ANCHORING TRANSITIONS OF NEMATIC LIQUID CRYSTALS ON … · ANCHORING TRANSITIONS OF NEMATIC LIQUID...
Transcript of ANCHORING TRANSITIONS OF NEMATIC LIQUID CRYSTALS ON … · ANCHORING TRANSITIONS OF NEMATIC LIQUID...
i
ANCHORING TRANSITIONS OF NEMATIC LIQUID CRYSTALS ON
LARGE ANGLE DEPOSITED SILICON OXIDE THIN FILMS
A dissertation submitted to Kent State University in partial
fulfillment of the requirements for the Degree of Doctor of Philosophy
By
Cheng Chen
August 2006
UMI Number: 3237852
32378522006
UMI MicroformCopyright
All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company 300 North Zeeb Road
P.O. Box 1346 Ann Arbor, MI 48106-1346
by ProQuest Information and Learning Company.
ii
Dissertation written by
Cheng Chen
B.S., Peking University, China. 2001
Ph. D., Kent State University, 2006
Approved by
Chair, Doctoral Dissertation Committee
, Philip J. Bos, Professor of Chemical Physics Interdisciplinary Program
Members, Doctoral Dissertation Committee
, John L. West, Professor of Chemistry Department
, Deng-Ke Yang, Professor of Chemical Physics Interdisciplinary Program
, David W. Allender, Professor of Chemical Physics Interdisciplinary Program
, Kenneth K. Laali, Professor of Chemistry Department
Accepted by
, Oleg D. Lavrentovich, Director, Chemical Physics Interdisciplinary Program
, John R.D. Stalvey, Dean, College of Arts and Sciences
iii
TABLE OF CONTENTS
TABLE OF CONTENTS................................................................................................... iii
LIST OF FIGURES ............................................................................................................ v
LIST OF TABLES........................................................................................................... xiii
ACKNOWLEDGEMENTS............................................................................................. xiv
Chapter 1 Introduction .............................................................................................................................1
1.1 Liquid Crystalline Materials................................................................................... 1
1.2 Liquid Crystal Displays......................................................................................- 3 -
1.3 Liquid Crystal Alignment and the Method to Achieve the Same ......................- 5 -
1.4 Overview of the Dissertation..............................................................................- 6 -
Chapter 2 Theory .......................................................................................................................................8
2.1 Introduction ............................................................................................................ 8
2.2 Review of Previous Theories ................................................................................. 9
2.2.1 Short Range Interactions ..................................................................................... 9
2.2.2 Long Range van der Waals Potential ................................................................ 10
2.2.3 Competition between Long Range and Short Range Forces............................. 11
2.2.4 Topography ....................................................................................................... 12
2.3 Our Theory ........................................................................................................... 15
2.4 Summary .............................................................................................................. 22
Chapter 3 Physical-chemical properties of LAD-SiOx thin films .............................................23
3.1 Introduction .......................................................................................................... 23
iv
3.2 Experimental Method........................................................................................... 24
3.2.1 Inorganic Alignment Layer Preparation............................................................ 24
3.2.2 Thin Film Characterization Method.................................................................. 26
3.3 Experimental Results and Discussions................................................................. 29
3.3.1 Surface Topography and Anisotropy ................................................................ 29
3.3.2 Stoichiometry and Surface Properties ............................................................... 30
3.4 Summary .............................................................................................................. 36
Chapter 4 Anchoring Transitions on LAD-SiOx Due to the Change in Liquid Crystal
Composition ..............................................................................................................................................37
4.1 Introduction .......................................................................................................... 37
4.2 Experimental Methods ......................................................................................... 38
4.2.1 Materials............................................................................................................ 38
4.2.2 Sample Preparation ........................................................................................... 39
4.2.3 General Examination Methods and Definition for Alignment Quality............. 39
4.2.4 Pretilt Measurement .......................................................................................... 40
4.2.5 Dielectric Anisotropy Measurement Method.................................................... 40
4.2.6 Birefringence Measurement Method................................................................. 42
4.2.7 Electro-Optical Curve and Response Time Measurement Methods ................. 42
4.3 Experimental Results............................................................................................ 46
4.3.1 The Effect of Large Longitudinal Dipole.......................................................... 46
4.3.2 The Effect of Large Lateral Dipole ................................................................... 49
4.3.3 The effect of varying the molecular structure of the additives ......................... 55
v
4.3.4 A Method to Make Improved Liquid Crystal Mixtures for Vertical Alignment
Applications. .............................................................................................................. 60
4.4 Discussions........................................................................................................... 67
4.4.1 The Effect of Large Longitudinal Dipole.......................................................... 67
4.4.2 The Effect of a Large Lateral Dipole ................................................................ 68
4.4.3 The effect of molecular structure on liquid crystal anchoring on SiOx............. 70
4.5 Summary .............................................................................................................. 74
Chapter 5 Temperature Dependence of the Anchoring Transitions on LAD-SiOx..............76
5.1 Introduction .......................................................................................................... 76
5.2 Experimental Methods ......................................................................................... 77
5.2.1 Cell Preparation and Characterization............................................................... 77
5.2.2 Surface Adsorption and Thermal Desorption.................................................... 77
5.3 Results .................................................................................................................. 80
5.3.1 Thermal Induced Anchoring Transitions .......................................................... 80
5.3.2 The Effect of Temperature on the Critical Concentration of 5CB.................... 80
5.3.3 Thermal Desorption........................................................................................... 85
