Chapter 3 Measurement Systems
3.5 ConoScope
Fig.3.5.The schematic diagram of ConoScope.
The ConoScope (autronic-MELCHERS GmbH) can be used for visual performance evaluation, such as luminance, contrast ratio, color shift, gray scale and many characteristics.
The basic operating principle of the ConoScope is described as follows. A typical scanning device (so-called “gonioscopic system”) would scan the half cone above the display to detect the variations of luminance and color for each specific direction. Plotting each azimuthal angle on a circle with the radius from the center additionally indicates the polar angle results in the so-called polar coordinate system. Using such a polar plot to mark for each direction luminance and /or color will result in (in the case of color) a colored figure.
An arrangement of several lenses (here represented by single lens) modifies the light propagation directions in a way, that all beams emerging from the sample in the same direction will meet in one spot in the focal plane (or within the “conoscopic figure”) corresponds to one specific direction of the viewing cone. The conoscopic figure directly shows color and luminance as they would have been plotted in a polar coordinate system as described under Fig.3.6. In our measurement, we used the collimated illumination as the backlight for the transmissive mode to analyze the luminance, contrast ratio, and light distribution of the FLCD.
Fig.3.6. The function of the ConoScopic lens system.
Chapter 4
Experiments and Results
4.1 Introduction
For the fast sub-millisecond switching FLCD, the major problem degrading the performance is the alignment defects that cause light leakage in the dark state and result in low contrast. In this thesis, three alignment issues that directly relate to the orientation of SSFLC were discussed. One of the issues is the sign of the surface polarity which could probably results in Ps attraction [33] or repulsion; another one is surface roughness which affects the effective surface pretilt [25][34]; the other one is the surface rubbing conditions, that is, rubbing direction and rubbing strength.
As mentioned in the section 2.2.3, the formation of horizontal chevron defect results from the random orientation of Ps which means that layer structures in opposite directions show up randomly. In the following sections, a new hybrid cell structure was applied to effectively orient the Ps and to obtain a uniform uni-direction layer. On the other hand, to achieve a zigzag free orientation the C2 uniform structure is preferred because the alignment is relatively simple and it is much easier to avoid twist state in C2 than in C1 [8]. In C2 state small surface roughness is an essential criteria, thus two methods were tried to improve the surface roughness.
4.1.1 Cell Fabrication Process
The flow chart shown as below is the fabrication process of our liquid crystal cell:
A. Each piece of glass is rinsed with detergent and rubbed by hands carefully. Then, wash the glass with DI water until the water flowing along the surface smoothly. After that, put the glass loaded with a holder into hot DI water and shake it with ultra sonic for 30 minutes.
B. Blow the surface of glass with nitrogen gas to rip off the water, and then bake for 1 hour at 110℃. Next, use UV-ozone to treat surface for 10 minutes (that is for better adhesion of alignment layer).
C. Put the clean ITO glass onto the spin coater. Drop the solution of alignment material (make sure the solution to cover the whole glass), then spin. The procedure of spin-coating is as follows:
2.5wt%PVA Polyimide
Speed Time Speed Time
1st spin 500rpm 10s 1st spin 500rpm 10s
2nd spin 1500rpm 20s 2nd spin 5000rpm 20s
After spin-coating, we need to bake the samples for full adhesion for a short time, which is 30 minutes at 110℃ for PVA and 30 minutes at 200℃ for polyimide. PVA (Mw Cleaning of
ITO glass
Coating of
alignment film Baking Rubbing
Spacer
13,000-23,000, 98% hydroldyzed) is from Sigma-Aldrich, and polyimide (PIA-X201-G01) is from Chisso. The film thickness, which is measured by AFM (from Digital Instruments), of 2.5wt% PVA is 1000Å, and the thickness of 50wt% polyimide is about 350Å.
D. Rub the samples with rubbing machine. The rubbing strength is different depending on the LC material, alignment film, and pretilt angle. The rubbing condition used is as below:
E. Drop a small amount of UV glue (NOA-65, from Norland) mixing with 1.6um spacer at the four corner of the bottom plate. Cover the top plate and press. Next, place the cell under a UV lamp for 5 minutes to fully cure the glue, and then an empty cell is done.
