advanced understanding of the optical …...advanced understanding of the optical properties in...
TRANSCRIPT
ADVANCED UNDERSTANDING OF THE OPTICAL PROPERTIES IN
PHASE COMPENSATED LIQUID CRYSTAL DEVICES
A dissertation submitted to Kent State University in partial
fulfillment of the requirements for the degree of Doctor of Philosophy
by
Yong-Kyu Jang
August 2007
ii
Approved by
Chair, Doctoral Dissertation Committee
Philip J. Bos , Professor, Chemical Physics Interdisciplinary Program
Members, Doctoral Dissertation Committee
Deng-Ke Yang , Professor, Chemical Physics Interdisciplinary Program
Liang-Chy Chien , Professor, Chemical Physics Interdisciplinary Program
John L. West , Professor, Chemistry Department
Robert Twieg , Professor, Chemistry Department
Accepted by
Oleg D. Lavrentovich , Director, Chemical Physics Interdisciplinary Program
Jerry Feezel , Dean, College of Arts and Sciences
Dissertation written by Yong-Kyu Jang
B.A., Chungnam National University, 1991 M.S., Korea University, 1993
Ph.D., Kent State University, 2007
iii
TABLE OF CONTENTS
LIST OF FIGURES……………………………………………………………………....xi
LIST OF TABLES……………………………………………………………………..xxiv
ACKNOWLEDGEMENTS………………………………………………………….....xxv
CHAPTER 1. INTRODUNTION
1-1. Motivations and objectives…………………………………………………...1
1-2. Overview of the dissertation………………………………………………….3
CHAPTER 2. ANALYSIS OF THE MULTI-REFLECTION EFFECTS IN
COMPENSATED LIQUID CRYSTAL DEVICES
2-1. Introduction…………………………………………………………………..7
2-2. Analytical calculations of multi-reflection effects in LCDs………………….9
2-3. Multi-reflection in an isotropic layer………………………………………..16
2-4. Multi-reflection in a liquid crystal device with no residual retardation in the
dark state (ECB-type LCD)………………………………………………....17
2-4-1. Phase analysis of the e-ray and o-ray………………………………...20
2-4-2. Analysis of the angle of light polarization…………………………...22
2-4-3. Analysis of transmittance without considering the dispersion properties
of refractive indices in ECB type LCDs……………………………..26
2-4-4. Analysis of transmittance with considering the dispersion properties of
refractive indices in ECB type LCDs………………………………...28
iv
2-5. Multi-reflection in a liquid crystal device with residual retardation in the dark
state (Pi-cell type LCD)……………………………………………………..31
2-5-1. Analysis of transmittance without considering the dispersion properties
of refractive indices in Pi-cell type LCDs……………………………35
2-5-2. Analysis of transmittance with considering the dispersion properties of
refractive indices in Pi-cell type LCDs………………………………37
2-5-3. Analysis of high and low frequency interference modes and their
effects on the optical properties of LCDs……………………………41
2-5-4. Refractive index mismatching effects in Pi-cell type LCDs…………47
2-6. Experimental and numerical calculation results of a Pi-cell………………..50
2-6-1. Compensating a Pi-cell by using a uniaxial crystal compensator……50
2-6-2. Compensating a Pi-cell by using hybrid-aligned negative C-plates…54
2-6-2-1. Multi-reflection effects in a real Pi-cell…………………….56
2-6-2-2. Multi-reflection effects in a real Pi-cell with additional crystal
polarizers……………………………………………………60
2-6-2-3. Refractive index mismatching effects in a Pi-cell………….63
2-6-3. Numerical calculations of the multi-reflection effects in a Pi-cell…...66
2-7. Summary…………………………………………………………………….70
CHAPTER 3. UNIVERSAL OFF-AXIS LIGHT TRANSMISSION PROPERTIES
OF THE BRIGHT STATE IN COMPENSATED LIQUID CRYSTAL
DEVICES
3-1. Introduction…………………………………………………………………71
v
3-2. Optical properties of general liquid crystal devices………………………...72
3-2-1. Calculation methods and compensations of the dark states of the LCDs
………………………………………………………………………..75
3-2-2. Optical parameters of the LCDs……………………………………...76
3-2-3. Calculation results of the off-axis light transmittance in dark states...79
3-2-4. Calculation results of the off-axis light transmittance in bright states.83
3-3. Universal simple models……………………………………………………87
3-3-1. Simple dark and bright state modeling……………………………….87
3-3-2. “3-layer” modeling…………………………………………………...90
3-4. Calculations of the optical properties of the simple models………………...92
3-4-1. Basic calculations…………………………………………………….92
3-4-2. Analyses of the effective phase retardation of the simple models…...95
3-4-2-1. Effective phase retardation in the director plane…………...96
3-4-2-2. Effective phase retardation out of the director plane……...101
3-4-3. Analyses of the off-axis light transmittance of the simple models…105
3-4-3-1. Off-axis light transmittance in a dark state………………..105
3-4-3-2. Off-axis light transmittance in a bright state………………105
3-5. Comparison of the viewing angle properties in all directions……………..109
3-6. Summary…………………………………………………………………...116
CHAPTER 4. OPTIMIZATION OF THE BRIGHT STATE DIRECTOR
CONFIGURATION IN COMPENSATED PI-CELL DEVICES
4-1. Introduction………………………………………………………………..118
vi
4-2. Optical Properties of Different Bright States in Pi-cells…………………..120
4-2-1. Optical compensations of the dark states of the Pi-cells……………120
4-2-2. Optical parameters of the Pi-cells…………………………………..120
4-2-3. Numerical calculations of the off-axis light transmittance in dark states
………………………………………………………………………125
4-2-4. Numerical calculations of the off-axis light transmittance in bright
states………………………………………………………………...130
4-3. Universal Bright State Model of Pi-cells…………………………………..138
4-3-1. Dark and bright state modeling……………………………………..138
4-3-2. Calculations…………………………………………………………141
4-3-3. Analyses of the effective birefringence in the director plane……….143
4-3-4. Analyses of the transmittance………………………………………149
4-3-5. Detailed analyses of the transmittance out of the director plane…...154
4-3-5-1. Contribution from the positive A-plates…………………..154
4-3-5-2. Contribution from the negative C-plates…………………..156
4-4. Summary…………………………………………………………………...163
CHAPTER 5. LUMINANCE AND COLOR PROPERTIES OF THE
COMPENSATED LIQUID CRYSTAL DEVICES IN THEIR
BRIGHT STATES
5-1. Introduction………………………………………………………………..165
5-2. Color Calculations…………………………………………………………166
5-2-1. Tristimulus values and color matching functions…………………..166
vii
5-2-2. Chromaticity diagrams……………………………………………...169
5-2-3. Color difference……………………………………………………..172
5-2-4. Illuminants…………………………………………………………..173
5-3. Analyses of the luminous transmittance (Y )……………………………...177
5-3-1. Thickness effects on luminous transmittance and cell parameters…177
5-3-2. Off-axis luminous transmittance……………………………………183
5-3-2-1. Luminous transmittance in the bright state model………...183
5-3-2-2. Luminous transmittance in real LCDs…………………….184
5-3-2-3. Viewing angle properties of luminous transmittances in all
directions…………………………………………………..185
5-4. Analyses of the color properties…………………………………………...197
5-4-1. Thickness effects on color properties……………………………….197
5-4-2. Chromaticity coordinates of the bright state model in the off-axis
viewing angles……………………………………………………...204
5-4-3. Chromaticity coordinates of the common bright state LCDs in the
off-axis viewing angles……………………………………………..213
5-4-4. Color difference of the bright state model in the off-axis viewing
angles……………………………………………………………….227
5-4-5. Color difference of the common bright state LCDs in the off-axis
viewing angles……………………………………………………...229
5-4-5-1. Detailed analyses of the color difference of the Pi-cell…...229
5-4-5-2. Detailed analyses of the color difference of the TN mode..230
viii
5-5. Summary…………………………………………………………………...241
5-5-1. Luminous transmittance of the bright state LCDs………………….241
5-5-2. Color properties of the bright state LCDs…………………………..242
CHAPTER 6. THE CONDITIONS AND THE LIMITATIONS OF THE PERFECT
PHASE COMPENSATION IN LIQUID CRYSTAL DISPLAYS
6-1. Introduction………………………………………………………………..244
6-2. Calculations of the phase difference……………………………………….244
6-2-1. “Complete” method…………………………………………………245
6-2-2. Approximate method………………………………………………..248
6-2-3. Comparison of both methods for calculating the phase difference…250
6-3. Conditions and limitations for the perfect phase compensation in the simple
director configurations……………………………………………………..253
6-3-1. Relationship of the parameters……………………………………...253
6-3-2. Thickness ratio for the perfect compensation………………………260
6-3-3. Thickness ratio for the different director configurations…………...261
6-3-4. Transmittances in the compensation system………………………..267
6-4. Applications………………………………………………………………..280
6-4-1. Uniform director configurations…………………………………….280
6-4-2. Non-uniform director configurations……………………………….300
6-4-2-1. A Pi-cell that has different bright state voltages
(thickness effects)……………………………………...…300
6-4-2-2. A Pi-cell that has different pretilt angles
ix
(thickness and director tilt angle effects)…………………301
6-5. Summary…………………………………………………………………...305
CHAPTER 7. CONCLUSIONS
7-1. Analysis of the multi-reflection effects in compensated liquid crystal devices
……………………………………………………………………………..308
7-2. Universal off-axis light transmission properties of the bright state in perfectly
compensated liquid crystal devices………………………………………...309
7-3. Optimization of the bright state director configuration for perfectly
compensated pi-cell devices……………………………………………….310
7-4. Luminance and color properties of the compensated liquid crystal devices in
their bright states…………………………………………………………...312
7-4-1. Luminous transmittance of the bright state LCDs………………….312
7-4-2. Color properties of the bright state LCDs ………………………….313
7-5. The conditions and the limitations of the perfect phase compensation in
liquid crystal displays……………………………………………………...315
APPENDIX A. CALCULATION OF THE 1-DIMENSIONAL LIQUID CRYSTAL
DIRECTOR CONFIGURATION BY THE VECTOR METHOD BASED
ON THE RELAXATION TECHNIQUE
A-1. Calculations……………………………………………………………….318
A-2. Program source codes (Matlab)…………………………………………...324
APPENDIX B. CALCULATION OF THE OPTICS IN LIQUID CRYSTAL
DEVICES BY EXTENDED 2×2 JONES MATRIX
x
B-1. Calculations………………………………………………………………..337
B-2. Program source codes (Matlab)…………………………………………...345
APPENDIX C. PROGRAM OVERVIEW (“LC Optics”)
C-1. “LC Optics” main screen………………………………………………….350
C-2. Popup menu of each layer…………………………………………………351
C-3. Layer properties…………………………………………………………...352
C-4. Variable rearrangement and plotting……………………………………...353
C-5. Plotting options……………………………………………………………354
C-6. Example of the 2-D Plotting………………………………………………355
C-7. Example of the 3-D Plotting………………………………………………356
C-8. Example of the Conoscopy Plotting………………………………………357
C-9. Example of the LC director configuration………………………………...358
REFERENCES………………………………………………………………………...359
xi
LIST OF FIGURES
Fig. 2-1. Simple layout of a compensated liquid crystal device and the definition of each
axis…….................................................................................................................15
Fig. 2-2. Director configurations of the liquid crystal in an ECB cell…………………...19
Fig. 2-3. Calculation result of the phases of 'eEr
and 'oEr
in ECB type LCDs……………21
Fig. 2-4. Effects on the angle [ )(λθ ] of (a) the index mismatching ( airo nn − ) and
(b) the relative difference of refractive indices ( oe nn − )……………………...24
Fig. 2-5. Effects on the transmittance (light leakage) of (a) the index mismatching
( airo nn − ) and (b) the relative difference of refractive indices ( oe nn − )…….25
Fig. 2-6. Transmittances of each mode and the total transmittance with the angle
variation in the dark state of an ECB type LCD without considering the dispersion
of refractive indices………………………………………………………………27
Fig. 2-7. Light wavelength dispersion of refractive indices ( on , en ) of a commercialized
liquid crystal (LC53)……………………………………………………………..29
Fig. 2-8. Transmittances of each mode and the total transmittance with the angle
variation in the dark state of an ECB type LCD with considering the dispersion of
refractive indices…………………………………………………………………30
Fig. 2-9. Director configurations of the liquid crystal in a Pi-cell……………………….33
Fig. 2-10. Calculation results of the effective birefringence at the normal direction
xii
in a Pi-cell………………………………………………………………………..34
Fig. 2-11. Transmittances of each mode and the total transmittance in the dark state
of a Pi-cell-type LCD without considering dispersion of refractive indices…….36
Fig. 2-12. Effective refractive indices of the liquid crystal (LC53) for the thickness
of 5.53 µm and effective residual birefringence of 100 nm……………………...39
Fig. 2-13. Transmittances of each mode and the total transmittance in the dark state
of a Pi-cell-type LCD with considering dispersion of refractive indices………...40
Fig. 2-14. Thickness effect on the total transmittance of a Pi-cell-type LCD in the dark
state………………………………………………………………………………43
Fig. 2-15. Phase retardation effect on the total transmittance of a Pi-cell-type LCD in the
dark state…………………………………………………………………………44
Fig. 2-16. Effective refractive indices ( effen ) of the liquid crystal (LC53) with different
thicknesses and a constant phase retardation ( effndΔ =100 nm)…………………45
Fig. 2-17. Effective refractive indices ( effen ) of the liquid crystal (LC53) with a constant
thickness ( d =5.53 μm) and different phase retardations………………………..46
Fig. 2-18. The effects on the dark state transmittance due to the refractive index
mismatching ( airo nn − )…………………………………………………………49
Fig. 2-19. Measurement setup of a Pi-cell compensation system and the cell structure
of the Pi-cell……………………………………………………………………...52
Fig. 2-20. Measured transmittance and the extracted minimum transmittance of the dark
state Pi-cell compensated with the compensator at each wavelength separately...53
xiii
Fig. 2-21. Basic structure of the compensation scheme of a Pi-cell by using WV-Film...55
Fig. 2-22. Optical stack configuration of a compensated real Pi-cell and the measurement
setup. All layers are combined as a single unit without any air gaps……………58
Fig. 2-23. Measurement results of the transmittance of a compensated real Pi-cell…….59
Fig. 2-24. Transmittances of the crossed sheet polarizers and the parallel sheet polarizers.
………………………………………………………………………………...61
Fig. 2-25. Measurement results of the transmittance of a compensated Pi-cell
with additional crystal polarizers and the experimental setup…………………...62
Fig. 2-26. Measurement results of the transmittances of a compensated Pi-cell
with additional crystal polarizers and air gaps, and the experimental setup……..65
Fig. 2-27. Measurement data of the real and imaginary parts of refractive indices……..68
Fig. 2-28. Measurement results and the numerical calculation results of the transmittance
of the Pi-cell compensated with the hybrid aligned negative-discotic films
for different applied voltages…………………………………………………….69
Fig. 3-1. Director configurations of the most common liquid crystal display modes…...74
Fig. 3-2. Stack configuration of the common liquid crystal devices…………………….78
Fig. 3-3. Numerical calculation results of the off-axis light transmission properties
of the common liquid crystal devices in their dark states………………………..80
Fig. 3-4. Director tilt angles of the dark state-liquid crystal layers……………………...81
Fig. 3-5. Effective phase retardation of the dark state-liquid crystal layers in the director
plane……………………………………………………………………………...82
Fig. 3-6. Numerical calculation results of the off-axis light transmission properties
xiv
of the common liquid crystal devices in their bright state……………………….84
Fig. 3-7. Director tilt angles of the bright state-liquid crystal layers…………………….85
Fig. 3-8. Effective phase retardation of the bright state-liquid crystal layers in the director
plane……………………………………………………………………………...86
Fig. 3-9. (a) Simplification of the liquid crystal layers of the various devices, (b) optical
compensation of the simplified liquid crystal layer and (c) simple dark and bright
state models………………………………………………………………………89
Fig. 3-10. (a) “3-layer” modeling of the most common LCD modes and (b) simple bright
state model……………………………………………………………………….91
Fig. 3-11. Projections of the optic axes of A and C-plates onto the plane perpendicular to
the light propagation vector ( Kr
) that lies (a) in the director plane (x-z plane) and
(b) out of the director plane (y-z plane)………………………………………….98
Fig. 3-12. Birefringence analyses of the bright state model of mode 1 in the director plane.
………………………………………………………………………………...99
Fig. 3-13. Birefringence analyses of the bright state model of mode 2 in the director
plane…………………………………………………………………………….100
Fig. 3-14. Birefringence analyses of the bright state model of mode 1 out of the director
plane…………………………………………………………………………….103
Fig. 3-15. Birefringence analyses of the bright state model of mode 2 out of the director
plane…………………………………………………………………………….104
Fig. 3-16. Numerical calculation results of the off-axis light transmission properties of the
simple models in their dark states………………………………………………107
xv
Fig. 3-17. Off-axis light transmission properties of the bright state models…………...108
Fig.3-18. Viewing angle properties of the common liquid crystal devices in their dark
states…………………………………………………………………………….110
Fig. 3-19. Viewing angle properties of the dark state models………………………….112
Fig. 3-20. Viewing angle properties of the common liquid crystal devices in their bright
states…………………………………………………………………………….113
Fig. 3-21. Viewing angle properties of the bright state models………………………...115
Fig. 4-1. Stack configuration of a Pi-cell……………………………………………….124
Fig. 4-2. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different bright state voltages (1.0-3.0V)…………126
Fig. 4-3. Director tilt angles of the dark state Pi-cells that have different bright state
voltages (1.0-3.0V)……………………………………………………………..127
Fig. 4-4. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different pretilt angles (2.0-30.0º)………………...128
Fig. 4-5. Director tilt angles of the dark state Pi-cells that have different pretilt angles
(2.0-30.0º)………………………………………………………………………129
Fig. 4-6. Numerical calculation results of the off-axis light transmission properties of the
bright state Pi-cells that have different bright state voltages (1.0-3.0V)……….132
Fig. 4-7. Director tilt angles of the bright state Pi-cells that have different bright state
voltages (1.0-3.0V)……………………………………………………………..133
Fig. 4-8. Conoscopic properties of the bright state transmittances with different bright
state voltages of (a) 1.0V, (b) 1.3V, (c) 2.0V, (d) 2.5V and (e) 3.0V…………..134
xvi
Fig. 4-9. Numerical calculation results of the off-axis light transmission properties of the
bright state Pi-cells that have different pretilt angles (2.0-30.0º)………………135
Fig. 4-10. Director tilt angles of the bright state Pi-cells that have different pretilt angles
(2.0-30.0º)………………………………………………………………………136
Fig. 4-11. Conoscopic properties of the bright state transmittances with different pretilt
angles of (a) 2.0º, (b) 5.5º, (c) 10.0º, (d) 20.0º and (e) 30.0º…………………...137
Fig. 4-12. (a) A Pi-cell whose dark state is perfectly compensated using hybrid-negative
C-plates and (b) simple dark and bright state models of the Pi-cell……………140
Fig. 4-13. Effective birefringence of each layer of the bright state model in the director
plane with a tilt angle (θ) of 50º………………………………………………...146
Fig. 4-14. Effective birefringence of the positive A-plates (PA1+PA2) of the bright state
model in the director plane……………………………………………………..147
Fig. 4-15. Effective birefringence of the total layers of the bright state model in the
director plane…………………………………………………………………...148
Fig. 4-16. Off-axis light transmission properties of the bright state model with different
tilt angles………………………………………………………………………..151
Fig. 4-17. Conoscopic properties of the transmittance of the bright state model with tilt
angle (a) 10º, (b) 30º, (c) 50º, (d) 65º, and (e) 70º……………………………...152
Fig. 4-18. Off-axis light transmission properties of the dark state model……………...153
Fig. 4-19. Off-axis light transmittances contributed separately from each layer of the
bright state model (tilt angle of 70º) out of the director plane………………….157
Fig. 4-20. Apparent azimuth angle (γ ) of the optic axis of each layer in the bright state
xvii
model with tilt angle 70º out of the director plane……………………………...158
Fig. 4-21. Director angle definition in (a) lab coordinate system (x’, y’, z’) and
(b) incident light frame ( KSPrrr
,, )………………………………………………159
Fig. 4-22. Difference of the apparent azimuth angle ( γΔ ) between the PA1 and
the PA2 in bright state model with various tilt angles (10º-70º)……………….160
Fig. 4-23. Effective birefringence of a positive A-plate (PA1 or PA2) out of the director
plane in the bright state model………………………………………………….161
Fig. 4-24. Subtotal effective birefringence of the negative C-plates (NC1+NC2) out of the
director plane in the bright state model…………………………………………162
Fig. 5-1. CIE 1931 2º color matching functions………………………………………..168
Fig. 5-2. CIE 1931 2º chromaticity diagram (2º viewing angle)……………………….170
Fig. 5-3. CIE 1976 ''vu chromaticity diagram (2º viewing angle)……………………..171
Fig. 5-4. Spectral power distributions of CIE illuminants……………………………...175
Fig. 5-5. Simple dark and bright state models (mode 1)………………………………..179
Fig. 5-6. Thickness effects on the luminous transmittance of the bright state model at the
normal direction………………………………………………………………...180
Fig. 5-7. Stack configuration of the common liquid crystal devices…………………...182
Fig. 5-8. Off-axis luminous transmittances (Y ) of the bright state model with different
Y values at the normal direction………………………………………………..187
Fig. 5-9. Off-axis luminous transmittances (Y ) of the bright state ECB mode with
different Y values at the normal direction………………………………………188
Fig. 5-10. Off-axis luminous transmittances (Y ) of the bright state VA mode with
xviii
different Y values at the normal direction………………………………………189
Fig. 5-11. Off-axis luminous transmittances (Y ) of the bright state Pi-cell mode with
different Y values at the normal direction………………………………………190
Fig. 5-12. Off-axis luminous transmittances (Y ) of the bright state TN mode with
different Y values at the normal direction………………………………………191
Fig. 5-13. Viewing angle properties of the luminous transmittances (Y ) in the bright state
model with different Y values at the normal direction…………………………192
Fig. 5-14. Viewing angle properties of the luminous transmittances (Y ) in the bright state
ECB mode with different Y values at the normal direction…………………….193
Fig. 5-15. Viewing angle properties of the luminous transmittances (Y ) in the bright state
VA mode with different Y values at the normal direction……………………...194
Fig. 5-16. Viewing angle properties of the luminous transmittances (Y ) in the bright state
Pi-cell mode with different Y values at the normal direction…………………..195
Fig. 5-17. Viewing angle properties of the luminous transmittances (Y ) in the bright state
TN mode with different Y values at the normal direction……………………...196
Fig. 5-18. ( 'u , 'v ) chromaticity coordinates of the bright state model in terms of
thicknesses at the normal direction……………………………………………..199
Fig. 5-19. Transmittances of the bright state model at the three major colors
[blue (λ =450 nm), green (λ =550 nm), and red (λ =650 nm)] at the normal
direction……………………………………………………………………...…200
Fig. 5-20. Phase retardation of the bright state model at the three major colors
[blue (λ =450 nm), green (λ =550 nm), and red (λ =650 nm)] at the normal
xix
direction………………………………………………………………………...201
Fig. 5-21. Lightness ( *L ) of the bright state model in function of thickness at the normal
direction………………………………………………………………………...202
Fig. 5-22. Color difference ( uvE *Δ ) of the bright state model in function of thickness at
the normal direction…………………………………………………………….203
Fig. 5-23. ( 'u , 'v ) chromaticity coordinates of the bright state model in functions of
viewing angles………………………………………………………………….206
Fig. 5-24. ( *u , *v ) chromaticity coordinates of the bright state model in functions of
viewing angles………………………………………………………………….207
Fig. 5-25. Transmittances of the bright state model with Y (%)=100 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm)…………………………….208
Fig. 5-26. Transmittances of the bright state model with Y (%)=90 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm)…………………………….209
Fig. 5-27. Transmittances of the bright state model with Y (%)=80 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm)…………………………….210
Fig. 5-28. Transmittances of the bright state model with Y (%)=70 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm)…………………………….211
Fig. 5-29. Off-axis phase retardation of the bright state model with Y (%)=80 in the three
major wavelengths (blue: 450 nm, green: 550 nm, red: 650 nm)………………212
Fig. 5-30. ( 'u , 'v ) chromaticity coordinates of the ECB in functions of viewing angles.
……………………………………………………………………………….215
Fig. 5-31. ( 'u , 'v ) chromaticity coordinates of the VA in functions of viewing angles.
xx
……………………………………………………………………………….216
Fig. 5-32. ( 'u , 'v ) chromaticity coordinates of the Pi-cell in functions of viewing angles.
……………………………………………………………………………….217
Fig. 5-33. ( 'u , 'v ) chromaticity coordinates of the TN in functions of viewing angles..218
Fig. 5-34. ( *u , *v ) chromaticity coordinates of the ECB in functions of viewing angles.
……………………………………………………………………………….219
Fig. 5-35. ( *u , *v ) chromaticity coordinates of the VA in functions of viewing angles.
……………………………………………………………………………….220
Fig. 5-36. ( *u , *v ) chromaticity coordinates of the Pi-cell in functions of viewing
angles…………………………………………………………………………...221
Fig. 5-37. ( *u , *v ) chromaticity coordinates of the TN in functions of viewing angles.
……………………………………………………………………………….222
Fig. 5-38. Transmittances of the TN mode with Y (%)=100 in the three major colors...223
Fig. 5-39. Transmittances of the TN mode with Y (%)=90 in the three major colors….224
Fig. 5-40. Transmittances of the TN mode with Y (%)=80 in the three major colors….225
Fig. 5-41. Transmittances of the TN mode with Y (%)=70 in the three major colors….226
Fig. 5-42. Off-axis color difference ( uvE *Δ ) of the bright state model………………..228
Fig. 5-43. Off-axis color difference ( uvE *Δ ) of the ECB mode……………………….233
Fig. 5-44. Off-axis color difference ( uvE *Δ ) of the VA mode………………………...234
Fig. 5-45. Off-axis color difference ( uvE *Δ ) of the Pi-cell mode……………………...235
Fig. 5-46. Off-axis color difference ( uvE *Δ ) of the TN mode…………………………236
xxi
Fig. 5-47. Transmittances of the Pi-cell with Y (%)=100 in the three major colors……237
Fig. 5-48. Transmittances of the Pi-cell with Y (%)=90 in the three major colors……..238
Fig. 5-49. Transmittances of the Pi-cell with Y (%)=80 in the three major colors……..239
Fig. 5-50. Transmittances of the Pi-cell with Y (%)=70 in the three major colors……..240
Fig. 6-1. Angle definitions with a uniaxial slab………………………………………...247
Fig. 6-2. Angle definitions with a uniaxial slab in the approximation method………...249
Fig. 6-3. Effective phase retardations of the nematic slabs (θ ,φ ) calculated by the
approximation method (“Simple”) and complete method
(“Recursive” and “2×2 matrix”)………………………………………………..251
Fig. 6-4. Angle definitions of a simple compensation system………………………….256
Fig. 6-5. Total effective phase retardation in functions of 'en and 'd in the compensation
system…………………………………………………………………………..257
Fig. 6-6. A pair of parameters ( 'en , 'd ) for the perfect compensation ( 0=Δ Totaleffnd ) of
the uniaxial slab………………………………………………………………...259
Fig. 6-7. Thickness ratio ( dd /' ) for the perfect compensation of the uniaxial slab
(θ =0, 30, 60, and 90º; d =1.0 μm) with the light wavelength of 550 nm……...263
Fig. 6-8. Three-dimensional thickness ratio ( dd /' ) for the perfect compensation of the
uniaxial slab (θ =0, 30, 60, and 90º; d =1.0 μm) with the light wavelength
of 550 nm……………………………………………………………………….265
Fig. 6-9. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 0º (director plane)………………………………………..268
xxii
Fig. 6-10. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 90º (out of the director plane)…………………………...270
Fig. 6-11. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 45º……………………………………………………….272
Fig. 6-12. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 0º (director plane)……………………...274
Fig. 6-13. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 90º (out of the director plane)………….276
Fig. 6-14. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 45º……………………………………...278
Fig. 6-15. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =0º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
……………………………………………………………………………….286
Fig. 6-16. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =30º) that
are perfectly compensated at the normal direction for the light
wavelength of 550 nm…………………………………………………………..288
Fig. 6-17. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =60º) that
are perfectly compensated at the normal direction for the light
wavelength of 550 nm…………………………………………………………..290
Fig. 6-18. Off-axis light transmittances of the compensation systems (θ =0º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
xxiii
……………………………………………………………………………….292
Fig. 6-19. Off-axis light transmittances of the compensation systems (θ =30º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
……………………………………………………………………………….294
Fig. 6-20. Off-axis light transmittances of the compensation systems (θ =60º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
……………………………………………………………………………….296
Fig. 6-21. Off-axis light transmittances in functions of the thickness and the director tilt
angle of the liquid crystal layer in the compensation systems that are perfectly
compensated at the normal direction for the light wavelength of 550 nm……...298
Fig. 6-22. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different pretilt angles (2.0-50.0º)………………...303
Fig. 6-23. Director tilt angles of the dark state Pi-cells that have different pretilt angles
(2.0-50.0º)………………………………………………………………………304
xxiv
LIST OF TABLES
Table 3-1. Cell parameters of the common liquid crystal (LC) devices we used………..77
Table 4-1. Cell parameters of the Pi-cells with different bright state voltages…………122
Table 4-2. Cell parameters of the Pi-cells with different pretilt angles………………...123
Table 4-3. Thickness of each layer of the bright state model with different tilt angles (θ).
………………………………………………………………………………145
Table 5-1. Tristimulus values and chromaticity coordinates of illuminants in
fields of view 2º………………………………………………………………...176
Table 5-2. Cell parameters of the common liquid crystal (LC) devices we used………181
Table 5-3. Thicknesses of the bright state model and the common LCDs……………..181
Table 6-1. Thicknesses of the uniformly aligned perfect compensation systems………285
Table 6-2. Cell parameters of the Pi-cells with different pretilt angles………………...302
xxv
ACKNOWLEDGEMENTS
First of all, I am sincerely thankful to my adviser, Dr. Philip J. Bos, for his great
dedication to my dissertation. Without his help, I would not been able to finish my
research at LCI, Kent State University. He always gave me the right directions to go
when I was stuck in something and encouraged me in difficult situations. He has not only
helped in the research, but has also taken care of my first American life and made me
very comfortable in a new culture. I really appreciate his efforts and considerations for
my family and me.
I would like to thank LCI faculty and staff members, Dr. Oleg Lavrentovich, Dr.
Philip Bos, Dr. Peter Palffy-Muhoray, Dr. Deng-Ke Yang, Dr. L. –C. Chien, Dr. Tony
Jakli, Dr. John West, Dr. Jack Kelly, Dr. Sergij Shiyanovskii, Dr. David Allender, Dr.
Satyendra Kumar, and Dr. Samuel Sprunt for their enthusiastic and very helpful lectures
about liquid crystals and various kind assistance. The knowledge that I have learned from
their classes will be the reference and the lighthouse in my future works.
I also want to thank to all my colleagues including my classmates at LCI,
especially Bohdan Senyuk, John Harden, Christopher Bailey, Clinton Braganza, Xiaoli
Zhou, Jeremy Neal, and Shouping Tang for their kind help and valuable discussions. I
also really thank to all BOSLAB members, particularly Dr. Bin Wang, Dr. Cheng Chen,
and Dr. Yanli Zhang for their kindness and diverse help.
xxvi
I could study here by virtue of Samsung Electronics Co., Ltd., especially S. M. D.
Hyung-Guel Kim, because they gave me this fabulous opportunity to study more in USA,
which was, actually, my dream come true. I know it was a hard decision for them to make,
but they did it for me. I am very grateful for their decision and efforts. Additionally, there
are lots of colleagues at my company supporting this opportunity, and I really appreciate
their efforts.
This dissertation is dedicated to my honorable parents, Chan-gak Jang and Bok-
gum Lee, my lovely wife, Sook-Ja Jang, my cute kids, Hayeri Jang and Byeol-Ha (Daisy)
Jang, and my brother, sisters and other relatives. Without their persistent patience,
support and encouragement, this dissertation would not have been possible. I deeply
appreciate their endeavors and much assistance.
1
CHAPTER 1
INTRODUCTION
1-1. Motivations and objectives
The viewing angle dependence of light transmittance in liquid crystal displays
(LCDs) is a well-known feature and one of the biggest problems. Especially, the dark
(black) state is very important because it is critical to the contrast ratio of the device or
the visibility of the images which are displayed on a screen. On this account, a great
amount of research has been done to reduce the dark state transmittance and its variation
at all viewing angles.
To minimize the dark state light leakage, the effective phase retardation at all
viewing angles should be minimized. There are several ways for the compensation of the
phase retardation in a dark state, such as the use of negative O-plates, positive O-plates,
and biaxial films. Although the methods are different, their final goals are the same, i.e.
making the total effective phase retardation of the dark state zero in all viewing directions.
Among the methods, using the negative O-plates, or recently it is called polymerized
discotic material (PDM) or hybrid aligned negative C-plates, is the most popular one
because of its effectiveness. By using this method, we can almost perfectly compensate
the phase retardation of the dark state in a LCD at the normal direction. Therefore, we
could obtain the dark transmittance that is limited only by the polarizer and the analyzer
used in conjunction with the liquid crystal device if there is no other effects.
2
However, in our calculations and experiments, the dark transmittance is much
larger than that of the crossed polarizers and changes the color of the dark state, even
though we optically compensate the dark state perfectly at the normal direction. We
analyzed this situation and found that it is related to the multi-reflected light from internal
interfaces within the liquid crystal device, and the multi-reflection strongly depends on
the residual birefringence of the dark state regardless of the phase compensation.
On the other hand, the bright (white) state determines the luminance and the
spectral variations of the transmittance as a function of the angle of the incident light.
This bright state is particularly important in the application of large size displays such as
LCD monitors, TVs, and signs. Therefore, this should be another vital factor when
evaluating the optical properties of the LCDs. However, we have not seen those studies
dealing with the off-axis light transmission properties (viewing angle properties) of the
bright state of general liquid crystal devices such as ECB, VA, TN, and Pi-cell.
Additionally, the viewing angle dependences of the luminance and the color properties
are essential in the optical properties of the devices. We have investigated those optical
properties of the bright states of the common LCDs, whose dark states are optically
optimized to give the minimum transmittance at the normal direction by using the hybrid
aligned negative C compensators, and found that there is universality in the optical
properties of the bright states no matter what liquid crystal modes are used.
Another important topic we want to consider is the phase compensation of the
dark states in LCDs. Recently, most of the commercialized liquid crystal displays use
phase compensators to improve the contrast ratio, viewing angle, and color properties of
3
the devices. However, it has not been reported what the exact conditions and limitations
are for the phase compensation of a liquid crystal layer. Also, the parameter relationship
affecting the limitations is obviously significant when we optically design a LCD.
As mentioned above, the hybrid aligned negative C-plate is suitable for the optical
compensation of a liquid crystal layer with the same optic axes. Seemingly, we could
perfectly compensate a uniformly aligned positive uniaxial slab by using a tilted negative
C-plate with adjusting the optic axis, refractive indices, and thickness. However,
according to our investigation, there is a limitation for the compensation of a uniaxial
layer even in the most simple director configuration. Additionally, we found very
important parameters that strongly affect the limitation of the phase compensation.
1-2. Overview of the dissertation
In Chapter 2, we consider perfectly compensated liquid crystal devices at the
normal direction and show the analytical and numerical calculation results and
experimental data showing the effects of the multi-reflections on the optical properties of
dark state LCDs. Specifically, we describe the contribution of the residual birefringence
and index mismatching to the extinction ratio and its wavelength dependence. According
to the analyses, there are two types of the interference. The first type of interference,
which has higher frequency, could affect the dark level almost equivalently for all visible
wavelengths, so it does not affect the color of the dark state. On the other hand, the
second type, which is closely related to the residual birefringence of the dark state, could
4
cause not only increasing the dark level but also a color shift because of the lower
frequency pattern of the interference in the wavelength space.
In Chapter 3, we investigate the off-axis light transmission characteristics of the
bright state of common liquid crystal device modes. The dark state of these device modes
is optically compensated to have minimum light transmittance at the normal direction.
Our research shows there is an unexpected universal shape of the off-axis light
transmission value in its bright state, regardless of what liquid crystal mode is used. To
understand this surprising fact, we build simple dark and bright state models that can be
applied to general liquid crystal devices and analyze them in terms of the effective
retardation and transmittance.
In Chapter 4, we study the off-axis light transmission properties of the bright state
in Pi-cell devices as a function of the bright state director configuration that is determined
by the applied bright state voltage and the pretilt angle of the device. We find that below
certain values of the voltage or pretilt angle, the off-axis light transmission properties are
insensitive to these parameters and can be described by a previously considered simple
model. However, above a critical pretilt angle or bright state voltage, the light
transmittance is a much stronger function of the incident angle of light. To understand the
facts, we develop a new model that explains this result and provides a description of the
basic issues affecting the optics of these types of devices.
In Chapter 5, we investigate the luminance and color properties of the bright state
simple model and the common LCDs such as ECB, VA, Pi-cell, and TN modes. We use
the same bright state simple model that we built in Chapter 3. The dark states of the
5
model and the common LCDs are optically compensated by using hybrid aligned-passive
type negative C-plates, as the same way that we used in Chapter 3. We analyze the
luminous transmittance of our bright state model in functions of the cell thickness and the
viewing angle, and after that we compare those results with that of real common LCDs.
These results will confirm the universality of the bright state viewing angle properties
and give an important conclusion: single domain LCDs can not escape from the
anisotropic shape of the bright state viewing angle properties. We also calculate and
analyze the off-axis color properties including color difference of the bright state model
and the common LCDs. From these results, we know that there are common features of
the color properties in the bright state of common LCDs.