5.4 Discussions........................................................................................................... 89
5.4.1 Thermal Induced Anchoring Transitions .......................................................... 89
5.4.2 The Effect of Temperature on the Critical Concentration of 5CB.................... 90
5.5 Summary .............................................................................................................. 95
Chapter 6 The Effect of LAD-SiOx Thickness on Liquid Crystal Anchoring .......................96
6.1 Introduction .......................................................................................................... 96
vi
6.2 Experimental Methods ......................................................................................... 97
6.2.1 LAD-SiOx Sample Preparation ......................................................................... 97
6.2.2 Polyimide Sample Preparation .......................................................................... 97
6.2.3 Pretilt Measurement .......................................................................................... 98
6.3 Experimental Results............................................................................................ 98
6.3.1 The Effect of LAD-SiOx Thickness on Liquid Crystal Alignment................... 98
6.3.2 The Effect of LAD-SiOx Thickness on the Critical Concentration of 5CB...... 99
6.3.3 Screening Effect .............................................................................................. 102
6.4 Discussions......................................................................................................... 106
6.4.1 The Effect of LAD-SiOx Thickness on the Alignment of Liquid Crystal....... 106
6.4.2 The Effect of LAD-SiOx Thickness on the Critical Concentration of 5CB.... 108
6.4.3 Screening Effect .............................................................................................. 113
6.5 Summary ............................................................................................................ 114
Chapter 7 Conclusions and Suggestions for Future Work.........................................................115
7.1 Summary of Dissertation Work.......................................................................... 115
7.2 Conclusions ........................................................................................................ 116
7.3 Suggestions for Future Work ............................................................................. 119
vii
LIST OF FIGURES
Figure 1: Illustration of Dubois-Violette and de Gennes’ model in which long range van
der Waals torque prefers planar alignment while short range forces prefer homeotropic
alignment................................................................................................................................................14
Figure 2: The preference in LC orientation by long-range/short-range forces......................18
Figure 3: The Working Principle of AFM........................................................................................28
Figure 4: The working principle of XPS ..........................................................................................28
Figure 5: AFM images of LAD-SiOx thermally evaporated at a medium angle. (a): 10µm
x 10µm tapping mode 3D image (b): 5µm x 5µm tapping mode 3D image (c): 3µm x
3µm contact mode 2D image of friction (d): Cross-section analysis....................................32
Figure 6: (a): RMS Roughness of LAD-SiOx surface as a function of layer thickness (b):
Anisotropy in surface roughness as a function of layer thickness .........................................33
Figure 7: XPS spectrum of thermally evaporated LAD-SiOx and e-beam evaporated
LAD- SiO2, measured at 45º take-off angle. Atomic ratio of Si and O of the sample can
be calculated from the corresponding area of the peak. Signal of carbon is from the
residual of CO2 or hydrocarbon contaminations on the sample surface. .............................34
Figure 8: XPS spectrum analysis of silicon (Si2p) in (a) e-beam evaporated LAD- SiO2
and (b) thermally evaporated LAD-SiOx. The blue line is the characteristic peak of Si in
SiO2; The cyanic line is the characteristic peak of Si in SiO; The magenta line is the
characteristic peak of Si in Si crystal; The black line is the measured Si peak; The red
line is the synthetic peak based on characteristic Si peak in SiO2, SiO and Si crystal. ...35
viii
Figure 9: Chemical structure of (a) 5CB and (b) C3.....................................................................45
Figure 10: Anchoring transitions from parallel to homeotropic to parallel again as the
concentration of 5CB in the mixture with LC1 decreases. From top left to bottom right:
pure 5CB, 50% 5CB, 25% 5CB, 10% 5CB, 5% 5CB, and pure LC1. Photo taken with
cells placed between crossed polarizers on a light table. .........................................................47
Figure 11: Anchoring transitions of liquid crystal mixtures (5CB/LC1) on LAD-SiOx due
to the change of the ratio of two components..............................................................................48
Figure 12: The addition of C3 into LC2 leads to an anchoring transition of liquid crystal
on LAD-SiOx from homeotropic to planar ...................................................................................51
Figure 13: The addition of 5CB into the mixture of C3 and LC2 causes an anchoring
transition from planar to homeotropic on LAD-SiOx................................................................52
Figure 14: The correlation between the concentration of C3 and the critical amount of
5CB that is needed to maintain homeotropic alignment of C3/5CB/LC2 mixture on
LAD-SiOx...............................................................................................................................................53
Figure 15: On E-beam evaporated SiO2, more C3 is needed than on thermally evaporated
SiOx to cause its mixture with LC2 to change from homeotropic alignment to planar
alignment ................................................................................................................................................54
Figure 16: Alignment of mixtures with different additives of LC1 on LAD-SiOx,
photographed between crossed polarizers on a light table. From top left to bottom right
cells are filled with: LC1; 10%C5-Ph-Ph-CN (5CB); 10%C5-Ph-Ph-O-C2; 5% C5-Ph-
Ph-Br, 10% C3-Cyclohexyl-Ph-O-C2 (PCH302); 10% C5-Ph-Ph; 10%C6-Ph-Ph-C5. .58
ix
Figure 17: The effect of cyano groups on the liquid crystal anchoring on LAD-SiOx. Left:
20% C7-Cyclohexyl-Ph-CN; Right: 5% C3 (C3- Cyclohexyl-COO-Ph(-2CN)-O-C2)..59
Figure 18: The addition of 5CB enables the mixture of LC2 and C3 to obtain uniform
vertical alignment on LAD-SiOx with a greater negative dielectric anisotropy. ...............62
Figure 19: The addition of 5CB also allows higher birefringence of the LC2/C3 mixture
to be used for vertical alignment applications on LAD-SiOx..................................................63
Figure 20: E-O curves of two identical LCoS devices filled with LC2 and improved
mixtures (88% LC2, 10% C3 and 2% 5CB) respectively........................................................64
Figure 21: Time response curves of two identical LCoS devices that used LAD-SiOx as
alignment layers and were filled with LC2 and improved mixtures (88% LC2, 10% C3
and 2% 5CB) respectively.................................................................................................................65
Figure 22: The addition of small amount of 5CB into a LC that has a large negative
dielectric anisotropy also helps to produce uniform vertical alignment on polyimide
alignment layers. Photo of SE-7511 coated cells purchased from EHC with ITO
patterns. Left cell was filled with LC1. Right cell was filled with 10% 5CB +90% LC1.