F. Measure the cell gap of an empty cell by spectrometer UV-Vis 650 (from Perkin Elmer) using interferometric method. The details will be explained in section 3.2.4.
G. Heat up the cell and LC material to the temperature above the clearing point. Inject the LC material from the edge of the cell until the LC is full of the gap.
H. Anneal the cell carefully with programmable hot plate from high temperature to room temperature. The annealing process depends on the phase sequence of the LC material.
I. Solder the wire at the ITO contact, and then the test cell is available for a mount of electro-optical measurements.
4.1.2 Measurement of Sign of Surface Polarity
Since we proposed the hybrid cell structure to align Ps in section 2.4, the sign of surface polarity played an important role. In order to check the sign of the surface polarity of
PVA Polyimide
Pile impression 0.23mm Pile impression 0.2mm Rotation speed 300rpm Rotation speed 300rpm Advancing speed 7.3mm/s Advancing speed 7.3mm/s
alignment layer, two crossed-rubbing cells were made. One of the cells has PVA coating on both substrates, as in Fig.4.1 (a), and the other one has PI coating on both substrates, as in Fig..4.1 (b). After filling the FLC material with positive Ps into the cross-rubbing cell, a twisted structure was formed. According to the sign of surface polarity, the Ps of molecules at the top and bottom surface directed inward or outward the cell. In addition, based on the switching nature of FLC, a specific layer direction should be observed in the cross-rubbing cell and layer normal of FLC in PVA cell would be perpendicular to one in PI cell. The FLC material used is R3206, with Ps~18 nC/cm2. After injecting the FLC and cooling down the cells to room temperature, we found that the layer was formed and the layer normal direction of PVA cell was perpendicular to PI cell, as shown in Fig.4.2. The results matched the assumption in Fig.4.1. According to the results, sign of the surface polarity was confirmed.
Fig.4.1. The illustration of twisted structure in crossed-rubbing (a) PVA cell and (b) PI cell.
R,S represent rubbing direction and sign of surface polarity.
(a)
(b)
Fig.4.2. Pictures of the cross-rubbing cells taken by POM. Pictures at the right are the magnified layer structures. (a) Layer structure in PVA cell. (b)Layer structure in PI cell.
4.1.3 Observation of CDR Type FLC Cells
Since the relationship between Ps and the sign of the surface polarity was revealed, the uni-direction layer model could be applied first on the test cell with CDR type FLC. Two FLC materials would be used─R2301 (from Clariant) and R3206 (from AZ Electronic Materials).
R2301 has phase sequence, I 86.8-84.8 N* 64.7 SmC* and R3206 has phase sequence, I 107-105.4 N* 78.4 SmC* -12(-6) Cr. The filling process was as below: First, the FLC material was injected to the hybrid cell by capillary force in isotropic state. Then the test cell was annealed in N* phase for 5 minutes and annealed in SmC* phase for 30 minutes. Finally, the test cell was cooled down to the room temperature.
OFF STATE
ON STATE
Fig.4.3. (a)(c) is the OFF and ON state of R2301 in hybrid cell, and (b)(d) are the OFF and ON state of R3206 in hybrid cell.
After the filling of the two FLC materials, the test cells were observed under polarizing optical microscope (POM) and it was found that a uni-direction layer structure was formed both with R2301 and R3206, as shown in Fig. 4.3. Fig. 4.3 (a)(b) showed the OFF states between cross-polarizer and Fig. 4.3 (c)(d) showed the ON states with application of 100 Hz square wave. As compared with Fig.4.3 (a), the microdomain structure in Fig.4.3 (b) was more obvious, although uni-direction layer was obtained in both cells.