In Chapter 6, we will study the conditions and the limitations of the phase
compensation in dark state LCDs. As an example compensator, we will use hybrid
aligned negative C-plates. We discuss the difference between an approximate method and
the “complete” method for the calculation of the phase retardation in a uniaxial material.
We analyze the relationship among the parameters ( ',',' dnn oe ) of a compensator for the
perfect compensation of a uniaxial slab. With this parameter relationship, we calculate
the thickness ratio of the compensator to uniaxial slab for the perfect compensation. We
also calculate and analyze the total effective phase retardations and the off-axis light
transmittances in functions of the director tilt angle, effective phase retardation, and the
thickness in the uniformly and non-uniformly aligned liquid crystal layers.
All calculations in this dissertation are fulfilled by our own GUI program, “LC
Optics”, which was made for the purpose of calculating the director configuration of a
6
liquid crystal layer and the optical properties of general LCDs such as transmittance,
reflectance, luminance, lightness, and color analyses. In Appendices A and B, we give the
basic calculations and program source codes related to the director calculation and the
optics calculation. Due to the huge size of the program, it is hard to attach the whole
source codes on this dissertation. Finally in Appendix C, we introduce the overview of
our program, “LC Optics”.
7
CHAPTER 2
ANALYSIS OF THE MULTI-REFLECTION EFFECTS IN COMPENSATED
LIQUID CRYSTAL DEVICES
2-1. Introduction
Many modes used in liquid crystal devices particularly those with an untwisted
structure1 have residual birefringence in their dark state due to the anchoring energy at
the alignment layers. For these devices, to obtain a true dark state, a passive phase
compensator is needed. Ideally, we can compensate the residual birefringence so that the
total effective birefringence will be zero at all wavelengths, and we can achieve an
extinction ratio that is limited by the polarizer and analyzer used in conjunction with the
liquid crystal device.
However, in our calculations and experiments, the dark transmittance is much
larger than that of the crossed polarizers and depends on the wavelength. We analyzed
this situation and found that it is related to multi-reflected light from internal interfaces
within the liquid crystal device. These reflections result in two types of interference. The
first is the interference of the extraordinary ray (e-ray) and ordinary ray (o-ray) by
themselves and the second one is the interference between e-ray and o-ray. The first type
has higher frequency in the wavelength space and is related to the optical path lengths of
the e-ray and o-ray independently. The second type has a lower frequency and depends
on the residual birefringence of the dark state. So, as the residual birefringence increases,
8
there is a greater contribution from the second type of interference. Considering the
optical properties of the liquid crystal display, the first type interference could affect the
dark level almost equivalently for all visible wavelengths, so it does not affect the color
of the dark state. On the other hand, the second type could cause a color shift because of
the lower frequency pattern of the interference in the wavelength space.
The numerical calculation of the light intensity including multi-reflection effects
in a reflection-mode image transducer utilizing a nematic liquid crystal with a 45° twist
has been reported2. That work covers the liquid crystal thickness dependency of the
output intensity with and without isotropic layers for a single wavelength. However, we
were not able to find reports concerning the interference effects related to the residual
birefringence and their contribution to the optical properties including the wavelength
dependence of the extinction ratio.
In this chapter, we will consider perfectly compensated liquid crystal devices at
the normal direction and show the analytical and numerical calculation results and
experimental data showing the effects of the multi-reflections on the optical properties.
Specifically, we will describe the contribution of the residual birefringence to the
extinction ratio and its wavelength dependence.
In Sec. 2-2, we will describe the effect of the multi-reflection analytically in a
dark state of a simple uniaxial type liquid crystal device that has one pair of interfaces,
but this could be expanded to a more complicated configuration easily using the same
concept. In Secs. 2-3~5 we will show the calculation results of light transmission when
multi-reflections are considered and analyze the effects. In Sec. 2-6, we will give the
9
experimental data with the numerical calculation including all necessary layers in an
example device.
2-2. Analytical calculations of multi-reflection effects in LCDs
In this section, we will consider a simple-uniaxial-type liquid crystal device with
its dark state compensated perfectly at the normal direction by using passive optical
retarders. Light is assumed to be incident at the normal direction and is reflected only at
the one pair of surfaces; on either side of the liquid crystal layer in air. [In actual liquid
crystal devices, there are no air layers inside the devices. However, in an actual LCD
there can be reflections from the multiple interfaces. Those reflections can provide a
similar (but perhaps smaller) effect as will be shown in Sec. 2-6] Figure 2-1 shows this
configuration with two crossed polarizers, and in this figure, δ and Γ are the phase
changes of the light at the liquid crystal and compensator, respectively. The electric field
after the polarizer is expressed as
)](exp[2
1ˆ wtkziExE in −=r
, (2-1)
where inE is the amplitude of electric field of incident light, and x is the unit vector
along the x-axis. In this equation, we assumed the polarizer and analyzer are ideal, i.e. the
incident light that is polarized along the transmittance axis of the polarizer can transmit it
perfectly, but the perpendicular component of incident light is blocked by polarizer. This
10
electric field ( )Er
that has wavelength λ in free space can be split into two eigenmodes,
extraordinary mode (e-mode) and ordinary mode (o-mode), in the liquid crystal layer, and
they will propagate with different phases. Each wave is also multi-reflected at the liquid
crystal–air interface and has the following form after the liquid crystal layer (time related
terms have been dropped here for simplicity):
epaeeeee
eapine tiriritEeE ...])5exp()3exp()[exp(
21ˆ 42 +++= δδδ
r (2-2)
...])4exp()2exp(1)[exp(21ˆ 42 +++= eeeee
epa
eapin iririttEe δδδ (2-3)
)2exp(1
1)exp(21ˆ
2ee
eepa
eapin ir
ittEeδ
δ−
= (2-4)
)2exp(1)exp()1(ˆ
2
2
ee
eee ir
irEe
δδ
−−
≡ . (2-5)
opaooooo
oapino tiriritEoE ...])5exp()3exp()[exp(
21ˆ 42 +++−= δδδ
r (2-6)
)2exp(1)exp()1(ˆ
2
2
oo
ooo ir
irEo
δδ
−−
−≡ , (2-7)
where e and o are the unit vectors along the electric field directions of the extraordinary
ray (e-ray) and ordinary ray (o-ray) [Fig. 2-1 (b)], respectively. The eapt and e
pat are the
transmission coefficients of the e-mode at the interface, air-liquid crystal, and er and
or are the reflection coefficients of the e-mode and o-mode, respectively at the same
11
interface. The eδ and oδ are the spatial phase changes of the e-mode and o-mode at the
liquid crystal layer. Those parameters are expressed using the refractive indices of e-ray
and o-ray ( oe nn , ) and the thickness ( d ) of the liquid crystal layer when light is incident
at a normal direction3,
22
⎟⎟⎠
⎞⎜⎜⎝
⎛+−
=aire
airee nn
nnr , (2-8)
22
⎟⎟⎠
⎞⎜⎜⎝
⎛+−
=airo
airoo nn
nnr , (2-9)
λπ
δdne
e2
= , (2-10)
λπ
δdno
o2
= . (2-11)
During the calculation, we used the fact that,
21 eepa
eap rtt −= , (2-12)
21 oopa
oap rtt −= , (2-13)
and we set,
inoe EEE21
≡= . (2-14)
12
After the liquid crystal layer, the e-ray and o-ray enter the compensator (refractive
indices '' , oe nn , thickness 'd ) and will have the phase change eΓ and oΓ , respectively. The
electric fields after the compensator are,
)](exp[)2exp(1
)1(ˆ2
2'
oeee
eee i
irr
EeE Γ+−
−= δ
δ
r, (2-15)
)](exp[)2exp(1
)1(ˆ2
2'
eooo
ooo i
irr
EoE Γ+−
−−= δ
δ
r, (2-16)
where
λπ ''2 dne
e =Γ , (2-17)
λπ ''2 dno
o =Γ . (2-18)
After the analyzer, the total electric field and transmittance can be calculated as follows:
)(2
1ˆ ''oeout EEyE +=
r, (2-19)
2
2
in
out
E
ET r
r
= (2-20)
eooe TTT ++≡ , (2-21)
13
where
)2cos(21)1(
21
24
222
eee
eee rr
rET
δ−+−
= , (2-22)
)2cos(21)1(
21
24
222
ooo
ooo rr
rET
δ−+−
= , (2-23)
)1)(1(21 22
oeoeeo rrEET −−−=
⎭⎬⎫
⎩⎨⎧
−−−+
−−−×
)]2exp(1)][2exp(1[1
)]2exp(1)][2exp(1[1
2222ooeeooee iriririr δδδδ
, (2-24)
where we used the fact that the difference of the phase between e-mode and o-mode at
the liquid crystal layer ( oe δδ − ) is the same with that of the compensator layer ( oe Γ−Γ )
because we assume the dark state of the liquid crystal device (LCD) is compensated
perfectly. We also set 1=inE for simplicity.
Equations (2-20~24) show that there are three contributions to the total
transmittance in a LCD compensation system; eT is the transmittance related to the
interference of the pure e-mode by itself due to the multi-reflection in a LCD, and oT is
the transmittance that came from the interference of the pure o-mode with the same
reason. The last term eoT is a coupled term between the e-mode and the o-mode. So, eT
and oT depend on the absolute light path lengths (λdne ,
λdno ) of the e-mode and the o-
14
mode, respectively, but the eoT is related to the relative difference of the light path length
(λλdndn oe − ) between the e-mode and the o-mode.
15
(a) Simple layout of a compensated liquid crystal device
(b) Definition of each axis
Fig. 2-1. Simple layout of a compensated liquid crystal device and the definition of each
axis. The compensator is designed to compensate the phase of the dark state perfectly in
the visible wavelength range.
45º
x
eo
y
16
2-3. Multi-reflection in an isotropic layer
In the simplest case, let us consider the liquid crystal layer is isotropic. Then, e-
ray and o-ray are not distinguishable. Therefore, we have these relationships among the
optical parameters,
δδδ ≡= oe , 0≡Γ=Γ oe , rrr oe ≡= ,
and can write the electric fields at the position between the liquid crystal and the analyzer
( 'eEr
, 'oEr
) as follows from Eqs. 2-15, 16:
)exp()2exp(1
)1(ˆ2
2' δ
δi
irrEeE ee −
−=
r, (2-25)
)exp()2exp(1
)1(ˆ2
2' δ
δi
irrEoE oo −
−−=
r. (2-26)
These two fields have the exactly same magnitude and phase. Consequently, in the
optical viewpoint, their vector summation is linearly polarized, and its direction is always
parallel to the direction of the absorption axis of the analyzer [ x -axis in Fig. 2-1(b)]
regardless of light wavelengths. Mathematically, from Eqs. 2-14, 19, the total electric
field after the analyzer is
0=outEr
, (2-27)
17
and the total transmittance is
0=T , (2-28)
for all light wavelengths.
2-4. Multi-reflection in a liquid crystal device with no residual retardation in the
dark state (ECB-type LCD)
Electrically controlled birefringence (ECB) is one of the well-known liquid
crystal display mode because of its simple structure. Figure 2-2 shows the illustration of
the director configurations of an ECB cell. The dark and bright states are exchangeable
according to the relative angles between the directions of liquid crystal molecules and
polarizers. In this illustration, we set the high voltage state as a dark state because this is a
usual way in real devices. In an ideal ECB cell, the residual birefringence at the high
voltage (dark state) is very low because most of the liquid crystal molecules align along
the external electric field if applied voltage is high enough. However, the molecules near
the surfaces are not completely aligned due to the anchoring energy at the interface
between the liquid crystal and the alignment layer. We assume here that there is no
residual birefringence at the high voltage for simplicity, but still the device has different
reflection coefficients ( oe rr , ) of the e-mode and o-mode at the interfaces,
18
δδδ ≡= oe , 0=Γ=Γ oe , oe rr ≠ .
Then, the electric fields (e-ray and o-ray) just before the analyzer, Eqs. 2-15 and 2-18, are
simplified to be
)exp()2exp(1
)1(ˆ2
2' δ
δi
irr
EeEe
eee −
−=
r, (2-29)
)exp()2exp(1
)1(ˆ2
2' δ
δi
irr
EoEo
ooo −
−−=
r. (2-30)
19
Fig. 2-2. Director configurations of the liquid crystal in an ECB cell.
Dark State Bright State
Voltage
20
2-4-1. Phase analysis of the e-ray and o-ray
We calculated the phases of the electric fields, 'eEr
and 'oEr
in function of light
wavelength (λ ), and the Fig. 2-3 shows the result. During the calculation, we set en and
on of liquid crystal at the air-liquid crystal interfaces are 1.656 and 1.5, respectively in all
wavelengths, and the thickness of the liquid crystal layer is 5.53 μm. As in Fig. 2-3, 'eEr
and 'oEr
are almost in the same phase, but their magnitudes depend on light wavelength
(λ ) from Eqs. 2-10, 11.
21
380 480 580 680 780-300
-200
-100
0
100
200
300 Ee' E
o'
Wavelength (nm)
Angl
e (d
eg.)
Fig. 2-3. Calculation result of the phases of 'eEr
and 'oEr
in ECB type LCDs.
22
2-4-2. Analysis of the angle of light polarization
The phases between e-ray and o-ray after the compensator are almost in phase in
ECB-type LCDs as in the result of the previous section. Therefore, the total electric field
( ''oe EErr
+ ) in front of the analyzer is almost linearly polarized at all wavelengths, but the
direction of the total field varies from the absorption axis of the analyzer periodically
with the same phase of 'eEr
and 'oEr
in wavelength space, and the angle variation [ )(λθ ]
between them is calculated from the formula 2-29 and 2-30,
)2exp(1)2exp(1
11
tan4
)( 2
2
2
21
δδπλθ
irir
rr
e
o
o
e
−−
×−−
−= − . (2-31)
When the direction of the total field is in the absorption axis of the analyzer ( 0=θ ), it
gives zero transmittance, and as the angle (θ ) increases, light leakage becomes bigger.
The amplitude of this angle variation is not only proportional to the relative
difference of the reflection coefficients ( oe rr − ), but also depends on the absolute value
of each of them. Therefore, from Eqs. 2-8, 9, and 2-31, the light leakage of a dark state
LCD becomes larger as the refractive index mismatching or the relative difference of the
refractive indices ( oe nn − ) of the e-mode and o-mode increases at the interfaces. Figure
2-4 is the calculation results of the angle [ )(λθ ] in order to see the effect of refractive
indices on the light leakage, and Fig. 2-5 is the corresponding calculation of
transmittance. In Figs. 2-4(a) and 2-5(a), we used the refractive index of air ( airn ) as a
23
variable ( airn =1.0~1.5, on =1.5, en =1.656, d =5.53 µm, and ndΔ =0), and in Fig. 2-4(b)
and 2-5(b), refractive index of extraordinary ray ( en ) was the variable ( airn =1.0, on =1.5,
en =1.5~2.0, d =5.53 µm, and ndΔ =0) while the other parameters are fixed values. From
Fig. 2-4(a), we can see that as the refractive index of air ( airn ) is close to the refractive
index of an ordinary ray ( on ) or the refractive index of an extraordinary ray ( en ), the
amplitude of the angle [ )(λθ ] decreases because the magnitude of index mismatching
( airo nn − ) is going down. This variation of the angle leads to the variation of the light
leakage as in Fig. 2-5(a). Figure 2-4(b) shows that as the relative difference of the
refractive indices ( oe nn − ) rises, the amplitude of the angle [ )(λθ ] increases because
the relative difference of the reflection coefficients ( oe rr − ) increases, and the light
leakage also has the same trend as in Fig. 2-5(b).
On the other hand, the amplitude of the )(λθ is independent of light wavelength
if there is no dispersion of the refractive indices, so the amplitude of the light leakage of a
dark state is the same in whole wavelength region, and it does not affect the dark color.
24
380 480 580 680 7800.0
0.5
1.0
1.5
2.0 n
air=1.0 n
air=1.1 n
air=1.2
nair
=1.3 nair
=1.4 nair
=1.5
Wavelength (nm)
θ (d
eg.)
(a) airn =1.0~1.5, on =1.5, en =1.656, d =5.53 µm, and ndΔ =0
380 480 580 680 780
0
1
2
3
4
5 n
e=1.5 n
e=1.6 n
e=1.7
ne=1.8 n
e=1.9 n
e=2.0
Wavelength (nm)
θ (d
eg.)
(b) airn =1.0, on =1.5, en =1.5~2.0, d =5.53 µm, and ndΔ =0
Fig. 2-4. Effects on the angle [ )(λθ ] of (a) the index mismatching ( airo nn − ) and (b) the
relative difference of refractive indices ( oe nn − ).
25
380 480 580 680 780
0.0000
0.0001
0.0002
0.0003 n
air=1.0 n
air=1.1 n
air=1.2
nair
=1.3 nair
=1.4 nair
=1.5
Wavelength (nm)
Tran
smitt
ance
(a) airn =1.0~1.5, on =1.5, en =1.656, d =5.53 µm, and ndΔ =0
380 480 580 680 780
0.000
0.001
0.002
0.003 n
e=1.5 n
e=1.6 n
e=1.7
ne=1.8 n
e=1.9 n
e=2.0
Wavelength (nm)
Tran
smitt
ance
(b) airn =1.0, on =1.5, en =1.5~2.0, d =5.53 µm, and ndΔ =0
Fig. 2-5. Effects on the transmittance (light leakage) of (a) the index mismatching
( airo nn − ) and (b) the relative difference of refractive indices ( oe nn − ).
26
2-4-3. Analysis of transmittance without considering the dispersion properties of
refractive indices in ECB type LCDs
Figure 2-6 shows the calculation results of the separate transmittance ( eT , oT , eoT )
and the total transmittance T with the angle variation (θ ) in a parameter condition:
airn =1.0, on =1.5, en =1.656 at the interface, d =5.53 µm, and ndΔ =0. As we mentioned
above, the e-mode and o-mode are almost in phase, and the valley and ridge points of the
total transmittance (T ) are constant as the wavelength changes because we assumed that
refractive indices are constant at all wavelengths. Also, the angle variation (θ ) is in the
same phase with that of the total transmittance exactly. Therefore, the minimum
transmittance at the dark state is limited by the amplitude of the interference pattern (T ),
and the dark color is hardly affected in this type of interference.
27
380 480 580 680 7800.00
0.04
0.08
0.12
0.16 Te To Teo
Wavelength (nm)
Te, T
o
-0.28
-0.24
-0.20
-0.16
-0.12
Teo
(a) Transmittances of each mode
380 480 580 680 7800.0000
0.0002
0.0004
0.0006
0.0008
0.0010
T θ
Wavelength (nm)
Tran
smitt
ance
0
1
2
3
4
Angl
e (d
eg.)
(b) Total transmittance and the angle variation
Fig. 2-6. Transmittances of each mode and the total transmittance with the angle variation
in the dark state of an ECB type LCD without considering the dispersion of refractive
indices. (Δnd = 0 nm, d=5.53 μm, ne=1.656, and no=1.5 at the surface).
28
2-4-4. Analysis of transmittance with considering the dispersion properties of
refractive indices in ECB type LCDs
In real situations, most materials, including liquid crystals, have wavelength
dispersion properties in their refractive indices. Figure 2-7 shows the dispersion property
of refractive indices of the commercialized liquid crystal material (LC53, Chisso Co.),
and we use this material in this dissertation. With considering the dispersion property of
this liquid crystal, we calculated the separate transmittance ( eT , oT , eoT ) and the total
transmittance T with the angle variation (θ ) as we did in Fig. 2-6, and the Fig. 2-8 is the
result. During the calculation, we used the same parameters as in Fig. 2-6 except the
refractive indices of liquid crystal. Comparing the results with Fig. 2-6, the amplitude of
the total transmittance depends on light wavelength, but the valley points of the
transmittance are still zero as the same as in Fig. 2-6. Therefore, if we consider the
dispersion property of refractive indices, the minimum transmittance of dark state could
be different for each light wavelength, and it could lead to the color shift of a dark state
even though the magnitude of the color shift is very small as we may neglect.
29
380 480 580 680 7801.45
1.50
1.55
1.60
1.65
1.70
1.75 n
e n
o
Wavelength (nm)
Ref
ract
ive
Indi
ces
Fig. 2-7. Light wavelength dispersion of refractive indices ( on , en ) of a commercialized
liquid crystal (LC53).
30
380 480 580 680 7800.00
0.04
0.08
0.12
0.16 Te To Teo
Wavelength (nm)
Te, T
o
-0.28
-0.24
-0.20
-0.16
-0.12
Teo
(a) Transmittances of each mode
380 480 580 680 7800.0000
0.0002
0.0004
0.0006
0.0008
0.0010
T θ
Wavelength (nm)
Tran
smitt
ance
0
1
2
3
4
Angl
e (d
eg.)
(b) Total transmittance with the angle variation
Fig. 2-8. Transmittances of each mode and the total transmittance with the angle variation
in the dark state of an ECB type LCD with considering the dispersion of refractive
indices. (Δnd = 0 nm, d=5.53 μm, LC: LC53).
31
2-5. Multi-reflection in a liquid crystal device with residual retardation in the dark
state (Pi-cell type LCD)
Most of the liquid crystal device modes have residual birefringence in their dark
states even though we apply very high voltage because of the surface anchoring energy at
the liquid-alignment layer interface. As an example mode, we will consider a Pi-cell4, 5
mode LCD. Figure 2-9 shows the illustration of the director configuration of liquid
crystal in a Pi-cell. The director configuration of the Pi-cell has bend structure in both
dark and bright states as in Fig. 2-9. Usually, other display modes such as an ECB, and a
twisted nematic (TN) have splay structure in their dark states. Commonly, the elastic
energy of a bend structure is higher than that of a splay structure because it has a bigger
elastic constant. For example, the elastic constants of the liquid crystal (LC53) that we
are using are as follows:
Splay ( 11K ): 10.5 pN ,
Twist ( 22K ): 7.2 pN ,
Bend ( 33K ): 15.7 pN .
Therefore, we have to apply a much bigger voltage in Pi-cells to achieve the same level
of residual birefringence of the other modes. That means the dark state of Pi-cells has
bigger residual birefringence than the other modes in a given similar condition, and that
fact requires the use of a passive type optical retardation film to compensate it. Figure 2-
10 shows the effective birefringence at normal direction with the incident light
32
wavelengths, 450, 550, and 650nm, in a Pi-cell [liquid crystal: LC53; dielectric constant
( εΔ of LC) of 9.4; cell thickness of 5.53 μm]. As in this figure, the Pi-cell still has ~100
nm retardation at around the dark voltage (5~6 V). Therefore, we have these relationships
in a Pi-cell,
oe δδ ≠ , oe Γ≠Γ , oe rr ≠ .
So, from Eqs. (2-15) and (2-26), we know that the e-mode and o-mode components of the
electric fields ( 'eEr
, 'oEr
) after the compensator are out of phase in time and wavelength
space. Therefore, the light after compensator is elliptically polarized, and the ellipticity
and the angle of the major axis are functions of the light wavelength. This causes the total
transmittance to have a beat frequency behavior after the analyzer.
33
Fig. 2-9. Director configurations of the liquid crystal in a Pi-cell.
Dark State
Voltage
Bright State
34
2 4 6 8 10
0
100
200
300
400 450nm 550nm 650nm
LC Voltage (V)
Δnd
(nm
)
Fig. 2-10. Calculation results of the effective birefringence at the normal direction in a Pi-
cell.
35
2-5-1. Analysis of transmittance without considering the dispersion properties of
refractive indices in Pi-cell type LCDs
Figure 2-11 shows the calculation results of the transmittance ( eT , oT , eoT )
contributed from each mode and the total transmittance T at the dark state in a Pi-cell,
respectively. The magnitude of the residual birefringence we used here is 100 nm at all
light wavelengths, and the compensator compensates the value exactly at all wavelengths.
The cell thickness in this calculation is 5.53 μm, and the refractive indices, on and en of
the liquid crystal at the reflection interfaces are 1.5 and 1.656 with no wavelength
dispersion, respectively. As seen in Fig. 2-11(a), eT and oT are not in phase and have
different amplitudes, which depend on the magnitude of reflection coefficients ( oe rr , ). In
Fig. 2-11(b), the minimum values (valley points) of the transmittance curve are not zero,
and their values vary as the wavelength changes. These features are unlike that of the
ideal ECB-type LCDs. These phenomena make the dark level of LCDs higher, and not
only limit the contrast ratio (luminance of bright state / luminance of dark state) that the
displays can reach (here 500-1000:1) but also lead to the color shift of the dark state of
the LCDs.
36
380 480 580 680 7800.00
0.04
0.08
0.12
0.16 Te To Teo
Wavelength (nm)
Te, T
o
-0.28
-0.24
-0.20
-0.16
-0.12
Teo
(a) Transmittances of each mode
380 480 580 680 7800.000
0.001
0.002
0.003
Wavelength (nm)
Tran
smitt
ance
(b) Total transmittance
Fig. 2-11. Transmittances of each mode and the total transmittance in the dark state of a
Pi-cell-type LCD without considering dispersion of refractive indices. ( en =1.656 and
on =1.5 at the reflection interfaces).
37
2-5-2. Analysis of transmittance with considering the dispersion properties of
refractive indices in Pi-cell type LCDs
For understanding the effects of wavelength dispersion properties of refractive
indices, we calculated transmittances in the similar condition as the previous section. The
magnitude of effective residual birefringence that we used here is 100 nm at a wavelength
550 nm of incident light, and the value is compensated by a compensator precisely at all
wavelengths. The cell thickness ( d ) in this calculation is 5.53 μm, and the liquid crystal
is LC53 that has refractive indices, )(λen and )(λon of extraordinary and ordinary rays,
respectively. We fully considered the dispersion of the refractive indices of the liquid
crystal with adjusting the effective value of the refractive index of extraordinary ray
( effen ) to meet the effective residual birefringence ( effndΔ ),
dnnnd oeffeeff )]()([)( λλλ −=Δ , (2-32)
at a wavelength 550 nm. During this adjusting process, we keep the ratio of
)( oe nnn −Δ to )( oeffeeff nnn −Δ constant at all wavelengths. However, when we calculate
the reflectance coefficients, we used real refractive indices ( en , on ) because the liquid
crystal molecules near the surfaces hardly change their orientations due to anchoring
energy. Figure 2-12 shows the effective refractive indices of the liquid crystal (LC53)
with the condition that has a thickness of 5.53 µm and effective residual birefringence
( effndΔ ), 100 nm at an incident light wavelength of 550 nm. By using these parameters,
38
we calculated the transmittances as in Fig. 2-13. As the same as in Fig. 2-11, eT and oT
are out of phase and have different amplitudes which depend on the magnitude of
reflection coefficients ( oe rr , ). In Fig. 2-13(b), the minimum transmittances at each
wavelength also are not zero, and their values vary as the wavelength changes. These
features are the same with in Fig. 2-11. However, additionally if we consider the
properties of wavelength dispersion of refractive indices, we can notice that the
amplitude of the transmittance depends on the light wavelength as it is in ECB-type
LCDs in the previous section, i.e. as the wavelength decreases, the amplitude increases.
This can be understandable because as the wavelength decreases, the refractive indices
increase as in Fig. 2-12, and the magnitude of refractive index mismatching at the liquid
crystal-air interfaces increases, and this causes the amplitude of the transmittance
increases as similar as in Fig. 2-5. From comparing these results with the result in Fig. 2-
11, we see that the effect of wavelength dispersion is relatively not big as much as it can
govern the color of the dark states of the LCDs, although it could aggravate the color
shift that caused by other reasons.
39
380 480 580 680 780
1.50
1.55
1.60
1.65
1.70
1.75 n
e n
o n
eeff
Wavelength (nm)
Ref
ract
ive
Indi
ces
Fig. 2-12. Effective refractive indices of the liquid crystal (LC53) for the thickness of
5.53 µm and effective residual birefringence of 100 nm.
40
380 480 580 680 7800.00
0.04
0.08
0.12
0.16 Te To Teo
Wavelength (nm)
Te, T
o
-0.28
-0.24
-0.20
-0.16
-0.12
Teo
(a) Transmittances of each mode
380 480 580 680 7800.000
0.001
0.002
0.003
Wavelength (nm)
Tran
smitt
ance
(b) Total transmittance
Fig. 2-13. Transmittances of each mode and the total transmittance in the dark state of a
Pi-cell-type LCD with considering dispersion of refractive indices.
41
2-5-3. Analysis of high and low frequency interference modes and their effects on
the optical properties of LCDs
The total transmittance [Fig. 2-13(b)] of a “Pi-cell type LCD” has two frequency
modes, roughly, high frequency and low frequency. In order to analyze the source of the
modes, we calculated the total transmittance at the two conditions. Firstly, the same phase
difference ( effndΔ = 100 nm) but different cell thicknesses ( d = 5.53, 10 μm), and the
result is shown in Fig. 2-14. Secondly, the constant cell thickness ( d = 5.53 μm) with
several different phase retardations ( effndΔ = 10, 50, 100, 300 nm), and Fig. 2-15 shows
the results. During the calculations, we considered fully the dispersion properties of the
refractive indices of liquid crystal (LC53) as the same as we did in Fig. 2-13, and the
corresponding effective refractive indices are in Figs. 2-16 and 2-17. From these results,
we know that the higher frequency mode is coming from the interference of the e-mode
and o-mode independently in a LCD and depends on the absolute light path lengths of
each mode. The lower frequency mode is caused by the interference between e-mode and
o-mode and affected by the relative phase difference of them. Considering the optical
properties, the lower frequency mode is much more critical, so we need to reduce the
residual birefringence to improve the dark quality of a LCD irrespective of the
compensation films. Another important thing is that the lower frequency mode is
governed by the optically anisotropic layers, so at which interface the multi-reflection
takes place is not so important in a real LCD because other layers except the liquid
crystal are mostly isotropic material. Therefore, the spacing between the interfaces
42
causing the reflections affects only on the higher frequency mode, as in Fig. 2-14 and
does not effect the color shift.
43
380 480 580 680 7800.000
0.001
0.002
0.003
d=5.53 μm d=10.0 μm
Wavelength (nm)
Tran
smitt
ance
Fig. 2-14. Thickness effect on the total transmittance of a Pi-cell-type LCD in the dark
state. Light path lengths of the e-mode and o-mode affect the high frequency mode.
( effndΔ =100 nm fixed).
44
380 480 580 680 7800.000
0.001
0.002
0.003
Δnd=0 nm Δnd=50 nm Δnd=100 nm Δnd=300 nm
Wavelength (nm)
Tran
smitt
ance
Fig. 2-15. Phase retardation effect on the total transmittance of a Pi-cell-type LCD in the
dark state. The low frequency mode is related to the phase difference between the e-mode
and o-mode. ( d =5.53 μm fixed).
45
380 480 580 680 780
1.50
1.55
1.60
1.65
1.70
1.75 n
e n
o
neeff (d=5.53μm)
neeff (d=10.0μm)
Wavelength (nm)
Ref
ract
ive
Indi
ces
Fig. 2-16. Effective refractive indices ( effen ) of the liquid crystal (LC53) with different
thicknesses and a constant phase retardation ( effndΔ =100 nm).
46
380 480 580 680 780
1.50
1.55
1.60
1.65
1.70
1.75 ne
no
neeff (Δnd=50 nm)
neeff (Δnd=100 nm)
neeff (Δnd=300 nm)
Wavelength (nm)
Ref
ract
ive
Indi
ces
Fig. 2-17. Effective refractive indices ( effen ) of the liquid crystal (LC53) with a constant
thickness ( d =5.53 μm) and different phase retardations.
47
2-5-4. Refractive index mismatching effects in Pi-cell type LCDs
There are about 15 to 20 layers in a real active matrix liquid crystal display
(AMLCD). Each of the interfaces between them can cause multi-reflection. Refractive
index mismatching at those interfaces are the source of multi-reflection in LCDs. We
calculated the effects of index mismatching on the light transmittance in dark states of
LCDs. The cell thickness ( d ) is 5.53 µm, and the effective residual retardation of the cell
( effndΔ ) is 100 nm at 550 nm of incident light wavelength. The liquid crystal is LC53,
and we fully considered the dispersion of the refractive indices of the liquid crystal as we
did in Sec. 2-5-2, and the refractive indices are in Fig. 2-12. However, when we calculate
the reflectance coefficients, we used real refractive indices ( en , on ) as the same reason.
Figure 2-18 shows the calculated results of the dark state transmittances in terms
of the magnitudes of index mismatching ( airo nn −)(λ ). As we can see in this figure, the
index mismatching strongly affects not only on the amplitude of interference pattern but
also on the absolute value of light leakage in dark states at all wavelengths. As expected,
the light leakage and the amplitude of interference due to multi-reflections sharply
decrease as the magnitude of index mismatching ( airo nn − ) falls.
As a rough calculation, let us say that we want to make a Pi-cell type LCD that
has at least a contrast ratio of 1000 (luminance of bright state / luminance of dark state).
We assume that polarizers are ideal, and there are no other sources that could give light
leakage in the dark state of the LCD. We also assume that the LCD is optically designed
to have transmittance of 0.35 at a normal direction. That means we have to achieve less
48
than 0.00035 of dark state transmittance. The refractive index of ordinary ray ( on ) of the
liquid crystal (LC 53) is about 1.5 at a wavelength of 550 nm. Therefore, from Fig. 2-18
we can say that we should control the magnitude of index mismatching ( airo nn − ) under
0.3. If the Pi-cell is contacting air interfaces directly, the magnitude of index mismatching
( airo nn − ) is about 0.5. Therefore, the maximum contrast ratio will be about 350 from
Fig. 2-18.
49
380 480 580 680 780
0.000
0.001
0.002
0.003 n
air=1.0
nair
=1.1 n
air=1.2
nair
=1.3 n
air=1.4
nair
=1.5
Wavelength (nm)
Tran
smitt
ance
Fig. 2-18. The effects on the dark state transmittance due to the refractive index
mismatching ( airo nn − ). ( d =5.53 μm, effndΔ =100 nm)
50
2-6. Experimental and numerical calculation results of a Pi-cell
2-6-1. Compensating a Pi-cell by using a uniaxial crystal compensator
To confirm the calculation results, we did experiments using a Pi-cell (liquid
crystal: LC53; cell thickness of 5.53 μm, pretilt angle of 5.5º on both sides). We used a
commercialized uniaxial compensator to compensate the phase retardation of the dark
state Pi-cell (applied voltage of 5.15 V) at the normal direction. The compensated and
measured wavelength range was 400-700 nm, and the step was 10 nm. However, the
phase difference of the LCD is hard to compensate perfectly over all of the wavelength
range simultaneously. Therefore, for each wavelength, we changed the retardation values
of the compensator to have minimum transmittance where the phase difference of the Pi-
cell is compensated exactly. Figure 2-19 (a) is the measurement layout, and Fig. 2-19 (b)
shows the structure of the Pi-cell. We used crystal polarizers to achieve high extinction
ratio in this experiment (Melles Griot, model 03PTO003/A, extinction ratio < 1/100 000).
The spectrophotometer (Model: USB2000 Ocean Optics) was used as a light detector.
Each optical component in the setup is separated physically, and the surfaces of the
crystal polarizers and the compensator are treated for anti-reflection in visible light
wavelength range. Therefore, the main sources of the multiple reflections are the either
sides of the Pi-cell and the layers inside the Pi-cell such as substrates, electrodes,
alignment layers, and liquid crystal.
Figure 2-20 (a) shows the measurement result of the dark state transmittance of
the Pi-cell whose phase retardations are compensated separately for each wavelength, and
Fig. 2-20 (b) is the minimum transmittance we take from the experimental raw data [Fig.
51
2-20 (a)]. As we can see in Fig. 2-20, there is a big light leakage in all of the wavelengths,
and especially in the blue region. This means the maximum contrast ratio of the Pi-cell is
limited by the multi-reflection effect, and also that effect could cause the blue shift in the
color of the dark state. In case of this experiment, the calculation results of the change of
the color coordinates [1931 CIE (International Commission on Illumination)] for the
standard D65 light source are
x = 0.3127, y = 3291 → x = 0.2722, y = 0.2802.
52
(a) Measurement setup of a Pi-cell compensation system
(b) Cell structure of the Pi-cell
Fig. 2-19. Measurement setup of a Pi-cell compensation system and the cell structure of
the Pi-cell. [Applied voltage of 5.15 V (dark state)].
Indium Tin Oxide (ITO) Polyimide
Liquid Crystal
Polyimide Indium Tin Oxide (ITO)
Glass
Glass
53
400 450 500 550 600 650 7000
1
2
3
4
5
Wavelength (nm)
Tran
smitt
ance
(%)
(a) Measured transmittance of the dark state Pi-cell
400 450 500 550 600 650 7000.00
0.01
0.02
Wavelength (nm)
Tran
smitt
ance
(b) Extracted minimum transmittance of the dark state Pi-cell
Fig. 2-20. Measured transmittance and the extracted minimum transmittance of the dark
state Pi-cell compensated with the compensator at each wavelength separately. Reference
(T=1): parallel crystal polarizers; LC volt: 5.15 V (dark state).
54
2-6-2. Compensating a Pi-cell by using hybrid-aligned negative C-plates
Hybrid-aligned negative C-plates or recently it is called polymerized discotic
material (PDM) was introduced by Mori et al.6, 7, 8 in 1997 to enlarge the viewing angle
and to reduce the driving voltage of a Pi-cell. The basic his idea is that the effect of each
layer of a positive birefringence material in a liquid crystal device can be optically
compensated by a layer of negative birefringence materials with the same optic axis
orientation. The PDM9 layer combined with a biaxial film [they are called wide view
(WV)-film or Fuji-film) is widely used now in many liquid crystal display modes
including twisted nematic (TN) and Pi-cells, and Fig. 2-21 shows the basic structure
when it is applied to a Pi-cell. As the figure, the PDM layers have negative birefringence
[ 0)( <−Δ oe nnn ] for optical compensation of the liquid crystal layer which has positive
birefringence [ 0)( >−Δ oe nnn ]. The optic axes of the PDM layer have similar angle
distribution compared with that of liquid crystal directors.
55
Fig. 2-21. Basic structure of the compensation scheme of a Pi-cell by using WV-Film.