...................................................................................................................................................................66
Figure 23: Dielectric anisotropy of 5CB/LCI mixtures as a function of 5CB concentration
...................................................................................................................................................................71
Figure 24: A cartoon showing the effect of adding 5CB into LC1. Green and orange rods
represent LC1 and 5CB molecules respectively. The blue surface represents the LAD-
SiOx..........................................................................................................................................................72
Figure 25: A cartoon that shows the interaction between the LAD-SiOx and the cyano
groups. .....................................................................................................................................................73
x
Figure 26: The working principle of a TDMS (thermal desorption mass spectroscopy) ...79
Figure 27: Microscopic images of a LAD-SiOx cell filled with 1/3 LC1 and 2/3 LC2 at
different temperatures. Left side photos were taken with crossed polarizers. Right side
photos were taken with parallel polarizers. ..................................................................................82
Figure 28: Intensity of transmitted light as a function of temperature. Samples were held
between crossed polarizers with evaporation direction 45º to the polarizer axis. All cells
have the same cell gap ~20µm.........................................................................................................83
Figure 29: Temperature dependence of the anchoring transitions of 5CB/LC1 mixtures on
LAD-SiOx...............................................................................................................................................84
Figure 30: XPS spectrum showing nitrogen atoms of 5CB on LAD-SiOx. On the
spectrum of the original sample and the sample that has been baked at 49.5ºC, a peak of
Nitrogen has been observed. This implies the existence of 5CB on the SiOx surface.
However on the spectrum of the sample that has been baked at 100ºC the nitrogen peak
no longer exists, indicating that the thermal deposption temperature of 5CB is between
49.5ºC and 100ºC.................................................................................................................................87
Figure 31: (a): Thermal desorption curve of 5CB (3 samples of 5CB absorbed on LAD-
SiOx were prepared by the same methods). (b): Thermal desorption curve of C3...........88
Figure 32: The critical concentration of 5CB in the planar-to-homeotropic anchoring
transition of 5CB/LC1 mixtures as a function of temperature................................................94
Figure 33: The effect of LAD-SiOx thickness on the alignment of liquid crystal. A
commercial liquid crystal mixture with a negative dielectric anisotropy was used in the
experiment. ..........................................................................................................................................100
xi
Figure 34: The anchoring transitions in a 5CB/LC1 mixture depend on the underlying
LAD-SiOx layer thickness...............................................................................................................101
Figure 35: The effect of LAD-SiOx layer thickness on the alignment of 5CB screened by
polyimide that prefers homeotropic anchoring. ........................................................................104
Figure 36: Anchoring Transitions induced by the screening effect of polyimide on top of
LAD-SiOx surface..............................................................................................................................105
Figure 37: Two infinite surfaces separated by distance D.........................................................111
Figure 38. The cross section of a half slab of a liquid crystal cell. .........................................111
Figure 39: The critical concentration of 5CB in the homeotropic-to-planar anchoring
transition of 5CB/LC1 mixtures (shown in Figure 34) depends on the thickness of
underlying LAD-SiOx layer ............................................................................................................112
xii
LIST OF TABLES
Table 1: The preference in LC orientation by long-range/short-range torques .....................17
Table 2: The refractive index and dielectric constant data of LC1 and LC2 .........................43
Table 3 General Composition of LC1 and LC2. Column 2 and 3 show the gas
chromatography retain time of LC1 and LC2. Void indicates the missing of this
component. Column 4 shows the molecular weight of the component. ..............................44
Table 4: Additives and their effects in determining the anchoring of their mixtures with
LC2 on LAD-SiOx. Here NUP, UVA, UP stand for non-uniform planar, uniform
vertical alignment (homeotropic) and uniform planar respectively................................ 57
xiii
To my family
xiv
ACKNOWLEDGEMENTS
This work is dedicated to my wife Rong Luo and my parents. Without their ceaseless
encouragement and support, the dissertation would not have been possible.
I also feel deeply grateful to Dr. Philip J. Bos who has been my advisor on the
dissertation work. His knowledge and enthusiasm have been a constant source of
motivation to me during this endeavor. Being always very considerate and helpful, Dr.
Bos has given me the most support, not only in my research but also in many other ways.
I would like to thank Dr. James E. Anderson for his collaboration. He has provided
me with numerous insightful suggestions and discussions.
I also want to thank all my colleagues at LCI for their kind help and valuable
discussions.
My committee members deserve special thanks for their willingness to participate
and for their valuable insights.
Funding for my research was provided by HANA Microdisplay Technologies, Inc.
- 1 -
Chapter 1
Introduction
1.1 Liquid Crystalline Materials
Thanks to the blooming LCD market, the phrase liquid crystal has become more and
more known to the public during the past decade. As told by its name, liquid crystal is an
intermediate phase between isotropic liquid and crystal. Everyday experience has shown
that materials undergo a single transition from solid to liquid. However, there are many
organic materials that exhibit mesophases where the molecular ordering lies between that
of a solid and that of an isotropic liquid. Of all the types of liquid crystal phases, nematic
is one of the most important and also by far the most widely used in the LCD industries.