Contrast and viewing angle
To see how much the contrast can be raised by the hybrid cell structure, we have measured the symmetrical cell 71 with R3206, hybrid cell 97 with R3206, symmetrical cell 59 with R2301 and hybrid cell 98 with R2301, as shown in Fig.4.4-Fig.4.7. The symmetrical cell 59 and 71 with horizontal chevron defect showed poor contrast (CR~29 for cell59;CR~18 for cell71). On the other hand, an improved contrast (CR~86) was shown in cell 97 with R3206, and the highest contrast (CR~144) was performed in cell 98 with R2301. As for viewing angle, because of the quasi in-plane switching and small surface pretilt the viewing angle could reach over ±70°.
Besides, the switching mechanism could be observed in conoscopic pictures. In Fig.4.5 and Fig.4.7, we saw the asymmetric light distribution of the ON state which was resulted from a tilted uni-layer structure, as shown in Fig.2.12 (a)(b), leading to an asymmetric switching cone. On the contrary, as shown in Fig.4.4 and Fig.4.6, the light distribution of the ON state showed symmetry which was considered as the averaged switching of two different layers in horizontal chevron. Thus, according to the conoscopic pictures, the uni-layer structure was also confirmed.
Fig.4.4. CR v.s. viewing direction of symmetrical cell 71 with R3206.
Fig.4.5. CR v.s. viewing direction of hybrid cell 97 with R3206.
Fig.4.6. CR v.s. viewing direction of symmetrical cell 59 with R2301.
Fig.4.7. CR v.s. viewing direction of hybrid cell 98 with R2301.
4.1.4 Observation of FGLC Series Cells
The self-made FGLC mixtures [37] show wide smectic C temperature range and good alignment ability. The 2%FGLC-1 mixture shows V-shaped switching and has SmA phase in the phase sequence. We filled 2%FGLC-1 mixture into hybrid (PVA/PI) cell, and the orientation of uni-direction layer was observed under POM as shown in Fig.4.8. Picture at the right is the ON state which is driven by a 100Hz 17V square wave, and the picture at the left is the OFF state under crossed polarizer. Although hybrid cell structure seemed successful on 2%FGLC-1 mixture, we found that after driving for a long time some stripe defect appeared, as shown in Fig.4.9, and the higher the driving voltage was applied the faster the stripe defect showed up. The stripe defect usually grows from the boundary between the two different domains, but sometimes grows from the center of the well-aligned domain. The growing direction of defects is parallel to the rubbing direction. The reason of defect formation will be discussed in section 4.4.
Fig.4.8. The alignment of 2%FGLC-1 in 1.78um hybrid cell. (a) Dark state. (b) Bright state.
Fig.4.9. (a) Before driving. (b) After driving, the stripe defect grows along the rubbing direction during driving.
4.1.5 EO Properties of FGLC Series Cells
To see if the hybrid cell structure affect the electro-optical properties of SSFLC or not, the response time of 2%FGLC-1 in 2um hybrid cell has been measured using various frequencies, as shown in Fig.4.11. Also the response time of 2%FGLC-1 in symmetric cell was listed in Table.4.2. During the measurement of response time, two conditions, as shown in Fig.4.10, were tried: 1) The positive field direction of applied square wave was “parallel”
to the Ps direction. 2) The positive field direction of applied square wave was “opposite” to the Ps direction.
Fig.4.10. Response time measurement conditions: (a) The positive field direction of applied square wave is parallel to Ps (b) The positive field direction of applied square wave is
opposite to Ps
The response time of 2%FGLC-1 was found faster in hybrid cell than in symmetric premade cell. And it was found that when the positive field of applied square wave is opposite to the direction of Ps, the fall time would be faster. We supposed that after removing the voltage the director returned to the initial state more quickly because the asymmetric polarity in hybrid cell provided a restoring force. In addition, if the positive field of applied square wave is opposite to the direction of Ps, the rise time would be slower, comparing to cell with positive field of applied square wave parallel to the direction of Ps. It was inferred that because of the asymmetric polarity of the cell, a dragging force slowed down the switching when field opposite to the direction of Ps was applied.
Table.4.1. Response time of 2%FGLC-1 in 2μm premade cell driven by 1kHz 30V square wave. Both of the two alignment films in premade cell are the same.
