Pi-cell was considered first, the relationship between the driving voltage and electric field can be simplified as E ~ U/d. The Gibbs free energy
Washing
Sealant with spacers Sealant dispensing
LC injection (capillary attraction)
Assembling & LC injection
Baking RMM-19B
+PGMEA Baking
UV lamp
of spay and bend cells can be calculated by Eq. (5-1) [13][71]: dielectric constant in vacuum and dielectric anisotropy, respectively. θ (z) is the LC tilt angle on z direction perpendicular to the glass substrate. The critical pretilt angle, αc, for a transition-free Pi-cell can be derived from Eq. (5-2) [13][72]:
(
K33−K11)
sin(2αc)+(
K33+K11)(
π −4αc)
=0 (5-2)Based on the calculation, the relationship between K33/K11 and critical pretilt angle is shown in Fig. 5-4. The critical pretilt angle is around 46°, which calculated from the LC used in experiment (LC: ZCE-5096XX, K11=9.8 pN, K33=11.8 pN at 20°C), for a symmetrical Pi-cell to form bend state configuration.
Fig. 5-4 The relationship between K33/K11 and critical pretilt angle.
0
5.2.3 Asymmetry Pretilt Angle Combination Simulations
By varying surfaces pretilt angle on both top and bottom surfaces for asymmetrical cells, the Voltage vs. Transmittance (V-T) curves simulated by DIMOS program suggested that the bend state configuration can also be achieved by surface pretilt angle of top/bottom substrates at 8°/47° and 10°/45° in a 4 μm cell. The V-T curves of the two combinations due to the Gibbs energy of bend state configuration ( 6.6 μJ/m2) lower than spray state are investigated in Fig. 5-5 and 5-6. As the results, the simulation result suggested that a transition-free Pi-cell can be prepared by asymmetrical cell with one regular 8° pretilt angle and the other surface greater than 47° when its cell gap smaller than 4 μm.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
0 2 4 6 8
Applied voltage (V)
Transmittance (a.u.)
8°/47°‐R 8°/47°‐G 8°/47°‐B 10°/45°‐R 10°/45°‐G 10°/45°‐B
Cell Gap~4μm
Fig. 5-5 The simulations of V-T curves of two pretilt angle combinations.
7.158 7.169
6.789
6.603 6.594 6.5
6.6 6.7 6.8 6.9 7 7.1 7.2 7.3
8°/8° 8°/35° 8°/45° 10°/45° 8°/47°
Pretilt Angle Combination (degree)
Gibbs Energy
2 (μJ/m)
Transition Free
Fig. 5-6 The Gibbs energies of different pretilt angle combinations.
5.3 The Investigation of RM films’ Retardations
The critical voltages to hold bend state measured at λ= 633nm were 0.6, 0.2 and 0 V for RM’s thickness of 80, 110 and 134 nm in 5 μm RMM-Pi-cells, respectively.
On the contrary, the state transition was not found even in the thin RM layer for all 3 μm cells. Above mentioned results were shown in Tab. 5-2. The experimental results confirmed the DIMOS simulation. Further investigation was carried out to determine the RM’s surface pretilt angle. RMM-19B’s ΔnRM (at λ=452nm), 0.1449, was obtained from a 1.58 μm homogenous planar cell. Based on the measured result (ΔnRM
=0.1449), the retardation of a 200 nm homogenous RM thin film should be in 28.98 nm. The thin RM film’s retardations, however, were observed in 0.17 and 0.09 nm for the thickness of 237 and 120 nm, respectively. The discrepancy suggested that the liquid crystalline molecular orientation in RM film was highly tilted. In order to
simplify the liquid crystal’s pretilt angle calculation, we assumed a uniform molecular tilt orientation for RM film. In the uniaxial RM case, the index ellipsoid of the RM, as illustrated in Fig. 5-7, is represented by the Eq. (5-3). In the Fig. 5-7, the cross section of the RM cut by a plane, which is perpendicular to the wave vector, shows an ellipse
simplify the liquid crystal’s pretilt angle calculation, we assumed a uniform molecular tilt orientation for RM film. In the uniaxial RM case, the index ellipsoid of the RM, as illustrated in Fig. 5-7, is represented by the Eq. (5-3). In the Fig. 5-7, the cross section of the RM cut by a plane, which is perpendicular to the wave vector, shows an ellipse