Generally speaking, a nematic liquid crystal is composed of rod-like organic molecules
trying to align parallel to each other. A nematic liquid crystal has long range orientational
order, but not positional order. The average direction of the molecules is labeled by a unit
vector nρ, called the director. A typical nematic liquid crystal molecule should have a
rigid elongated core and one or two flexible tails.
2
The combination of molecular orientational order and fluidity in a single phase
results in remarkable properties unique to liquid crystals. [1] Due to the anisotropy of
nematic liquid crystal molecular shape, and the long range orientational order, the
macroscopic dielectric anisotropy and optical birefringence are prevented from being
averaged to zero. And because of the fluidity (within certain temperature ranges), nematic
molecules are able to realign in an electric field to minimize the free energy. These two
features make nematic liquid crystals very useful in making electrically switchable
optical devices such as LCDs.
Depending on the sign of the dielectric anisotropy, ⊥−=∆ εεε || , nematic liquid
crystals can be divided into two categories. A nematic with a positive dielectric
anisotropy has greater polarizability along the director axis than in the direction
perpendicular to it, and the director tends to align in the direction of the external electric
field. On the contrary, a nematic with a negative dielectric anisotropy is more polarizable
in the direction perpendicular to the director axis, and its director tends to align
perpendicular to the direction of the external electric field.
In regards of optical anisotropy, ∆n = n|| − n⊥ , most nematic liquid crystals are
positive, i.e., light sees a higher refractive index for the electric field of the light along the
director direction than perpendicular to the director direction. When polarized light
passes through a liquid crystal layer it splits into two parts: ordinary light and
extraordinary light. These two may experience different optical retardation because of
3
the birefringence of liquid crystals. Also, the output state can be controlled by electrically
adjusting the liquid crystal orientation. Therefore, both phase and amplitude modulation
can be achieved using electrically addressed liquid crystal films.
1.2 Liquid Crystal Displays
Liquid crystals have found applications in many electronic devices because of their
unique electro-optical properties. Among all the applications, the Liquid Crystal Display
(LCD) is no doubt the most famous. Usually an LCD is composed of a thin layer of liquid
crystalline material sandwiched between two glass plates with transparent electrodes. By
controlling the voltage on the electrodes we can control the amount of light transmitted or
reflected by each pixel on the display. Thus, images/text can be produced.
Several liquid crystal modes are commonly used in LCD industries, including TN
(twisted nematic), STN (super twisted nematic), ECB (electrically controlled
birefringence), Pi-Cell, VA (vertical alignment), IPS (in-plane switching) and others.
These names refer to specific liquid crystal director orientation (alignment)
configurations that will be introduced in the next section.
Two very important characteristics for all LCDs are Contrast Ratio and Response
Time. Contrast ratio refers to the ratio of light intensity between a bright state and a dark
state of a LCD. Response time is essentially the time needed to switch the liquid crystal
between bright and dark states. Those two characteristics have been proven to be critical
to the performance of a LCD.
4
An LCD can be either transmissive or reflective, or as a combination – transflective.
Direct-view flat panel LCDs in the market are usually transmissive while LCDs on wrist
watches and in Rear Projection TVs (RP-TVs) are reflective. Mobile devices such as cell
phones, MP3 players and PDAs are designed for both indoor and outdoor use so most
likely transflective LCDs are used. For the purpose of this dissertation, I want to
emphasize a type of LCDs called LCoS (Liquid Crystal on Silicon). LCoS is a technology
that incorporates reflective LCD technology onto a silicon chip with a CMOS
(Complimentary Metal Oxide Semiconductor) active matrix lying underneath. LCoS may
enable the industry to manufacture high resolution RP-TVs with lower cost and better
performance. For LCoS technology, a high contrast ratio, a fast response and a long
lifetime with high light throughput are critical. Currently, TN and VA technologies are
the most widely used in LCoS.
In a vertically aligned nematic liquid crystal (VAN) cell, liquid crystals with a
negative dielectric anisotropy are utilized. The inner surfaces of the cell are pretreated
with alignment layers that give a liquid crystal orientation normal to the surface. In a
Normally Black mode, a VAN cell is sandwiched between two crossed polarizers.
Without voltage, light that goes in normal to the surface will not be affected by
birefringence. So, the black state can be really black. With voltage, the director falls
down trying to be perpendicular to the electric field and the effective birefringence
increases. The polarization of the light will be changed when passing through the cell so
that light will pass through the analyzer.
5
One major advantage of the VAN mode is its superior high on-axis contrast ratio
even without a retarder. With the help of a negative C plate, a high contrast ratio over big
viewing angle can also be achieved. In most designs, VAN requires a small pretilt angle
from the surface normal to prevent the formation of disclination lines, which seriously
lower the display quality. A larger pretilt angle also allows the liquid crystal device to
work at an increased speed. However, the pretilt angle must be small enough not to
degrade the black state and hence the contrast ratio of the display. So, the pretilt angle has
to be carefully chosen and controlled so that it balances both properties.
1.3 Liquid Crystal Alignment and the Method to Achieve the Same
Traditionally, liquid crystal alignment is achieved by unidirectional rubbing of
polyimide thin films on the surface of the electrodes. Polymer chains are believed to
align along the rubbing direction and provide an anisotropy that aligns the liquid crystal
director. Depending on the type of polyimide used, both planar and vertical alignment
can be obtained. This technique has been widely adopted in LCD manufacturing.
However, rubbing is at the same time thought to be dirty and not preferable in the
clean room because it generates a lot of particles. Rubbing may also produce cosmetic
defects such as scratches on the surface. This is very important to microdisplay
applications where any defect will be magnified, sometimes with a factor of more than 40
when projected. What’s more, the organic nature of the polyimide alignment layer makes
it susceptible to damage from strong light intensity, especially when UV light is
6
considered. This leaves the lifetime of the device questionable. Because of all the reasons
above, a rub-free, inorganic alignment layer is highly desired.