Response time Material
Rise time(τ10-90) Fall time(τ90-10)
2%FGLC-1 680us 1.1ms
(a) (b)
0
Fig.4.11. Response times for the (a) rising process and the (b) falling process of 2%FGLC-1 in 2μm hybrid cell driven by a 30V bipolar square wave with various frequency. Blue dot is the data measured when the direction of positive electric field is opposite to the direction of Ps. Pink square is the data measured when the direction of positive electric field is parallel to
the direction of Ps.
4.2 Improvement of Surface Roughness
To fulfill the criterion of C2 uniform state, it is known that the pretilt angle of the alignment film must be small [38]. However, even if the pretilt angle of alignment film is small, a uniform C2 structure would still not be obtained with a rough surface because the
(a)
(b)
local surface variation changes the surface pretilt [25][34]. H. Furue et al. obtained a zigzag free C2 structure by using a very smooth alignment film [25]. Based on this concept, we wondered to achieve a smooth and uniform alignment surface by coating an additional layer which could flatten the surface and has alignment ability.
4.2.1 High T
gLC Coating
The first material we used was oligoflourene [39], F(Pr)5F(MB)2, which has a high glass transition temperature, Tg, at 149 °C and a nematic-to-isotropic transition temperature, Tc, at 366 °C. The coating of high Tg LC had two purposes: 1) since oligoflourene could achieve monodomain [39], the well aligned surface would probably provide a smooth surface.
2) Liquid crystal with high Tg could preserve its alignment at temperature higher than the nematic-to-isotropic temperature of the FLC materials. Thus, the alignment of oligoflourene would maintain during FLC filling process.
Thin film of oligofluorene was prepared by spincoating from 0.25 wt% solution in chloroform at a spin rate of 1500 rpm. The oligoflourene film coated on rubbed polyimide had thickness about 20nm to 30nm. Afterward, annealing at a temperature slightly above Tg was performed to well align the molecules. In Fig.4.12 the surface roughness was reduced due to additional layer coating. The UV-ozone treatment is needed to rip off the residual organics on the surface and to enhance the adhesion of alignment layer when coating. After coating PI, the surface mean roughness was reduced from 5.917nm to 0.692nm. Moreover, the coating of oligoflourene reduced the mean roughness to 0.345nm. The surface morphology was measured by Atomic Force Microscope (AFM).
Fig.4.12. The surface morphology of layer (a) with no treatment, (b) treated by UV-ozone, (c) coating with polyimide after UV-ozone treatment, and (d) coating with oligoflourene on
rubbed polyimide.
We used the oligoflourene-coated substrate to replace one side or both sides of the cell and filled the FLC materials. The cell was observed by POM and it seemed the alignment was improved somewhere, as shown in Fig.4.13 (a). However, it was found that a part of the FLC materials would dissolve oligoflourene and mixed together, which destroyed the alignment as shown in Fig.4.13 (b). The reason of the mixing may be the π-π interaction between benzene rings of the FLC material and oligoflourene.
Fig.4.13. Pictures of alignment of FLC in oligoflourene coated cell. (a) The area which had better alignment. (b) Mixing of oligoflourene and FLC.
4.2.2 Photo-reactive LC coating
To avoid the dissolution between FLC material and the additional layer, we chose a photo-reactive LC 90519 (from ITRI, Taiwan) as the additional coating layer. The photo-reactive LC 90519 is sensitive at 365nm and dissolves in toluene. After coating 90519 on rubbed PI, the substrate was annealed at 90℃ for 1 minute for drying. Then the substrate was exposed under UV irradiation with intensity of 8mW/cm2 for 1 minute under N2
atmosphere. The 90519 was first aligned by rubbed PI and cross-linked by unpolarized UV exposure. The bonding fixed the molecules not to dissolve in the FLC materials. After the preparation of substrates, the surface mean roughness of 90519 of different concentration was measured by AFM, as shown in Fig.4.14, and the thickness was also measured, as shown in
Fig.4.15. The surface of 2wt% 90519 was smoother than one of 3wt% and its surface roughness reached the lowest 0.699 nm. As for 1wt% 90519 coating, the insufficient concentration resulted in a broken surface which produced a higher roughness. After making a FLC cell, no mixing happened in the cell. However, alignment was not improved because the surface was not smooth enough.