The PDM layers have negative birefringence for optical compensation of the liquid
crystal that has positive birefringence. The optic axes of the PDM layer have similar
angle distribution compared with that of liquid crystal directors.
Pi-Cell Δn > 0
Biaxial Film
Polarizer
Polarizer
Biaxial Film
PDM Δn < 0
PDM Δn < 0
WV-Film
WV-Film
56
2-6-2-1. Multi-reflection effects in a real Pi-cell
We used the WV-film that is designed for a Pi-cell to compensate the phase
retardation of the dark state of a Pi-cell. The optical stack configuration of the
compensated real Pi-cell and the measurement setup are shown in Fig. 2-22. In this
experiment, we used sheet polarizers in place of the crystal polarizers with the same
angles. The reference of transmittance that gives 100% of transmittance was the
transmittance of the parallel sheet polarizers. All films and the Pi-cell are combined as a
single unit so that we can achieve minimum index mismatching at the interfaces unlike
the previous experiment. We applied voltage ranging from 4.9 to 5.7 V to liquid crystal
layer to get the minimum transmittance at each wavelength where the phase retardation
of the liquid crystal is compensated by the compensation films (PDM layers) perfectly
because the Pi-cell and the films are acting like the crossed uniaxial layers.
Figure 2-23 shows the measurement results of transmittances. From this
experiment, we see that there is a big multi-reflection effect in the dark state of the real
Pi-cell, and its effect is much more severe in the blue region. In this figure, we also see
that there is light leakage above the light wavelength of 700 nm. This is coming from the
light leakage of the crossed sheet polarizers as we can see in Fig. 2-24(a). An important
fact to note is that the minimum transmittances we achieved at each wavelength are
bigger than the transmittance of the crossed polarizers, and that difference increases in
the blue region even though we compensated the phase of the liquid crystal separately in
each wavelength. This experimental fact agrees well with our calculation results in Sec.
57
2-5. Therefore, this situation should lead to the blue shift of the dark color of the Pi-cell,
and the maximum contrast ratio we can get is around 1000 in this condition.
58
(a) Optical stack configuration of a compensated real Pi-cell
(b) Measurement setup
Fig. 2-22. Optical stack configuration of a compensated real Pi-cell and the measurement
setup. All layers are combined as a single unit without any air gaps.
Halogen Lamp
Pi-cell combined with all films
Spectro-photometer
Indium Tin Oxide (ITO) Polyimide
Liquid Crystal Polyimide
Indium Tin Oxide (ITO) Glass
Compensator (PDM) Biaxial film Polarizer
Triacetyl Cellulose (TAC)
Triacetyl Cellulose (TAC) Polarizer
Biaxial film Compensator (PDM)
Glass
Light Source
59
400 450 500 550 600 650 700 7500.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0 4.9 V 5.0 V 5.1 V 5.2 V 5.3 V 5.4 V 5.5 V 5.6 V 5.7 V
Wavelength (nm)
Tran
smitt
ance
(%)
Fig. 2-23. Measurement results of the transmittance of a compensated real Pi-cell. All
films including polarizers and compensators are attached to the Pi-cell as a single unit.
Reference transmittance (T=100%): transmittance of parallel sheet polarizers.
60
2-6-2-2. Multi-reflection effects in a real Pi-cell with additional crystal polarizers
To remove the effects of sheet polarizers that have relatively large transmittance
in their crossed state, we did another experiment. We used the same setup and Pi-cell but
inserted additional crystal polarizers (with the same angles of sheet polarizers, Melles
Griot, model 03PTO003/A, extinction ratio < 1/100 000 at visible wavelengths) outside
the sheet polarizers to get rid of the light leakage that takes place in crossed sheet
polarizers. In this experiment, the reference of transmittance that gives 100% of
transmittance was the parallel crystal polarizers. It means the measurement results
include the absorption effect of sheet polarizers. Figure 2-25 (a), (b) show the
measurement result and experimental setup, respectively. In this result, there in no big
light leakage above the light wavelength of 700 nm unlikely in Fig. 2-23, and this is
because we used almost ideal additional polarizers. Another different point from Fig. 2-
23 is that the light leakage in blue wavelengths (about 400-480 nm) decreases as the
wavelength decreases, and this is opposite direction compared with the result in Fig. 2-23.
This result is understandable if we see the transmittance curve of parallel sheet polarizers
[Fig. 2-24(b)]. As we mentioned above, the transmittance in Fig. 2-25 includes the
absorption effects of sheet polarizers, and they have large absorption in blue region
regardless of the polarization of incident light as we can see in Fig. 2-24(b). From this
experiment, we can say that there is light leakage in a Pi-cell even though we compensate
the phase retardation of the dark state perfectly and use ideal polarizers. This light
leakage is coming from the multiple reflection effect in a LCD, and it depends on the
light wavelength. Therefore, it could affect the color of the dark state of a LCD.
61
400 450 500 550 600 650 700 7500.00
0.05
0.10
0.15
0.20
0.25
0.30
Wavelength (nm)
Tran
smitt
ance
(%)
(a) Crossed sheet polarizers
400 450 500 550 600 650 700 7500
20
40
60
80
100
Wavelength (nm)
Tran
smitt
ance
(%)
(b) Parallel sheet polarizers
Fig. 2-24. Transmittances of the crossed sheet polarizers and the parallel sheet polarizers.
Reference transmittance (T=100%): transmittance in air.
62
400 450 500 550 600 650 700 7500.0
0.2
0.4
0.6
0.8
1.0 4.9 V 5.0 V 5.1 V 5.2 V 5.3 V 5.4 V 5.5 V 5.6 V 5.7 V
Wavelength (nm)
Tran
smitt
ance
(%)
(a) Measurement results of the transmittances
(b) Experimental setup
Fig. 2-25. Measurement results of the transmittances of a compensated Pi-cell with
additional crystal polarizers and the experimental setup. Reference transmittance
(T=100%): transmittance of parallel crystal polarizers.
Halogen Lamp
Crystal Polarizer
Pi-cell combined with
all films
Spectro-photometer
Crystal Polarizer
63
2-6-2-3. Refractive index mismatching effects in a Pi-cell
As we described in Sec. 2-5-4, refractive index mismatching is a very important
factor in multi-reflections. In order to see the effects experimentally, we did another
similar experiment. All methods and samples including a Pi-cell and compensation films
are the same things that we used in the previous section. The difference is that we
separated the compensation films from the Pi-cell to make big index mismatching
between the Pi-cell and compensation films by inserting an air gap. The measurement
results and the experimental setup are shown in Fig. 2-26 (a), (b), respectively.
Comparing with Fig. 2-25, overall interference patterns are the same, but the magnitude
of light leakage increases more than three times. This trend agrees with our analytical
calculation results (Fig. 2-18). Therefore, we can say that the refractive index
mismatching should be one of the important factors when we design a LCD; otherwise, it
could not only increase the light leakage but also make color shift (possible blue shift) of
a dark state.
In Fig. 2-25, the maximum contrast ratio that we can achieve is about 100/0.1 =
1000 as a rough calculation at an incident light wavelength of 550nm. The refractive
index of the ordinary ray ( on ) of the liquid crystal (LC 53) is about 1.5 at a wavelength
of 550 nm. The layer that has the maximum refractive index in LCD layers is Indium Tin
Oxide (ITO) that has about 1.8 of refractive index at a wavelength of 550nm. It means the
magnitude of refractive index mismatching ( ITOo nn − ) is about 0.3. This result of a
rough calculation based on the experimental data agrees well with that of the analytical
calculation in Sec. 2-5-4.
64
If we do similar calculation in Fig. 2-26 that is measured with bigger refractive
index mismatching, the maximum contrast ratio is about 100/0.3≈330 at a wavelength of
550nm. In this case, the magnitude of index mismatching ( airo nn − ) is about 0.5. This
experimental result is also very consistent with that of the analytical calculation in Sec. 2-
5-4.
65
400 450 500 550 600 650 700 7500.0
0.2
0.4
0.6
0.8
1.0 4.9 V 5.0 V 5.1 V 5.2 V 5.3 V 5.4 V 5.5 V 5.6 V 5.7 V
Wavelength (nm)
Tran
smitt
ance
(%)
(a) Measurement results of the transmittances
(b) Experimental setup
Fig. 2-26. Measurement results of the transmittances of a compensated Pi-cell with
additional crystal polarizers and air gaps, and the experimental setup. Reference
transmittance (T=100%): transmittance of parallel crystal polarizers.
Halogen Lamp
Crystal Polarizer
Pi-cell
Spectro-photometer
Crystal Polarizer
Sheet polarizers and compensators (WV-films)
66
2-6-3. Numerical calculations of the multi-reflection effects in a Pi-cell
We simulated these experiments by using a numerical relaxation technique to
calculate the director field in the Pi-cell and by using the Berreman 4×4 method10, 11, 12,
13 to calculate the optical properties. One of the problems when we use the Berreman 4x4
method is that there is a Fabry-Perot effect in the calculation results due to the multiple
reflections in a cell. This effect is coming from the assumption that the incident light has
an infinite coherence length14. However, most of light sources have a finite coherence
length in real situations. To remove the Fabry-Perot effect, there are several methods
such as spectrum averaging of transmission light, apodization method15, 16, restricting the
number of multi-reflections 17, and changing the refractive index of outside media 18.
During our optical calculations, we used the spectrum averaging method with 1.0 nm of
averaging bandwidth and 0.01 nm of calculation interval. We made this calculation
program by ourselves.
In these calculations, we used exactly the same optical stack as in Fig. 2-22 and
considered the dispersion of the refractive indices of all layers. The optical parameters of
the layers that we did not have from the manufacturer such as the glass, ITO (Indium Tin
Oxide), and polyimide, were measured by ourselves at a resolution of 1 nm by using a
spectroscopic ellipsometry (model: WVASE 32, J.A. Woollam. Co., Inc.), and the results
are in Fig. 2-27.
Figure 2-28 (a), (b) shows the measurement and numerical calculation results,
respectively. The only measured parameter that was adjusted to acquire Fig. 2-28 (b) was
the thickness of the liquid crystal layer. The thickness of the liquid crystal layer was
67
measured before filling the liquid crystal as 5.53 µm, but we adjusted the value to 5.18
µm during the calculation to achieve the best fitting to the experimental data. We think
this is acceptable because the thickness of the liquid crystal layer could be changed
during the cell-making process.
68
380 480 580 680 7801.0
1.5
2.0
2.5 Glass ITO Polyimide
Wavelength (nm)
Ref
ract
ive
Inde
x, n
(a) Real parts of the refractive indices
380 480 580 680 780
0.000
0.004
0.008
0.012
0.016
0.020 Glass ITO Polyimide
Wavelength (nm)
Ref
ract
ive
Inde
x, k
(b) Imaginary parts of the refractive indices
Fig. 2-27. Measurement data of the real and imaginary parts of refractive indices.
Instrument: WVASE 32 spectroscopic ellipsometry (J.A. Woollam. Co., Inc.)
69
400 450 500 550 600 650 7000.000
0.005
0.010
0.015
0.020 5.0V 5.2V 5.4V 5.6V
Wavelength (nm)
Tran
smitt
ance
(a) Measurement results
400 450 500 550 600 650 7000.000
0.005
0.010
0.015
0.020 5.0V 5.2V 5.4V 5.6V
Wavelength (nm)
Tran
smitt
ance
(b) Numerical calculation results
Fig. 2-28. Measurement results and the numerical calculation results of the transmittance
of the Pi-cell compensated with the hybrid aligned negative-discotic films for different
applied voltages. Reference (T=1): parallel sheet polarizers.
70
2-7. Summary
We calculated the multi-reflection effects analytically and numerically in the dark
state of a compensated liquid crystal device and compared the results with the measured
transmittance of an example device.
According to our analysis, there are two types of interference in devices with
significant residual retardation in the dark state that is compensated by a passive optical
retarder. The first one is due to the pure e-ray and pure o-ray by themselves, and the
second one is coming from the coupling between the e-mode and o-mode. The first type
has higher frequency in the wavelength space and is related to the optical path length of
the e-ray and o-ray and is independent of their difference. Most of the modes used in
liquid crystal devices have this type of interference. The second type of interference has
lower frequency than that of the first one and depends on the residual birefringence of
dark state. So, as the residual birefringence increases, the second type of interference
becomes more significant. In the viewpoint of the optical properties of a liquid crystal
device, the first type of interference could affect the dark level and extinction ratio almost
equivalently for visible wavelength region. On the other hand, the second type could
cause a wavelength dependence of the extinction ratio, or a color shift of the dark state
because of the lower frequency pattern of the interference in the wavelength space.
71
CHAPTER 3
UNIVERSAL OFF-AXIS LIGHT TRANSMISSION PROPERTIES OF THE
BRIGHT STATE IN COMPENSATED LIQUID CRYSTAL
DEVICES
3-1. Introduction
The viewing angle dependence of the light transmittance in liquid crystal devices
is a well-known feature and one of the biggest problems. The dark (black) state is
especially important because it is critical to the contrast ratio of the device. On this
account, a great deal of research has been done to reduce the dark state transmittance and
its variation at all viewing angles.
On the other hand, the bright (white) state determines the luminance and the
spectral variations of the transmittance as a function of the angle of the incident light.
This bright state is especially important in the applications of large size displays such as
LCD monitors, TVs, and signs. Therefore, this should be another vital factor when
deciding the optical properties of the devices. However, we have not seen those studies
dealing with the off-axis light transmission properties (viewing angle properties) of the
bright state of general liquid crystal devices.
In this work, we investigate the off-axis light transmission properties (viewing
angle properties) of the bright state in the most common liquid crystal devices whose
dark states are optically compensated to have minimum transmittance at the normal
72
direction. According to our results, there is an interesting universality in the off-axis light
transmission properties of the bright state of liquid crystal devices, which is independent
of the display modes. In order to explain these interesting facts, we make simple dark and
bright state models that can be applied to general liquid crystal devices and analyze them
in terms of the effective retardation and transmittance.
In Sec. 3-2, we will show the off-axis light transmission properties of the bright
state in the most commonly used liquid crystal devices. In Sec. 3-3, we will introduce
simple dark and bright state models, and in Sec. 3-4, we will calculate the angular
dependencies of the effective retardation and transmittance of the models and then
compare the results with those of Sec. 3-2. Finally, we will compare the viewing angle
properties of our models with those of the real liquid crystal devices in all viewing
directions in Sec. 3-5.
3-2. Optical properties of general liquid crystal devices
Figure 3-1 shows the cartoons of director configurations of the most common
liquid crystal display (LCD) modes, such as the electrically controlled birefringence
(ECB), vertical alignment (VA), Pi-cell19, 20 (or OCB a, 21, 22), and symmetric splay-cell.
With the twisted nematic (TN), these modes cover most of the popular concepts of LCDs
that use nematic liquid crystal materials. The bright and dark states of each mode can be
a Pi-cell was invented in 1983 by Bos. Subsequently, Uchida used it with a single biaxial film as a
compensator with normally black mode and called it OCB in 1993. Their structures of the liquid crystal and
polarizers are same.
73
exchangeable by changing the optical design-parameters such as polarizer angles and
compensation films. In this dissertation, we set the bright and dark states as in Fig. 3-1
because those methods are widely used now to improve the dark state qualities.
74
Fig. 3-1. Director configurations of the most common liquid crystal display modes. The
dark and bright states are exchangeable by optical design.
Dark Bright Dark Bright
ECB VA
Pi-cell Symmetric splay
Dark Bright Bright Dark
75
3-2-1. Calculation methods and compensations of the dark states of the LCDs
We calculated the off-axis light transmission (or viewing angle) properties of
these LCD modes in their dark and bright states with an incident light wavelength of 550
nm. During the calculation, we used the numerical relaxation technique to get the director
field of the liquid crystal layer. We also used the 2×2-matrix method23, 24, 25 to calculate
the optical properties of the devices. We made these calculation programs by ourselves.
The dark state of each mode is optically compensated via two (above and below
the liquid crystal layer) compensators with the hybrid-aligned negative C plates as
discussed in Sec. 2-6-2. They have exactly the same angular distribution as the directors
of the liquid crystal layer of the dark state, and their extraordinary and ordinary refractive
indices are the same as the ordinary and extraordinary indices of the liquid crystal,
respectively. This is in accordance with Mori’s argument26, 27 that the effect of each layer
of a positive birefringence material in a liquid crystal device can be optically
compensated by a layer of negative birefringence materials with the same optic axis
orientation. Other methods, for example, using positive O-plates28, 29 or biaxial material30
can be used for optical compensation of the dark state. However, their final destinations
are the same, i.e. we want to make the total effective retardation of liquid crystal and
compensator layers zero in all viewing directions and wavelengths if we use ideal
polarizers. Therefore, the more we perfectly compensate the dark state, the more the final
compensator effects are the same, no matter what compensation schemes are used.
76
3-2-2. Optical parameters of the LCDs
The thicknesses of the liquid crystal layer and the compensators of each display
mode are determined by meeting two conditions. Firstly, it gives the minimum
transmittance at all viewing angles in the dark state. Secondly, the transmittance of the
bright state at the normal direction has a specific value ( oT ). This value can be chosen
from zero to maximum transmittance that we can achieve under the crossed polarizers,
but a bit smaller than the maximum value is usually used to achieve high transmittance
and to escape gray scale inversions and color shift in the off axis viewing directions. In
this dissertation, we set the value to ¾ of the maximum transmittance because this is a
similar condition in real liquid crystal devices, and this value is not a critical factor in
determining the physical concept of devices.
The detailed cell specifications and the stack configuration are in Table 3-1 and
Fig. 3-2, respectively. Where, the liquid crystal we used is LC53, and we assumed the
polarizers are ideal, i.e. the light that polarized along the absorption axis of a polarizer is
absorbed completely, but the light that is polarized along the transmission axis of a
polarizer is transmitted perfectly except surface reflections. The refractive indices of the
glass, indium tin oxide (ITO), and polyimide are in Fig. 2-27, and their thicknesses are
0.7 mm, 400 Å, and 500 Å. We want to make clear that the specific numbers are only for
the purposes of producing a graph that demonstrates the features of the general concepts
considered here.
77
Table 3-1. Cell parameters of the common liquid crystal (LC) devices we used. These
specific numbers are only for the purposes of producing a graph that demonstrates the
features. (LC: en =1.6644 and on =1.5070 at λ=550 nm; Δε=-9.4 for VA, +9.4 for other
devices).
Thickness (µm)
Easy axis (°)
Voltage (V)
Pretilt angle(°) Device
modes LC Compensator Top Bottom Bright Dark LC Pi-cell 5.037 2.255 90 90 1.3 5.0 5.5 ECB 1.423 0.637 90 -90 0.0 5.0 5.5 VA 1.455 0.541 90 -90 5.0 0.0 84.5 TN 2.038 0.832 45 -45 0.0 5.0 5.5
Splay-cell 1.87 0.934 90 90 5.0 0.0 5.5
78
Fig. 3-2. Stack configuration of the common liquid crystal devices.
ITO Polyimide
Liquid Crystal
Polyimide ITO
Glass
Compensator (PDM)
Polarizer (+45º)
Polarizer (-45º)
Compensator (PDM)
Glass
Light Source
79
3-2-3. Calculation results of the off-axis light transmittance in dark states
Figure 3-3 depicts the numerical calculation results of the off-axis light
transmission properties of the common liquid crystal devices in their dark states, whose
phase retardations are compensated as described in Sec. 3-2-1. We added the off-axis
light transmittance curves of the just crossed polarizers for the comparison. We also
calculated the director tilt angles and effective phase retardation of the dark state-liquid
crystal layer of each device, as in Fig. 3-4 and Fig. 3-5, respectively. These figures show
that the dark state transmittances of the common LCDs have the similar level of the
crossed polarizers in main viewing directions. Therefore, we can say that the effective
phase retardation of each dark state-liquid crystal layer of the common LCDs is optically
compensated well.
Another thing we want to note here is that there is relatively big light leakage at
the off-axis viewing angles, as in Fig. 3-3 (a) and (b), although we used ideal crossed
polarizers. This light leakage is caused by the fact that the absorption planes of the
crossed polarizers are not perpendicular with each other at off-axis viewing angles except
the planes that include the absorption or transmission axis of the one of the polarizers.
This fact was firstly pointed out by Chen et al.31 in 1998, and the problem can be curable
by using an A-plate and a C-plate or single biaxial film. To improve the dispersion
properties of light wavelength, using two-biaxial films methods were introduced32, 33, 34,
and, recently, wide viewing-angle polarizers having relaxed manufacturing tolerances
were presented35.
80
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Pi-cell ECB VA TN Splay-cell Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane (Φ=0º)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Pi-cell ECB VA TN Splay-cell Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane (Φ=90º)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Pi-cell ECB VA TN Splay-cell Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(c) Φ=135º (Most of the curves are overlapped.)
Fig. 3-3. Numerical calculation results of the off-axis light transmission properties of the
common liquid crystal devices in their dark states.
81
0.0 0.2 0.4 0.6 0.8 1.0
0
50
100
150
200 Pi-cell ECB VA Splay-cell
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 3-4. Director tilt angles of the dark state-liquid crystal layers.
82
-80 -60 -40 -20 0 20 40 60 80-50
0
50
100
150
200
250
300
350
400
450
500 Pi-cell ECB VA Splay-cell
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-5. Effective phase retardation of the dark state-liquid crystal layers in the director
plane. (λ=550 nm).
83
3-2-4. Calculation results of the off-axis light transmittance in bright states
We numerically calculated the off-axis light transmission properties of the
common liquid crystal devices in their bright states, and the Fig. 3-6 shows the results.
From these transmittance figures, we notice that even though the director configurations
and effective birefringence of the liquid crystal layers are completely different from each
other, as in Fig. 3-7 and Fig. 3-8, the off-axis light transmission properties of the
compensated liquid crystal devices, amazingly, have unified shapes. (We cannot define
the birefringence out of the director plane because the optic axes of the directors are
apparently twisted, so we calculated it only in the director plane. For the same reason,
only the off-axis light transmission properties are calculated in TN mode.) In the viewing
angle of the director plane [Fig. 3-6 (a)], all the transmittance curves have similar “bell”
shapes, i.e. the transmittance constantly decreases as the viewing angle increases from the
normal direction no matter what liquid crystal modes are used. On the other hand, the
transmittance out of the director plane [Fig. 3-6 (b)] rises as the viewing angle increases
from the normal direction and then falls after passing the specific angles (about ± 50° in
these calculations), regardless of the director configurations of the liquid crystal in their
bright states. These results show that the off-axis light transmission properties of the
bright state in single domain liquid crystal devices have a common shape, independent of
their modes, as long as their dark states are optically compensated to give the lowest
transmittance for all viewing angles. This surprising fact motivated us to investigate the
reason for this in the next section.
84
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5 Pi-cell ECB VA TN Splay-cell
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Pi-cell ECB VA TN Splay-cell
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 3-6. Numerical calculation results of the off-axis light transmission properties of the
common liquid crystal devices in their bright state.
85
0.0 0.2 0.4 0.6 0.8 1.0-100
-50
0
50
100
150
200 Pi-cell ECB VA Splay-cell
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 3-7. Director tilt angles of the bright state-liquid crystal layers.
86
-80 -60 -40 -20 0 20 40 60 800
100
200
300
400
500 Pi-cell ECB VA Splay-cell
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-8. Effective phase retardation of the bright state-liquid crystal layers in the director
plane. (λ=550 nm).
87
3-3. Universal simple models
Because we have seen universality in the bright state of many display modes that
have an almost perfectly compensated dark state, we are looking for a very basic
explanation of the bright state viewing angle properties that does not include highly
detailed considerations. In this section, we will go to two different ways, but finally, we
will achieve the same simple bright state model.
3-3-1. Simple dark and bright state modeling
Let us consider the liquid crystal director configurations of several non-twisted
common liquid crystal devices (LCDs) such as ECB, VA, Pi-cell, and symmetric splay-
cell. Figure 3-9 (a) shows cartoons of the liquid crystal director configurations of each
device. Each of the devices has two states, state 1 and state 2, depending on applied
voltage, and either of them is bright state and the other is dark state. In these two states,
we can notice that there are simply two parts in a liquid crystal layer: the static part
(empty director shape in the figure) and dynamic part (filled director shape). In the static
part, most of the liquid crystal directors only slightly change their orientation between the
two states. On the contrary, the directors in the dynamic part are very sensitive to the
applied voltage, so the orientation of the directors is completely different between the
state 1 and state 2 of each device as in Fig. 3-4 and Fig. 3-7. Considering the director-tilt
angles, both states have vertical and horizontal components, but state 1 has a larger
vertical component than state 2, and state 2 has a larger horizontal component than state 1.
We can conceptually think of each liquid crystal device as being composed of two types
88
of layers, static and dynamic, as shown in Fig. 3-9 (a). The director configurations of the
static layers are the same in both state 1 and 2. Therefore, the only difference between the
two states is in the dynamic layers.
We consider compensating one of the two states perfectly to make it the dark state
by using passive, negative type optical compensators (negative A and C-plates). In order
to do that, we set the direction of the optic axes of the negative plates to be exactly the
same as those of the positive plates of the dark state. There are two ways: mode 1 and
mode 2. In mode 1, we use the state 1 as the dark state and the state 2 as the bright state,
and they are reversed in mode 2, as shown in Fig. 3-9 (b). In this figure, we can see that
the static layers of liquid crystal in mode 1 and mode 2 are always optically completely
compensated in both dark and bright states, simultaneously by their corresponding
components (empty shapes in the figure) of the compensators. Therefore, the contribution
of the static layers to the optical transmittance is canceled by the compensators, and
finally we achieve the simple dark and bright state models of mode 1 and 2 as in Fig. 3-9
(c). In these simple models, the dark states are optically neutralized in both modes, but
the bright states have a net birefringence that is a function of the viewing direction.
89
Fig. 3-9. (a) Simplification of the liquid crystal layers of the various devices, (b) optical
compensation of the simplified liquid crystal layer and (c) simple dark and bright state
models. The long axis of the ellipses indicates the optic axis orientation of the liquid
crystal with a positive birefringence. The symmetry axis of the disks indicates the optic
axis of the compensator with a negative birefringence. The shaded ellipses represent
directors that change, going from state 1 to state 2, and the shaded disks represent the
component of the compensator that compensates for them in one of the states.
90
3-3-2. “3-layer” modeling
As the second way, we will consider a simple “3-layer” model. This type of
model was used in early calculation of LCDs before detailed calculations were easy to
perform. The “3-layer” model of the common display modes is shown in Fig. 3-10 (a). In
each case, the display mode is shown along with a compensator that is assumed to
compensate the dark state of the device. The director distribution of the compensator is
shown as outlined directors to indicate that the material has a negative birefringence.
From this figure, we can see that the net optical effect of the compensator and the
bright state of the liquid crystal layer is only seen in the mid-layer of each device.
Furthermore, the difference between the compensator and the bright state of the liquid
crystal is the same in all display modes. As a result, we get the simple bright state model
composed of a positive A-plate and a negative C-plate as in Fig. 3-10 (b). This simple
model is exactly the same as the mode 1 in Fig. 3-9 (c).
From this bright state model, we can see that simple interaction between a
uniform negative C-plate and a positive A-plate might be expected to describe the
viewing angle properties of the bright state of most compensated liquid crystal devices.
This implies that the physics of the device is very simple and might be able to be
explained by a very simple and intuitive optical model.
91
Fig. 3-10. (a) “3-layer” modeling of the most common LCD modes and (b) simple bright
state model.
(b)
Bright Dark
Bright
ECB
Dark Bright Dark
Pi-cell
VA TN(a)
92
3-4. Calculations of the optical properties of the simple models
3-4-1. Basic calculations
The light transmittance (T ) of the simple model [Fig. 3-9 (c)] under ideal crossed
polarizers can be written as follows:
2sin
2sin
21 2
max2 Γ
≡Γ
= TTTT paap , (3-1)
where apT and paT are the transmittances with considering the surface reflections at the
interfaces between air and polarizers in the incident and exit media, respectively. They
can be calculated through the Fresnel equations36,
2
coscos
ijii
jjij t
nn
Tθθ
= , (3-2)
where in and jn are the refractive indices of the incident and refracted media, iθ and jθ
are the incident and refraction angles, respectively, and ijt is the transmission coefficient.
The p -polarization and s -polarization components of ijt are expressed in Eqs. (3-3) and
(3-4), respectively,
93
jiij
iipij nn
nt
θθθcoscos
cos2+
= , (3-3)
jjii
iisij nn
nt
θθθcoscos
cos2+
= . (3-4)
The Γ in Eq. (3-1) is the phase difference between e-ray and o-ray in the anisotropic
layers and written in terms of incident light wavelength (λ ) and effective retardation
( effndΔ ) of media when we assume the e-ray and o-ray have the same path in the media,
effndΔ=Γλπ2 . (3-5)
[The exact expression of the phase difference (Γ ) is dKK ozez )( −=Γ , where ezK and
ozK are z components of the wave vectors of the e-ray and o-ray, respectively. For
simplicity, we assume the e-ray and o-ray have the same path in this dissertation, which
is reasonable in a real situation.] In the normal direction, let us say the transmittance of
the bright state is oT , and the corresponding phase difference is oΓ . Then, as indicated by
Eq. (3-1),
21
max
1sin2 ⎟⎟⎠
⎞⎜⎜⎝
⎛=Γ −
TTo
o . (3-6)
94
This phase difference is purely caused by the positive A-plate (bottom layer) of the bright
state model in mode 1 of Fig. 3-9 (c) because the optical axis of the negative C-plate (top
layer) is in the normal direction. With similar reasoning, the phase difference is
exclusively related to the negative A-plate (top layer) of the bright state model in mode 2.
From these facts and Eqs. (3-5) and (3-6), we can determine the thicknesses of the
positive A(C)-plate (bottom layer) in mode 1 and the negative A-plate (top layer) in mode
2 that gives the bright state transmittance oT for a given material in the normal direction.
In mode 1, the incident light-angle-dependence of the effective retardation
( effndΔ ) of the positive A(C)-plate (bottom layer) and negative C-plate (top layer) are
calculated in Eqs. (3-7) and (3-8), respectively,
απλ
ψψ cos2sincos 2222 nn
nn
nnnd o
o
oe
oepositiveeff Δ
Γ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ , (3-7)
'cos''
'sin''cos'
''2222 αψψ
dnnn
nnnd o
oe
oenegativeeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ , (3-8)
where )'( ee nn and )'( oo nn are the refractive indices of the e-ray and o-ray, )'(ψψ is the
angle between the optic axis of the director and light propagation vector, )'(αα is the
polar angle of incident light in the medium, and 'd is the thickness of the negative C-
plate. In a given assumption that the transmittance of the dark state should be minimized,
95
the total retardation of the dark state model ( darkΔ ) should be zero at a given light
incident angle,
0=Δ+Δ=Δdark
negativeeffdark
positiveeffdark ndnd . (3-9)
From Eqs. (3-7), (3-8) and (3-9), the thickness of the negative C-plate ( 'd ) of mode 1
can be decided. In the same way, we can also determine the thickness of the positive
A(C)-plate (bottom layer) in mode 2.
3-4-2. Analyses of the effective phase retardation of the simple models
In this section, we are going to calculate the total effective retardation of the
bright state by way of the formula in the previous section in the two viewing planes:
director plane [x-z plane in Fig. 3-9 (c)] and out of the director plane [y-z plane in Fig. 3-
9 (c)]. The extraordinary and ordinary refractive indices of the positive (negative) plates
are 1.6 (1.5) and 1.5 (1.6) at the incident light wavelength of 550 nm, respectively. The
thicknesses of the positive (bottom) and negative (top) plates in mode 1 are 1.852 and
1.736 µm and are 1.852 µm in both plates of mode 2. These values were chosen to meet
the ideal dark and bright state conditions in Sec. 3-2-2 under crossed polarizers. We use
these parameters to calculate the phase retardation and transmittance of our simple
models in this chapter. All calculations are performed at the incident light wavelength of
550 nm.
96
3-4-2-1. Effective phase retardation in the director plane
With the viewing direction in the director plane of our bright state model in mode
1 [Fig. 3-9 (c)], the projection of the optic axes of the positive A-plate and negative C-
plate onto the plane perpendicular to the light propagation vector ( Kr
) are parallel to each
other, as in Fig. 3-11 (a). Consequently, the total effective phase retardation ( ||whiteΔ ) of
the bright state is the summation of both values,
negativeeff
positiveeffwhite ndnd Δ+Δ=Δ|| . (3-10)
Figure 3-12 illustrates how the effective retardation changes, as calculated from Eqs. (3-
7), (3-8) and (3-10), in the director plane of our bright state model in mode 1. During the
calculations, we considered the refraction of the incident light at the air interface. In the
normal direction, the birefringence of the negative C-plate is zero as expected, so it does
not contribute to the total retardation. As the viewing angle increases from the normal
direction, the effective retardation of the positive A-plate and the negative C-plate always
decreases. The total effective retardation of both plates then decreases steeply as the
viewing angle increases in the director plane.
Figure 3-13 is the corresponding figure for mode 2. It is almost the mirror image
of Fig. 3-12. Intuitively, this feature of mirror symmetry is expected from the director
structures in mode 1 and 2, i.e. positive A-plate in mode 1 to negative A-plate in mode 2
and negative C-plate in mode 1 to positive C-plate in mode 2. In this mode, unlike mode
1, the effective retardation of each layer (positive C-plate and negative A-plate) and their
97
total increase as the viewing angles increase from the normal direction, but the sign of the
total effective birefringence is negative. Consequently, the absolute value of the total
effective retardation decreases as the viewing angle increases, and this is the same trend
as in mode 1.
98
Fig. 3-11. Projections of the optic axes of A and C-plates onto the plane perpendicular to
the light propagation vector ( Kr
) that lies (a) in the director plane (x-z plane) and (b) out
of the director plane (y-z plane).
99
-80 -60 -40 -20 0 20 40 60 80-200
-100
0
100
200
300
400
PA NC PA & NC
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-12. Birefringence analyses of the bright state model of mode 1 in the director plane.
(PA: Positive A-plate; NC: Negative C-plate).
100
-80 -60 -40 -20 0 20 40 60 80-400
-300
-200
-100
0
100
200 PC NA PC & NA
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-13. Birefringence analyses of the bright state model of mode 2 in the director
plane. (PC: Positive C-plate; NA: Negative A-plate).
101
3-4-2-2. Effective phase retardation out of the director plane
Unlike the director plane, for viewing angles out of the director plane of our
bright state model in mode 1, the projection of the optic axes of the positive A-plate and
negative C-plate onto the plane perpendicular to the light propagation vector ( Kr
) are
perpendicular to each other, as in Fig. 3-11 (b). Therefore, the total effective retardation
( ⊥Δwhite ) of the bright state in this direction should be the difference between them,
negativeeff
positiveeffwhite ndnd Δ−Δ=Δ⊥ . (3-11)
The retardation variations of the bright state model in mode 1 [Fig. 3-9 (c)] out of the
director plane are calculated from Eqs. (3-7), (3-8) and (3-11) and shown in Fig. 3-14.
The contribution of the negative C-plate is the same as when in the director plane.
However, in this direction, the retardation of the positive A-plate rises continuously as
the viewing angle increases due to the increase in effective thickness. The total effective
retardation of the bright state is the difference between the retardation values of the
positive A-plate and negative C-plate, as mentioned above. Consequently, the value
increases as the viewing angle increases from the normal direction.
Figure 3-15 shows the calculation results for mode 2 by using the same method as
mode 1. This one is also a near-mirror image of the figure in mode 1 [Fig. 3-14]. The
total effective retardation decreases as the viewing angle increases, and it has a negative
sign. Therefore, the absolute value of the total effective retardation always increases as
the viewing angle increases out of the director plane. We saw this same result in mode 1,
102
so we can expect that the optical properties of the bright states are similar between mode
1 and mode 2, even though their director structures are different.
103
-80 -60 -40 -20 0 20 40 60 80-200
-100
0
100
200
300
400 PA NC PA & NC
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-14. Birefringence analyses of the bright state model of mode 1 out of the director
plane. (PA: Positive A-plate; NC: Negative C-plate).
104
-80 -60 -40 -20 0 20 40 60 80-400
-300
-200
-100
0
100
200 PC NA PC & NA
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 3-15. Birefringence analyses of the bright state model of mode 2 out of the director
plane. (PC: Positive C-plate; NA: Negative A-plate).
105
3-4-3. Analyses of the off-axis light transmittance of the simple models
We numerically calculated the off-axis light transmittance of our simple model in
dark and bright states. In these calculations, we inserted the dark and bright state models
[Fig. 3-9 (c)] between two crossed ideal polarizers. The optic axes of the positive A-plate
in mode 1 and negative A-plate in mode 2 made 45° relative to the one of the polarizer
axes. We used the incident light wavelength of 550 nm. All other parameters are the same
as we used when we calculated the effective phase retardation of our models, as in Sec. 3-
4-2.
3-4-3-1. Off-axis light transmittance in a dark state
Figure 3-16 shows the numerical calculation results of the off-axis light
transmission properties of our simple models in their dark states, whose phase
retardations are compensated as described in Sec. 3-3. We added the off-axis light
transmittance of the crossed polarizers for the comparison. These figures show that the
dark state transmittances of the simple models have almost the same level of the crossed
polarizers in main viewing directions. Therefore, we can say that the effective phase
retardations of the dark state-simple models are optically compensated well.
3-4-3-2. Off-axis light transmittance in a bright state
From the analyses of effective phase retardation in Sec. 3-4-2, we know that
although we use an ideal compensator for perfect optical compensation of the dark state,
it adversely changes the retardation of the bright state for all viewing angles. Figure 3-17
106
shows the numerical calculation results of the off-axis light transmission properties of the
bright state model in mode 1 and 2 in the director and out of the director planes. As
expected from the retardation analyses and Eq. (3-1), the transmittance curves of mode 1
and mode 2 are almost the same in the director plane and out of the director plane, even
though their detailed structures are different. Another thing to be noted is that the
transmittance in the director plane decreases as the viewing angle increases from the
normal direction, and it is consistent with the retardation analyses. On the other hand, the
transmittance out of the director plane rises until the total effective retardation reaches the
half wavelength of the incident light and then falls because it has passed the maximum
transmittance point (± 50° in this calculation, and it depends on oT ). The properties of
fallen transmittance in the large viewing angles in both viewing planes are also partially
coming from surface reflection effects.