In 1971, John L. Janning first reported that obliquely deposited inorganic layers are
able to align liquid crystals. [2] Ever since, the topic has been extensively studied by
numerous researchers. The scope of the research covers many inorganic materials (such
as SiO, SiO2, CaF2, MgF2, metals) and many deposition techniques (such as thermal
evaporation, e-beam evaporation, sputtering, ion-beam etching, and chemical vapor
deposition). The resulting alignments include planar, high pretilt and vertical alignment.
The advantage of using inorganic alignment layers is not limited to a cleaner process and
better UV stability. It also provides a reliable method to produce alignment that is very
difficult to obtain using PI (such as 45° tilt and 3° pretilt of VA). Big efforts have been
spent to understand the mechanism of the alignment, which has been found to be rather
complicated. A detailed literature review of vertical alignment on inorganic layers is
provided in Chapter 2.
1.4 Overview of the Dissertation
In this dissertation we will first review previous work of the alignment of liquid
crystal on inorganic thin films. This includes the methods to produce an alignment layer,
the liquid crystal alignment behavior on inorganic alignment layers, and the mechanism
of the alignment.
7
Following the literature review will be a theory section in which we followed and
expanded a model proposed by Dubois-Violette and de Gennes to discuss the competition
between long range van der Waals forces and short range dipolar forces in determining
the liquid crystal alignment on SiOx.
After the discussion of the theory, experimental data on this topic will be presented.
The first part is a study on the physical-chemical properties of SiOx thin films and the
effects on liquid crystal alignment. The second part discusses the how two types of
materials in liquid crystal mixtures affect the alignment by shifting the balance between
long range van der Waals interactions and short range dipolar interactions. Experimental
results on the anchoring transitions caused by the shift of competition balance will be
shown.
The third part reports the temperature dependence of the observed anchoring
transition. Surface adsorption and thermal desorption is believed to cause the change in
short range interaction strength hence the balance between long range van der Waals
potential. Surface thermal desorption experiments were conducted and results are used to
add to our theory to explain the temperature dependence of anchoring transitions.
The dependence of anchoring transition on SiOx thickness is also studied and
explained by the correlation between van der Waals potential and alignment layer
thickness.
Finally, we will summarize all the experimental data and discuss how the theory
explains the phenomena we have considered.
- 8 -
Chapter 2
Theory
2.1 Introduction
Obliquely evaporated silicon oxide (SiOx) thin films have been of great interest in
the past decades for its use as liquid crystal (LC) alignment layers. It is produced by
evaporating silicon oxide source onto the target surface in vacuum. The obtained silicon
oxide thin film may vary in its Si/O ratio as well as its chemical state so it’s generally
called SiOx. Compared to the traditional rubbed polyimides (PI), SiOx is obtained using a
non-contact method that produces less cosmetic defects as well as fewer particles that can
contaminate the alignment surface. It is more UV stable. It has been found capable of
producing a wide range of pretilt angles. These advantages have caused SiOx to be
considered or implemented in applications like microdisplays and telecommunications
devices. Particularly, Large-Angle-Deposited SiOx (LAD-SiOx) has attracted interests for
its capability of producing vertical (homeotropic) alignment of liquid crystals.
9
Along with the increased interest have been efforts to understand the mechanism of
the alignment. In this chapter we will give a review of some work in that area. Based on
those previous theories and the assumptions that we found valid in our particular case
we propose an expansion to a previous model and use it to explain the mechanisms of
liquid crystal anchoring transitions on LAD-SiOx.
2.2 Review of Previous Theories
2.2.1 Short Range Interactions
Surface short range interactions have been found important in liquid crystal
alignment. For example, it was discovered that the 5 degree (shallow angle deposition)
SiOx column structures that stick out from the surface, are important for LC anchoring.
Also important are molecular groups on SiOx that has been coated with alcohol, silane or
other organic materials. The surface has been pictured as a comb with liquid crystal
molecules embedded between the molecular groups sticking out of the surface. 1,2,3 Wu et
al. have reported an interesting alignment phenomenon observed on SiOx and
successfully explained it using this theory. 4
From a more chemical-physical point of view, some other researchers have
demonstrated that the strong interfacial interactions between the surface and the surface
liquid crystal molecules give rise to the anchoring energy that determines the bulk
orientation.5,6 Those interfacial interactions may include steric interaction, charge-charge
interaction, charge-dipole interaction, dipole-dipole interaction, hydrogen bonding or
10
even chemical bonding. Since liquid crystals are often polar materials, the coupling
between the permanent dipole of a liquid crystal molecule and the surface dipoles/charges
can be significant. As a result, dipole moments may tend to be normal to the interface to
maximize their interaction. 7 This effect is essentially short range and never goes beyond
a few tens of angstroms but it could be a big contribution to the LC anchoring.
2.2.2 Long Range van der Waals Potential
The van der Waals potential between liquid crystals and an anisotropic medium has
been reviewed by previous researchers. In two classic papers8,9 Dubois-Violette and de
Gennes have shown that the more polarizable axis of liquid crystal will align parallel to
the more polarizable directions of the surface and the angular dependence can be
separated out in the potential by using a simple expression:
θ20 sinUU = (1)
Here U0 denotes the van der Waals potential with liquid crystal aligned in its
preferred direction, and θ is the angle between the two more-polarizable axes. Many
other researchers have followed the same θ2sin model10,11 or P2(cosθ) model12, 13.