1.0 1.5 2.0 2.5 3.0
Fig.4.14. The surface roughness of various percentages of 90519.
1.0 1.5 2.0 2.5 3.0
Fig.4.15. Trend of thickness regarding to different concentration. The red curve is 2nd order polynomial fitting curve.
4.3 Influence of Surface Rubbing Conditions 4.3.1 Rubbing Direction
Generally speaking, regarding to the alignment by rubbing method, most people adopt parallel rubbing, that is, same rubbing directions for both substrates, as shown in Fig.. Since the chevron structure was introduced in parallel rubbing cell, we wondered that if the rubbing direction was opposite, as shown in Fig.4.16, would a uniform layer structure be formed or not. It was found that antiparallel rubbing hardly obtained a well aligned domain but produced lots of stripe defects over the whole cell, as shown in Fig.4.17.
Fig.4.16. Alignment directions: (a) parallel rubbing (b) antiparallel rubbing
Fig.4.17. Alignment of antiparallel rubbing cell with 2%FGLC-1.
4.3.2 Rubbing Strength
Rubbing is the alignment method adopted in this research, which is generally carried out by moving the substrate with a constant velocity under a rotating roller covered with a velvet cloth. The rubbing force are varied with the rubbing process parameters such as the roller diameter and the rotation speed, the substrate advancing speed, number of rubbings, and the pile impression (the depth of the rubbing cloth pressed down by roller). The rubbing strength (RS) defined by Seo et al. [38] is as following
(
2 / −1)
= NM rn V
RS π
Where N is the number of rubbings (N=1 in our experiments), M is the depth of the deformed fibers of the cloth (mm), n is the rotation speed of the roller (300/60 s-1), V is the advancing speed of the substrate (7.3 mm/s), and r is the radius of the roller (22.5 mm). The RS is given in mm unit.
In order to find the best alignment condition, we changed several depths of pile impression for PVA and PI surface. Because the alignment defect appearing in nematic phase would transmit to the alignment in SmC* phase, a monodomain is required in nematic phase.
In Fig.4.18, three cells with different rubbing strength showed their alignment during cooling process. The upper pictures were the alignment in N* phase, and pictures at the bottom were the alignment in SmC* phase. Within this experiment, it was found that the pile impression was very critical and the best pile impression was 0.20mm for PI and 0.23mm for PVA. A little change would lead to poor alignment, as shown in Fig.4.18. The tolerance of the pile impression might less than 0.02mm.
Fig.4.18. The alignment of 2%FGLC-1 in 1.6μm PVA cell due to different depth of pile impression. (a)0.19mm; (b)0.23mm; (c)0.25mm.
4.4 Discussions
The hybrid cell succeeded in achieving uni-direction layer structure with R2301 and R3206. However, the hybrid cell failed with 2%FGLC-1 mixture due to the stripe defect. The appearance of stripes indicated that the layer orientation may be unstable. The reason for the formation of the stripes was inferred as the following: the smectic layer in V shaped switching generally is perpendicular to the rubbing direction (see Fig.4.19 (a)), not tilted as shown in the uni-direction model in Fig.4.19 (c), but the smectic layer was forced to tilt due to the
The hybrid cell succeeded in achieving uni-direction layer structure with R2301 and R3206. However, the hybrid cell failed with 2%FGLC-1 mixture due to the stripe defect. The appearance of stripes indicated that the layer orientation may be unstable. The reason for the formation of the stripes was inferred as the following: the smectic layer in V shaped switching generally is perpendicular to the rubbing direction (see Fig.4.19 (a)), not tilted as shown in the uni-direction model in Fig.4.19 (c), but the smectic layer was forced to tilt due to the