When we compare the results of our bright state model to the optical properties of
the real common liquid crystal devices in Sec. 3-2, Fig. 3-6, they have very similar curve
shapes for both main viewing-angle-directions. In other words, the transmittance of the
bright state falls as the viewing angle increases in the director plane; while, out of the
director plane, the transmittance rises first and then falls after it reaches the peak point if
the liquid crystal layer is optically designed so that the transmittance of the normal
direction is lower than the maximum value.
107
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Mode 1 Mode 2 Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane (Φ=0º)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Mode 1 Mode 2 Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane (Φ=90º)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03 Mode 1 Mode 2 Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(c) Φ=135º
Fig. 3-16. Numerical calculation results of the off-axis light transmission properties of the
simple models in their dark states. (Most of the curves are overlapped.)
108
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5 Mode 1 Mode 2
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Mode 1 Mode 2
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane (mode 1 and mode 2 are overlapped)
Fig. 3-17. Off-axis light transmission properties of the bright state models.
109
3-5. Comparison of the viewing angle properties in all directions
Until now, we described the optical properties only in the two main viewing
planes: in the director and out of the director planes, but the similarities of the optical
properties between our simple model and the real liquid crystal devices are maintained
over all viewing angles. We numerically calculated the transmittances in all viewing
directions at an incident light wavelength of 550 nm. The calculation methods and optical
parameters are the same that we used in previous sections.
Figures 3-18 and 3-19 show the conoscopic figures of the dark state
transmittances of the real liquid crystal devices and our simple models, respectively.
They have almost the same figure shapes, and this is manifest because we compensated
the phase retardation of their dark states almost perfectly in all viewing angles.
Figures 3-20 and 3-21 illustrate the conoscopic figures of the bright state
transmittances of the real, common liquid crystal devices and our simple models,
respectively. Though the symmetries are a little bit different depending on the detailed
director angle distribution of each display mode, the overall figure shapes are the same
and have the common features, as explained above.
110
(a) Pi-cell (b) ECB
(c) VA (d) TN
(e) Symmetric splay-cell: next page
Fig.3-18. Viewing angle properties of the common liquid crystal devices in their dark
states. (Azimuth angles: 0 ~ 360º; Polar angle: 0 ~ 80º).
111
(e) Symmetry splay cell
112
(a) Mode 1
(b) Mode 2
Fig. 3-19. Viewing angle properties of the dark state models.
(Azimuth angles: 0 ~ 360º; Polar angle: 0 ~ 80º).
113
(a) Pi-cell (b) ECB
(c) VA (d) TN
(e) Symmetric splay-cell: next page
Fig. 3-20. Viewing angle properties of the common liquid crystal devices in their bright
states. (Azimuth angles: 0 ~ 360º; Polar angle: 0 ~ 80º).
114
(e) Symmetry splay cell
115
(a) Mode 1
(b) Mode 2
Fig. 3-21. Viewing angle properties of the bright state models.
(Azimuth angles: 0 ~ 360º; Polar angle: 0 ~ 80º).
116
3-6. Summary
We calculated the off-axis light transmission properties of the bright state of most
common liquid crystal devices, such as ECB, VA, TN, Pi-cell, and symmetric splay-cell,
whose dark states were optically compensated to have minimum transmittances for all
viewing angles. From the results of these calculations, we found that their bright states
have a universal viewing angle shape in spite of completely different director structures
in their liquid crystal layers.
In order to understand this strange phenomenon, we made simple dark and bright
state models describing general liquid crystal devices and analyzed them in terms of
effective retardation and transmittance. In accordance with these analyses, the total
effective retardation in the director plane constantly falls as the viewing angle increases
(“bell shape”). On the contrary, the total effective retardation out of the director plane
consistently rises in the same situation (“reversed bell shape”). These retardation
changes cause the transmittance changes. In the director plane, the transmittance
decreases as the viewing angle becomes larger because the birefringence decreases in that
direction. On the other hand, the transmittance out of the director plane increases first and
then falls after the specific viewing angle if the liquid crystal layer is optically designed
so that the transmittance of the normal direction is lower than the maximum value.
These viewing angle features of our bright state models agree well with the
properties of most common liquid crystal devices, not only in the two main viewing
planes, but also for all viewing directions. Therefore, we can say that our simple model
can reasonably describe the optical properties of the real liquid crystal devices considered
117
here. Accordingly, our simple model can be used to analytically understand and predict
the optical properties, such as transmittance, luminance distribution and color analyses of
current LCDs or possible candidates of new display modes because usually analytical
methods for optical calculations are almost impossible in real devices.
Based on these results, we can say that the single domain LCD modes, considered
here, whose dark states are optically compensated to give minimum transmittance,
inevitably have asymmetric shapes of the off-axis light transmission properties between
the director and out of the director planes in their bright states. Therefore, in order to
achieve isotropic shapes of the bright state viewing angle properties, multi-domain liquid
crystal modes are necessary.
118
CHAPTER 4
OPTIMIZATION OF THE BRIGHT STATE DIRECTOR CONFIGURATION
IN COMPENSATED PI-CELL DEVICES
4-1. Introduction
One of the popular modes in liquid crystal displays (LCDs) is the twisted nematic
mode (TN), and it has been used in many applications because of its simple and low cost
manufacturing process. In spite of those merits, it is hard to use the mode in large size
monitor or TV applications due to its slow response time and limited viewing angle
properties. To overcome these problems, various other LCD modes based on multi-
domain technology such as patterned vertical alignment (PVA), multi-domain vertical
alignment (MVA), in-plane switching (IPS), and advanced super view (ASV) have been
developed, commercialized and shown outstanding improvement recently. However,
their transmittances are sacrificed to improve viewing angle properties, and their response
times are not satisfactory to drive fast moving pictures.
To achieve the desired characteristics of wide viewing angle and fast response,
the optimization of the Pi-cell 37 , 38 mode is currently of high interest. (Pi-cell was
invented in 1983 by Bos. Subsequently, Uchida used it with a single biaxial film as a
compensator with normally black mode and called it OCB39, 40 in 1993. Their structures
of the liquid crystal and polarizers are the same.). The dark state of Pi-cell is especially
very important to improve visibility of the display, so a great amount of work has been
119
done to achieve high contrast ratio without color variation at all viewing angles41, 42, 43, 44,
45, 46. On the other hand, the bright state determines the luminance of the display at all
angles of the incident light. Therefore, it must be another crucial factor when evaluating
the optical properties of a device. However, we have not seen those studies dealing with
the off-axis light transmission properties of the bright states of Pi-cells.
In this chapter, we investigate the off-axis light transmission properties of the
differently director configured bright states in Pi-cells whose dark states are optically
compensated almost perfectly. To make various director configurations of the bright state,
we applied different bright state voltages and used diverse pretilt angles. During the
calculations, we adjusted the thicknesses of the liquid crystal layer and compensators to
achieve the same transmittance of the bright state at normal direction. According to our
results, their bright states have a curious universality in the off-axis light transmission
properties and are insensitive to the parameters if the bright state voltage or pretilt angle
is not so big, and these features agree well with our fist simple bright state model47.
However, in bright state with high tilt angle of director configuration, the off-axis light
transmittance fluctuates as the incident light angle varies, and it makes the viewing angle
properties of a Pi-cell worse. In order to explain these interesting facts, we make a new
dark and bright state model and analyze it in terms of effective birefringence,
transmittance and angular distribution of directors.
In Sec. 4-2, we will show the off-axis light transmission properties of the bright
states in Pi-cells that have various director configurations. In Sec. 4-3, we will build a
dark and bright state model, calculate the angular dependencies of the effective
120
birefringence and transmittance of the model and then analyze them after comparing the
results with those of the Sec. 4-2.
4-2. Optical Properties of Different Bright States in Pi-cells
We numerically calculated the off-axis light transmission properties of Pi-cells at
an incident light wavelength of 550 nm. Their bright states are implemented using
different applied voltages ranging from 1 to 3 V and pretilt angles of 2-30º with all the
same dark voltage, 5 V. During the calculations, we used numerical relaxation techniques
to get the director field in the liquid crystal layer and the 2×2-matrix method 48 to
calculate the optical properties of the devices.
4-2-1. Optical compensations of the dark states of the Pi-cells
The dark state of each mode is optically compensated via two (top and bottom of
the liquid crystal layer) compensators with the hybrid-negative C structure. The hybrid-
negative C structure has exactly the same angular distribution as the directors of liquid
crystal layer of dark state, and their extraordinary, ordinary refractive indices are the
same as the ordinary, extraordinary indices of the liquid crystal, respectively. We used
this same method in chapter 3 for the compensation of the dark state in a LCD.
4-2-2. Optical parameters of the Pi-cells
The thicknesses of the liquid crystal layer and the compensator of each display
mode are determined to meet two conditions. Firstly, it gives the minimum transmittance
121
at all viewing angles in the dark state. Secondly, the transmittance of the bright state at
normal direction has a specific value ( oT ). This value can be chosen from zero to
maximum transmittance that we can achieve under the crossed polarizers, but a bit
smaller than the maximum value is usually used to achieve high transmittance and to
escape color shift at the off axis viewing directions. In this chapter, we set the value to ¾
of the maximum transmittance because this is a similar condition in real liquid crystal
devices and is not a critical factor in determining physical concept of the devices. The
detailed specifications are in Table 4-1 and Table 4-2 for the different bright state
voltages and pretilt angles, respectively. The liquid crystal (LC) we used here is LC53
( en =1.6644, on =1.5070 at λ=550 nm; Δε=9.4), and the cell layout is shown in Fig. 4-1.
Other optical parameters are the same as we used in Sec. 3-2.
122
Table 4-1. Cell parameters of the Pi-cells with different bright state voltages.
(Pretilt angle of 5.5°, dark state voltage of 5.0 V).
Thickness (µm) Bright state voltage (V) LC Compensator 1.0 4.598 2.06 1.3 5.037 2.255 2.0 6.875 3.078 2.5 9.273 4.151 3.0 13.25 5.932
123
Table 4-2. Cell parameters of the Pi-cells with different pretilt angles.
(Bright state voltage of 1.3 V, dark state voltage of 5.0 V).
Thickness (µm) Pretilt angle (°) LC Compensator 2.0 4.745 2.131 5.5 5.037 2.255 10.0 5.474 2.438 20.0 6.769 2.963 30.0 8.785 3.752
124
Fig. 4-1. Stack configuration of a Pi-cell.
ITO Polyimide
Liquid Crystal
Polyimide ITO
Glass
Compensator (PDM)
Polarizer (-45º)
Polarizer (+45º)
Compensator (PDM)
Glass
Light Source
125
4-2-3. Numerical calculations of the off-axis light transmittance in dark states
Figure 4-2 shows the numerical calculation results of the off-axis light
transmission properties of the dark state Pi-cells, whose bright states are optimized to
have several different voltages, but have the same dark state voltage of 5.0 V and pretilt
angle of 5.5º. The phase retardations of their dark states are compensated separately, as
described in Sec. 4-2-1. We added the off-axis light transmittance curves of the just
crossed polarizers for the comparison. We also calculated the director tilt angles of the
dark state-liquid crystal layer, as in Fig. 4-3.
We did similar calculations for the Pi-cells that have different pretilt angles, but
they have the same dark and bright states voltages of 5.0 V and 1.3 V, respectively.
Figures 4-4 and 4-5 illustrate the results of off-axis light transmission properties and the
director configurations of liquid crystal layers in their dark states.
These figures show that the dark state transmittances of the Pi-cells are not far
from the level of the crossed polarizers in main viewing directions, although there is a
relatively large difference at the viewing angle of about ±70º when the bright state
voltages or the pretilt angles are big enough. Actually, these results hint that there might
be a limitation in the phase compensation of a dark state by using the layers of hybrid-
negative C structure. We will deal with this topic in chapter 6.
From these numerical calculation results, we can say that the effective phase retardation
of each dark state-liquid crystal layer of the Pi-cells is optically compensated reasonably.
126
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1.0 V 1.3 V 2.0 V 2.5 V 3.0 V Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1.0 V 1.3 V 2.0 V 2.5 V 3.0 V Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 4-2. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different bright state voltages (1.0-3.0V).
(Dark state voltage: 5.0V; pretilt angle: 5.5º).
127
0.0 0.2 0.4 0.6 0.8 1.0
0
50
100
150
200
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 4-3. Director tilt angles of the dark state Pi-cells that have different bright state
voltages (1.0-3.0V). (Dark state voltage: 5.0V; pretilt angle: 5.5º).
128
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
2.0o 5.5o
10.0o 20.0o
30.0o Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
2.0o 5.5o
10.0o 20.0o
30.0o Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 4-4. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different pretilt angles (2.0-30.0º).
(Bright state voltage: 1.3 V; dark state voltage: 5.0 V).
129
0.0 0.2 0.4 0.6 0.8 1.0
0
50
100
150
200 2.0o 5.5o
10.0o 20.0o
30.0o
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 4-5. Director tilt angles of the dark state Pi-cells that have different pretilt angles
(2.0-30.0º). (Bright state voltage: 1.3 V; dark state voltage: 5.0V).
130
4-2-4. Numerical calculations of the off-axis light transmittance in bright states
We numerically calculated the off-axis light transmission properties of the bright
state Pi-cells that have several different bright state voltages (1.0-3.0 V), but they have
the same dark state voltage of 5.0 V and pretilt angle of 5.5º. Figures 4-6 and 4-7 show
the results of the transmittances and the director tilt angles of liquid crystal layer
calculated in the same conditions, and Fig. 4-8 shows the conoscopic figures of the bright
state transmittances of the Pi-cells.
We also did similar calculations for the Pi-cells that have different pretilt angles
(2-30º), but the same dark and bright states voltages of 5.0 V and 1.3 V, respectively.
Figures 4-9 and 4-10 depict the results of off-axis light transmission properties in two
main directions and the director configurations of the liquid crystal layers in their bright
states, respectively. Figure 4-11 shows the conoscopic figures of the bright state
transmittances of the Pi-cells, which are calculated in the same conditions.
From these transmittance figures, we can notice that the off-axis light
transmission properties of the bright state in a compensated Pi-cell, amazingly, have
unified shape and are relatively less sensitive to the parameters if the bright state voltage
or pretilt angle is below certain value (about 2.0 V and 20º in these figures). In the
viewing angle of the director plane [Fig. 4-6 (a) and Fig. 4-9 (a)], all the transmittance
curves have similar “bell” shapes i.e. the transmittance constantly decreases as the
viewing angle increases from the normal direction no matter what director configurations
(bright state voltage or pretilt angle) of liquid crystal are used as a bright state. The
transmittance out of the director plane [Fig. 4-6 (b) and Fig. 4-9 (b)] rises as the viewing
131
angle increases from the normal direction and then falls monotonically after passing the
specific angles (about ± 50° in these calculations) if the Pi-cell has the director
configuration of low tilt angle (low bright state voltage or low pretilt angle). These
features of universality and insensitiveness of transmittance to the director configuration,
exactly, comply with our first simple bright state model, as described in chapter 3 and
give room to the optical design of a Pi-cell.
However, in the high tilted director configuration of the bright state out of the
director plane, the transmittance is a much strong function of incident light angle, and it
leads to narrow viewing angle properties. Therefore, it does not follow our first simple
bright state model anymore. Additionally, these results mean that the more the director
configuration of the bright state becomes similar to that of dark state, the worse the off-
axis light transmission property of the bright state when its dark state is optically
optimized. We analyzed these interesting facts by building another more general bright
state model that can not only include our first simple model but also stand for all director
configurations of bright states in Pi-cells, and the results are in next section.
132
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5 1.0 V 1.3 V 2.0 V 2.5 V 3.0 V
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
1.0 V 1.3 V 2.0 V 2.5 V 3.0 V
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 4-6. Numerical calculation results of the off-axis light transmission properties of the
bright state Pi-cells that have different bright state voltages (1.0-3.0V).
(Dark state voltage: 5.0V; pretilt angle: 5.5º).
133
0.0 0.2 0.4 0.6 0.8 1.00
50
100
150
200 1.0 V 1.3 V 2.0 V 2.5 V 3.0 V
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 4-7. Director tilt angles of the bright state Pi-cells that have different bright state
voltages (1.0-3.0V). (Dark state voltage: 5.0V; pretilt angle: 5.5º).
134
Fig. 4-8. Conoscopic properties of the bright state transmittances with different bright
state voltages of (a) 1.0V, (b) 1.3V, (c) 2.0V, (d) 2.5V and (e) 3.0V. (Dark state voltage:
5.0V; pretilt angle: 5.5º).
135
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5 2.0o 5.5o 10.0o
20.0o 30.0o
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
2.0o 5.5o
10.0o 20.0o
30.0o
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of director plane
Fig. 4-9. Numerical calculation results of the off-axis light transmission properties of the
bright state Pi-cells that have different pretilt angles (2.0-30.0º).
(Bright state voltage: 1.3 V; dark state voltage: 5.0V).
136
0.0 0.2 0.4 0.6 0.8 1.0
0
50
100
150
200
2.0o 5.5o
10.0o 20.0o
30.0o
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 4-10. Director tilt angles of the bright state Pi-cells that have different pretilt angles
(2.0-30.0º). (Bright state voltage: 1.3 V; dark state voltage: 5.0V).
137
Fig. 4-11. Conoscopic properties of the bright state transmittances with different pretilt
angles of (a) 2.0º, (b) 5.5º, (c) 10.0º, (d) 20.0º and (e) 30.0º. (Bright state voltage: 1.3 V;
dark state voltage: 5.0V).
138
4-3. Universal Bright State Model of Pi-cells
4-3-1. Dark and bright state modeling
Let us consider a Pi-cell whose dark state is optically optimized to give the
minimum transmittance at the normal direction by means of passive type phase
compensators, hybrid-negative C-plates. The angle distribution of the optic axes of the
negative C plates should be the same as that of the liquid crystal directors for the best
compensation of the phase difference between ordinary ray (o-ray) and extraordinary ray
(e-ray) in the liquid crystal layer. Figure 4-12 (a) shows the cartoon of this director
configuration of the device. In this figure, we can notice that there are simply two parts in
a liquid crystal layer: static part (empty director shape in the figure) and dynamic part
(filled director shape). In the static part, most of the liquid crystal directors hardly change
their orientations, or the changes are very small if they do between the dark and bright
state because of the anchoring energy at the liquid crystal–alignment layer interfaces. On
the contrary, the directors in the dynamic part are relatively sensitive to the applied
voltage, so the orientations of the directors are completely different between the dark and
bright state. In the aspect of the total system composed of liquid crystal and passive type
compensators, the static part of the liquid crystal layer is optically neutralized by the
corresponding part of the compensators (empty director shape in the figure) in both dark
and bright state, simultaneously, and it does not contribute to the optical properties of the
device. Therefore, we can simplify the compensated device, conceptually, and get a
simple dark and bright state model, which made up of two symmetrically tilted positive
139
A-plates (PA1, 2. In this chapter, an O-plate is referred to as a tilted A-plate, as is
common usage) and two same normal negative C-plates (NC1, 2), as Fig. 4-12 (b). In this
simple model, the dark state is still optically neutralized, but the bright state has net
birefringence in any viewing directions. Intuitively, we can also see that the tilt angle (θ )
of a positive A-plate in the bright state model is related to the director configuration of
the bright state in real liquid crystal device, e.g. as the voltage or pretilt angle of bright
state in Pi-cell increases, the tilt angle (θ ) of the A-plate also increases.
140
Fig. 4-12. (a) A Pi-cell whose dark state is perfectly compensated using hybrid-negative
C-plates and (b) simple dark and bright state models of the Pi-cell.
141
4-3-2. Calculations
The light transmittance (T ) of the simple model [Fig. 4-12 (b)] under the ideal
crossed polarizers can be written as follows:
2sin
2sin
21 2
max2 Γ
≡Γ
= TTTT paap , (4-1)
where apT and paT are the transmittances with considering the surface reflections at the
interfaces between air and polarizers in the incident and exit media, respectively. The
Γ in Eq. (4-1) is the phase difference between e-ray and o-ray in the anisotropic layers
and written in terms of the incident light wavelength (λ ) and effective birefringence
( effndΔ ) of the media when we assume the e-ray and o-ray have the same path in the
media,
effndΔ=Γλπ2 . (4-2)
[The exact expression of the phase difference (Γ ) is dKK ozez )( −=Γ , where ezK and
ozK are the z components of the wave vectors of the e-ray and o-ray, respectively. For
simplicity in this chapter, we assume the e-ray and o-ray have the same path, and this is
very reasonable in a real situation. We will compare both methods in Chap. 6.] The
142
incident light angle dependence of the effective birefringence ( effndΔ ) in each tilted
positive A-plate and negative C-plate is calculated in Eqs. (4-3) and (4-4), respectively,
αψψ cossincos 2222
dnnn
nnnd o
oe
oepositiveeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ , (4-3)
'cos''
'sin''cos'
''2222 αψψ
dnnn
nnnd o
oe
oenegativeeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ , (4-4)
where )'( ee nn and )'( oo nn are the refractive indices of e-ray and o-ray, )'(ψψ is the
angle between the optic axis of the director and light propagation vector, )'(αα is the
polar angle of incident light in the medium, and )'(dd is the thickness of a tilted positive
A-plate (negative C-plate). Where, d is determined so that oTT = in normal direction. In
a given assumption that the transmittance of the dark state should be minimized, the total
birefringence of the dark state model ( darkΔ ) should be zero at a given light incident
angle,
022 =Δ×+Δ×=Δdark
negativeeffdark
positiveeffdark ndnd . (4-5)
From Eqs. (4-3), (4-4) and (4-5), the thickness of a negative C-plate ( 'd ) can be decided.
143
4-3-3. Analyses of the effective birefringence in the director plane
We cannot define the birefringence out of the director plane [x-z plane in Fig. 4-
12 (b)] in the bright state model or in a real Pi-cell, because the projection of the directors
onto a plane perpendicular to the light propagation vector ( Kr
) are not co-linear.
Therefore, in this section, we are going to analyze the total effective birefringence of the
bright state model only in the director plane [y-z plane in Fig. 4-12 (b)]. All calculations
are performed at an incident light wavelength of 550 nm. During the calculations, we
considered the refraction of the incident light at the air interface.
The extraordinary and ordinary refractive indices of the positive (negative) plates
are 1.6 (1.5) and 1.5 (1.6), respectively. The thickness of each positive (PA1, 2) and
negative (NC1, 2) plate for the various tilt angles of the positive A-plate are on Table 4-3.
These values were taken to meet the ideal dark and the bright state conditions under
crossed polarizers, as described in Sec. 4-2-2. We use these parameters in all calculations
of our model in this chapter.
In the viewing direction on the director plane of our bright state model, the
projections of the optic axes of the tilted positive A-plates (PA1, 2) and negative C-plates
(NC1, 2) onto the plane perpendicular to the light propagation vector ( Kr
) are parallel to
each other [Fig. 3-11 (a)]. Therefore, the total effective birefringence of the bright state
model in the director plane ( ||brightΔ ) is the summation of each value,
2121|| NCeff
NCeff
PAeff
PAeffbright ndndndnd Δ+Δ+Δ+Δ=Δ . (4-6)
144
Figure 4-13 shows the effective birefringence of each layer calculated from Eqs. (4-3),
(4-4) and (4-6), in the director plane of our bright state model with a tilt angle (θ) of 50º.
In the normal direction, the birefringence of the negative C-plates (NC1, 2) is zero as
expected, so it does not contribute to the total birefringence. As the viewing angle
increases from the normal direction, the effective birefringence of the negative C-plates
always decreases because of its molecular structure. Meanwhile, the effective
birefringence of the tilted positive A-plates (PA1, 2) depends on the tilt angles (θ) and
symmetrically varies each other.
Figure 4-14 illustrates the summation of effective birefringence contributed from
the positive A-plates (PA1+PA2), which have tilt angles of 10-70º. The curve shape of
this summation depends on the tilt angle (θ) of the positive A-plate as the figure. As we
can see in this figure, if the tilt angle is small, the subtotal effective birefringence
(PA1+PA2) decreases as the viewing angle increases. On the contrary, if the tilt angle is
relatively large, the subtotal value increases as the viewing angle increases from the
normal direction.
Figure 4-15 is the total effective birefringence of the bright state model, which is
composed of the tilted positive A-plates (PA1, 2) and negative C-plates (NC1, 2) with
various tilt angles ranging from 10º to 70º. Amazingly, the curves are all “bell” shapes
and their detailed values are very similar to each other no matter what tilt angles are used
in the bright state model. This universality of the birefringence in the director plane
agrees well with our first simple bright state model.
145
Table 4-3. Thickness of each layer of the bright state model with different tilt angles (θ).
Thickness (µm) Tilt angle (°) PA1,2 NC1,2
10.0 0.958 0.898 30.0 1.267 1.188 50.0 2.376 2.228 65.0 5.618 5.267 70.0 8.627 8.088
146
-80 -60 -40 -20 0 20 40 60 80-300
-200
-100
0
100
200
300
400
[PA1+PA2]PA2
Total
PA1
NC1,2[NC1+NC2]
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 4-13. Effective birefringence of each layer of the bright state model in the director
plane with a tilt angle (θ) of 50º.
147
-80 -60 -40 -20 0 20 40 60 800
200
400
600
800
1000
1200
10o 30o 50o
65o 70o
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 4-14. Effective birefringence of the positive A-plates (PA1+PA2) of the bright state
model in the director plane. [tilt angle (θ): 10, 30, 50, 65, and 70º]
148
-80 -60 -40 -20 0 20 40 60 800
50
100
150
200
250 10o 30o 50o
65o 70o
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 4-15. Effective birefringence of the total layers of the bright state model in the
director plane. [tilt angle (θ): 10, 30, 50, 65, and 70º]
149
4-3-4. Analyses of the transmittance
Figure 4-16 is the numerical calculation results showing the off-axis light
transmission properties of the bright state model with various tilt angles (10-70º) in the
director and out of the director planes, and Fig. 4-17 shows their conoscopic properties
under the same situation. We also calculated the off-axis light transmission properties of
the dark state model to make clear that we compensated the phase retardation of the dark
state completely, and the results are in Fig. 4-18. In these calculations, we inserted the
bright state models [Fig. 4-12 (b)] between two crossed polarizers, and the azimuth angle
of an optic axis of a tilted positive A-plate made 45° angle relative to the one of the
polarizer axes, as usual liquid crystal devices. We used a light wavelength of 550 nm
during the calculations.
In the director plane [Fig. 4-16 (a)], the light transmittance curves are almost the
same regardless of the tilt angles of positive A-plate in the bright state model, and their
values decrease as the viewing angle increases from the normal direction. These results
are consistent with the birefringence analyses and are the same shapes as the
transmittance curves of real Pi-cells [Fig. 4-6 (a) and Fig. 4-9 (a)]. This is the universal
viewing angle properties of liquid crystal devices whose dark states are optically
optimized.
On the other hand, the light transmittance out of the director plane [Fig. 4-16 (b)]
looks somewhat complicated, but their shapes are very similar to those of the real Pi-cells
[Fig. 4-6 (b) and Fig. 4-9 (b)]. Therefore, we can say our new bright state model can
describe the basic properties of bright states of real Pi-cells well. With a low tilt angle of
150
the positive A-plate in the bright state model, the transmittance rises until the total
effective retardation reaches a half-wavelength of the incident light and then falls because
it has passed the maximum transmittance point (± 50° in this calculation and it depends
on oT ). These features with a low tilt angle (low bright state voltage or low pretilt angle)
are the same as the properties of our first simple bright state model, as in Chap. 3, and
can be understandable because, in this case, the projections of the directors onto a plane
perpendicular to the light propagation vector ( Kr
) are roughly co-linear. In a different
way, the transmittance of the bright state with a high tilt angle fluctuates as the viewing
angle varies, and it makes viewing angle properties worse. In the hope of understanding
the detailed reasons for the fluctuation, we analyze the transmittance out of the director
plane in next section.
151
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
10o 30o 50o
65o 70o
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
10o 30o
50o 65o
70o
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 4-16. Off-axis light transmission properties of the bright state model with different
tilt angles.
152
Fig. 4-17. Conoscopic properties of the transmittance of the bright state model with tilt
angle (a) 10º, (b) 30º, (c) 50º, (d) 65º, and (e) 70º.
153
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
10o 30o
50o 65o
70o Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
10o 30o
50o 65o
70o Crossed polarizers
Viewing angle (deg.)
Tran
smitt
ance
(b) Out of director plane
Fig. 4-18. Off-axis light transmission properties of the dark state model. All curves are
overlapped each other, and their values are the same level of the crossed polarizers.
154
4-3-5. Detailed analyses of the transmittance out of the director plane
We calculated two more transmittance curves contributed separately from each
layer (PA: positive A-plates only and NC: negative C-plates only) of the bright state
model (tilt angle of 70º) out of the director plane, as in Fig. 4-19. Roughly, we can say
from this figure that the fluctuation of the total transmittance is mainly coming from the
effects of each separate layer. Therefore, we are going to analyze the fluctuations of each
separate transmittance to understand the total transmittance.
4-3-5-1. Contribution from the positive A-plates
Firstly, we want to know what causes the fluctuation of the transmittance
contributed from positive A-plates only. Figure 4-20 shows the apparent azimuth angle
(γ ) of the optic axis of each layer with respect to the direction of P-polarization ( Pr
) of
the incident light. In this calculation, the bright state model has a tilt angle of 70º, and the
light incident direction is out of the director plane. The detailed angle definitions are in
Fig. 4-21, and the angles, γ and ψ are calculated from Eqs.49 (4-7) and (4-8),
θαφθαφθγ
sinsincoscoscossincostan−
= , (4-7)
θαφθαψ sincoscoscossincos += , (4-8)
where θ and φ are the tilt and azimuth angles of a director, α is the light incident angle
in a medium, γ is the angle between Pr
and er , and ψ is the angle from the light
155
propagation vector ( Kr
) to the director ( nr ). From Fig. 4-20, we can see that as the
viewing angle increases from the normal direction, the azimuth angle difference ( γΔ )
between PA1 and PA2 in incident light frame ( KSPrrr
,, ) varies from zero to bigger than
90º. The maximum angle difference it can reach depends on the magnitude of the tilt
angle (θ ) of the positive A-plates, as in Fig. 4-22. If the angle difference ( γΔ ) is 90º, the
total effective birefringence of the positive A-plates is canceled regardless of their
magnitudes of birefringence, so they do not contribute to the transmittance. This is the
reason why the transmittance contributed from the A-plates is almost zero at the viewing
angle around 33º (1st minimum point of PA in Fig. 4-19). After this first minimum point,
the transmittance increases and then falls again. This feature can be understood from Fig.
4-23 that shows the effective birefringence of a positive A-plate (PA1 or PA2) out of the
director plane in the bright state model, which is calculated from Eqs. (4-3) and (4-8).
The magnitudes of the birefringence of the PA1 and PA2 are the same at all viewing
angles out of the director plane because of their symmetry. However, their main
directions are different, so they can not be added. As we can see in this figure, the
birefringence of the A-plate (PA1 or PA2) whose tilt angle (θ) is 70º goes to one
wavelength of incident light ( λ =550 nm) as the viewing angle increases to 80º.
Therefore, the transmittance contributed from the A-plates (PA1, PA2) goes to zero due
to Eqs. (4-1) and (4-2), and it makes second minimum point at the viewing angle around
80º, as in Fig. 4-19.
156
4-3-5-2. Contribution from the negative C-plates
Secondly, we are going to analyze the fluctuation of the transmittance coming
from the negative C-plates (NC in Fig. 4-19) of the bright state model [Fig. 4-12 (b)] with
70º tilt angle of positive A-plates. In contrast to the positive A-plates, the optic axes of
the negative C-plates are in the same direction, so the total effective birefringence of
them is summation of each magnitude. Additionally, the angles between their optic axes
and transmission axes of the polarizers remain almost ±45º at all off-axes out of the
director plane, as in Fig. 4-20. It means only the effective birefringence of the negative C-
plates affects the transmittance of the bright state model. Figure 4-24 shows the subtotal
effective birefringence of the negative C-plates (NC1 and NC2) out of the director plane
with light wavelength (λ ) of 550 nm. In this figure, with tilt angle 70º of A-plates, the
absolute value of effective birefringence increases as viewing angle increases, and it
passes through the first maximum point ( 2λ=Δ effnd around 35º of viewing angle), the
first minimum point ( λ=Δ effnd around 55º) and the second maximum point
( 23λ=Δ effnd around 70º) of transmittance, which are calculated from Eqs. (4-1) and (4-
2). This variation of the birefringence causes the fluctuation of the transmittance
contributed from the negative C-plates (NC in Fig. 4-19). In the viewing angles out of the
director plane, we can say conclusively that these features of each separate transmittance
make the total transmittance fluctuate in large tilt angles of positive A-plates in our bright
state model (or director configuration of bright state with high voltage or high pretilt
angle in a real Pi-cell).
157
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5 PA NC Total
Viewing angle (deg.)
Tran
smitt
ance
Fig. 4-19. Off-axis light transmittances contributed separately from each layer of the
bright state model (tilt angle of 70º) out of the director plane.
(PA: positive A-plates only and NC: negative C-plates only).
158
-80 -60 -40 -20 0 20 40 60 80-50
0
50
100
150
200
PA1
Polarizer
Analyzer
PA2
NC1,2
Viewing angle (deg.)
γ (d
eg.)
Fig. 4-20. Apparent azimuth angle (γ ) of the optic axis of each layer in the bright state
model with tilt angle 70º out of the director plane.
159
Fig. 4-21. Director angle definition in (a) lab coordinate system ( ',',' zyx ) and (b)
incident light frame ( KSPrrr
,, ).
160
-80 -60 -40 -20 0 20 40 60 800
50
100
150
70o
65o
50o
30o
10o
Viewing angle (deg.)
Δγ (d
eg.)
Fig. 4-22. Difference of the apparent azimuth angle ( γΔ ) between the PA1 and the PA2
in bright state model with various tilt angles (10º-70º).
161
-80 -60 -40 -20 0 20 40 60 800
100
200
300
400
500
600
10o 30o 50o
65o 70o
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 4-23. Effective birefringence of a positive A-plate (PA1 or PA2) out of the director
plane in the bright state model. [Tilt angle of A-plates (θ): 10º~70º; Light wavelength (λ):
550 nm].
162
-80 -60 -40 -20 0 20 40 60 80-1200
-1000
-800
-600
-400
-200
0
200
10o 30o
50o 65o
70o
Viewing angle (deg.)
Effe
ctiv
e Δn
d (n
m)
Fig. 4-24. Subtotal effective birefringence of the negative C-plates (NC1+NC2) out of the
director plane in the bright state model. [Tilt angle of A-plates (θ): 10º~70º; Light
wavelength (λ): 550 nm].
163
4-4. Summary
We calculated the off-axis light transmission properties of the differently director
configured bright states in Pi-cells whose dark states were optically compensated to have
minimum transmittances at the normal direction. From the results of these calculations,
we found that the off-axis light transmission properties of the bright states surprisingly
had unified shapes and were relatively insensitive to the variation of the parameters
considered when the voltage of the bright state or the pretilt angle were below some
particular value. On the other hand, the transmittance in highly tilted director
configuration of bright state out of the director plane varies more significantly and makes
the viewing angle properties of a Pi-cell worse.
In order to understand these curious phenomena, we made a new dark and bright
state model describing Pi-cells and analyzed it in terms of effective birefringence,
transmittance and angular distribution of directors. According to the analyses, the total
effective birefringence in the director plane constantly falls as the viewing angle
increases (“bell shape”) regardless of the director configuration, and this result is
consistent with transmittance analyses and describes the optical properties of real Pi-cells
well.
Out of the director plane, we cannot define the total birefringence because the
projection of the directors on the plane perpendicular to the light propagation vector ( Kr
)
are not co-linear, so we analyzed transmittance in this viewing plane. In the director
configuration of low tilt angle in bright state model, the transmittance rises first and then
falls after the specific viewing angle if the liquid crystal layer is optically designed so that
164
the transmittance of the normal direction is lower than the maximum value. On the other
hand, the transmittance with a high tilt director configuration fluctuates as the viewing
angle varies as in real Pi-cells. To understand the causes of the fluctuation of the total
transmittance, we calculated the transmittances, which are contributed separately from
positive A-plates and negative C-plates of the bright state model, and analyzed them
using the apparent angle distribution of directors in incident light coordinate system and
the birefringence of each layer. In accordance with the analyses, both the apparent angle
distribution of directors and the birefringence of each layer play a key role in the
fluctuation of the total transmittance.
165
CHAPTER 5
LUMINANCE AND COLOR PROPERTIES OF THE COMPENSATED LIQUID
CRYSTAL DEVICES IN THEIR BRIGHT STATES
5-1. Introduction
In this chapter, we investigate the luminance and color properties of the bright
state simple model and the common LCDs such as ECB, VA, Pi-cell, and TN modes. We
use the same bright state simple model that we built in Chapter 3. The dark states of the
model and the common LCDs are optically compensated almost perfectly by using hybrid
aligned-passive type negative C-plates, as the same way that we used in Chapter 3.
During the calculation, we use the numerical relaxation technique to get the director field
of the liquid crystal layer and the 2×2-matrix method to calculate the light transmittances
for each wavelength, ranging from 380 to 780 nm. After calculating the transmittance at
each wavelength, we calculate and analyze the luminous transmittances and the color
properties of the devices.
In Sec. 5-2, we give the basic calculations for the luminance and color analyses of
the LCDs. Based on these calculations, we analyze the luminous transmittance of our
bright state model in functions of the cell thickness and the viewing angle, and after that
we compare those results with that of real common LCDs in Sec. 5-3. In Sec. 5-4, we
calculate and analyze the off-axis color properties including color difference of the bright
state model and the common LCDs.