More recently Lu14, Vithana15, and Kang16 et al. have shown that LC with a positive
∆ε (dielectric anisotropy) prefers parallel alignment (also called planar) on LAD-SiOx,
while LC with a negative ∆ε prefers perpendicular alignment (also called homeotropic).
Lu et al. have explained the effect by considering the difference in van der Waals
11
potential between parallel and perpendicular states, caused by the dielectric anisotropy of
LC.
2.2.3 Competition between Long Range and Short Range Forces
In reference [9], E. Dubois-Violette and P. G. de Gennes discussed the local
Fredericks transitions. A solid/nematic interface was considered, where long range van
der Waals torques favor perpendicular anchoring, while short range effects tend to induce
a parallel anchoring. The final anchoring depends on the relative strength of short range
and long range interaction. The authors proposed to use equation (2.2) to express the total
energy.
0222 sin2)(2sin)(2 θθθ
δ δ
WdzdzdKdzzuF ++−= ∫ ∫
∞ ∞
(2.2)
As shown in Figure 1, z is the distance from surface, δ is a small isotropic gap to
prevent the energy from diverging, which is in the magnitude of the size of a liquid
crystal molecule. 0θ is the angle that director deviates from the short range torque
preferred direction (surface normal) on the interface, θ is the actual anchoring angle. F is
the total free energy; )(zu is the van der Waals potential, K is the elastic constant of
liquid crystal and W is the surface anchoring energy that corresponds to short range
interactions.
Two anchoring transitions were predicted: parallel <—> conical (tilted) and conical
<—> perpendicular. These anchoring transitions are called local Fredericks transitions
12
because they are caused by local (short range) forces. Sonin et al. have successfully
demonstrated local Fredericks transitions on mica cleavages covered by an amorphous
film.17, 18
2.2.4 Topography
Topography is an important factor in liquid crystal alignment. A classic view
describes the surface of SiOx as porous columns or periodic structures. LC molecules are
believed to align parallel to the surface everywhere and the orientation of the director is
determined when the elastic distortion energy is minimized.19,20,21,22 A more recent study
by Papanek and Martinot-Lagarde has shown that other factors such as order electricity
are important in the case of porous SiOx surface.23
However, studies have shown that the porous surface morphology exists only when
the evaporation angle (the angle between the SiOx beam and substrate surface) is small.
Evaporation at a medium or larger angle (e.g. >30º) results in a more compact structure
and a smooth surface. 24 , 25 We have confirmed this using AFM (Atomic Force
Microscopy). In this paper we restrict our attention to the particular case of Large Angle
Deposited SiOx (LAD-SiOx), where we found that the elastic energy resulted from the
topography is at least one order of magnitude smaller than the measured anchoring
energy. In this case topography is unlikely to have a significant effect on the liquid
crystal alignment.
13
The fact that different liquid crystal materials may choose completely different
orientation on the same SiOx substrate also indicates a mechanism that cannot be
explained solely by the elastic distortions of the director.
14
Figure 1: Illustration of Dubois-Violette and de Gennes’ model in which long range van
der Waals torque prefers planar alignment while short range forces prefer homeotropic
alignment.
0θ
δ
Z
15
2.3 Our Theory
Our ideas are based on the model proposed by de Gennes and Dubois-Violette in
reference [9]. The theory is related to anchoring transitions that are seen on smooth LAD-
SiOx, and is made with three assumptions that have previously been accepted by many
others as discussed in the last section:
a) Short range dipolar interactions tend to align dipole moments perpendicular
to the SiOx surface.
b) Long range van der Waals interaction tends to align the more polarizable
direction of the liquid crystal with those of the alignment layer
c) We can neglect surface topography and resulting steric forces for the case of
large angle deposited SiOx alignment layers used in this study.
From the first assumption it follows that for an LC with a positive ∆ε (the dipole is
more or less along the long molecular axis); a perpendicular boundary condition is
preferred by short range dipolar interactions, while for an LC with a negative ∆ε a
parallel boundary condition is preferred because the dipole is more or less perpendicular
to the long molecular axis.
The second assumption gives the long range force preference of bulk LC orientation
as a function of dielectric anisotropy. We assume here that the in-plane polarizability of
LAD-SiOx is greater than the out-of-plane polarizability. This assumption is consistent
with the molecular structure of SiOx thin films. According to Philipp26,27 and Hohl et al.28
the molecular structure of SiOx can be described in a Random Binding Model. In the
16
model every silicon atom is combined with four other atoms (either oxygen or another
silicon) to form a matrix. Considering the dimensions of this matrix, electrons should be
easier to move in-plane than out-of-plane. Therefore, LAD-SiOx should be more
polarizable in the surface plane than along its normal direction.
As a result, a liquid crystal with a positive ∆ε tends to align parallel to the surface
but a liquid crystal with a negative ∆ε tends to align perpendicular to the surface. In both
cases the electrically more polarizable direction of the liquid crystal is parallel to the
more polarizable direction of SiOx.
The third assumption holds true in our particular case of large angle deposited SiOx.
This allows us to neglect the elastic energy distortion on SiOx surfaces.
Based on the above assumptions we can list the orientational preferences of both the
long range van der Waals forces and short range dipolar forces in Table 1. A cartoon
illustration is also shown in Figure 2. It is clear that long range van der Waals forces and
short range dipolar forces have opposite preference in the liquid crystal orientation
direction. The final liquid crystal anchoring on SiOx is determined by the competition
between the long range van der Waals forces and short range surface dipolar forces.