166
5-2. Color Calculations
In this dissertation, we will use the CIE (International Commission on
Illumination or Commission International de l’Éclairage) standard colorimetric system to
describe the color properties of LCDs because it is commonly used in industrial
companies and many other quantitative applications.
5-2-1. Tristimulus values and color matching functions
The tristimulus values ( X , Y , Z ) are kinds of roots for color calculations.
According to the CIE, the values are defined as follows50:
∫= λλλλ dxTSkX )()()( , (5-1)
∫= λλλλ dyTSkY )()()( , (5-2)
∫= λλλλ dzTSkZ )()()( , (5-3)
where λ is the light wavelength, )(λT is the transmission spectrum of a LCD, )(λS is
the light source spectrum, and k is the proportionality associated with choice of units.
For example, if we use 683 lumens per watt, known as the maximum luminous
efficiency 51 , as a k value, then Y is a luminance having a unit, 2/ mcd . Another
conventional way is using k as a normalization constant defined as
167
∫=
λλλ dySk
)()(100 . (5-4)
This definition of k means the Y tristimulus value is 100 for a perfect transmission
system in all light wavelengths, i.e. 1)( =λT or for a light source. In this case, the
tristimulus value Y is called the luminance factor or the luminous transmittance for a
transmission light and has no unit. We will use this latter definition in this dissertation. In
Eqs. (5-1), (5-2), and (5-3), the integration is taken in the visible wavelength region
(about 380-780 nm), and the x , y , and z are the color matching functions. There are
two types of the color matching functions depending on fields of view: 2º and 10º. The
former was introduced in 1931 and recommended by the CIE when the viewing angle is
1-4º, and the latter was introduced in 1964 and recommended for a viewing angle
exceeding 4º. In this dissertation, we will use 2º fields of view, and the Fig. 5-1 shows the
corresponding color matching functions.
168
380 480 580 680 7800.0
0.5
1.0
1.5
2.0
x y z
Wavelength (nm)
Tris
timul
us v
alue
s
Fig. 5-1. CIE 1931 2º color matching functions
169
5-2-2. Chromaticity diagrams
There are several chromaticity diagrams to indicate a color on a two-dimensional
plane, such as xy , uv , and ''vu diagrams. The xy chromaticity diagram (Fig. 5-2) was
firstly introduced in 1931 by the CIE and has been used until now, but the uniformity of
the diagram is not good. In order to improve the uniformity, the uv diagram was
proposed by MacAdam and recommended by the CIE in 1960. In 1976, the CIE modified
the uv chromaticity diagram according to Eastwood’s report, and recommended a new
''vu chromaticity diagram 52 (Fig. 5-3). In this dissertation, we use ''vu chromaticity
diagram because it has the best uniformity so far and widely used. These all chromaticity
diagrams are expressed in terms of the tristimulus values ( X , Y , Z ) as follows:
)/( ZYXXx ++= , (5-5)
)/( ZYXYy ++= , (5-6)
)315/(4 ZYXXu ++= , (5-7)
)315/(6 ZYXYv ++= , (5-8)
)315/(4' ZYXXu ++= , (5-9)
)315/(9' ZYXYv ++= . (5-10)
170
Fig. 5-2. CIE 1931 2º chromaticity diagram (2º viewing angle).
171
Fig. 5-3. CIE 1976 ''vu chromaticity diagram (2º viewing angle).
172
5-2-3. Color difference
In 1976, the CIE recommended CIE 1976 *** vuL Color Space (abbreviated
CIELUV) for applications such as TVs. That is expressed by the following three-
dimensional orthogonal coordinates, *L (lightness), *u (redness-greenness), and *v
(yellowness-blueness) as follows:
16)/(116* −= nYYfL , (5-11)
)''(*13* nuuLu −= , (5-12)
)''(*13* nvvLv −= , (5-13)
where
3/1)/()/( nn YYYYf = if 3)116/24(/ >nYY , (5-14)
116/16)/)(108/841()/( += nn YYYYf if 3)116/24(/ ≤nYY , (5-15)
where Y , 'u , and 'v represent the tristimulus value Y of the object under examination
and the chromaticity coordinates obtained according to Eqs. (5-9) and (5-10). nY , 'nu ,
and 'nv represent the tristimulus value Y and the chromaticity coordinates 'u and 'v of a
suitable reference under the same illuminant. The values are normalized so that 100=nY .
For investigating the angular dependence of the color in LCDs, 'u , 'v , Y are the ( 'u , 'v )
coordinates and luminous transmittance of the pixel under examination, and 'nu , 'nv , nY
173
are the ( 'u , 'v ) coordinates and luminous transmittance of the backlight at normal
direction53. In this dissertation, we use the backlight power distribution as a reference
when we calculated the 'nu , 'nv , and nY .
In order to quantitatively compare the color difference between a pair of samples or two
interesting points, ( *1L , *1u , *1v ), ( *2L , *2u , *2v ), the color difference ( uvE *Δ ),
defined in the CIE 1976 *** vuL Color Space, is usually used, and the expression is
212
212
212 *)*(*)*(*)*(* vvuuLLE uv −+−+−=Δ . (5-16)
For analyzing the angular dependence of the color in LCDs, *2u , *2v , 2Y are the ( *u ,
*v ) chromaticity coordinates and luminous transmittance of the pixel under examination
and *1u , *1v , 1Y are the ( *u , *v ) chromaticity coordinates and luminous transmittance
of the same pixel at the normal direction.
5-2-4. Illuminants
A number of spectral power distributions have been defined by the CIE for use in
describing color, such as illuminant A , equivalent to a blackbody radiator with a color
temperature of 2856 K , illuminant C , and 65D . Natural daylight is defined by the D
illuminants. In this dissertation, we use 65D , CIE daylight with a correlated color
temperature of 6500 K . Figure 5-4 shows the spectral power distribution of those CIE
174
illuminants, and Table 5-1 shows the tristimulus values and chromaticity coordinates of
the illuminants.
175
380 480 580 680 7800.0
0.5
1.0
1.5
2.0
A C D65
Wavelength (nm)
Rel
ativ
e po
wer
dis
tribu
tion
Fig. 5-4. Spectral power distributions of CIE illuminants.
176
Table 5-1. Tristimulus values and chromaticity coordinates of illuminants in fields of
view 2º.
Tristimulus values Chromaticity coordinates Illuminants X Y Z x y 'u 'v
A 109.850 100.0 35.585 0.4476 0.4074 0.2560 0.5243 C 98.07 100.0 118.23 0.3101 0.3163 0.2009 0.4610 65D 95.047 100.0 108.883 0.3127 0.3291 0.1978 0.4684
177
5-3. Analyses of the luminous transmittance (Y )
Figure 5-5 shows the simple dark and bright state models (mode 1) that we built
in Chapter 3. In this chapter, we use the same model for analysis of the luminous
transmittance (Y ) and color properties of LCDs. The dark states of the model and LCDs
were compensated almost perfectly by using passive type negative C-plates, as the same
way that we used in Chapter 3. During the calculation, we used the numerical relaxation
technique to get the director field of the liquid crystal layer and the 2×2-matrix method to
calculate the light transmittances for each wavelength, ranging from 380 to 780 nm. After
calculation of the transmittance at each wavelength, we calculated the luminous
transmittance (Y ) of the devices according to the equations in Sec. 5-2.
5-3-1. Thickness effects on luminous transmittance and cell parameters
We calculated the luminous transmittance (Y ) of the bright state model at the
normal direction in function of the thickness of the positive A-plate, and the result is in
Fig. 5-6. In these calculations, the bright state model is located between ideal crossed
polarizers. Figure 5-6 (b) shows the normalized luminous transmittance ( Y %) with
respect to the maximum value of Y . We also did the same calculations for the common
real liquid crystal devices such as ECB, VA, Pi-cell, and TN modes. During the
calculations, the dark states of each different thickness are compensated separately for
each thickness. The detailed cell specifications, thickness information and the stack
configuration are in Table 5-2, Table 5-3, and Fig. 5-7, respectively. Where, the liquid
crystal we used is LC53 whose refractive indices and elastic constants are shown in Fig.
178
2-7 and Sec. 2-5, respectively. We assumed the polarizers are ideal, i.e. the light that
polarized along the absorption axis of a polarizer is absorbed completely, but the light
that is polarized along the transmission axis of a polarizer is transmitted perfectly except
surface reflections. The refractive indices of the glass, indium tin oxide (ITO), and
polyimide are shown in Fig. 2-27, and their thicknesses are 0.7 mm, 400 Å, and 500 Å.
We want to make clear that the specific numbers are only for the purposes of producing a
graph that demonstrates the features of the general concepts considered here.
179
Fig. 5-5. Simple dark and bright state models (mode 1)
Bright Dark
Positive A-plate
Negative C-plate
180
0.0 0.5 1.0 1.5 2.00
10
20
30
40
50
Thickness (μm)
Y
(a) Raw data
0.0 0.5 1.0 1.5 2.00
20
40
60
80
100
120
Thickness (μm)
Y (%
) of M
ax. Y
(b) Normalized data
Fig. 5-6. Thickness effects on the luminous transmittance of the bright state model at the
normal direction.
181
Table 5-2. Cell parameters of the common liquid crystal (LC) devices we used.
(LC: LC53; Δε=-9.4 for VA, +9.4 for other devices).
Easy axis (°)
Voltage (V)
Pretilt angle (°) Device
modes Top Bottom Bright Dark LC Pi-cell 90 90 1.3 5.0 5.5 ECB 90 -90 0.0 5.0 5.5 VA 90 -90 5.0 0.0 84.5 TN 45 -45 0.0 5.0 5.5
Table 5-3. Thicknesses of the bright state model and the common LCDs.
[PA: positive A-plate; NC: negative C-plates (compensator); LC: liquid crystal]
Model ECB VA Pi-cell TN Y (%) PA NC LC NC LC NC LC NC LC NC 100 1.742 0.780 2.055 0.919 2.1 0.780 7.274 3.255 3.187 1.28190 1.381 0.618 1.628 0.728 1.665 0.619 5.766 2.581 2.393 0.96280 1.222 0.547 1.442 0.645 1.474 0.548 5.106 2.285 2.089 0.83970 1.093 0.489 1.29 0.577 1.318 0.490 4.567 2.044 1.851 0.744
182
Fig. 5-7. Stack configuration of the common liquid crystal devices.
ITO Polyimide
Liquid Crystal
Polyimide ITO
Glass
Compensator (PDM)
Polarizer (+45º)
Polarizer (-45º)
Compensator (PDM)
Glass
Illuminant ( 65D )
183
5-3-2. Off-axis luminous transmittance
We numerically calculated the off-axis luminous transmittance (Y ) of the bright
state model and the common LCDs (ECB, VA, Pi-cell, and TN) for different cell
thicknesses, which have four different normalized Y (%) values at the normal direction:
100, 90, 80, and 70 %. The thicknesses of the liquid crystal and compensator layers,
which give the almost perfect compensation in their dark states, are tabulated in Table 5-
3.
5-3-2-1. Luminous transmittance in the bright state model
Figure 5-8 shows the results of the bright state model in two main directions: the
director and the out of the director planes. These results are consistent with that of
Chapter 3, where we calculated and analyzed the effective phase retardations and the off-
axis light transmittances at a single wavelength of 550 nm and at a single thickness of a
liquid crystal layer corresponding about 75% of Y (%) in this chapter.
In the director plane, the luminous transmittances (Y ) always decrease as the
viewing angle increases. This is because, as the viewing angle increases, the total
effective retardation at each wavelength decreases monotonically (“bell shape”), as we
analyzed in Chapter 3.
On the other hand out of the director plane, the luminous transmittances (Y )
increase first and then fall after specific viewing angles when the cell is optically
designed so that Y (%) value is less than 100% at the normal direction. If the cell is
designed as Y (%) has 100% at the normal direction, the luminous transmittance
184
decreases constantly as the viewing angle increases, as in Fig. 5-8 (b). These results also
agree well with the analyses in Chapter 3, i.e. as the viewing angle increases, the total
effective retardation out of the director plane always increases in all wavelengths
(“reversed bell shape”).
5-3-2-2. Luminous transmittance in real LCDs
We did the same calculations for the real LCDs (ECB, VA, Pi-cell, and TN) and
the results are in Fig. 5-9~12. All figures are, amazingly, very similar to that of the bright
state model except Y 100% of the Pi-cell out of the director plane, as in Fig. 5-11(b). In
this case, the luminous transmittance (Y ) fluctuates as the viewing angle increases from
the normal direction. We remember these kinds of the fluctuations in Chapter 4.
According to the results, in the highly tilted director configuration of a bright state, the
off-axis light transmittance varies significantly as the viewing angle increases out of the
director plane, and it makes the viewing angle properties of a Pi-cell worse. In
accordance with the analyses in Chapter 4, both the apparent angle distribution of
directors and the effective phase retardation of each layer play a key role in the
fluctuation of the total transmittance. We think the fluctuation of the luminous
transmittance in Fig. 5-11 (b) is also coming from the similar reasons. This is reasonable
because, as Y (%) value increases at the normal direction, the magnitudes of the effective
retardations of both liquid crystal and compensator layers increase at off-axis viewing
angles, and this is the similar situation in highly tilted director configurations. If we use
the highly tilted director configuration in bright state, the cell thickness of the device
185
should be increased to maintain the magnitude of effective retardation at the normal
direction, when we use the same liquid crystal material. It means the magnitudes of
effective retardations of both liquid crystal and compensator layers increase at off-axis
viewing angles.
5-3-2-3. Viewing angle properties of luminous transmittances in all directions
We also numerically calculated the luminous transmittances of the bright state
model and the real LCDs in all viewing directions, and the Fig. 5-13~17 show the
conoscopic figures. As we can notice in these results, the figure shapes are the same as
that of in Chapter 3 if the Y (%) value is less than 100%. That means, although we
calculated the off-axis light transmittance at a single wavelength of 550 nm in Chapter 3,
it is reasonable and can describe the luminance properties of LCDs well. Another thing
we earned in this section is that the analyses of the phase retardation in Chapter 3 can
predict the luminance properties of LCDs, excellently, even though the cell thicknesses
are changed.
All the conoscopic figures look like similar as we saw in two main directions in
previous section. It confirms again that if the dark states of LCDs are optically optimized
to have minimum transmittance, there is universality in the optical properties of the bright
state regardless of the display modes.
Another important thing is that all figures are anisotropic shapes, although the
figure of Y 100% of each display mode is similar to isotropic shape. Usually, we
optically design a LCD to have a little lower luminance than its maximum value at the
186
normal direction on the purposing of escaping gray scale inversions and to minimize
color shift at off-axis viewing angles. With this in mind, we can say that the single
domain LCD modes, considered here, whose dark states are optically compensated to
give the minimum luminous transmittance, inevitably have asymmetric shapes of the off-
axis luminous transmittances between the director and out of the director planes in their
bright states. Conclusively, we can say that in order to achieve isotropic shapes of the
bright state viewing angle properties, multi-domain methods or asymmetric distribution
of backlight intensity is mandatory.
187
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(b) Out of the director plane
Fig. 5-8. Off-axis luminous transmittances (Y ) of the bright state model with different
Y values at the normal direction.
188
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(b) Out of the director plane
Fig. 5-9. Off-axis luminous transmittances ( Y ) of the bright state ECB mode with
different Y values at the normal direction.
189
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(b) Out of the director plane
Fig. 5-10. Off-axis luminous transmittances ( Y ) of the bright state VA mode with
different Y values at the normal direction.
190
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(b) Out of the director plane
Fig. 5-11. Off-axis luminous transmittances (Y ) of the bright state Pi-cell mode with
different Y values at the normal direction.
191
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(a) Director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
Y100% Y90% Y80% Y70%
Viewing angle (deg.)
Y
(b) Out of the director plane
Fig. 5-12. Off-axis luminous transmittances ( Y ) of the bright state TN mode with
different Y values at the normal direction.
192
(a) Y =100% (b) Y =90%
(c) Y =80% (d) Y =70%
Fig. 5-13. Viewing angle properties of the luminous transmittances (Y ) in the bright state
model with different Y values at the normal direction.
193
(a) Y =100% (b) Y =90%
(c) Y =80% (d) Y =70%
Fig. 5-14. Viewing angle properties of the luminous transmittances (Y ) in the bright state
ECB mode with different Y values at the normal direction.
194
(a) Y =100% (b) Y =90%
(c) Y =80% (d) Y =70%
Fig. 5-15. Viewing angle properties of the luminous transmittances (Y ) in the bright state
VA mode with different Y values at the normal direction.
195
(a) Y =100% (b) Y =90%
(c) Y =80% (d) Y =70%
Fig. 5-16. Viewing angle properties of the luminous transmittances (Y ) in the bright state
Pi-cell mode with different Y values at the normal direction.
196
(a) Y =100% (b) Y =90%
(c) Y =80% (d) Y =70%
Fig. 5-17. Viewing angle properties of the luminous transmittances (Y ) in the bright state
TN mode with different Y values at the normal direction.
197
5-4. Analyses of the color properties
In this section, we calculate and analyze the color properties of the bright state
model (Fig. 5-5) and the real LCD modes (ECB, VA, PI-cell, and TN). The dark states of
these devices are almost perfectly compensated by using passive type negative C-plates,
as the same way that we used in Chapter 3. During the calculation, we used the numerical
relaxation technique to get the director field of the liquid crystal layer and the 2×2-matrix
method to calculate the light transmittances for each wavelength, ranging from 380 to
780 nm. After calculation of the transmittance at each wavelength, we calculated the
color properties of the devices according to the equations in Sec. 5-2 with the illuminant
(or light source), 65D whose specifications are in Fig. 5-4 and Table 5-1.
5-4-1. Thickness effects on color properties
In order to investigate the thickness (or phase retardation) effects on the color
properties of LCDs, we calculated the chromaticity coordinates ( 'u , 'v ) of the bright state
model in function of the thickness of the positive A-plate at the normal direction, and the
result is in Fig. 5-18. This figure shows that as the thickness (or phase retardation)
increases from zero, the chromaticity coordinates ( 'u , 'v ) goes from blue color to yellow
color first and then revolve around the point of the light source color ( 'u =0.1978,
'v =0.4684).
To understand this color variation, we calculated the transmittances and phase
retardations (Γ ) of the bright state model at the three major colors [blue (λ =450 nm),
green (λ =550 nm), and red (λ =650 nm)] at the normal direction, and the Figs. 5-19 and
198
5-20 show the results, respectively. As expected from Eqs. (3-1), (3-5), and Fig. 5-20, the
transmittances are sinusoidal functions in thickness space, and the pitch of blue color is
smaller than that of red color, and the pitch of green color is larger than that of blue color.
These different pitches make different colors for different thicknesses (or phase
retardations).
Comparing Figs. 5-18 and 5-19 shows that at a thickness about 2.6 μm, the blue
transmittance is zero, and at the same thickness, the chromaticity diagram gives the
maximum yellow color. At a thickness about 3.6 μm, the green transmittance goes to
zero and it lead to the maximum blue color in the chromaticity diagram, as in Fig. 5-18.
This color variation with the change of lightness ( *L ), as in Fig. 5-21, cause the
color difference ( uvE *Δ ) fluctuate as in Fig. 5-22.
199
Fig. 5-18. ( 'u , 'v ) chromaticity coordinates of the bright state model in terms of
thicknesses at the normal direction. (Thickness of the A-plate: 0.2-5.0 μm, 0.2 μm step)
d=0.2 μm
d=5.0 μm
d=4.0 μm
d=3.0 μm d=2.0 μm
200
0 1 2 3 4 50.0
0.1
0.2
0.3
0.4
0.5 Blue Green Red
Thickness (μm)
Tran
smitt
ance
Fig. 5-19. Transmittances of the bright state model at the three major colors [blue
(λ =450 nm), green (λ =550 nm), and red (λ =650 nm)] at the normal direction.
201
0 1 2 3 4 50.0
0.5
1.0
1.5
2.0 Blue Green Red
Thickness (μm)
Phas
e re
tard
atio
n (λ
)
Fig. 5-20. Phase retardation of the bright state model at the three major colors [blue
(λ =450 nm), green (λ =550 nm), and red (λ =650 nm)] at the normal direction. (Liquid
crystal: LC53)
202
0 1 2 3 4 50
20
40
60
80
Thickness (μm)
L*
Fig. 5-21. Lightness ( *L ) of the bright state model in function of thickness at the normal
direction. (Reference: illuminant 65D )
203
0 1 2 3 4 50
20
40
60
80
100
120
Thickness (μm)
ΔE* uv
Fig. 5-22. Color difference ( uvE *Δ ) of the bright state model in function of thickness at
the normal direction. (Reference point: thickness of 0.2 μm)
204
5-4-2. Chromaticity coordinates of the bright state model in the off-axis viewing
angles
We calculated the off-axis chromaticity coordinates, ( 'u , 'v ) and ( *u , *v ), of the
bright state model that has different Y (%) values (100, 90, 80, and 70 %) at the normal
direction, and the results are in Figs. 5-23 and 5-24, respectively. The illuminant (or light
source) we used is 65D , and the Eqs. 5-9, 5-10, 5-12, and 5-13 are used for the
calculations. The detailed specifications are in Tables 5-1, 5-2, and 5-3.
Firstly, these figures show that the absolute chromaticity coordinates at the
normal direction (red dots in the figures) move from the blue to yellow colors as the
Y (%) value (or cell thickness) increases, and this result is consistent with the thickness
effects of the previous section.
Secondly, as the Y (%) value increases, the total variation of the chromaticity
coordinates (from the normal direction to the viewing polar angle of 80º in these figures)
widens in both director (Φ=90, 270º) and out of the director (Φ=0, 180º) planes. If we
compare the magnitudes of the variations between the director plane and the out of the
director plane, the variation out of the director plane is bigger. This implies that when we
optically design a LCD to have higher transmittance at the normal direction, the display
could have bigger color variation, especially out of the director plane. The reason of this
fact can be understandable from the off-axis light transmittances of the red, green, and
blue colors. Figures 5-25, 5-26, 5-27, and 5-28 show the calculation results of the off-axis
light transmittances of the bright state model, which has Y (%) value of 100, 90, 80, and
70% at the normal direction. In these transmittance curves, we see that as the Y (%) value
205
increases, the transmittances of the blue, green, and red colors more significantly change
at off-axis viewing angles compared with normal direction, and also this phenomenon is
more severe in the viewing angles out of the director plane than in the director plane.
This transmittance variation associated with phase retardation causes the color change at
the off-axis viewing angles.
Finally, we can see that as the viewing angle increases, the color coordinates of
the bright state model, initially, move to yellow direction out of the director plane. On the
other hand, the color goes to the blue direction first in the director plane. These facts are
always true regardless of Y (%) value at the normal direction and related to the variation
of the phase retardation at off-axis directions. Figure 5-29 is the calculated result of the
off-axis phase retardation of the bright state model with Y (%)=80. This effective phase
retardation is calculated from Eqs. 3-7, 3-8, 3-10, and 3-11. As we can see in this figure,
the effective phase retardations increase as the viewing angle increases out of the
direction plane for all light wavelengths. On the contrary, the effective phase retardations
fall as the angle of incident light increases for all light colors. This universality is
consistent with the result of the Chapter 3. With these facts and the analysis of the
thickness effects on the color properties in Sec. 5-4-1, we can understand the reasons. As
the viewing angle increases out of the director plane, the effective phase retardation of
the bright state model increases, and this leads to color shift to yellow initially. In
contrast in the director plane, the effective phase retardation falls as the viewing angle
increases, and this initially causes a color shift to blue.
206
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-23. ( 'u , 'v ) chromaticity coordinates of the bright state model in functions of
viewing angles. [0, 180: out of the director plane (overlapped each other); 90, 270:
director plane (overlapped each other); red dot: normal direction; viewing polar angles
are spanned from 0 to 80º with 5º step]
207
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-24. ( *u , *v ) chromaticity coordinates of the bright state model in functions of
viewing angles. [0, 180: out of the director plane (overlapped each other); 90, 270:
director plane (overlapped each other); red dot: normal direction; viewing polar angles
are spanned from 0 to 80º with 5º step]
208
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-25. Transmittances of the bright state model with Y (%)=100 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm).
209
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-26. Transmittances of the bright state model with Y (%)=90 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm).
210
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-27. Transmittances of the bright state model with Y (%)=80 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm).
211
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-28. Transmittances of the bright state model with Y (%)=70 in the three major
colors (blue: 450 nm, green: 550 nm, red: 650 nm).
212
-80 -60 -40 -20 0 20 40 60 800.0
0.2
0.4
0.6
0.8
1.0 Blue Green Red
Viewing angle (deg.)
Phas
e re
tard
atio
n (λ
)
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.2
0.4
0.6
0.8
1.0 Blue Green Red
Viewing angle (deg.)
Phas
e re
tard
atio
n (λ
)
(b) Director plane
Fig. 5-29. Off-axis phase retardation of the bright state model with Y (%)=80 in the three
major wavelengths (blue: 450 nm, green: 550 nm, red: 650 nm).
213
5-4-3. Chromaticity coordinates of the common bright state LCDs in the off-axis
viewing angles
We did the similar color calculations for the real LCDs (ECB, VA, Pi-cell, and
TN modes), which are optically designed to have several different Y (%) values (100, 90,
80, and 70 %) at the normal direction. We used 65D illuminant again, and the detailed
cell specifications are in Tables 5-1, 5-2, and 5-3. Figures 5-30~33 and 5-34~37 show the
calculation results of the off-axis chromaticity coordinates, ( 'u , 'v ) and ( *u , *v ),
respectively. Overall, the figure shapes are in good agreement with that of our simple
bright state model except the TN mode.
The absolute chromaticity coordinates at the normal direction move from blue to
yellow as Y (%) value increases in all display modes.
As Y (%) value increases, the total variation of the chromaticity coordinates
increases in both main viewing planes, and the magnitudes of the variations out of the
director plane are bigger than that of the director plane.
As the viewing angle increases, the color coordinates of the bright state move to
the yellow direction first in the viewing angles out of the director plane, but the color
shifts to the blue direction first in the director plane.
In the TN mode, the magnitude of the color variation out of the director plane
(refer to the mid-layer of directors) is the similar level with that of other display modes.
However, in the director plane, the color variation is much bigger than that of other
display modes, and the variation becomes large as Y (%) value decreases. This is the
opposite direction compared with our bright state model and other display modes
214
considered here. To understand these interesting facts, we calculated the off-axis
transmittances of TN mode for three light wavelengths: blue (450 nm), green (550 nm),
and red (650 nm), and the results are in Figs. 5-38~41. From these figures, we can notice
that as the Y (%) value decreases, the off-axis transmittances, especially red and green,
sharply go to zero at large viewing angles compared with that of the bright state model
(Figs. 5-25~28). This makes the bright state of a TN mode have a strong bluish color at
large viewing angle in the director plane (refer to mid-layer). We think this kind of
behavior of TN modes is related to the twist effect of liquid crystal directors. From these
results, we can say that the conventional “3-layer” model for the TN mode, which is
considered in Chapter 3, is partially incorrect when we calculate the optical properties of
the device.
215
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-30. ( 'u , 'v ) chromaticity coordinates of the ECB in functions of viewing angles.
[0, 180: out of the director plane (almost overlapped each other); 90, 270: director plane
(almost overlapped each other); red dot: normal direction; viewing polar angles are
spanned from 0 to 80º with 5º step]
216
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-31. ( 'u , 'v ) chromaticity coordinates of the VA in functions of viewing angles. [0,
180: out of the director plane (almost overlapped each other); 90, 270: director plane
(almost overlapped each other); red dot: normal direction; viewing polar angles are
spanned from 0 to 80º with 5º step]
217
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-32. ( 'u , 'v ) chromaticity coordinates of the Pi-cell in functions of viewing angles.
[0, 180: out of the director plane; 90, 270: director plane (almost overlapped each other);
red dot: normal direction; viewing polar angles are spanned from 0 to 80º with 5º step]
218
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-33. ( 'u , 'v ) chromaticity coordinates of the TN in functions of viewing angles. [0,
180: out of the director plane (almost overlapped each other); 90, 270: director plane
(almost overlapped each other); red dot: normal direction; viewing polar angles are
spanned from 0 to 80º with 5º step]
219
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-34. ( *u , *v ) chromaticity coordinates of the ECB in functions of viewing angles.
[0, 180: out of the director plane (almost overlapped each other); 90, 270: director plane
(almost overlapped each other); red dot: normal direction; viewing polar angles are
spanned from 0 to 80º with 5º step]
220
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-35. ( *u , *v ) chromaticity coordinates of the VA in functions of viewing angles.
[0, 180: out of the director plane (almost overlapped each other); 90, 270: director plane
(almost overlapped each other); red dot: normal direction; viewing polar angles are
spanned from 0 to 80º with 5º step]
221
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-36. ( *u , *v ) chromaticity coordinates of the Pi-cell in functions of viewing
angles. [0, 180: out of the director plane; 90, 270: director plane (almost overlapped each
other); red dot: normal direction; viewing polar angles are spanned from 0 to 80º with 5º
step]
222
(a) Y (%)=100 (b) Y (%)=90
(c) Y (%)=80 (d) Y (%)=70
Fig. 5-37. ( *u , *v ) chromaticity coordinates of the TN in functions of viewing angles.
[0, 180: out of the director plane (almost overlapped each other); 90, 270: director plane;
red dot: normal direction; viewing polar angles are spanned from 0 to 80º with 5º step]
223
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-38. Transmittances of the TN mode with Y (%)=100 in the three major colors
(blue: 450 nm, green: 550 nm, red: 650 nm).
224
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-39. Transmittances of the TN mode with Y (%)=90 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
225
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-40. Transmittances of the TN mode with Y (%)=80 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
226
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-41. Transmittances of the TN mode with Y (%)=70 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
227
5-4-4. Color difference of the bright state model in the off-axis viewing angles
In this section, we investigate the viewing angle dependence of the color
difference ( uvE *Δ ) in the bright state model. We chose the color coordinates of the
normal direction as the reference point ( *1L , *1u , *1v ) for the calculations. Figure 5-42
shows the results calculated from Eq. 5-16.
In the viewing directions out of the director plane, as the Y (%) value increases,
the off-axis color difference ( uvE *Δ ) increases. This result is consistent with the color
variation as a function of Y (%) value in Sec. 5-4-2, and coming from the facts that the
off-axis transmittances of the red, green, and blue are different more and more as the
Y (%) value increases.
In the viewing angles of the director plane, the off-axis color difference ( uvE *Δ )
increases as the Y (%) value increases, but the variation is very small except the
condition,Y (%)=100. This fact can be understandable from the transmittance curves of
red, green, and blue in Figs. 5-25~28. As we see from these figures, the transmittances of
red, green, and blue smoothly change as the Y (%) value increases, compared with that of
out of the director plane. In the cell condition, Y (%)=100, the color difference sharply
increases at the off-axis angles, and this is coming from the sort of the fluctuation of the
blue transmittance, as we can see in Fig. 5-25 (b). These features of transmittances are
correlated to the effective phase retardation at the off-axis angles (Fig. 5-29).
228
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(b) Director plane
Fig. 5-42. Off-axis color difference ( uvE *Δ ) of the bright state model.
(Reference point: normal direction).
229
5-4-5. Color difference of the common bright state LCDs in the off-axis viewing
angles
We did the same calculations for the common LCDs (ECB, VA, Pi-cell, and TN
modes), and the Figs. 5-43~46 show the results. Comparing the results with that of the
bright state model in previous section, the figure shapes are very similar each other but
the TN mode.
In the directions out of the director plane, as the Y (%) value increases, the off-
axis color difference ( uvE *Δ ) increases, as the same way of the bright state model. In the
director plane, the off-axis color difference ( uvE *Δ ) very slowly increases as the Y (%)
value increases if the LCD is designed to have less than 100% of Y (%) at the normal
direction. In the condition that the Y (%) value approaches 100, the color difference
sharply increases at the off-axis viewing angles, with the same reason described in
previous section.
5-4-5-1. Detailed analyses of the color difference of the Pi-cell
The color difference ( uvE *Δ ) of the Pi-cell out of the director plane fluctuates at
large viewing angles when the cell is designed to have high transmittance at the normal
direction. We think this is one of the strong points of the Pi-cell because we can suppress
the color shift of the device better than other LCD modes at large viewing angles. This
sort of the fluctuation of the color difference can be imagined from the analyses of the
chromaticity coordinates of the Pi-cell in Sec. 5-4-3. If we see the Figs. 5-30~37, the
chromaticity coordinates of the Pi-cell turn back to their start position as the viewing
230
angle increases out of the director plane. This feature of the Pi-cell could provide the
optical design of a LCD with room to achieve high transmittance at the normal direction
without big color shift at large viewing angles.
The reasons for this phenomenon in the Pi-cell can be understood from the
separated transmittances of the red, green, and blue colors. Figures 5-47~50 show the off-
axis transmittance curves calculated under the same conditions. In the director plane, the
transmittances of the red, green, and blue colors vary smoothly as the Y (%) value
increases, and this gives the similar color differences for diverse Y (%) values, as in Fig.
5-45 (b). On the contrary, out of the director plane, the transmittances of the red, green,
and blue colors more significantly change as the Y (%) value increases, and these results
in various curve shapes of the color difference, as in Fig. 5-45 (a). Another important
thing, in the viewing direction out of the director plane, is that the transmittances,
especially the blue color, fluctuate as the Y (%) value increases at large viewing angles.
We think this fluctuation leads to the fluctuation of the color difference, and it suppresses
the color shift at large viewing angles.
5-4-5-2. Detailed analyses of the color difference of the TN mode
Another interesting fact we can notice from the calculations of the color
differences in the bright state LCDs is in the TN mode (Fig. 5-46). As we can see from
this figure, the color difference ( uvE *Δ ) of the TN mode is relatively insensitive to the
Y (%) value in both main viewing planes. Especially, the color difference in the condition,
Y (%)=100, is excellent and even better than the others. This surprising feature in TN
231
mode is very noticeable because, in our bright state model and in other display modes
considered here, the color differences ( uvE *Δ ) are the worst in the condition that has
100% of Y (%) value.
This feature of the TN mode can be understood from the transmission properties
of the major wavelengths in Fig. 5-38. As we see, the transmittance curves have the
different shapes, especially for the blue wavelength, compared with that of the bright
state model (Fig. 5-25). All three major colors (blue: 450 nm, green: 550 nm, and red:
650 nm) have very similar transmittances not only in the normal direction, but also in the
off-axis angles out of the director plane. This phenomenon becomes much stronger as the
Y (%) value increases.
These features of the TN mode remind us the Mauguin condition54, 55,
ndΔ<<Φλπ2 , (5-17)
where Φ is the total twist angle, d is the thickness of the liquid crystal layer, nΔ is the
birefringence of the liquid crystal material, and λ is the light wavelength. The Mauguin
condition states that if the director configuration of the liquid crystal satisfies the
condition, the liquid crystal acts like a wave-guide for incident linear polarized light,
whose direction is along the easy axis of the liquid crystal. This means the light
transmittance is insensitive to the light wavelengths.
232
With this knowledge in mind, now we can understand the surprising facts of the
TN modes. As the Y (%) value increases, and as the viewing angle increases out of the
director plane, and as the light wavelength decreases, the effective retardation becomes
larger, and there are much more chances to meet the Mauguin condition. Therefore, the
light transmittance becomes insensitive to the incident light wavelengths, and this reduces
the color difference ( uvE *Δ ) of the device. One of the possible problems, which might
happen, is the gray scale inversions in TN mode when we design the device to have the
maximum luminance at the normal direction.
233
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(b) Director plane
Fig. 5-43. Off-axis color difference ( uvE *Δ ) of the ECB mode. (Reference point: normal
direction).
234
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(b) Director plane
Fig. 5-44. Off-axis color difference ( uvE *Δ ) of the VA mode. (Reference point: normal
direction).
235
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(b) Director plane
Fig. 5-45. Off-axis color difference ( uvE *Δ ) of the Pi-cell mode. (Reference point:
normal direction).
236
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800
10
20
30
40
50
60
70 Y100% Y90% Y80% Y70%
Viewing angle (deg.)
ΔE* uv
(b) Director plane
Fig. 5-46. Off-axis color difference ( uvE *Δ ) of the TN mode. (Reference point: normal
direction).
237
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-47. Transmittances of the Pi-cell with Y (%)=100 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
238
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-48. Transmittances of the Pi-cell with Y (%)=90 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
239
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-49. Transmittances of the Pi-cell with Y (%)=80 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
240
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(a) Out of the director plane
-80 -60 -40 -20 0 20 40 60 800.0
0.1
0.2
0.3
0.4
0.5
Blue Green Red
Viewing angle (deg.)
Tran
smitt
ance
(b) Director plane
Fig. 5-50. Transmittances of the Pi-cell with Y (%)=70 in the three major colors (blue:
450 nm, green: 550 nm, red: 650 nm).
241
5-5. Summary
We numerically calculated and analyzed the off-axis luminous transmittance (Y )
and color properties of the bright state model with the common LCDs (ECB, VA, Pi-cell,
and TN). Their cell thicknesses are adjusted to have four different Y (%) values at the
normal direction: 100, 90, 80, and 70 %. Their dark states are almost perfectly
compensated by using the hybrid aligned negative C-plates.
5-5-1. Luminous transmittance of the bright state LCDs
According to the results, in both the bright state simple model and the real LCDs,
the luminous transmittances (Y ) always decrease as the viewing angle increases in the
director plane. On the other hand out of the director plane, the luminous transmittances
(Y ) increase first and then fall after specific viewing angles when the cell is optically
designed so that Y (%) value is less than 100% at the normal direction. These results are
in good agreement with the birefringence analyses.
We also calculated the luminous transmittances (Y ) in all viewing directions, and
compared them by using conoscopic figures. All the figures look similar. It confirms
again that if the dark states of LCDs are optically optimized, there is universality in the
optical properties of the bright state regardless of the display modes.