This hypothesis may explain many effects that were hard to explain before. For
example, it has been found that the orientation of the first layer of liquid crystal can differ
appreciably from the orientation in the bulk. Resinikov et al. reported that the first layer
(or a monolayer) of 5CB aligns perpendicularly at the liquid crystal/quartz surface, but
17
the bulk of 5CB shows parallel anchoring.29 Similar phenomena have been reported on
other substrates like polymers, crystals and glass.30,31
Table 1: The preference in LC orientation by long-range/short-range torques
Liquid crystal
dielectric
anisotropy
Long range van der
Waals force preferred
liquid crystal orientation
Short range dipolar force
preferred liquid crystal
orientation
Positive Parallel to the interface Perpendicular to the interface
Negative Perpendicular to the
interface Parallel to the interface
18
Figure 2: The preference in LC orientation by long-range/short-range forces
+ LC - LC
Short range forces
+ LC - LC
van der Waals force
19
Following the model that Dubois-Violette and de Gennes proposed in reference [9]
we start with equation (2.2) to described the free energy in the situation where long range
van der Waals torque prefers parallel anchoring while short range torques prefer
perpendicular anchoring.
We limit the consideration to either planar or perpendicular anchoring ( θθ =0 ) so
that the more complicated conical situation can be excluded. We further assume that
there’s no deformation of liquid crystal director orientation to eliminate the elastic energy.
This assumption may not be completely true but it should give us a fairly good
approximation since the short range interaction only works on the first layer of liquid
crystal. Therefore, the formula is simplified to:
θθδ
22 sinsin)(2 WdzzuF +−= ∫∞
(2.3)
Let us define ∫∞
=δ
dzzuU )( then
θθ 22 sinsin2 WUF +−= (2.4)
Here θ can only be 0 (perpendicular) or 2/π (parallel) from the surface normal.
Let us use superscript + and – to denote the material with positive and negative
dielectric anisotropy respectively. Now consider the following situations:
a) An LC with a positive ∆ε When θ = 0, 0=+F ; when θ = π /2, +++ −= UWF2
20
So, when ++ < U W , the system has lower energy in the parallel state and when
++ >U W perpendicular anchoring gives lower energy.
b) An LC with a negative ∆ε Similar to equation 2.4, for a liquid crystal that has a negative ∆ε the total energy
can be written as
θθ 22 coscos2 −−− +−= WUF (2.5)
to reflect the preference of long range and short range torque.
When 0=θ , −−− −= UWF2 , when 2/πθ = , 0=−F
So if the short range interaction is strong enough, i.e., −− > UW , a planar anchoring
is preferred. On another hand if −− < UW and van der Waals wins, a perpendicular
anchoring is preferred.
c) A mixture containing both negative and positive ∆ε LCs In a mixture that contains liquid crystals with both positive and negative ∆ε we
have to take into consideration the distribution of each component in the bulk and on the
surface. A simplified model would be two active components (one positive and one
negative) in a neutral base. Here we use x to denote the concentration of one component
in the mixture.
)/( neutralmmmmx ++= −+++ (2.6)
21
)/( neutralmmmmx ++= −+−− (2.7)
Here m is the amount of the component in the mixture.
In a liquid crystal mixture sandwiched between two LAD-SiOx, for any component,
it is safe to assume that the bulk concentration in the cell is the same as x . However, the
surface concentration can deviate from x appreciably. The surface concentration of a
component can be represented by its surface coverage ratio Θ defined as
+++ =Θ Nn / (2.8)
−−− =Θ Nn / (2.9)
Here n is the number of adsorbed molecules and N is the maximum number of the
molecules of this component that can be adsorbed, i.e., the total available sites for this
particular component. As can be seen we have assumed that the total available sites could
be different for different components because of their very different properties.
Therefore, the total energy can be expressed as
θθθθ 2222 sinsincoscos2 ++++−−−− Θ+−Θ+−= WUxWUxF (2.10)
The difference in energy between perpendicular anchoring and parallel anchoring is
++++−−−− Θ−+Θ+−=−=∆ WUxWUxFFF )]2/()0([22 π (2.11)
An anchoring transition takes place at the critical point when 0=∆F , i.e.,
++++−−−− Θ−=Θ− WUxWUx (2.12)
22
2.4 Summary
In this chapter we have reviewed some important work regarding the liquid crystal
alignment on SiOx. With certain assumptions we showed that long range van der Waals
forces and short range dipolar interactions have opposite preference in liquid crystal
alignment directions. We expressed the competition between long range and short range
interactions in the form of a model proposed by de Gennes et al. Further we expanded
this model to the case where multiple components were present with different dielectric
anisotropies. The contribution to the energy by long range and short range interactions of
each active component is assumed to be proportional to its bulk concentration and surface
coverage ratio respectively. As a result, change of the concentration or surface adsorption
properties of any component may shift the balance between long range van der Waals
interactions and short range dipolar interactions, leading to anchoring transitions. The
point where an anchoring transition happens has been given in the model as a state where
no energy difference exists between homeotropic alignment and planar alignment.
- 23 -
Chapter 3
Physical-chemical properties of LAD-SiOx thin films
3.1 Introduction
The history of SiOx as a liquid crystal alignment material started with John Janning’s
report in 1971 that obliquely evaporated SiO films caused 5CB and MBBA to align in a
preferred direction. Later it was discovered that the composition of the resulted thin film
may deviate from SiO and become SiOx where x can be between 1 and 2. Janning’s
discovery inspired great interest in the research of SiOx thin films as alignment layers
both in applications and in scientific understanding. More recently, LAD-SiOx alignment
layers found application in producing high quality VAN (vertically aligned nematic)
microdisplays for rear projection TVs. Companies such as Sony and JVC are using this
technique in mass production of products. Many other companies are trying to develop
new technologies and products using SiOx. After 20 years, SiOx alignment layers have
become a hot spot of research in the display industry.