Another important thing is that all conoscopic figures are anisotropic shapes. This
means that the single domain LCD modes, considered here, inevitably have asymmetric
shapes of the luminance between the director and out of the director planes in their bright
states. Therefore, we can say that in order to achieve isotropic shapes of the bright state
242
viewing angle properties, multi-domain methods or asymmetric distribution of backlight
intensity is mandatory.
5-5-2. Color properties of the bright state LCDs
Firstly, we investigated the thickness (or phase retardation) effects on the color
properties of LCDs. These calculations show that as the thickness (or phase retardation)
increases from zero, the chromaticity coordinates ( 'u , 'v ) initially goes from blue to
yellow colors and then revolve around the point of the light source color. We analyzed
this fact by using the transmittances and the phase retardations of the three major colors.
Secondly, we calculated the off-axis chromaticity coordinates of the bright state
model and real LCDs such as ECB, VA, PI-cell, and TN modes. Overall figure shapes of
the real LCDs are in good agreement with that of our simple bright state model except the
TN mode. The absolute chromaticity coordinates at the normal direction move from blue
to yellow as Y (%) value increases in all display modes. As Y (%) value increases, the
total variation of the chromaticity coordinates increases in both main viewing planes, and
the magnitudes of the variations out of the director plane are bigger than that of the
director plane. As the viewing angle increases, the color coordinates of the bright state
move to the yellow direction first in the viewing angles out of the director plane, but the
color shifts to the blue direction first in the director plane.
In the TN mode, the magnitude of the color variation out of the director plane
(refer to the mid-layer of directors) is the similar level with that of other display modes.
However, in the director plane, the color variation is much bigger than that of other
243
display modes, and the variation becomes large as Y (%) value decreases. This is the
opposite direction compared with our bright state model and other display modes
considered here.
Finally, we investigate the viewing angle dependence of the color difference
( uvE *Δ ) in the bright state model and the real LCDs. When we compare the results, the
color differences of the real LCDs have the same curve shapes as that of the bright state
model except the TN mode. In the viewing directions out of the director plane, as the
Y (%) value increases, the off-axis color difference ( uvE *Δ ) increases, and this is coming
from the facts that the off-axis transmittances of the red, green, and blue are different
more and more as the Y (%) value increases. In the viewing angles of the director plane,
the off-axis color difference ( uvE *Δ ) increases as the Y (%) value increases, but the
variation is very small. This result is coming from the properties of the blue, green, and
red transmittances.
In the Pi-cell, out of the director plane, the color difference ( uvE *Δ ) fluctuates at
large viewing angles when the cell is designed to have high transmittance at the normal
direction. We think this is one of the strong points of the Pi-cell because we can suppress
the color shift of the device better than other LCD modes at large viewing angles.
In the TN mode, the color difference ( uvE *Δ ) is relatively insensitive to the
Y (%) value. Especially, the color difference in the condition, Y (%)=100, is excellent
and even better than the others. This surprising feature in TN mode is related to the
Mauguin condition.
244
CHAPTER 6
THE CONDITIONS AND THE LIMITATIONS OF THE PERFECT PHASE
COMPENSATION IN LIQUID CRYSTAL DISPLAYS
6-1. Introduction
In this chapter, we will study the conditions and the limitations of the phase
compensation for the dark state of LCDs. As an example compensator, we will use hybrid
aligned negative C-plates.
In Sec. 6-2, we discuss the difference between an approximate method and a
“complete” method for the calculation of the phase retardation in a uniaxial material. In
Sec. 6-3, we analyze the relationship among the parameters ( ',',' dnn oe ) of a
compensator for the perfect compensation of a uniaxial slab. With this parameter
relationship, we discuss the thickness ratio of the compensator to uniaxial slab for the
perfect compensation. In Sec. 6-4, we calculate and analyze the total effective phase
retardation and the off-axis light transmittances as a function of the director tilt angle and
the thickness in the uniformly and non-uniformly aligned liquid crystal layers.
6-2. Calculations of the phase difference
In this section, we investigate the methods for calculating the phase difference of
the extraordinary and ordinary rays in a uniaxial medium.
245
6-2-1. “Complete” method
Let us consider a uniaxial slab (ordinary and extraordinary refractive indices of
on and en , respectively) whose optic axis or director ( nr ) is aligned to have the tilt angle
of θ and the azimuth angle of φ in the laboratory coordinate system ( zyx ,, ). Incident
light with the wavelength of λ in free space propagates in the zx − plane. Figure 6-1
shows the definition of these quantities. The phase difference ( Γ ) between the
extraordinary ray ( e -ray) and the ordinary ray ( o -ray) after propagating the uniaxial slab
is expressed as follows56, 57:
dKK ozez )( −=Γ , (6-1)
where ezK and ozK are the z components of the light propagation vectors of the e -ray
and the o -ray, respectively, and they are written as,
'cos2e
effeez nK θ
λπ
= , (6-2)
'cos2oooz nK θ
λπ
= , (6-3)
where effen is the effective refractive index of the e -ray and expressed as58,
246
ψψ 2222 sincos oe
oeeffe
nn
nnn
+= , (6-4)
where ψ is the angle between the propagation vector of the e -ray and the director ( nr )
of the uniaxial slab, and it can be calculated from59,
θθφθθψ sin'coscoscos'sincos ee += , (6-5)
where 'eθ is the angle of the light propagation vector from z -axis in the medium.
247
(a)
(b)
Fig. 6-1. Angle definitions with a uniaxial slab.
θ
'eθ
eKr
ψ
z
),( φθnr
x
φ y
d'oθ
'eθ
e -ray
oα
o -ray
oKr
Uniaxial slab
( oe nn , ) x
z
248
6-2-2. Approximate method
Let us say that the refractive indices of the e -ray and the o -ray in the uniaxial
slab are similar, i.e. oe nn ≅ . Then, we can assume that the optical paths of the e -ray and
the o -ray are the same in the medium, as in Fig. 6-2. Therefore, we have this relationship,
αθθ ≡= '' oe , (6-6)
where α is the angle of the light propagation vector ( e -ray and the o -ray) from the z -
axis. With this assumption, the phase difference (Γ ) between the e -ray and the o -ray
after propagating the uniaxial slab is written as follows60:
effndΔ=Γλπ2
αλ
πcos
)(2 dnn oeffe −= , (6-7)
where effen is the effective refractive index of the e -ray as expressed in Eq. (6-4), and the
ψ is rewritten as,
θαφθαψ sincoscoscossincos += . (6-8)
249
(a)
(b)
Fig. 6-2. Angle definitions with a uniaxial slab in the approximation method.
dα
e -ray
oα
o -ray
oKr
Uniaxial slab
( oe nn , ) x
z
θ
α
Kr
ψ
z
),( φθnr
x
φ y
250
6-2-3. Comparison of both methods for calculating the phase difference
In order to see the difference between two methods, described in previous
sections, we calculate the effective phase retardation ( effndΔ ) of a nematic slab by using
both methods. The ordinary ( on ) and extraordinary ( en ) refractive indices of the slab are
1.5 and 1.6, respectively, and the thickness is 2.75 μm. The light is incident in zx −
plane with the wavelength of 550 nm, and the incident angle ( oα ) in air is range from -
80º to +80º. In both methods, the refraction is considered at the air interface. In the
simple approximation method, the average refractive index is used to calculate the
refraction angle. In the “complete” method, we have to know the effen and 'eθ ,
simultaneously. To find both values, we tried two ways. First, we used the recursive
program. Second, we used the eigenvalues of the 2×2 matrix method.
Figure 6-3 shows the results for the several different director configurations
(θ ,φ ). As we can see, they are very similar, and the two ways used in the “complete”
method are the same. At the normal direction, the approximation method and the
“complete” method are the same as expected. As the incident light angle ( oα ) increases,
the difference between two methods increases, but the magnitudes of the differences are
extremely small. Therefore, in this chapter, without hurting the generality, we will use the
simple approximation method to calculate and analyze the effective phase retardations of
the uniaxial type anisotropic materials.
251
-80 -60 -40 -20 0 20 40 60 800
50
100
150
200
250
300
350
400 Simple Recursive 2x2 matrix
Light incident angle, αo (deg.)
Effe
ctiv
e Δn
d (n
m)
(a) θ =0º, φ =0º (director plane)
-80 -60 -40 -20 0 20 40 60 800
50
100
150
200
250
300
350
400
Simple Recursive 2x2 matrix
Light incident angle, αo (deg.)
Effe
ctiv
e Δn
d (n
m)
(b) θ =0º, φ =90º (out of the director plane)
(c) and (d): next page
Fig. 6-3. Effective phase retardations of the nematic slabs (θ ,φ ) calculated by the
approximation method (“Simple”) and complete method (“Recursive” and “2×2 matrix”).
“Recursive” and “2×2 matrix” are overlapped each other in all the figures.
252
-80 -60 -40 -20 0 20 40 60 800
50
100
150
200
250
300
350
400
Simple Recursive 2x2 matrix
Light incident angle, αo (deg.)
Effe
ctiv
e Δn
d (n
m)
(c) θ =90º, φ =0º
-80 -60 -40 -20 0 20 40 60 800
50
100
150
200
250
300
350
400
Simple Recursive 2x2 matrix
Light incident angle, αo (deg.)
Effe
ctiv
e Δn
d (n
m)
(d) θ =45º, φ =45º
253
6-3. Conditions and limitations for the perfect phase compensation in the simple
director configurations
6-3-1. Relationship of the parameters
Let us consider a simple compensation system, as in Fig. 6-4. We have a uniaxial
slab that has the thickness of d , the refractive indices of ( en , on ) and the director angles
of (θ , φ ). We are going to compensate the uniaxial slab with a layer of negative type
compensator, which has the thickness of 'd , the refractive indices of ( 'en , 'on ) and the
same director angles of the uniaxial slab. Using the same director angles is intuitively
reasonable.
The effective phase retardations of the uniaxial slab ( LCeffndΔ ) and the
compensator ( NCeffndΔ ) are expressed, respectively as follows:
αψψ cossincos 2222
dnnn
nnnd o
oe
oeLCeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ , (6-9)
'cos''
'sin''cos'
''2222 αψψ
dnnn
nnnd o
oe
oeNCeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ . (6-10)
The total effective phase retardation ( TotaleffndΔ ) of the uniaxial slab and the compensator
is the summation of both values because their optic axes are parallel with each other,
254
NCeff
LCeff
Totaleff ndndnd Δ+Δ=Δ . (6-11)
For the perfect compensation of the uniaxial slab, the total effective phase retardation
( TotaleffndΔ ) should be zero,
0=Δ+Δ=Δ NCeff
LCeff
Totaleff ndndnd . (6-12)
If we think about the parameters that we need for the perfect compensation of a uniaxial
slab, we have to decide the values, 'en , 'on , and 'd of the compensator. To understand
the relationship between the parameters, we calculate the total effective retardation of a
compensation system. The uniaxial slab has the ordinary and extraordinary refractive
indices of 1.5 and 1.6, respectively, and its thickness is 1.0 μm. The director tilt angle (θ )
of the slab split to 0, 30, and 60º, but their azimuth angles (φ ) are the same, 0º. We fixed
the ordinary refractive index ( 'on ) of the compensator as 1.6. The extraordinary index
( 'en ) and the thickness ( 'd ) of the compensator are spanned as 1.4~1.54 and 0.4~1.6 μm,
respectively. The light is incident with the wavelength of 550 nm, and the incident angle
( oα ) is zero.
Figure 6-5 shows the calculated total effective retardation ( TotaleffndΔ ) in functions of 'en
and 'd of the compensation system. From this figure, we know that for a given value of
255
'en , we can always choose the corresponding 'd that gives zero total effective
retardation. Figure 6-6 shows a pair of parameters ( 'en , 'd ) for the perfect compensation
( 0=Δ Totaleffnd ) of the uniaxial slab. These results show that only two variables among 'en ,
'on , and 'd are independent for the perfect compensation of a given uniaxial slab. For
example, if 'en , 'on are given, we can perfectly compensate a uniaxial slab in a given
viewing angle by adjusting only the thickness ( 'd ) of the negative type compensator.
This means that we do not have to think about the detailed values of 'en and 'on .
256
(a)
(b)
Fig. 6-4. Angle definitions of a simple compensation system.
Compensator ( ',' oe nn )
d
'd
θ
),( φθnr
Uniaxial slab ( oe nn , )
θ
α
Kr
ψ
z
),( φθnr
x
φ y
257
(a) θ =0º
(b) and (c): next page
Fig. 6-5. Total effective phase retardation in functions of 'en and 'd in the compensation
system. (φ =0º, oα =0º, en =1.6, on =1.5, d =1.0 μm, 'on =1.6)
258
(b) θ =30º
(c) θ =60º
259
1.40 1.42 1.44 1.46 1.48 1.50 1.52 1.54
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
θ=0o
θ=30o
θ=60o
ne'
d' (μ
m)
Fig. 6-6. A pair of parameters ( 'en , 'd ) for the perfect compensation ( 0=Δ Totaleffnd ) of
the uniaxial slab. (φ =0º, oα =0º, en =1.6, on =1.5, d =1.0 μm, 'on =1.6)
260
6-3-2. Thickness ratio for the perfect compensation
With these facts in mind, let us set the ordinary and extraordinary refractive
indices of the negative type compensator the same as the extraordinary and ordinary
refractive indices of the uniaxial slab, respectively, i.e. oe nn =' , eo nn =' . Then, the Eq.
(6-10) is written as,
αψψ cos'
sincos 2222
dnnn
nnnd e
eo
oeNCeff ⎟
⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=Δ . (6-13)
From the fact that the total effective phase retardation ( TotaleffndΔ ) should be zero for the
perfect compensation of a uniaxial slab [Eq. (6-12)], we have this thickness relationship
that satisfies the condition for the perfect phase compensation of a uniaxial slab,
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛−
+−
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛−
+=
e
eo
oe
o
oe
oe
nnn
nn
nnn
nn
dd
ψψ
ψψ
2222
2222
sincos
sincos' (6-14)
effNC
effLC
nnΔ−Δ
≡ .
This equation means that in a given director configuration of a uniaxial slab (θ , φ ), we
can always compensate it completely in an incident light condition (α , λ ). However, the
261
magnitude of dd /' varies in functions of θ , φ , and α . Therefore, it implies that it is
almost impossible to compensate an anisotropic system exactly for all light wavelengths
and viewing angles, simultaneously. If the anisotropic system is composed of the layers
that have different director configurations, the perfect compensating system becomes
much more difficult. This means we need to adjust the thickness of each sub-layer of the
compensator, separately, with considering the director configuration of each sub-layer of
the uniaxial slab for better phase compensation.
6-3-3. Thickness ratio for the different director configurations
In this section, we calculate the thickness ratio ( dd /' ) for the perfect
compensation of the uniaxial slab (Fig. 6-4), which has several different director
configurations. We use the refractive indices of the commercialized liquid crystal, LC53
(Fig. 2-7), as the refractive indices of the uniaxial slab. We fixed the thickness of the
uniaxial slab to be 1.0 μm for the convenience. We followed the description of the
previous sections for the refractive indices and the director angle of the compensator. All
angle definitions are complied with that in Fig. 6-4. Figure 6-7 shows the calculated
thickness ratios ( dd /' ) for the perfect compensation of the uniaxial slab (θ =0, 30, 60,
and 90º) with the light wavelength of 550 nm, and Fig. 6-8 shows the three-dimensional
figure for the same light wavelength. These figures show that as stated in the previous
section, the thickness ratio ( dd /' ) for the perfect compensation of the uniaxial slab is the
functions of θ , φ , and α for a given material. The light wavelength (λ ) dependence on
the thickness ratio is relatively small if the wavelength dispersion of the uniaxial slab is
262
not so big. However, the thickness ratio is very sensitive to the director configuration (θ ,
φ ) and the light incident angle ( oα ).
263
-80 -60 -40 -20 0 20 40 60 80
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
d'/d
(a) φ =0º (director plane), λ =550 nm
(b) and (c): next page
Fig. 6-7. Thickness ratio ( dd /' ) for the perfect compensation of the uniaxial slab (θ =0,
30, 60, and 90º; d =1.0 μm) with the light wavelength of 550 nm.
264
-80 -60 -40 -20 0 20 40 60 80
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
d'/d
(b) φ =90º (out of the director plane), λ =550 nm
-80 -60 -40 -20 0 20 40 60 80
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
d'/d
(c) φ =45º, λ =550 nm
265
(a) φ =0º (director plane), λ =550 nm
(b) and (c): next page
Fig. 6-8. Three-dimensional thickness ratio ( dd /' ) for the perfect compensation of the
uniaxial slab (θ =0~90º; d =1.0 μm) with the light wavelength of 550 nm.
266
(b) φ =90º (out of the director plane), λ =550 nm
(c) φ =45º, λ =550 nm
267
6-3-4. Transmittances in the compensation system
In this section, we numerically calculate the off-axis light transmittance of the
compensation system that we used in the previous section (Sec. 6-3-3). We insert the
system between the ideal crossed polarizers, where the transmission axis of the bottom
polarizer makes -45º angles relative to the easy axis of the uniaxial slab. We calculate the
off-axis light transmittance for three different thickness ratios ( dd /' ), 0.8, 0.9, and 1.0
with the light wavelength of 550 nm. In this calculation, we also set the thickness of the
uniaxial slab as 1.0 μm for all different tilt angles (θ ). Figures 6-9, 6-10, and 6-11 show
the results for the different azimuth angles of the uniaxial slab (φ ), 0, 90, and 45º,
respectively, and Figs. 6-12~14 are the corresponding three-dimensional figures. In these
figures, the transmittance is 1 if the transmitted light has the same intensity as that of
incident light. In the viewing directions of the director plane (φ =0º) and the out of the
director plane (φ =90º), the light transmittances include the light leakage that comes from
the geometrical reasons of the crossed polarizers, as described in Sec. 3-2-3. However,
the transmittance in the direction, φ =45º, does not include the geometrical polarizer
effects. With this knowledge and the results in previous section in mind, we can see that
as the thickness ratio ( dd /' ) varies, the points of the perfect compensation (or the
minimum light transmittance) move toward the points that satisfy the condition, dd /' . In
the meanwhile, in the points that does not satisfy the condition, dd /' , the light
transmittance (or leakage) could increase depending on the director configuration and the
light incident angle. These transmittance results are in good agreement with the results in
Sec. 6-3-3.
268
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-9. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 0º (director plane).
(θ =0, 30, 60, and 90º; d =1.0 μm, λ =550 nm).
269
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) dd /' =0.9
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) dd /' =1.0
270
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-10. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 90º (out of the director plane).
(θ =0, 30, 60, and 90º; d =1.0 μm, λ =550 nm).
271
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) dd /' =0.9
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) dd /' =1.0
272
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-11. Off-axis light transmission properties of the compensation system with the
azimuth angle (φ ) of 45º.
(θ =0, 30, 60, and 90º; d =1.0 μm, λ =550 nm).
273
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) dd /' =0.9
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05
θ=0o
θ=30o
θ=60o
θ=90o
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) dd /' =1.0
274
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-12. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 0º (director plane).
(θ =0~90º; d =1.0 μm, λ =550 nm).
275
(b) dd /' =0.9
(c) dd /' =1.0
276
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-13. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 90º (out of the director plane).
(θ =0~90º; d =1.0 μm, λ =550 nm).
277
(b) dd /' =0.9
(c) dd /' =1.0
278
(a) dd /' =0.8
(b) and (c): next page
Fig. 6-14. Three-dimensional off-axis light transmission properties of the compensation
system with the azimuth angle (φ ) of 45º.
(θ =0~90º; d =1.0 μm, λ =550 nm).
279
(b) dd /' =0.9
(c) dd /' =1.0
280
6-4. Applications
6-4-1. Uniform director configurations
As an application, let us consider a uniformly aligned liquid crystal layer. The tilt angle
(θ ) of the liquid crystal director is split into 3 types, 0, 30, and 60º. The thickness of each
uniformly tilted liquid crystal layer is determined so that each liquid crystal layer has four
different effective phase retardations ( LCeffndΔ : 50, 100, 200, and 300 nm at a light
wavelength of 550 nm) at the normal direction. Now, we want to compensate the liquid
crystal layers perfectly at the normal direction with the light wavelength of 550 nm. We
use a negative type compensator for the phase compensation of the liquid crystal layer
with the same optic axis as that of liquid crystal. The thickness of the compensator is
determined to meet the thickness ratio ( dd /' ) for the perfect phase compensation at the
normal direction. The thickness ratios, here we use, are the same as the values of the
normal direction in Fig. 6-7. The detailed thicknesses are calculated as in Table 6-1.
Where, we used the optical parameters of the commercialized liquid crystal, LC53 (Fig.
2-7), for the liquid crystal layer of the compensation system. The ordinary and
extraordinary refractive indices of the compensator are the extraordinary and ordinary
refractive indices of the liquid crystal, respectively, as we did in Sec. 6-3.
Firstly, we calculate and analyze the total effective phase retardation ( TotaleffndΔ ) of
the compensation systems. Figures 6-15, 6-16, and 6-17 show the calculation results for
the different tilt angles (θ ) of 0, 30, and 60º, respectively. Each figure contains the total
281
effective phase retardation in the viewing directions (φ ) of 0º (director plane), 90º (out of
the director plane), and 45º.
As we can see in these figures, the total effective phase retardation at the normal
direction is zero at all figures because we compensated perfectly at the normal direction
for each director configuration. All figure shapes are consistent with the analyses of the
thickness ratio ( dd /' ) in Fig. 6-7. If we compare between the director plane (φ = 0º) and
the out of the director plane (φ = 90º), the variation of the total effective retardation in the
director plane is bigger than that of the out of the director plane. This is because the
variation of the thickness ratio ( dd /' ) in the director plane [Fig. 6-7 (a)] is larger than
that of the out of the director plane [Fig. 6-7 (b)]. With the similar reason, all curve
shapes of the total effective retardations out of the director plane are symmetric.
In a given tilt angle (θ ) of the liquid crystal director, as the effective phase
retardation of the liquid crystal layer ( LCeffndΔ ) increases, the magnitude of the off-axis
total effective retardation ( TotaleffndΔ ) also increases. On the other hand, with the constant
effective phase retardation of the liquid crystal layer, the magnitude of the off-axis total
effective retardation increases as the tilt angle of the liquid crystal director increases.
These important facts cause the light leakage at the off-axis angles.
Secondly, we calculate the off-axis light transmittances of the compensation
systems with the ideal crossed polarizers. The transmission axis of the bottom polarizer
makes -45º angles with the easy axis of the liquid crystal layer. Figures 6-18, 6-19, and 6-
282
20 are the numerical calculation results of the transmittances for the light wavelength of
550 nm. We used 2×2 matrix method for the calculations.
In these results, the effects of the light leakage coming from the geometrical
reason of the crossed polarizers are added to the off-axis transmittances in the director
(φ =0º) and out of the director (φ =90º) planes, as described in Sec. 3-2-3. With the
similar geometrical reason, the apparent relative angle between the transmission axis of
the polarizer and the director of the liquid crystal layer varies as the viewing polar angle
( oα ) increases in the viewing azimuth angles (φ ) of 45º and 90º (out of the director
plane). These additional effects make the analyses of the off-axis transmittance, based on
the birefringence analyses, complicated. As an example, if we compare Fig. 6-16(b) with
Fig. 6-19(b) and Fig. 6-17(b) with Fig. 6-20(b), which have the same set of the effective
retardation of the liquid crystal layer ( LCeffndΔ ), the transmittance out of the director plane
is not the same order as that of the total effective phase retardation ( TotaleffndΔ ) at the off-
axis angles, i.e. the transmittance out of the director plane does not exactly follow the
total effective phase retardation. We think this phenomenon in the viewing direction out
of the director plane is coming from the geometrical reasons.
In other viewing directions, as expected, the off-axis light transmittances agree
well with the analyses of the total effective phase retardation ( TotaleffndΔ ). In a given
director tilt angle ( θ ), the off-axis light transmittance increases as the effective
retardation of the liquid crystal layer ( LCeffndΔ ) at the normal direction increases. These
results mean that as the residual birefringence increases with the constant director
283
configuration, the dark state light leakage of a LCD increases. In a similar way, as the tilt
angle of the director increases with the constant phase retardation of the liquid crystal
layer, the off-axis light leakage rises.
These facts contradict each other. According to the results, if we want to have low
light leakage in a dark state, the effective phase retardation of the liquid crystal layer at
the normal direction should be reduced regardless of the director tilt angle. In addition,
the director tilt angle of the liquid crystal layer should be decreased for better phase
compensation. However, if the director tilt angle decreases, it makes the liquid crystal
layer have higher phase retardation. Therefore, the effective phase retardation of the
liquid crystal layer competes with the director tilt angle for the best performance of the
phase compensation of the liquid crystal layer.
In order to make this issue clear, we calculate the off-axis light transmittances as a
function of the thickness and the director tilt angle of the liquid crystal layer with the
light wavelength of 550 nm. In this calculation, we compensate the liquid crystal layer
perfectly at the normal direction at each thickness and tilt angle, as the previous
calculation. We set the light incident angle ( oα ) as -70º because this direction is the most
sensitive to the transmittance, as in Figs.6-18~20. Figure 6-21 shows the results. In this
figure also, the geometrical polarizer effects are in the director (φ =0º) and out of the
director (φ =90º) planes, and the geometrical director angle effects are out of the director
plane ( φ =90º) and the diagonal plane ( φ =45º). The fluctuation of the off-axis
transmittance out of the director plane in Fig. 6-21 (b) is coming from the same effects.
284
This calculation result confirms that as the cell thickness with a given liquid
crystal material (or effective phase retardation) decreases, the light leakage of a dark state
LCD decreases independent of the director configuration. In a given cell thickness, the
light leakage is the function of the director tilt angle of the liquid crystal layer. In this
calculation, as the tilt angle approaches both sides, θ =0º and θ =90º, the light leakage
falls in all conditions of the cell thickness. This is the result of the competition between
the effective phase retardation and the director tilt angle for the best phase compensation
of the liquid crystal layer, as described above.
285
Table 6-1. Thicknesses of the uniformly aligned perfect compensation systems (at the
normal direction). (unit: μm; LC: liquid crystal; NC: negative C-type compensator;
λ =550 nm).
θ =0º ( dd /' =1.0) θ =30º ( dd /' =0.9281) θ =60º ( dd /' =0.7998)LCeffndΔ
(nm) LC NC LC NC LC NC 50 0.3175 0.3175 0.4405 0.4088 1.4225 1.1377 100 0.6351 0.6351 0.8809 0.8176 2.8451 2.2754 200 1.2701 1.2701 1.7618 1.6351 5.6901 4.5509 300 1.9052 1.9052 2.6428 2.4527 8.5352 6.8263
286
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-15. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =0º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
287
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(c) φ =45º
288
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-16. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =30º) that
are perfectly compensated at the normal direction for the light wavelength of 550 nm.
289
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
-30
-20
-10
0
10
20
30
40 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(c) φ =45º
290
-80 -60 -40 -20 0 20 40 60 80
-50
0
50
100
150
200
250 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-17. Total phase retardation ( TotaleffndΔ ) of the compensation systems (θ =60º) that
are perfectly compensated at the normal direction for the light wavelength of 550 nm.
291
-80 -60 -40 -20 0 20 40 60 80
-50
0
50
100
150
200
250 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
-50
0
50
100
150
200
250 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tota
l Effe
ctiv
e Δn
d (n
m)
(c) φ =45º
292
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-18. Off-axis light transmittances of the compensation systems (θ =0º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
293
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) φ =45º
294
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-19. Off-axis light transmittances of the compensation systems (θ =30º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
295
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.01
0.02
0.03
0.04
0.05 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) φ =45º
296
-80 -60 -40 -20 0 20 40 60 80
0.00
0.05
0.10
0.15
0.20
0.25
0.30 Δnd
effLC=50 nm
Δndeff
LC=100 nm Δnd
effLC=200 nm
Δndeff
LC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(a) φ =0º (director plane)
(b) and (c): next page
Fig. 6-20. Off-axis light transmittances of the compensation systems (θ =60º) that are
perfectly compensated at the normal direction for the light wavelength of 550 nm.
297
-80 -60 -40 -20 0 20 40 60 80
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Δndeff
LC=50 nm Δnd
effLC=100 nm
Δndeff
LC=200 nm Δnd
effLC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(b) φ =90º (out of the director plane)
-80 -60 -40 -20 0 20 40 60 80
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Δndeff
LC=50 nm Δnd
effLC=100 nm
Δndeff
LC=200 nm Δnd
effLC=300 nm
Light incident angle, αo (deg.)
Tran
smitt
ance
(c) φ =45º
298
(a) Viewing direction: φ =0º (director plane), oα =-70º
(b) and (c): next page
Fig. 6-21. Off-axis light transmittances in functions of the thickness and the director tilt
angle of the liquid crystal layer in the compensation systems that are perfectly
compensated at the normal direction for the light wavelength of 550 nm.
299
(b) Viewing direction: φ =90º (out of the director plane), oα =-70º
(c) Viewing direction: φ =45º, oα =-70º
300
6-4-2. Non-uniform director configurations
In Chapter 4, Sec. 4-2, we numerically calculated the off-axis light transmission
properties of Pi-cells, whose dark states are optically optimized at the normal direction by
using hybrid aligned negative C-plates. Their bright states are implemented by using
different applied voltages and different pretilt angles with all the same dark voltage. In
this section, we will investigate their dark state transmittances in the same situations,
based on our knowledge in this chapter. The detailed information related to the phase
compensation, the cell parameters, and the structure is in Sec. 4-2.
6-4-2-1. A Pi-cell that has different bright state voltages (thickness effects)
Figure 4-2 is the numerical calculation results of the off-axis light transmission
properties of the dark state Pi-cells, whose bright states are optimized to have several
different voltages, but they have the same dark state voltage of 5.0 V and pretilt angle of
5.5º. Therefore, their dark states have the same director configuration, as in Fig. 4-3, but
their cell thicknesses (or effective phase retardation at the normal direction) are different,
as in Table 4-1. Figure 4-2 shows that as the bright state voltage (or effective phase
retardation at the normal direction) increases, the off-axis light transmittance of the dark
state increases. This result is in good agreement with the results in Sec. 6-4-1 where we
considered uniformly aligned director configurations. Consequently, we can say that as
the effective phase retardation of the liquid crystal layer increases, the perfect phase
compensation becomes harder, and the dark state light leakage increases at the off-axis
301
viewing directions. This is true in both uniformly and non-uniformly aligned liquid
crystal displays.
6-4-2-2. A Pi-cell that has different pretilt angles (thickness and director tilt angle
effects)
We did similar calculations for the Pi-cells that have different pretilt angles, but
they have the same dark and bright states voltages of 5.0 V and 1.3 V, respectively. In
addition to the results in Sec. 4-2, we calculated two more pretilt angles, 40.0º and 50.0º
for more investigation. The detailed cell parameters are in Table 6-2. Figures 6-22 and 6-
23 show the results of the off-axis light transmission properties with the light wavelength
of 550 nm and the director configurations of the liquid crystal layers in their dark states,
respectively. In this situation, as the pretilt angle increases, the director tilt-angle as well
as the cell thickness increases, as in Table 6-2 and Fig. 6-23. Figure 6-22 shows that as
the pretilt angle (or cell thickness) increases, the off-axis light transmittance of the dark
state increases very much. In other words, as the pretilt angle increases in a Pi-cell, the
average tilt angle and the cell thickness increase to meet the same transmittance of the
bright state at the normal direction, and this makes the phase compensation of the dark
state of the liquid crystal layer difficult at the off-axis viewing angles. This result also
agrees well with the results in Sec. 6-4-1.
302
Table 6-2. Cell parameters of the Pi-cells with different pretilt angles.
(Bright state voltage of 1.3 V, dark state voltage of 5.0 V).
Thickness (µm) Pretilt angle (°) LC Compensator 2.0 4.745 2.131 10.0 5.474 2.438 20.0 6.769 2.963 30.0 8.785 3.752 40.0 12.150 5.036 50.0 18.366 7.371
303
-80 -60 -40 -20 0 20 40 60 80
0.00
0.05
0.10
0.15
0.20
0.25
Pretilt 2o
Pretilt 10o
Pretilt 20o
Pretilt 30o
Pretilt 40o
Pretilt 50o
Viewing angle, αo (deg.)
Tran
smitt
ance
(a) Director plane
-80 -60 -40 -20 0 20 40 60 80
0.00
0.05
0.10
0.15
0.20
0.25
Pretilt 2o
Pretilt 10o
Pretilt 20o
Pretilt 30o
Pretilt 40o
Pretilt 50o
Viewing angle, αo (deg.)
Tran
smitt
ance
(b) Out of the director plane
Fig. 6-22. Numerical calculation results of the off-axis light transmission properties of the
dark state Pi-cells that have different pretilt angles (2.0-50.0º).
(Bright state voltage: 1.3 V; dark state voltage: 5.0 V; liquid crystal: LC53).
304
0.0 0.2 0.4 0.6 0.8 1.00
20
40
60
80
100
120
140
160
180 Pretilt 2o Pretilt 30o
Pretilt 10o Pretilt 40o
Pretilt 20o Pretilt 50o
Normalized cell thickness
Tilt
angl
e (d
eg.)
Fig. 6-23. Director tilt angles of the dark state Pi-cells that have different pretilt angles
(2.0-50.0º). (Bright state voltage: 1.3 V; dark state voltage: 5.0V; liquid crystal: LC53).
305
6-5. Summary
We compared the calculation methods of the phase difference. At the normal
direction, the approximate method and the “complete” method are the same as expected.
As the light incident angle increases, the difference between them increases, but the
magnitudes of the differences are very small.
We analyzed the parameters for the perfect phase compensation of a uniaxial slab.
To understand the relationship between the parameters, we calculated the total effective
phase retardation of a compensation system. From this, we knew that only two variables
among 'en , 'on , and 'd are independent for the perfect compensation of a given uniaxial
slab.
We found that there is a constant thickness ratio ( dd /' ) for the perfect phase
compensation of a uniaxial slab. In a given director configuration of a uniaxial slab (θ ,
φ ), we can always compensate it completely in an incident light condition (α , λ ).
However, the magnitude of dd /' varies in functions of θ , φ , and α . Therefore, it
implies that it is almost impossible to compensate an anisotropic system exactly for all
light wavelengths and viewing angles, simultaneously. If the anisotropic system is
composed of the layers that have different director configurations, the perfect
compensating the system becomes much more difficult. This means we need to adjust the
thickness of each sub-layer of the compensator, separately, with considering the director
configuration of each sub-layer of the uniaxial slab for better phase compensation.
The light wavelength dependence on the thickness ratio is relatively small if the
wavelength dispersion of the uniaxial slab is not so big. However, the thickness ratio is
306
very sensitive to the director configuration and the light incident angle. We confirmed
these facts by calculating the off-axis transmittance of the compensation systems.
As the first application, we considered a uniformly aligned liquid crystal layer
that had different tilt angles and effective phase retardations. We calculated the off-axis
phase retardation of the compensation system, which was perfectly compensated at the
normal direction. According to the results, in a given tilt angle of the liquid crystal
director, as the effective phase retardation of the liquid crystal layer increases, the
magnitude of the off-axis total effective retardation also increases. On the other hand,
with the constant effective phase retardation of the liquid crystal layer, the magnitude of
the total effective retardation increases as the tilt angle of the liquid crystal director
increases. These results mean that as the residual birefringence increases with the
constant director configuration, the dark state light leakage of a LCD increases. In a
similar way, as the tilt angle of the director increases with the constant phase retardation
of the liquid crystal layer, the off-axis light leakage rises.
Another important fact we found is that the effective phase retardation of the
liquid crystal layer competes with the director tilt angle for the best compensation of the
liquid crystal layer. To make this fact clear, we calculated the off-axis light
transmittances in functions of the thickness and the director tilt angle of the liquid crystal
layer. These calculations confirm that as the cell thickness decreases with a given liquid
crystal material, the light leakage of a dark state LCD decreases independent of the
director configuration. In a given cell thickness, the light leakage is the function of the
307
director tilt angle of the liquid crystal layer. As the tilt angle approaches both sides, θ =0º
and θ =90º, the light leakage falls in all conditions of the cell thickness.
As the second application, we considered a non-uniformly aligned Pi-cell that has
several different bright state voltages and pretilt angles. According to the results, as the
cell thickness (or effective phase retardation at the normal direction) or the director tilt
angle increase, the phase compensation of the dark state becomes harder, and this makes
the off-axis light leakage increases in the dark state. These results are consistent with the
analyses of the uniformly aligned compensation system.
308
CHAPTER 7
CONCLUSIONS
7-1. Analysis of the multi-reflection effects in compensated liquid crystal devices
We calculated the multi-reflection effects analytically and numerically in the dark
state of a compensated liquid crystal device and compared the results with the measured
transmittance of an example device. According to our analysis, there are two types of
interference in devices with significant residual retardation in the dark state that is
compensated by a passive optical retarder. The first one is due to the pure e-ray and pure
o-ray by themselves, and the second one is coming from the coupling between the e-
mode and o-mode. The first type has higher frequency in the wavelength space and is
related to the optical path length of the e-ray and o-ray and is independent of their
difference. Most of the modes used in liquid crystal devices have this type of interference.
The second type of interference has lower frequency than that of the first one and
depends on the residual birefringence of dark state. So, as the residual birefringence
increases, the second type of interference becomes more significant. In the viewpoint of
the optical properties of a liquid crystal device, the first type of interference could affect
the dark level and extinction ratio almost equivalently for visible wavelength region. On
the other hand, the second type could cause a wavelength dependence of the extinction
ratio, or a color shift of the dark state because of the lower frequency pattern of the
interference in the wavelength space.
309
7-2. Universal off-axis light transmission properties of the bright state in
compensated liquid crystal devices
We calculated the off-axis light transmission properties of the bright state of most
common liquid crystal devices, such as ECB, VA, TN, Pi-cell, and symmetric splay-cell,
whose dark states were optically compensated to have minimum transmittances for all
viewing angles. From the results of these calculations, we found that their bright states
have a universal viewing angle shape in spite of completely different director structures
in their liquid crystal layers.
In order to understand this strange phenomenon, we made simple dark and bright
state models describing general liquid crystal devices and analyzed them in terms of
effective retardation and transmittance. In accordance with these analyses, the total
effective retardation in the director plane constantly falls as the viewing angle increases
(“bell shape”). On the contrary, the total effective retardation out of the director plane
consistently rises in the same situation (“reversed bell shape”). These retardation
changes cause the transmittance changes. In the director plane, the transmittance
decreases as the viewing angle becomes larger because the birefringence decreases in that
direction. On the other hand, the transmittance out of the director plane increases first and
then falls after the specific viewing angle if the liquid crystal layer is optically designed
so that the transmittance of the normal direction is lower than the maximum value.
These viewing angle features of our bright state models agree well with the
properties of most common liquid crystal devices, not only in the two main viewing
planes, but also for all viewing directions. Therefore, we can say that our simple model
310
can reasonably describe the optical properties of the real liquid crystal devices considered
here. Accordingly, our simple model can be used to analytically understand and predict
the optical properties, such as transmittance, luminance distribution and color analyses of
current LCDs or possible candidates of new display modes because usually analytical
methods for optical calculations are almost impossible in real devices.
Based on these results, we can say that the single domain LCD modes, considered
here, whose dark states are optically compensated to give minimum transmittance,
inevitably have asymmetric shapes of the off-axis light transmission properties between
the director and out of the director planes in their bright states. Therefore, in order to
achieve isotropic shapes of the bright state viewing angle properties, multi-domain liquid
crystal modes are necessary.
7-3. Optimization of the bright state director configuration in compensated pi-cell
devices
We calculated the off-axis light transmission properties of the differently director
configured bright states in Pi-cells whose dark states were optically compensated
perfectly to have minimum transmittances in all viewing angles. From the results of these
calculations, we found that the off-axis light transmission properties of the bright states
surprisingly had unified shapes and were relatively insensitive to the variation of the
parameters considered when the voltage of the bright state or the pretilt angle were below
some particular value. On the other hand, the transmittance in highly tilted director
311
configuration of bright state out of the director plane varies more significantly and makes
the viewing angle properties of a Pi-cell worse.
In order to understand these curious phenomena, we made a new dark and bright
state model describing Pi-cells and analyzed it in terms of effective birefringence,
transmittance and angular distribution of directors. According to the analyses, the total
effective birefringence in the director plane constantly falls as the viewing angle
increases (“bell shape”) regardless of the director configuration, and this result is
consistent with transmittance analyses and describes the optical properties of real Pi-cells
well.
Out of the director plane, we cannot define the total birefringence because the
projection of the directors on the plane perpendicular to the light propagation vector ( Kr
)
are not co-linear, so we analyzed transmittance in this viewing plane. In the director
configuration of low tilt angle in bright state model, the transmittance rises first and then
falls after the specific viewing angle if the liquid crystal layer is optically designed so that
the transmittance of the normal direction is lower than the maximum value. On the other
hand, the transmittance with a high tilt director configuration fluctuates as the viewing
angle varies as in real Pi-cells. To understand the causes of the fluctuation of the total
transmittance, we calculated the transmittances, which are contributed separately from
positive A-plates and negative C-plates of the bright state model, and analyzed them
using the apparent angle distribution of directors in incident light coordinate system and
the birefringence of each layer. In accordance with the analyses, both the apparent angle
312
distribution of directors and the birefringence of each layer play a key role in the
fluctuation of the total transmittance.
7-4. Luminance and color properties of the compensated liquid crystal devices in
their bright states
We numerically calculated and analyzed the off-axis luminous transmittance (Y )
and color properties of the bright state model with the common LCDs (ECB, VA, Pi-cell,
and TN). Their cell thicknesses are adjusted to have four different Y (%) values at the
normal direction: 100, 90, 80, and 70 %. The dark states of them are almost perfectly
compensated by using the hybrid aligned negative C-plates.
7-4-1. Luminous transmittance of the bright state LCDs
According to the results, in both the bright state simple model and the real LCDs,
the luminous transmittances (Y ) always decrease as the viewing angle increases in the
director plane. On the other hand out of the director plane, the luminous transmittances
(Y ) increase first and then fall after specific viewing angles when the cell is optically
designed so that Y (%) value is less than 100% at the normal direction. These results are
in good agreement with the birefringence analyses.
We also calculated the luminous transmittances (Y ) in all viewing directions, and
compared them by using conoscopic figures. All the figures look like similar. It confirms
again that if the dark states of LCDs are optically compensated perfectly, there is
universality in the optical properties of the bright state regardless of the display modes.
313
Another important thing is that all conoscopic figures are anisotropic shapes. This
means that the single domain LCD modes, considered here, inevitably have asymmetric
shapes of the luminance between the director and out of the director planes in their bright
states. Therefore, we can say that in order to achieve isotropic shapes of the bright state
viewing angle properties, multi-domain methods or asymmetric distribution of backlight
intensity is mandatory.
7-4-2. Color properties of the bright state LCDs
Firstly, we investigated the thickness (or phase retardation) effects on the color
properties of LCDs. These calculations show that as the thickness (or phase retardation)
increases from zero, the chromaticity coordinates ( 'u , 'v ) initially goes from blue to
yellow colors and then revolve around the point of the light source color. We analyzed
this fact by using the transmittances and the phase retardations of the three major colors.
Secondly, we calculated the off-axis chromaticity coordinates of the bright state
model and real LCDs such as ECB, VA, PI-cell, and TN modes. Overall figure shapes of
the real LCDs are in good agreement with that of our simple bright state model except the
TN mode. The absolute chromaticity coordinates at the normal direction move from blue
to yellow as Y (%) value increases in all display modes. As Y (%) value increases, the
total variation of the chromaticity coordinates increases in both main viewing planes, and
the magnitudes of the variations out of the director plane are bigger than that of the
director plane. As the viewing angle increases, the color coordinates of the bright state
314
move to the yellow direction first in the viewing angles out of the director plane, but the
color shifts to the blue direction first in the director plane.
In the TN mode, the magnitude of the color variation out of the director plane
(refer to the mid-layer of directors) is the similar level with that of other display modes.
However, in the director plane, the color variation is much bigger than that of other
display modes, and the variation becomes large as Y (%) value decreases. This is the
opposite direction compared with our bright state model and other display modes
considered here.
Finally, we investigate the viewing angle dependence of the color difference
( uvE *Δ ) in the bright state model and the real LCDs. When we compare the results, the
color differences of the real LCDs have the same curve shapes as that of the bright state
model except the TN mode. In the viewing directions out of the director plane, as the
Y (%) value increases, the off-axis color difference ( uvE *Δ ) increases, and this is coming
from the facts that the off-axis transmittances of the red, green, and blue are different
more and more as the Y (%) value increases. In the viewing angles of the director plane,
the off-axis color difference ( uvE *Δ ) increases as the Y (%) value increases, but the
variation is very small. This result is coming from the properties of the blue, green, and
red transmittances.
In the Pi-cell, out of the director plane, the color difference ( uvE *Δ ) fluctuates at
large viewing angles when the cell is designed to have high transmittance at the normal
direction. We think this is one of the strong points of the Pi-cell because we can suppress
the color shift of the device better than other LCD modes at large viewing angles.
315
In the TN mode, the color difference ( uvE *Δ ) is relatively insensitive to the
Y (%) value. Especially, the color difference in the condition, Y (%)=100, is excellent
and even better than the others. This surprising feature in TN mode is related to the
Mauguin condition.
7-5. The conditions and the limitations of the perfect phase compensation in liquid
crystal displays
We compared the calculation methods of the phase difference. At the normal
direction, the approximate method and the “complete” method are the same as expected.
As the light incident angle increases, the difference between them increases, but the
magnitudes of the differences are very small.
We analyzed the parameters for the perfect phase compensation of a uniaxial slab.
To understand the relationship between the parameters, we calculated the total effective
phase retardation of a compensation system. From this, we knew that only two variables
among 'en , 'on , and 'd are independent for the perfect compensation of a given uniaxial
slab.
We found that there is a constant thickness ratio ( dd /' ) for the perfect phase
compensation of a uniaxial slab. In a given director configuration of a uniaxial slab (θ ,
φ ), we can always compensate it completely in an incident light condition (α , λ ).
However, the magnitude of dd /' varies in functions of θ , φ , and α . Therefore, it
implies that it is almost impossible to compensate an anisotropic system exactly for all
light wavelengths and viewing angles, simultaneously. If the anisotropic system is
316
composed of the layers that have different director configurations, the perfect
compensating the system becomes much more difficult. This means we need to adjust the
thickness of each sub-layer of the compensator, separately, with considering the director
configuration of each sub-layer of the uniaxial slab for better phase compensation.
The light wavelength dependence on the thickness ratio is relatively small if the
wavelength dispersion of the uniaxial slab is not so big. However, the thickness ratio is
very sensitive to the director configuration and the light incident angle. We confirmed
these facts by calculating the off-axis transmittance of the compensation systems.
As the first application, we considered a uniformly aligned liquid crystal layer
that had different tilt angles and effective phase retardations. We calculated the off-axis
phase retardation of the compensation system, which was perfectly compensated at the
normal direction. According to the results, in a given tilt angle of the liquid crystal
director, as the effective phase retardation of the liquid crystal layer increases, the
magnitude of the off-axis total effective retardation also increases. On the other hand,
with the constant effective phase retardation of the liquid crystal layer, the magnitude of
the total effective retardation increases as the tilt angle of the liquid crystal director
increases. These results mean that as the residual birefringence increases with the
constant director configuration, the dark state light leakage of a LCD increases. In a
similar way, as the tilt angle of the director increases with the constant phase retardation
of the liquid crystal layer, the off-axis light leakage rises.
Another important fact we found is that the effective phase retardation of the
liquid crystal layer competes with the director tilt angle for the best compensation of the
317
liquid crystal layer. To make clear this fact, we calculated the off-axis light
transmittances in functions of the thickness and the director tilt angle of the liquid crystal
layer. These calculations confirm that as the cell thickness decreases with a given liquid
crystal material, the light leakage of a dark state LCD decreases independent of the
director configuration. In a given cell thickness, the light leakage is the function of the
director tilt angle of the liquid crystal layer. As the tilt angle approaches both sides, θ =0º
and θ =90º, the light leakage falls in all conditions of the cell thickness.
As the second application, we considered a non-uniformly aligned Pi-cell that has
several different bright state voltages and pretilt angles. According to the results, as the
cell thickness (or effective phase retardation at the normal direction) or the director tilt
angle increase, the phase compensation of the dark state becomes harder, and this makes
the off-axis light leakage increases in the dark state. These results are consistent with the
analyses of the uniformly aligned compensation system.
318
APPENDIX A
CALCULATION OF THE 1-DIMENSIONAL LIQUID CRYSTAL DIRECTOR
CONFIGURATION BY THE VECTOR METHOD BASED ON THE
RELAXATION TECHNIQUE
A-1. Calculations
Let us consider a uniaxial liquid crystal cell that has a pair of parallel substrates,
up and down of liquid crystal material. We assume the strong anchoring at the liquid
crystal-substrate interfaces, and apply external voltage (V) across the cell (thickness of
d ) along the z -direction in Cartesian coordinates. The liquid crystal director ( nr )
responds to the applied electric field ( Er
), which is uniform in x , y -directions, and
expressed as follows:
zyx nznynxzn ˆˆˆ)( ++=r , (A-1)
Where x , y , and z are the unit vectors of the directions, x , y , and z , and xn , yn , zn
are the x , y , z components of the director ( nr ), respectively. The director ( nr ) is a
function of z , but independent of x , y -directions.
The elastic energy density or it is called the Flank-Osceen elastic energy density
( OFf . ) of this liquid crystal cell is expressed as,
319
( ) ( ) ( )233
222
211. 2
121
21 nnKqnnKnKf oOF
rrrrr×∇×++×∇•+•∇= , (A-2)
where 11K , 22K , and 33K are the elastic constants of the splay, twist, and bend
deformations of the liquid crystal director, respectively, and oq is written as,
pqo
π2= , (A-3)
where p is the natural twisted pitch of the liquid crystal material and has a positive
(negative) sign for the right (left) handed twisted structure.
The electric energy density ( ef ) of this liquid crystal cell is expressed as follows:
EDfe
rr•=
21 , (A-4)
where Dr
is the electric displacement. With in mind that the applied electric field ( Er
) has
the only z -component ( zE ), Eq. (A-4) can be written as,
])([21
2||
2
zo
ze n
Df
⊥⊥ −+=
εεεε, (A-5)
320
where oε is the electric permittivity in free space, and ||ε and ⊥ε are the parallel and
perpendicular components of the dielectric constant of the liquid crystal material,
respectively. zD is the z -component of the electric displacement vector ( Dr
) and is
constant in the liquid crystal medium from the Maxwell equation,
0=•∇ Dr
. (A-6)
Let us say the liquid crystal layer is divided into N sub layers, and the liquid crystal
molecules are aligned uniformly within each layer. Then, zD can be calculated from this
formula,
∑= ⊥⊥ −+
= N
iiz
oz
ndVN
D
12
|| )(1
1
εεε
ε, (A-7)
where izn is the z -component of the liquid crystal director ( nr ) of the i th sub layer.
The total free energy density ( f ) of this liquid crystal cell consists of the elastic
energy ( OFf . ) and the electric energy ( ef ), and can be written from Eqs. (A-1), (A-2),
and (A-5) as follows at a constant charge (Q ) condition:
2
1121
⎟⎠⎞
⎜⎝⎛=
dzdnKf z
321
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
+⎟⎠
⎞⎜⎝
⎛+
⎟⎟⎠
⎞⎜⎜⎝
⎛−⎟
⎠
⎞⎜⎝
⎛⎟⎟⎠
⎞⎜⎜⎝
⎛−⎟
⎠
⎞⎜⎝
⎛+⎟⎟⎠
⎞⎜⎜⎝
⎛
+2
22
22
22
2
22
21
ox
yo
yxo
xyyx
xy
yx
qdz
dnnq
dzdn
nqdz
dndz
dnnn
dzdn
ndz
dnn
K
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛⎟⎠
⎞⎜⎝
⎛+⎟⎠
⎞⎜⎝
⎛+⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟
⎠
⎞⎜⎝
⎛+2
22
22
22
233 2
21
dzdn
ndz
dndz
dnnn
dzdn
ndz
dnn
dzdn
nK yy
yxyx
xx
yz
xz
])([21
2||
2
zo
z
nD
⊥⊥ −++
εεεε. (A-8)
The total free energy ( F ) of the liquid crystal cell will be minimized at an
equilibrium state. We can achieve this equilibrium state via the dynamic equations,
⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂
=∂∂
−zxx
x
nf
dzd
nf
tn
,
γ , (A-9)
⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂
=∂
∂−
zyy
y
nf
dzd
nf
tn
,
γ , (A-10)
⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂
=∂∂
−zzz
z
nf
dzd
nf
tn
,
γ , (A-11)
where γ is the rotational viscosity of the liquid crystal. If we set the time variation small
enough, Eqs. (A-9), (A-10), and (A-11) can be written as follows with Eq. (A-8):
322
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂Δ
−=Δzxx
x nf
dzd
nftn
,γ
( ) ( )⎥⎥⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
⎟⎟⎠
⎞⎜⎜⎝
⎛−+⎟⎟
⎠
⎞⎜⎜⎝
⎛+++
⎟⎠⎞
⎜⎝⎛⎟⎠⎞
⎜⎝⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛⎟⎠⎞
⎜⎝⎛+
⎟⎟⎠
⎞⎜⎜⎝
⎛−−⎟
⎠⎞
⎜⎝⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛
Δ=
2
2
22332
22
332
222
33
3322
2
3322
2
3322
22
)2(2
dznd
nnKKdz
ndnKnKnK
dzdn
dzdn
nKdz
dndz
dnnK
dzdn
nKKdz
dnnK
dzdn
qK
t
yyx
xzyx
zxz
yxy
yx
xx
yo
γ, (A-12)
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂Δ
−=Δzyy
y nf
dzd
nftn
,γ
( ) ( )⎥⎥⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
⎟⎟⎠
⎞⎜⎜⎝
⎛−+⎟
⎟⎠
⎞⎜⎜⎝
⎛+++
⎟⎠⎞
⎜⎝⎛⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛⎟⎠⎞
⎜⎝⎛+
⎟⎠⎞
⎜⎝⎛−−⎟⎟
⎠
⎞⎜⎜⎝
⎛+⎟
⎠⎞
⎜⎝⎛−
Δ=
2
2
22332
22
332
332
22
3322
2
3322
2
3322
22
)2(2
dznd
nnKKdz
ndnKnKnK
dzdn
dzdn
nKdz
dndz
dnnK
dzdn
nKKdz
dnnK
dzdn
qK
t
xyx
yzyx
zyz
yxx
xy
yy
xo
γ, (A-13)
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
−∂∂Δ
−=Δzzz
z nf
dzd
nftn
,γ
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
−+
−+
⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛−⎟
⎠⎞
⎜⎝⎛−
Δ=
⊥⊥
⊥
22||
||2
2
2
11
2
33
2
33
])([
)(
zo
zz
zyz
xz
n
nD
dzndK
dzdn
nKdz
dnnK
t
εεεε
εεγ. (A-14)
323
From these equations, we can calculate the new state-director configuration ( Newjn ) after
time interval, tΔ , from the initial director configuration ( Oldjn ) as follows:
jOldj
Newj nnn Δ+= , (A-15)
with the boundary condition of,
( ) ( ) ( )222 Newz
Newy
Newx
NewjNew
j
nnn
nn
++= , (A-16)
where the subscript, j , represents x , y , and z . By repeating this relaxation process, we
can achieve the director configuration of each sub layer at the equilibrium state. When we
do the calculation under the constant voltage condition, we need recalculate the zD at
each relaxation process by using the given applied voltage and the director configuration
at the instant.
324
A-2. Program source codes (Matlab)
function LcAngle = LcConfigCalVectorFcn(DirectorStruct, InitAngle) %LCCONFIGCALVECTORFCN Calculate LC director configuration by using the VECTOR method %LcAngle: Director configuration return values, format: [Thickness, theta_xy, phi], micro, degree units %DirectorStruct: Structure that includes the parameters needed for the director calculation %DirectorStruct.[K11, K22, K33, EpsPar, EpsPer, Pitch, Thickness, TopRubbing, BotRubbing, TopPretilt, BotPretilt, Twist,... % DirectorTypeRadioButton, Volt, PeripheralSumDoverEps, CalculationMethodRadioButton,... % ConstantModeRadioButton, InitialDirectorRadioButton, LayerNum, MaxIterationNum, StoppingCondition] %InitAngle: Initial director angles, format: [Thickness, theta_xy, phi], micro, degree units, Same format with the "LcAngle" % if InitAngle==[]: Use program default values as an Initial angles to calculate director configuration %theta: LC director polar angle from the Z-axis in RADIAN unit (0 deg.: Homeotropic, 90deg.:In plane Alignment) DURING THE CALCULATION %theta_xy_deg : LC director Tilt angle from the X-Y plane in DEGREE unit (0 deg.: In plane, 90deg.:Homeotropic Alignment) AFTER CALCULATION <= RETURN VALUE %phi : LC azimuthal angle from the X-axis in RADIAN unit <= DURING THE CALCULATION %phi_deg : LC azimuthal angle from the X-axis in DEGREE unit <= RETURN VALUE %Units: MKSA %December 22, 2005 %Yong-Kyu Jang (Liquid Crystal Institute, Kent State University) %Common Parameters eps_0 = 8.8542 * 1e-12; %Permittivity in vacuum (MKSA unit) %LC Parameters K11 = DirectorStruct.K11 * 1e-12; %Splay elastic constant(N) K22 = DirectorStruct.K22 * 1e-12; %Twist elastic constant(N) K33 = DirectorStruct.K33 * 1e-12; %Bend elastic constant(N) eps_par = DirectorStruct.EpsPar; %Dielectric constant (Parallel component to director) eps_per = DirectorStruct.EpsPer; %Dielectric constant (Perpendicular component to director) %Cell Parameters d = DirectorStruct.Thickness * 1e-6; %Cell thickness(m), Positive value (>0) if isempty(DirectorStruct.Pitch) %Infinite Pitch (No Chirality) d_over_p = 0; else
325
d_over_p = d/(DirectorStruct.Pitch * 1e-6); %d/p ("+" sign: Right handed twist, "-" sign: Left handed twist) end top_pretilt_angle = DirectorStruct.TopPretilt * pi/180; %LC pretilt angle from the X-Y plane, Positive value (0 deg.: in plane) bot_pretilt_angle = DirectorStruct.BotPretilt * pi/180;%LC pretilt angle from the X-Y plane, Positive value (0 deg.: in plane) top_rub_angle = DirectorStruct.TopRubbing * pi/180; %Top glass(C/F) rubbing angle from X-axis in assembled LCD bot_rub_angle = DirectorStruct.BotRubbing * pi/180; %Bottom glass(TFT) rubbing angle from X-axis in assembled LCD twist_angle = DirectorStruct.Twist * pi/180; %LC total twist angle in MANUAL mode (+:Right handed twist, -: Left handed twist) V = DirectorStruct.Volt; %LC Voltage peripheral_d_eps = DirectorStruct.PeripheralSumDoverEps;%Sum of the Thickness over Dielectric constant(d/eps) of the adjacent layers of LC that cause LC Voltage drop %Simulation Parameters N = DirectorStruct.LayerNum; %Number of the LC sub layers if DirectorStruct.InitialDirectorRadioButton == 1 twist_mode = 0; %0: AUTO mode to find initial twist angle else if DirectorStruct.DirectorTypeRadioButton == 1 twist_mode = 1; %1: SPLAY/TWIST structure MANUAL mode (Use given twist angle) else twist_mode = 2; %2(else): BEND structure MANUAL mode (Use given twist angle) end end constant_mode = DirectorStruct.ConstantModeRadioButton; %1: Constant VOLTAGE (V) mode, 2(else): Constant CHARGE (Q) mode creep_up = 1.02; %1.02 max_ratio = 1.04; %1.02; max_iteration_num = DirectorStruct.MaxIterationNum; %If iteration number (m) is bigger than this value, relaxation process stop avg_delta_f_limit = DirectorStruct.StoppingCondition; %10^-10 ~ 10^-7 %If average-delta-free energy is less than this value, relaxation process stop %LC director-initialization Parameters theta_0 = 5 * pi/180; %Relaxation starting value of the theta(from X-Y plane) in the mid layer(z=d/2) when pretilt angle = 0 deg %Parameter recalculation & initialziation delta_eps = eps_par - eps_per; %Difference of the dielectric constants delta_d = d/N; %Thickness of the sub layers real_volt = V; %Used to optimize delta_n_limit phi_nat = 2*pi*d_over_p; %Natural twist angle due to the chiral dopant
326
top_pretilt_angle = abs(top_pretilt_angle); %LC pretilt angle from the X-Y plane, Positive value (0 deg.: in plane) bot_pretilt_angle = abs(bot_pretilt_angle); %LC pretilt angle from the X-Y plane, Positive value (0 deg.: in plane) top_rub_angle = rem(top_rub_angle, 2*pi); %Rubbing angle rearrangement: -2*pi < top_rub_angle < +2*pi bot_rub_angle = rem(bot_rub_angle, 2*pi); %Rubbing angle rearrangement: -2*pi < bot_rub_angle < +2*pi q = 2 * pi * d_over_p / d; %2¥ð/P <- Natural pitch, "+" sign: Right handed twist, "-" sign: Left handed twist z = linspace(0, d, N); %Position of the cell, <- Array n = 1 : N; %Array of the LC sub layers d_theta_d_z = zeros(1, N); %1'st derivative of theta to z (d¥È/dz) d2_theta_d_z2 = zeros(1, N); %2'nd derivative of theta to z (d2¥È/dz2) d_phi_d_z = zeros(1, N); %1'st derivative of phi to z (d¥Õ/dz) d2_phi_d_z2 = zeros(1, N); %2'nd derivative of phi to z (d2¥Õ/dz2) dnxdz = zeros(1, N); %1'st derivative of nx to z (dnx/dz) dnydz = zeros(1, N); %1'st derivative of ny to z (dny/dz) dnzdz = zeros(1, N); %1'st derivative of nz to z (dnz/dz) d2nxdz2 = zeros(1, N); %2'st derivative of nx to z (d2nx/dz2) d2nydz2 = zeros(1, N); %2'st derivative of ny to z (d2ny/dz2) d2nzdz2 = zeros(1, N); %2'st derivative of nz to z (d2nz/dz2) delta_t_over_gamma = 0.001; %¥Ät/¥ã <- Determine the period of the relaxation Dz_temp = eps_0 * V; %Temporary Dz real_V_temp = d / eps_0 / N; %Temporary real voltage theta_init = zeros(6, N); %LC director's polar angle from the z-axis phi_init = zeros(6, N); %LC director's azimuthal angle from the X-axis F = zeros(1, 6); %Total free energy per unit area in the LCD cell (J) f_old = zeros(1, N); termination_flag = 0; %0: Not ready to stop relaxation, 1: ready to stop relaxation (avg_delta_f < avg_delta_f_limit) avg_delta_nx_old = 1e10; %Used for stability check avg_delta_ny_old = 1e10; %Used for stability check avg_delta_nz_old = 1e10; %Used for stability check max_delta_nx_old = 1e10; %Used for stability check max_delta_ny_old = 1e10; %Used for stability check max_delta_nz_old = 1e10; %Used for stability check edge_limit_m = 0; %Used to check stopping point old_edge_limit_m = 0; %Used to check stopping point edge_limit_num = 0; %Used to check stopping point if isempty(InitAngle) %if InitAngle==[]: Use program default values as an Initial angles to calculate director configuration %*************************************************************************% %Calculate the top rubbing angle in the coordinate system where the top sub layer's director angle is zero.
327
temp_top_rub_angle = rem((top_rub_angle - (bot_rub_angle + phi_nat)), 2*pi); if temp_top_rub_angle < 0 temp_top_rub_angle = 2*pi + temp_top_rub_angle; end %Calculate the initial twist angle, theta and phi, theta_init: LC polar angle from the Z-axis if twist_mode == 0 %AUTO mode to calculate initial twist angle if (0 <= temp_top_rub_angle) & (temp_top_rub_angle < pi) %Right handed twist(¥Õ>=0), Head to Head (top_pretilt_angle < 0) SPLAY twist_angle = phi_nat + temp_top_rub_angle; theta_init(1, n) = pi/2 - (bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); %Used to calculate elastic free energy at initial state phi_init(1, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Used to calculate free energy and initial director polar angles %Right handed twist(¥Õ>0), Head to Tail (top_pretilt_angle > 0) HOMOGENEOUS twist_angle = phi_nat + temp_top_rub_angle + pi; theta_init(2, n) = pi/2 - (bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(2, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Tail (top_pretilt_angle > 0) HOMOGENEOUS twist_angle = phi_nat + temp_top_rub_angle - pi; theta_init(3, n) = pi/2 - (bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(3, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Head (top_pretilt_angle < 0) SPLAY twist_angle = phi_nat + temp_top_rub_angle - 2*pi; theta_init(4, n) = pi/2 - (bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(4, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Right handed twist(¥Õ>=0), Head to Head (top_pretilt_angle > 0) BEND twist_angle = phi_nat + temp_top_rub_angle; theta_init(5, n) = pi/2 - (bot_pretilt_angle + (pi - top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(5, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Head (top_pretilt_angle > 0) BEND twist_angle = phi_nat + temp_top_rub_angle - 2*pi; theta_init(6, n) = pi/2 - (bot_pretilt_angle + (pi - top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(6, n) = bot_rub_angle + twist_angle * (n-1)/(N-1);
328
elseif (pi <= temp_top_rub_angle) & (temp_top_rub_angle < 2*pi) %Right handed twist(¥Õ>0), Head to Head (top_pretilt_angle < 0) SPLAY twist_angle = phi_nat + temp_top_rub_angle; theta_init(1, n) = pi/2 - (bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(1, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Right handed twist(¥Õ>=0), Head to Tail (top_pretilt_angle > 0) HOMOGENEOUS twist_angle = phi_nat + temp_top_rub_angle - pi; theta_init(2, n) = pi/2 - (bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(2, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Tail (top_pretilt_angle > 0) HOMOGENEOUS twist_angle = phi_nat + temp_top_rub_angle - 3*pi; theta_init(3, n) = pi/2 - (bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(3, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Head (top_pretilt_angle < 0) SPLAY twist_angle = phi_nat + temp_top_rub_angle - 2*pi; theta_init(4, n) = pi/2 - (bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(4, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Right handed twist(¥Õ>0), Head to Head (top_pretilt_angle > 0) BEND twist_angle = phi_nat + temp_top_rub_angle; theta_init(5, n) = pi/2 - (bot_pretilt_angle + (pi - top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(5, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); %Left handed twist(¥Õ<0), Head to Head (top_pretilt_angle > 0) BEND twist_angle = phi_nat + temp_top_rub_angle - 2*pi; theta_init(6, n) = pi/2 - (bot_pretilt_angle + (pi - top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); phi_init(6, n) = bot_rub_angle + twist_angle * (n-1)/(N-1); else beep errordlg('Input rubbing angles ERROR in auto mode', 'Rubbing Angles Error', 'on') LcAngle = []; return end
329
%Calculate elastic free energy to find optimum initial director configuration for p = 1 : 6 d_theta_d_z(2:N-1) = (theta_init(p, 3:N) - theta_init(p, 1:N-2)) / (2 * delta_d); %1'st derivative of the theta d2_theta_d_z2(2:N-1) = (theta_init(p, 3:N) + theta_init(p, 1:N-2) - 2*theta_init(p, 2:N-1)) / (delta_d)^2; %2nd derivative of the theta d_phi_d_z(2:N-1) = (phi_init(p, 3:N) - phi_init(p, 1:N-2)) / (2 * delta_d); %1'st derivative of the phi d2_phi_d_z2(2:N-1) = (phi_init(p, 3:N) + phi_init(p, 1:N-2) - 2*phi_init(p, 2:N-1)) / (delta_d)^2; %2nd derivative of the phi s_t2 = sin(theta_init(p, :)).^2; c_t2 = cos(theta_init(p, :)).^2; %Calculation of the Average free energy density in a layer f = 1/2 * (K11 * s_t2 .* d_theta_d_z .^2 + K22 * (s_t2 .^2 .* d_phi_d_z .^2 - 2 * q * s_t2 .* d_phi_d_z + q^2)... + K33 * (c_t2 .* d_theta_d_z .^2 + c_t2 .* s_t2 .* d_phi_d_z .^2)); %Array of the free energy density in each layers (J/m3) F(p) = sum(f) * delta_d; %Total free energy per unit area in the LCD cell (J/m2) end %Find minimum free energy and initialize the theta & phi by using the minimum free energy condition [min_F, min_index] = min(F); % Find minimum free energy and it's index if (min_index == 1) | (min_index == 4) %Head to Head, SPLAY directer structure theta = pi/2 - (theta_0 * sin(2*pi*(n-1)/(N-1)) + bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle)*(n-1)/(N-1)); %Polar angle of the LC director from the Z-axis elseif (min_index == 2) | (min_index == 3) %Head to Tail, HOMOGENEOUS directer structure theta = pi/2 - (theta_0 * sin(pi*(n-1)/(N-1)) + bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle)*(n-1)/(N-1)); %Polar angle of the LC director from the Z-axis else %Head to Head, BEND directer structure theta = theta_init(min_index, :); end phi = phi_init(min_index, :); %Azimuthal angle of the LC director clear theta_init phi_init F %*************************************************************************%
330
elseif twist_mode == 1 %SPLAY or TWIST structure MANUAL mode %Calculate the top rubbing angle in the coordinate system where the bottom rubbing angle is zero. temp_top_rub_angle = rem((top_rub_angle - bot_rub_angle), 2*pi); if temp_top_rub_angle < 0 temp_top_rub_angle = 2*pi + temp_top_rub_angle; end rem_twist_angle = rem(twist_angle, 2*pi); if abs(rem_twist_angle - temp_top_rub_angle)<1e-6 | abs(rem_twist_angle - (temp_top_rub_angle - 2*pi))<1e-6 theta = pi/2 - (theta_0 * sin(2*pi*(n-1)/(N-1)) + bot_pretilt_angle + (-top_pretilt_angle - bot_pretilt_angle)*(n-1)/(N-1)); elseif abs(rem_twist_angle - (temp_top_rub_angle + pi))<1e-6 | abs(rem_twist_angle - (temp_top_rub_angle - pi))<1e-6 | abs(rem_twist_angle - (temp_top_rub_angle - 3*pi))<1e-6 theta = pi/2 - (theta_0 * sin(pi*(n-1)/(N-1)) + bot_pretilt_angle + (top_pretilt_angle - bot_pretilt_angle)*(n-1)/(N-1)); else beep errordlg('Input rubbing angles ERROR in manual mode', 'Rubbing Angles Error', 'on') LcAngle = []; return end phi = bot_rub_angle + twist_angle * (n-1)/(N-1); %*************************************************************************% else %BEND structure MANUAL mode %Calculate the top rubbing angle in the coordinate system where the bottom rubbing angle is zero. temp_top_rub_angle = rem((top_rub_angle - bot_rub_angle), 2*pi); if temp_top_rub_angle < 0 temp_top_rub_angle = 2*pi + temp_top_rub_angle; end rem_twist_angle = rem(twist_angle, 2*pi); if abs(rem_twist_angle - temp_top_rub_angle)<1e-6 | abs(rem_twist_angle - (temp_top_rub_angle - 2*pi))<1e-6 theta = pi/2 - (bot_pretilt_angle + (pi - top_pretilt_angle - bot_pretilt_angle) * (n-1)/(N-1)); else beep errordlg('Input rubbing angles ERROR in manual mode', 'Rubbing Angles Error', 'on') LcAngle = []; return end phi = bot_rub_angle + twist_angle * (n-1)/(N-1); end %End of the initialization
331
%=========================================================================% else %if InitAngle ~= []: Use given values as an Initial angles to calculate director configuration theta = ((90 - InitAngle(:, 2)) * pi/180)'; phi = (InitAngle(:, 3) * pi/180)'; end %Change the initial conditions from theta, phi to nx, ny, nz format nx = sin(theta) .* cos(phi); ny = sin(theta) .* sin(phi); nz = cos(theta); %Used for strong anchoring assumption during the relaxation nx_bot = nx(1); ny_bot = ny(1); nz_bot = nz(1); nx_top = nx(N); ny_top = ny(N); nz_top = nz(N); %=========================================================================% %Start of the relaxation loop m = 1; %Number of the iterations while(1) if m == 1 %Calculation of the Dz eps_n = eps_per + delta_eps .* nz .^2; %z component of the epsilon (epsilon_zz) at each layer, <- Array Dz = Dz_temp / (peripheral_d_eps + d/N*sum(1 ./ eps_n)); %z component of the displacement vector D %Calculation of the nx, ny, nz derivatives dnxdz(2:N-1) = (nx(3:N) - nx(1:N-2)) / (2 * delta_d); %1'st derivative of the nx dnydz(2:N-1) = (ny(3:N) - ny(1:N-2)) / (2 * delta_d); %1'st derivative of the ny dnzdz(2:N-1) = (nz(3:N) - nz(1:N-2)) / (2 * delta_d); %1'st derivative of the nz d2nxdz2(2:N-1) = (nx(3:N) + nx(1:N-2) - 2*nx(2:N-1)) / (delta_d)^2; %2nd derivative of the nx d2nydz2(2:N-1) = (ny(3:N) + ny(1:N-2) - 2*ny(2:N-1)) / (delta_d)^2; %2nd derivative of the ny d2nzdz2(2:N-1) = (nz(3:N) + nz(1:N-2) - 2*nz(2:N-1)) / (delta_d)^2; %2nd derivative of the nz end %Calculation of the delta_nx, delta_ny and delta_nz
332
delta_nx = delta_t_over_gamma.*(2.*K22.*q.*dnydz+K33.*nx.*dnxdz.^2-(2.*K22-K33).*nx.*dnydz.^2+2.*K22.*ny.*dnxdz.*dnydz+2.*K33.*nz.*dnxdz.*dnzdz+(K33.*nx.^2+K22.*ny.^2+K33.*nz.^2).*d2nxdz2+(K33-K22).*nx.*ny.*d2nydz2); delta_ny = delta_t_over_gamma.*(-2.*K22.*q.*dnxdz+K33.*ny.*dnydz.^2-(2.*K22-K33).*ny.*dnxdz.^2+2.*K22.*nx.*dnxdz.*dnydz+2.*K33.*nz.*dnydz.*dnzdz+(K22.*nx.^2+K33.*ny.^2+K33.*nz.^2).*d2nydz2+(K33-K22).*nx.*ny.*d2nxdz2); delta_nz = delta_t_over_gamma.*(-K33.*nz.*dnxdz.^2-K33.*nz.*dnydz.^2+K11.*d2nzdz2+Dz.^2.*delta_eps.*nz./eps_0./(eps_per+delta_eps.*nz.^2).^2); %Calculation of the real LC-Voltage at Constant Charge mode if constant_mode == 2 real_volt = real_V_temp * Dz * sum(1 ./ eps_n); end %Definition of the limit values if real_volt < 1.5 avg_delta_n_limit = 0.001; max_delta_n_limit = 0.002; elseif real_volt >= 1.5 & real_volt < 2 avg_delta_n_limit = 0.002; max_delta_n_limit = 0.003; elseif real_volt >= 2 & real_volt < 3 avg_delta_n_limit = 0.002; max_delta_n_limit = 0.004; else avg_delta_n_limit = 0.005; max_delta_n_limit = 0.01; end %Stability check of the nx, ny, nz %If the maximum delta_nx, delta_ny, delta_nz are larger than the limit, then, reduce the delta_t/¥ã to the 1/2 of the current value. %Max delta_nx and ratio stability check max_delta_nx = max(abs(delta_nx)); max_delta_nx_ratio = max_delta_nx/max_delta_nx_old; while((max_delta_nx > max_delta_n_limit) | (max_delta_nx_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /2; delta_nx = delta_nx /2; delta_ny = delta_ny /2; delta_nz = delta_nz /2; max_delta_nx = max(abs(delta_nx)); max_delta_nx_ratio = max_delta_nx/max_delta_nx_old; edge_limit_m = m; end max_delta_nx_old = max_delta_nx + (max_delta_nx < eps) * eps; %Max delta_ny and ratio stability check max_delta_ny = max(abs(delta_ny));
333
max_delta_ny_ratio = max_delta_ny/max_delta_ny_old; while((max_delta_ny > max_delta_n_limit) | (max_delta_ny_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /2; delta_nx = delta_nx /2; delta_ny = delta_ny /2; delta_nz = delta_nz /2; max_delta_ny = max(abs(delta_ny)); max_delta_ny_ratio = max_delta_ny/max_delta_ny_old; edge_limit_m = m; end max_delta_ny_old = max_delta_ny + (max_delta_ny < eps) * eps; %Max delta_nz and ratio stability check max_delta_nz = max(abs(delta_nz)); max_delta_nz_ratio = max_delta_nz/max_delta_nz_old; while((max_delta_nz > max_delta_n_limit) | (max_delta_nz_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /2; delta_nx = delta_nx /2; delta_ny = delta_ny /2; delta_nz = delta_nz /2; max_delta_nz = max(abs(delta_nz)); max_delta_nz_ratio = max_delta_nz/max_delta_nz_old; edge_limit_m = m; end max_delta_nz_old = max_delta_nz + (max_delta_nz < eps) * eps; %If the average value of the delta_nx, delta_ny, delta_nz are larger than the limit, then, reduce the delta_t/¥ã to the 1/3 of the current value. %Average delta_nx and ratio stability check avg_delta_nx = mean(abs(delta_nx)); %Average value of the delta_nx, sum |new nx - old nx| / N avg_delta_nx_ratio = avg_delta_nx/avg_delta_nx_old; while((avg_delta_nx > avg_delta_n_limit) | (avg_delta_nx_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /3; delta_nx = delta_nx /3; delta_ny = delta_ny /3; delta_nz = delta_nz /3; avg_delta_nx = mean(abs(delta_nx)); avg_delta_nx_ratio = avg_delta_nx/avg_delta_nx_old; edge_limit_m = m; end avg_delta_nx_old = avg_delta_nx + (avg_delta_nx < eps) * eps; %Average delta_ny and ratio stability check avg_delta_ny = mean(abs(delta_ny)); avg_delta_ny_ratio = avg_delta_ny/avg_delta_ny_old; while((avg_delta_ny > avg_delta_n_limit) | (avg_delta_ny_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /3;
334
delta_nx = delta_nx /3; delta_ny = delta_ny /3; delta_nz = delta_nz /3; avg_delta_ny = mean(abs(delta_ny)); avg_delta_ny_ratio = avg_delta_ny/avg_delta_ny_old; edge_limit_m = m; end avg_delta_ny_old = avg_delta_ny + (avg_delta_ny < eps) * eps; %Average delta_nz and ratio stability check avg_delta_nz = mean(abs(delta_nz)); avg_delta_nz_ratio = avg_delta_nz/avg_delta_nz_old; while((avg_delta_nz > avg_delta_n_limit) | (avg_delta_nz_ratio > max_ratio)) delta_t_over_gamma = delta_t_over_gamma /3; delta_nx = delta_nx /3; delta_ny = delta_ny /3; delta_nz = delta_nz /3; avg_delta_nz = mean(abs(delta_nz)); avg_delta_nz_ratio = avg_delta_nz/avg_delta_nz_old; edge_limit_m = m; end avg_delta_nz_old = avg_delta_nz + (avg_delta_nz < eps) * eps; %Calculation & update of the new nx, ny, nz values nx = nx + delta_nx; ny = ny + delta_ny; nz = nz + delta_nz; %Assume strong anchoring nx(1) = nx_bot; ny(1) = ny_bot; nz(1) = nz_bot; nx(N) = nx_top; ny(N) = ny_top; nz(N) = nz_top; %Normalization of the LC director sqrt_n = sqrt(nx.^2 + ny.^2 + nz.^2); nx = nx ./ sqrt_n; ny = ny ./ sqrt_n; nz = nz ./ sqrt_n; %Creep up of the delta_t/gamma delta_t_over_gamma = delta_t_over_gamma * creep_up; %1.01 %*********************************************************************% %Calculation of the Average free energy density in a layer %Calculation of the Dz eps_n = eps_per + delta_eps .* nz .^2; %z component of the epsilon (epsilon_zz) at each layer, <- Array
335
if constant_mode == 1 %1: Constant VOLTAGE (V) mode, 2(else): Constant CHARGE (Q) mode Dz = Dz_temp / (peripheral_d_eps + d/N*sum(1 ./ eps_n)); %z component of the displacement vector D end %Calculation of the nx, ny, nz derivatives dnxdz(2:N-1) = (nx(3:N) - nx(1:N-2)) / (2 * delta_d); %1'st derivative of the nx dnydz(2:N-1) = (ny(3:N) - ny(1:N-2)) / (2 * delta_d); %1'st derivative of the ny dnzdz(2:N-1) = (nz(3:N) - nz(1:N-2)) / (2 * delta_d); %1'st derivative of the nz d2nxdz2(2:N-1) = (nx(3:N) + nx(1:N-2) - 2*nx(2:N-1)) / (delta_d)^2; %2nd derivative of the nx d2nydz2(2:N-1) = (ny(3:N) + ny(1:N-2) - 2*ny(2:N-1)) / (delta_d)^2; %2nd derivative of the ny d2nzdz2(2:N-1) = (nz(3:N) + nz(1:N-2) - 2*nz(2:N-1)) / (delta_d)^2; %2nd derivative of the nz %Calculation of the free energy density f = 1./2.*K11.*dnzdz.^2+1./2.*K22.*(nx.^2.*dnydz.^2+ny.^2.*dnxdz.^2-2.*nx.*ny.*dnydz.*dnxdz-2.*q.*nx.*dnydz+2.*q.*ny.*dnxdz+q.^2)+1./2.*K33.*(nz.^2.*dnxdz.^2+nz.^2.*dnydz.^2+nx.^2.*dnxdz.^2+2.*nx.*ny.*dnydz.*dnxdz+ny.^2.*dnydz.^2)+1./2.*Dz.^2./eps_0./(eps_per+delta_eps.*nz.^2); avg_delta_f = mean(abs(f - f_old)); %Average change of the free energy density between the relaxations f_old = f; %Store the current free energy density to the old one %Termination control if avg_delta_f < avg_delta_f_limit %If average-delta-free energy is smaller than avg_delta_f_limit, it's ready to stop termination_flag = 1; end if (termination_flag == 1) & (old_edge_limit_m ~= edge_limit_m) edge_limit_num = edge_limit_num + 1; old_edge_limit_m = edge_limit_m; end if ((termination_flag == 1) & (m == edge_limit_m + 10)) | edge_limit_num >= 10 break end if m >= max_iteration_num %If iteration number (m) is bigger than max_iteration_num, relaxation process stop str = sprintf('Maximum iteration number (%d) OVER!!!', max_iteration_num); h1 = warndlg(str,'Iteration number warning!');
336
%set(h1, 'WindowStyle', 'modal') break end m = m + 1; end %End of the relaxation (while loop) %=========================================================================% %Change the output nx, ny, nz to the theta, phi format theta = acos(nz); phi = acos(nx ./ (sqrt(nx.^2 + ny.^2) + (sqrt(nx.^2 + ny.^2) < eps) * eps)); for p = 1 : N if ny(p) < 0 phi(p) = -phi(p); end end if twist_mode == 2 for p = 1 : N if nx(p) < 0 theta(p) = -theta(p); phi(p) = phi(p) - pi; end end end %Change the definition of the LC theta(theta) theta_xy_deg = (pi/2 - theta) * 180/pi; %Redefine theta again, Angle from X-Y plane in degree unit phi_deg = phi * 180/pi; LcAngle(:, 1) = z * 1e6; LcAngle(:, 2) = theta_xy_deg; LcAngle(:, 3) = phi_deg;
337
APPENDIX B
CALCULATION OF THE OPTICS IN LIQUID CRYSTAL DEVICES BY
EXTENDED 2×2 JONES MATRIX61, 62, 63
B-1. Calculations
Let us consider a uniaxial liquid crystal cell described in Appendix A with a pair
of polarizers; a plane wave is incident on the LCD at an oblique angle of kθ from the
normal direction; and the light propagation vector ( Kr
) lies on the zx − plane. That is,
koko kzkxK θθ cosˆsinˆ +=r
, (B-1)
where λπω /2/ == cko , and λ is the light wavelength in free space. The LCD can be
divided into N layers. The first ( 1=n ) and the last ( Nn = ) layers are the entrance and
exit polarizers, respectively, the second ( 2=n ) and the ( 1−N )th layers are the substrates,
and the liquid crystal layers ( 2~3 −= Nn ) are located between them. Each of these N
layers is characterized by a dielectric constant tensor:
⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛
=
zzzyzx
yzyyyx
xzxyxx
εεεεεεεεε
ε , (B-2)
338
where each element of the dielectric constant tensor for the uniaxial material can be
calculated from
jiijij nn)( || ⊥⊥ −+= εεδεε . (B-3)
Let us assume the plane wave in each layer, which is a solution of the Maxwell equation,
as follows:
)](exp[)](exp[),( tzkxkiEtrKiEtrE zxoo ωω −+=−•=rrrrrr
, (B-4)
)](exp[)](exp[),( tzkxkiHtrKiHtrH zxoo ωω −+=−•=rrrrrr
. (B-5)
From Eqs. (B-4), (B-5) and the Maxwell equation, we have the wave equation as below:
0=+⎟⎟⎠
⎞⎜⎜⎝
⎛×× oo
oo
EEkK
kK rr
rr
ε , (B-6)
where oEr
is expressed as,
zoyoxoo EzEyExE ˆˆˆ ++=r
. (B-7)
339
From Eqs. (B-1), (B-6), and (B-7), we obtain these three equations,
022
2
=⎟⎟⎠
⎞⎜⎜⎝
⎛+++⎟
⎟⎠
⎞⎜⎜⎝
⎛+− zoxz
o
zxyoxyxoxx
o
z Ek
kkEE
kk
εεε , (B-8)
02
2
2
2
=+⎟⎟⎠
⎞⎜⎜⎝
⎛+−−+ zoyzyoyy
o
z
o
xxoyx EE
kk
kk
E εεε , (B-9)
02
2
2 =⎟⎟⎠
⎞⎜⎜⎝
⎛+−++⎟
⎟⎠
⎞⎜⎜⎝
⎛+ zozz
o
xyozyxozx
o
zx Ekk
EEk
kkεεε . (B-10)
In these equations, there is a nontrivial solution only when
0
2
2
2
2
2
2
2
22
2
=
+−+
+−−
++−
zzo
xzyzx
o
zx
yzyyo
z
o
xyx
xzo
zxxyxx
o
z
kk
kkk
kk
kk
kkk
kk
εεε
εεε
εεε
. (B-11)
From this equation, we obtain four eigen values for a uniaxial material as follows:
21
2
221
⎟⎟⎠
⎞⎜⎜⎝
⎛−=
o
xo
o
z
kk
nkk
, (B-12)
340
21
2
222
2
222 sincos1
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛ −−−+−=
o
x
e
oezz
zz
eo
o
x
zz
xz
o
z
kk
nnnnn
kk
kk
φθεεε
ε, (B-13)
o
z
o
z
kk
kk 13 −= , (B-14)
o
z
o
z
kk
kk 24 −= , (B-15)
where θ and φ are the director tilt and azimuth angles, and en , on are the refractive
indices of the extraordinary, ordinary rays of the uniaxial material, respectively.
Equations (B-12) and (B-13) are the eigen values corresponding to the transmission eigen
waves, and (B-14) and (B-15) are those of reflection light. In this dissertation, we assume
only the transmitted eigen waves are propagating in the medium.
In Eqs. (B-8), (B-9), and (B-10), only two of them are linearly independent
because there is a constraint of Eq. (B-11). Therefore, we can express two components of
the electric field in terms of the other component. For eigen wave 1, we express 1yE in
functions of 1xE . From Eqs. (B-9) and (B-10), we obtain these equations,
111
111 xy
x
xyy Me
eE
eE == , (B-16)
where
341
zyyzzzo
xyy
o
x
o
zx k
kkk
kk
e εεεε −⎟⎟⎠
⎞⎜⎜⎝
⎛−⎟
⎟⎠
⎞⎜⎜⎝
⎛−+= 2
2
2
2
2
21
1 , (B-17)
⎟⎟⎠
⎞⎜⎜⎝
⎛++⎟
⎟⎠
⎞⎜⎜⎝
⎛−= zx
o
zxyzzz
o
xyxy k
kkkk
e εεεε 21
2
2
1 , (B-18)
)](exp[ 11
10
1
11 tzkxki
eE
eE
M zxx
x
x
xx ω−+== . (B-19)
With the same way, we obtain these equations for the eigen wave 2 from Eqs. (B-8) and
(B-10) as follows:
222
222 yx
y
yxx Me
eE
eE == , (B-20)
where
⎟⎟⎠
⎞⎜⎜⎝
⎛+−⎟
⎟⎠
⎞⎜⎜⎝
⎛+−= xz
o
zxzyzz
o
xxyx k
kkkk
e εεεε 22
2
2
2 , (B-21)
⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟
⎟⎠
⎞⎜⎜⎝
⎛++⎟
⎟⎠
⎞⎜⎜⎝
⎛−⎟
⎟⎠
⎞⎜⎜⎝
⎛+−= xz
o
zxzx
o
zxzz
o
xxx
o
zy k
kkkkk
kk
kk
e εεεε 22
22
2
2
2
22
2 , (B-22)
)](exp[ 22
20
2
22 tzkxki
eE
eE
M zxy
y
y
yy ω−+== . (B-23)
342
Using equations (B16~23), the E -vector represented by the tangential components of the
total electric field at any point can be expressed in terms of mode vector as follows:
⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛=⎟⎟
⎠
⎞⎜⎜⎝
⎛
2
2
1
1
y
x
y
x
y
x
EE
EE
EE
⎟⎟⎠
⎞⎜⎜⎝
⎛=
2
1
y
x
MM
S , (B-24)
where
⎟⎟⎠
⎞⎜⎜⎝
⎛=
21
21
yy
xx
eeee
S . (B-25)
Using Eqs. (B-19) and (B-23), the mode vector propagates from the bottom of n th layer
to the top of the layer by
0,2
1
,2
1
ny
xn
dnny
x
MM
GMM
⎟⎟⎠
⎞⎜⎜⎝
⎛=⎟⎟
⎠
⎞⎜⎜⎝
⎛, (B-26)
where the propagation matrix ( nG ) is expressed as
nz
zn dik
dikG ⎟⎟
⎠
⎞⎜⎜⎝
⎛=
)exp(00)exp(
2
1 . (B-27)
343
From Eqs. (B-24) and (B-26), the E -vector at the top of the n th layer is expressed by the
E -vector at the bottom of the n th layer as,
0,, ny
xn
dnny
x
EE
JEE
⎟⎟⎠
⎞⎜⎜⎝
⎛=⎟⎟
⎠
⎞⎜⎜⎝
⎛, (B-28)
where
1−= nnnn SGSJ . (B-29)
With the boundary condition that the tangential components of the electric field are
continuous at each layer interface, we obtain the E -vector relationship between the input
light and the output light of the LCD system composed of N layers as follows:
Iny
x
Outy
x
EE
JEE
⎟⎟⎠
⎞⎜⎜⎝
⎛=⎟⎟
⎠
⎞⎜⎜⎝
⎛, (B-30)
where the extended Jones matrix, J is expressed as
121 ...... JJJJJJ nNN −= . (B-31)
344
Until now, we have not considered reflections, especially at surfaces. However, the
reflections at the air interfaces are significant. To take into account the reflections, we
rewrite the extended Jones matrix as follows:
EntnNNExtEntExt JJJJJJJJJJJ 121 ......' −== , (B-32)
where the reflection matrices, ExtJ and EntJ can be calculated from
⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜
⎝
⎛
+
+=
ppk
k
pkp
p
Ent
n
nJ
θθθ
θθθ
coscoscos2
0
0coscos
cos2
, (B-33)
⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜
⎝
⎛
+
+=
kpp
pp
kpp
kp
Ext
nn
nn
J
θθθ
θθθ
coscoscos2
0
0coscos
cos2
, (B-34)
where pn and pθ are the average refractive index and refraction angle in the polarizer (or
outer most layer), respectively.
The total light transmittance (T ) of the LCD system is calculated by
222
222
cos
cosinyk
inx
outyk
outx
EE
EET
θ
θ
+
+= . (B-35)
345
B-2. Program source codes (Matlab)
function T = OpticsCal2by2Fcn(OpticsStruct) %OPTICSCAL2BY2FCN Calculation of the Transmittance by using the EXTENDED JONES MATRIX (2X2 MATRIX) (Liquid Crystals. 1997. Vol. 22. No. 2. 171-175) %T: Transmittance return value %OpticsStruct: Structure that includes the parameters for the Optics calculation %OpticsStruct.[nx, ny, nz, d, Theta, Phi, Psi, Lambda, ni, nt, ThetaV, PhiV, PolarizationAngle, Epi, Esi, Total_N] %X, Y, Z: Lab Coordinate system %x', y', z': Principal Coordinate system %x, y, z: "Incident Light Coordinate system" (Z = z) %Light incidents in x-z plane and %Polar angle: theta_i from the Z(z)-axis (= Viewing Polar Angle if n_i == n_t) %Azimuthal angle: phi_i from the X-axis (=Viewing Azimuthal Angle) %All of the axis of the birefringence material is refered to optic axis (z'-axis) %theta : Tilt angle from the X-Y plane in RADIAN unit (0 deg.: In plane, 90deg.:Homeotropic Alignment) %phi : Azimuthal angle from the X-axis in RADIAN unit %Unit: MKSA %January 02, 2006 %Yong-Kyu Jang (Liquid Crystal Institute, Kent State University) %Simulation parameter lambda = OpticsStruct.Lambda; %Wavelength of the incident light(cm) N = OpticsStruct.Total_N; %The number of the total layers n_i = OpticsStruct.ni; %Refractive index of the incident medium n_t = OpticsStruct.nt; %Refractive index of the transmitted medium theta_t = OpticsStruct.ThetaV; %Transmitted Polar angle of the light from the Z-axis (0 deg.: normal transmittance, radian unit) phi_i = OpticsStruct.PhiV; %Transmitted(=Incident) Azimuthal angle of the light from the X-axis (0 deg.: X-Z plane transmittance, radian unit) sin_theta_i = n_t/n_i * sin(theta_t); %Incident Polar angle of the light from the Z-axis (0 deg.: normal incident, radian unit) cos_theta_i = sqrt(1 - sin_theta_i^2); cos_theta_t = cos(theta_t); if (cos_theta_i <= 0 | cos_theta_t <= 0) beep errordlg('Incident and transmitted light angle should be less than 90 degree', 'Angle definition Error', 'on') T = []; return end
346
if isempty(OpticsStruct.Epi) %Linear polarization Exi = cos(OpticsStruct.PolarizationAngle - phi_i) * cos_theta_i; Eyi = sin(OpticsStruct.PolarizationAngle - phi_i); else Exi = OpticsStruct.Epi * cos_theta_i; %Definition of the polarization of the incident light (X-Y coordinate) Eyi = OpticsStruct.Esi; %Definition of the polarization of the incident light (X-Y coordinate) end %Find the 1'st non-zero thickness layer for p = 1 : N if OpticsStruct.d(p) ~= 0 n_pi = (real(OpticsStruct.nx(p)) + real(OpticsStruct.ny(p)) + real(OpticsStruct.nz(p)))/3; %Average refractive index of the 1'st layer (It is needed to consider the reflection at the interface between air and 1'st layer) break else n_pi = n_t; end end %Find the last non_zero thickness layer for p = N : -1 : 1 if OpticsStruct.d(p) ~= 0 n_pt = (real(OpticsStruct.nx(p)) + real(OpticsStruct.ny(p)) + real(OpticsStruct.nz(p)))/3; %Average refractive index of the last layer (It is needed to consider the reflection at the interface between air and last layer) break else n_pt = n_t; end end cos_theta_pi = sqrt(1 - (n_i/n_pi)^2*sin_theta_i^2); %Polar angle of the light in the 1'st layers (0 deg.: normal incident) cos_theta_pt = sqrt(1 - (n_i/n_pt)^2*sin_theta_i^2); %Polar angle of the light in the last layers (0 deg.: normal incident) K_0 = 2 * pi / lambda; %Magnitude of the wavevector in free space K_x = K_0 * n_i * sin_theta_i; %X component of the wave vector in free space (K_x is conserved in all of the layer) Kx_K0 = n_i * sin_theta_i; Kx_K0_2 = Kx_K0^2; %Calculation of the Entrance & Exit matrix J_ent = zeros(2); %Initialization of the Entrance matrix (Reflection at the entrance interface) J_ext = zeros(2); %Initialization of the Eixt matrix (Reflection at the exit interface)
347
J_ent(1,1) = 2 * n_i * cos_theta_pi / (n_pi * cos_theta_i + n_i * cos_theta_pi); %X-polarization component of the Entrance matrix J_ent(2,2) = 2 * n_i * cos_theta_i / (n_i * cos_theta_i + n_pi * cos_theta_pi); %Y-polarization component of the Entrance matrix J_ext(1,1) = 2 * n_pt * cos_theta_t / (n_t * cos_theta_pt + n_pt * cos_theta_t); %X-polarization component of the Exit matrix J_ext(2,2) = 2 * n_pt * cos_theta_pt / (n_pt * cos_theta_pt + n_t * cos_theta_t); %Y-polarization component of the Exit matrix G_n = zeros(2); %Initialization of the phase matrix %Calculation of the J matrix J = eye(2); %Initialization of the J matrix to the identity matrix for n = 1 : N %Assignment of the optical parameters of each layer nx = OpticsStruct.nx(n); ny = OpticsStruct.ny(n); nz = OpticsStruct.nz(n); h = OpticsStruct.d(n); %cm unit theta = OpticsStruct.Theta(n); %radian unit phi = OpticsStruct.Phi(n); %radian unit psi = OpticsStruct.Psi(n); %radian unit %Trick to escape the uncalculable situation phi_phi_i = phi - phi_i; if abs(sin(phi_phi_i)) < 1e-4 phi_phi_i = phi_phi_i + 1e-4; end if abs(cos(phi_phi_i)) < 1e-6 phi_phi_i = phi_phi_i + 1e-6; end if abs(cos(theta)) < 1e-6 theta = theta + 1e-6; end %Calculation of the dielctric constant tensor s_t = sin(theta); c_t = cos(theta); s_p = sin(phi_phi_i); c_p = cos(phi_phi_i); if isequal(nx, ny) %Isotropic & Uniaxial materials (nx=ny~=nz) nz2_nx2 = nz^2 - nx^2; eps_xx = nx^2 + nz2_nx2 * c_t^2 * c_p^2; eps_xy = nz2_nx2 * c_t^2 * s_p * c_p; eps_xz = nz2_nx2 * s_t * c_t * c_p; eps_yy = nx^2 + nz2_nx2 * c_t^2 * s_p^2; eps_yz = nz2_nx2 * s_t * c_t * s_p; eps_zz = nx^2 + nz2_nx2 * s_t^2;
348
%Calculation of the eigenvalues (Kz1_K0, Kz2_K0) Kz1_K0 = sqrt(nx^2 - Kx_K0_2); Kz2_K0 = -eps_xz / eps_zz * Kx_K0 + nx * nz / eps_zz * sqrt(eps_zz - (1 - nz2_nx2 / nz^2 * c_t^2 * s_p^2) * Kx_K0_2); else %Biaxial materials (nx~=ny~=nz) a2 = -sin(psi) * s_t * c_p - cos(psi) * s_p; a3 = c_t * c_p; b2 = -sin(psi) * s_t * s_p + cos(psi) * c_p; b3 = c_t * s_p; c2 = sin(psi) * c_t; c3 = s_t; ny2_nx2 = ny^2 - nx^2; nz2_nx2 = nz^2 - nx^2; eps_xx = nx^2 + ny2_nx2 * a2^2 + nz2_nx2 * a3^2; eps_yy = nx^2 + ny2_nx2 * b2^2 + nz2_nx2 * b3^2; eps_zz = nx^2 + ny2_nx2 * c2^2 + nz2_nx2 * c3^2; eps_xy = ny2_nx2 * a2 * b2 + nz2_nx2 * a3 * b3; eps_xz = ny2_nx2 * a2 * c2 + nz2_nx2 * a3 * c3; eps_yz = ny2_nx2 * b2 * c2 + nz2_nx2 * b3 * c3; %Calculation of the Delta matrix Delta(1,1) = -Kx_K0 * eps_xz/eps_zz; Delta(1,2) = 1 - Kx_K0^2/eps_zz; Delta(1,3) = -Kx_K0 * eps_yz/eps_zz; Delta(2,1) = eps_xx - eps_xz^2/eps_zz; Delta(2,2) = Delta(1,1); Delta(2,3) = eps_xy - eps_xz * eps_yz/eps_zz; Delta(3,4) = 1; Delta(4,1) = Delta(2,3); Delta(4,2) = Delta(1,3); Delta(4,3) = eps_yy - eps_yz^2/eps_zz - Kx_K0^2; %Calculation of the eigenvalues of the Biaxial Film D = eig(Delta); D_temp = real(D); index = find(D_temp > 0); if length(index) == 2 Kz2_K0 = D(index(1)); Kz1_K0 = D(index(2)); else beep errordlg('Can not calculate Eigenvalues of the Biaxial Film', 'Biaxial Film Error', 'on') T = []; return end end %of the Dielectric constant calculation %Calculation of the G_n matrix (Phase matrix) G_n(1,1) = exp(i * K_0 * Kz1_K0 * h); G_n(2,2) = exp(i * K_0 * Kz2_K0 * h);
349
%Calculation of the S_n matrix e_x1 = (Kz1_K0^2 + Kx_K0_2 - eps_yy) * (Kx_K0_2 - eps_zz) - eps_yz * eps_yz; e_y1 = eps_xy * (Kx_K0_2 - eps_zz) + eps_yz * (Kx_K0 * Kz1_K0 + eps_xz); e_x2 = eps_xy * (-Kx_K0_2 + eps_zz) - eps_yz * (Kx_K0 * Kz2_K0 + eps_xz); e_y2 = (-Kz2_K0^2 + eps_xx) * (Kx_K0_2 - eps_zz) + (Kx_K0 * Kz2_K0 + eps_xz) * (Kx_K0 * Kz2_K0 + eps_xz); if e_y2 == 0 e_y2 = e_y2 + eps; end if e_x1 == 0 e_x1 = e_x1 + eps; end S_n = [1, e_x2 / e_y2; e_y1 / e_x1, 1]; % while (det(S_n) == 0) % S_n = [S_n(1) + eps, e_x2 / e_y2; e_y1 / e_x1, 1]; % end %Calculation of the J_n matrix J = S_n * G_n / S_n * J; end %End of n loop (Calculation of the J matrix) %Calculation of the J' matrix %J_prime = J; %Neglect reflection at the glass surface J_prime = J_ext * J * J_ent; %Calculation of the Output Electric Field Ext = J_prime(1,1)*Exi + J_prime(1,2)*Eyi; Eyt = J_prime(2,1)*Exi + J_prime(2,2)*Eyi; %Calculation of the Transmittance T = n_t * cos_theta_i * (abs(Ext)^2 + cos_theta_t^2 * abs(Eyt)^2) / (n_i * cos_theta_t * (abs(Exi)^2 + cos_theta_i^2 * abs(Eyi)^2));
350
APPENDIX C
PROGRAM OVERVIEW (“LC Optics”)
C-1. “LC Optics” main screen
This program uses the director calculation and the optics calculation explained in
previous two appendices. In this main screen, we build a stack configuration of a LCD by
using “Add”, “Delete”, and “Property” push buttons. This stack configuration including
all information of each layer can be saved in “File” menu as the configuration file (*.cfg).
351
C-2. Popup menu of each layer
By the popup menu in each layer of the stack configuration, we can choose a
Polarizer, Glass, ITO, Polyimide, Liquid Crystal, Uniaxial Film, Biaxial Film,
Inhomogeneous Film, Isotropic Layer, Anisotropic Layer, and Metal. These layers cover
most of the possible layers that we can use in LCDs. As similar to the main layers, we
can also choose the light source types such as the standard illuminants A, B, C, D65, E,
and any kind of user light source.
352
C-3. Layer properties
The “Property” push button of each layer pops up a window corresponding to the user-
selected layer, which includes all the possible information that we need in calculations.
This information can be saved as a file (*.mat) for the use in other LCDs. As an example,
the bottom figure shows the Liquid Crystal property. We can set the refractive indices by
the three ways and set the material, manufacturing, and simulation parameters. We can
choose the calculation methods such as “Vector”, “Theta-Phi”, “Constant V”, and
“Constant Q”. We can also select “Auto” or “Manual” mode as the initial director
configuration. In “Auto” mode, the program searches the stable director configuration.
Each parameter can be designated as a variable by checking the “Var” check box.
353
C-4. Variable rearrangement and plotting
After building and setting the stack configuration of a LCD, the selected variables show
up in this window automatically. We can set the “Start”, “Step”, “End”, and “Number” of
each variable for the calculation, and each variable can be coupled each other. In this
screen, we can choose the calculation methods, 4×4 or 2 ×2 matrices with the detailed
spectrum averaging methods. We can also calculate the contrast ratio if there is a liquid
crystal layer. After the calculation, we can plot 2-D, 3-D, and conoscopy figures in terms
of the director angle, transmittance, reflectance, luminance, lightness, color coordinates,
and color difference for various color spaces and viewing fields. After calculation, we
can save not only the data but also all the configuration including data also (*.ykj).
354
C-5. Plotting options
If we want to change the default setting of the plotting, we can do that in this
window. We can control the 2-D, 3-D, and conoscopy, separately, and change the
detailed figure shape. We can also do the data interpolation with controlling the method.
355
C-6. Example of the 2-D Plotting
356
C-7. Example of the 3-D Plotting
357
C-8. Example of the Conoscopy Plotting
358
C-9. Example of the LC director configuration
359
REFERENCES
1 S. Saito, Mol. Cryst. and Liq. Cryst. 138, 187 (1986).
2 G. Paul Montgomery, SPIE 202, 103 (1979).
3 M. Born, and E. Wolf, Principles of Optics, 7th edition, Chap. 1, p.42.
4 P. Bos, SID Int. Symp. Digest Tech. Papers 1983, 30.
5 P. Bos, Mol. Cryst. and Liq. Cryst. 113, 329 (1984).
6 H. Mori, Y. Itoh, Y. Nishiura, T. Nakamura, Y. Shinagawa, SID Int. Symp. Digest Tech.
Papers 1997, 941.
7 H. Mori and P. Bos, Jpn. J. Appl. Phys. 38, 2837 (1999).
8 H. Mori and P. Bos, IDRC (SID) 1997, M88.
9 Y. Ito, R. Matsubara, R. Nakamura, M. Nagai, S. Nakamura, H. Mori, and K.
Mihayashi, SID Int. Symp. Digest Tech. Papers 2005, 986.
10 Dwight W. Berreman, J. Opt. Soc. Am. 62, 502 (1972).
11 H. Wohler, J. Opt. Soc. Am. A 5, 1554 (1988).
12 C. J. Chen, A. Lien, and M. I. Nathan, Jpn. J. appl. Phys. 35, L1204 (1996).
13 Y. Huang, T. X. Wu, and S. T. Wu, J. Appl. Phys. 93, 2490 (2003).
14 E. Hecht, Optics, 4th edition, Chap. 7, p.314.
15 D. W. Berreman, SID Int. Symp. Digest Tech. Papers 1993, 101.
16 Tae-Hoon Yoon, and Jae Chang Kim, Jpn. J. Appl. Phys. 37, 529 (1998).
360
17 H. L. Ong, Jpn. J. Appl. Phys. Part 1 33, 1085 (1994).
18 K. H. Yang, J. Appl. Phys. 68, 1550 (1990).
19 P. Bos, SID Int. Symp. Digest Tech. Papers 1983, 30.
20 P. Bos, Mol. Cryst. Liq. Cryst. 113, 329 (1984).
21 Y. Yamaguchi, T. Miyashita, T. Uchida, SID Int. Symp. Digest Tech. Papers 1993,
277.
22 C.-L. Kuo, T. Miyashita, M. Suzuki, T. Uchida, SID Int. Symp. Digest Tech. Papers
1994, 927.
23 A. Lien, Appl. Phys. Lett. 57, 2767 (1990).
24 A. Lien, Liquid Crystals 22, 171 (1997).
25 A. Lien and C.-J. Chen, Jpn. J. Appl. Phys. Part 2 35, L1200 (1996).
26 H. Mori, Y. Itoh, Y. Nishiura, T. Nakamura, Y. Shinagawa, SID Int. Symp. Digest
Tech. Papers 1997, 941.
27 H. Mori and P. Bos, Jpn. J. Appl. Phys. 38, 2837 (1999).
28 P. Van De Witte, S. Stallinga, and J. A. M. M. Van Haaren, Jpn. J. Appl. Phys. 39, 101
(2000).
29 Y. Satoh, H. Mazaki, E. Yoda, T. Kaminade, T. Toyooka, and Y. Kobori , SID Int.
Symp. Digest Tech. Papers 2000, 347.
30 C.-L. Kuo, T. Miyashita, M. Suzuki, and T. Uchida, Jpn, J. Appl. Phys. Part 2 34,
L1362 (1995).
361
31 J. Chen, K. H. Kim, J. J. Jyu, J. H. Souk, J. R. Kelly, and P. Bos, SID Int. Symp. Digest
Tech. Papers 1998, 315.
32 T. Ishinabe, T. Miyashita, and T. Uchida, SID Int. Symp. Digest Tech. Papers 2000,
1094.
33 T. Ishinabe, T. Miyashita, T. Uchida, and Y. Fujimura, Proc. 21st Int. Display Research
Conf. Conjunction with the 8th Int. Display Workshops, 2001, 485.
34 T. Ishinabe, T. Miyashita, and T. Uchida, Jpn. J. Appl. Phys. Part 1 41, 4553 (2002).
35 X.-D. Mi, T. Ishikawa, A. F. Kurtz, and D. Kessler, SID Int. Symp. Digest Tech.
Papers 2006, 1535.
36 M. Born, and E. Wolf, Principles of Optics, 7th edition (Cambridge University Press,
2002), Chap. 1, p.40.
37 P. Bos, P. Johnson Jr. and K. Rickey Koehler/Beran: SID Int. Symp. Dig. Tech. Pap.
1983, p. 30.
38 P. Bos and K. Rickey Koehler/Beran: Mol. Cryst. Liq. Cryst. 113 (1984) 329.
39 Y. Yamaguchi, T. Miyashita, T. Uchida: SID Int. Symp. Dig. Tech. Pap. 1993, p. 277.
40 C.-L. Kuo, T. Miyashita, M. Suzuki, T. Uchida: SID Int. Symp. Dig. Tech. Pap. 1994,
p. 927.
41 M. Flynn and P. Bos: Mol. Cryst. Liq. Cryst. 263 (1995) 377.
42 T. Miyashita, C.-L. Kuo, M. Suzuki, T. Uchida: SID Int. Symp. Dig. Tech. Pap. 1995,
p. 797.
362
43 C.-L. Kuo, T. Miyashita, M. Suzuki, and T. Uchida: Jpn. J. Appl. Phys. 34 (1995)
L1362.
44 H. Mori, Y. Itoh, Y. Nishiura, T. Nakamura, and Y. Shinagawa: SID Int. Symp. Dig.
Tech. Pap. 1997, p. 941.
45 H. Mori and P. Bos: Jpn. J. Appl. Phys. 38 (1999) 2837.
46 Y.–K. Jang and P. Bos: J. Appl. Phys. 101 (2007) 033131.
47 Y.-K Jang and P. Bos, J. Appl. Phys. 102, 013101 (2007).
48 A. Lien: Liquid Crystals 22 (1997) 171.
49 P. Van De Witte, S. Stallinga and J. A. M. M. Van Haaren, Jpn. J. Appl. Phys. 39
(2000) 101.
50 G. Wyszecki, and W. S. Stiles, Color Science: Concepts and Methods, Quantitative
Data and Formulas (Wiley, 1967)
51 Roy S. Berns, Billmeyer and Saltzman's Principles of Color Technology (3rd edition
Wiely Interscience, Chap. 2, p.62.)
52 Noboru Ohta, and Alan R. Robertson, Colorimetry Fundamentals and Applications
(Wiley, Chap. 4, p.120.)
53 Pochi Yeh, and Claire Gu, Optics of Liquid Crystal Displays (Wiely Interscience,
Chap. 6, p.278.)
54 Pochi Yeh, and Claire Gu, Optics of Liquid Crystal Displays (Wiley Interscience,
Chap. 5, p.164.
55 Ch. Mauguin, Bull. Soc. Fr. Mineral. Cristallogr 34, 71 (1911).
363
56 P. Yeh and C. Gu, Optics of Liquid Crystal Displays (Wiley, New York, 1999), Chap.
4, p.137.
57 P. Bos, Lecture note of the class "Liquid Crystal Displays I" in Liquid Crystal Institute,
Kent State University.
58 Maurice Kleman, and Oldg D. Lavrentovich, Soft Matter Physics (Springer-Verlag
New York Inc., Chap. 3, p.93.)
59 P. Van De Witte, S. Stallinga and J. A. M. M. Van Haaren, Jpn. J. Appl. Phys. 39
(2000) 101.
60 M. Born and E. Wolf, Principles of Optics, 7th edition (Cambridge University,
Cambridge, 2001), Chap. 15, p.827.
61 A. Lien, Appl. Phys. Lett. 57, 2767 (1990).
62 A. Lien, Liquid Crystals 22, 171 (1997).
63 A. Lien and C.-J. Chen, Jpn. J. Appl. Phys. Part 2 35, L1200 (1996).