24
SiOx alignment layers possess many unique merits when compared to other
alignment layers. For instance, SiOx layers are able to produce a wide range of pretilt
angle that are extremely difficult to produce on traditional polyimide alignment layers.
Another desirable feature of SiOx alignment layers is that the deposition process is
“clean”. Compared to rubbing polyimides, SiOx deposition doesn’t generate so many
particles that contaminate the alignment surface. A rub-free process also prevents the
devices from cosmetic defects such as scratches that can be disastrous to microdisplay
applications. Thanks to its inorganic nature, SiOx alignment layers are also less sensitive
to UV. Because of these unique advantages, SiOx has been widely considered in
applications such as STN, VAN, pi-cell and dual frequency liquid crystal devices.
For VAN applications, silicon oxide films evaporated at a relatively large angle
(>30º w.r.t. the surface) are normally used. Though applications have been successful, the
properties of the LAD-SiOx alignment layers and their effects on liquid crystal alignment
are still poorly understood.
In this chapter we will discuss the properties of LAD-SiOx thin films used in our
experiments.
3.2 Experimental Method
3.2.1 Inorganic Alignment Layer Preparation
Two types of silicon oxide films were used: thermally evaporated SiOx and e-beam
evaporated SiO2. For the purpose of simplicity, I will hereafter name them as SiOx and
25
SiO2 respectively. But they should be strictly differentiated for reasons that will be
discussed later.
SiOx alignment layers were prepared by thermally evaporating silicon monoxide
(SiO) powders (purchased from Kurt J. Lesker Company) onto substrates. Though the
equipment is able to do oblique evaporation with any angle to the substrate surface we
did all our depositions at a large angle of incidence (usually 40°-50° w.r.t. the substrate
surface). This particular range of angles has been shown by previous researchers to be
effective in producing vertical alignment of liquid crystal. The thickness of coating is
measured in-situ by an oscillating quartz crystal thickness monitor. The reading of the
thickness monitor has been calibrated by ellipsometry measurement data. The deposition
rate was controlled to be 2~3Å/s. Residual pressure in the deposition chamber was
controlled by back-bleeding air through a needle valve.
Electron beam (e-beam) provides a source of heat with much higher temperature. So
instead of silicon monoxide, silicon dioxide is typically used as the source of evaporation.
In our experiments, SiO2 films were prepared by evaporating quartz pellets by e-beam
using the same process parameters as those used for thermal evaporation. The e-beam
evaporator we used was of the same basic geometry as the thermal evaporator. Thus the
two evaporators have almost the same geometry and should produce substrates for a fair
comparison. E-beam evaporated SiO2 thin films were used only in a few cases in our
study, mainly to compare with SiOx.
26
3.2.2 Thin Film Characterization Method
3.2.2.1 AFM
AFM is a widely used technique for surface characterization. It consists of a micro
scale cantilever with a sharp tip (probe) that is used to scan the specimen surface. When
the tip is brought into proximity of a sample surface, forces between the tip and the
sample lead to a deflection of the cantilever according to Hooke’ Law. Typically, the
deflection is measured using a laser spot reflected from the top of the cantilever into an
array of photodiodes. The sample is mounted on a piezoelectric tube that can move the
sample in the z direction for maintaining a constant force, and the x and y directions for
scanning the sample. The resulting map of s(x,y) represents the topography of the sample.
Usually there are two types of scan methods: contact mode and tapping mode. The
former uses static probe while the latter uses probe oscillating at close to its resonance
frequency.
In our measurements we used both contact and tapping modes to scan fresh thin film
samples in order to obtain clear images of the surface morphology. Then the images were
analyzed by software to obtain cross-section plots and statistical information such as
surface roughness, average horizontal domain size, average peak-to-peak height, etc. In
anisotropy measurement, samples were first scanned along the evaporation direction, then
along the direction perpendicular to it. For each scan, surface roughness was calculated.
Roughness anisotropy is defined as the difference between the results of the two scans.
27
3.2.2.2 XPS
The XPS technique is based on the photoelectric effect that electrons eject from a
surface when photons impinge upon it. Al Kα (1486.6eV) or Mg Kα (1253.6eV) are
often the photon energies of choice. The energy of the photoelectrons leaving the sample
is determined using a Concentric Hemispherical Analyzer and this gives a spectrum with
a series of photoelectron peaks. The binding energies of the peaks are characteristic of
each element and its local environment. The peak areas can be used (with appropriate
sensitivity factors) to determine the composition of the material’s surface. The shape of
each peak and the binding energy can be slightly altered by the chemical state of the
emitting atom. Hence, XPS can provide chemical bonding information as well.
The XPS technique is highly surface specific due to the short range of the
photoelectrons that are ejected from the solid. By using different incident angles of X-ray,
photoelectrons excited from different depths under the surface can be collected. Thus a
depth profile of the sample can be obtained using an Angular Resolved XPS.
In our study samples of LAD-SiOx and LAD-SiO2, thin films were deposited on
glass substrates and measured using Al Kα as the photon source. Spectra were analyzed
to give chemical state, atomic ratio and other information. Other than the depth profiling,
all measurements were done using a 45° angle.
28
Figure 3: The Working Principle of AFM
Figure 4: The working principle of XPS
Electron Energy Analyzer
X-ray source
Pump
Sample
29
3.3 Experimental Results and Discussions
3.3.1 Surface Topography and Anisotropy
Surface topography of obliquely evaporated LAD-SiOx was examined by AFM
(Atomic Force Microscopy). Figure 5 shows some typical AFM images of LAD-SiOx
thermally coated at a large angle of incidence. A few points can be seen from the images.
First, the surface topography suggests that LAD-SiOx thin films are possibly
composed of densely packed column structures and the direction of column growth is
close to the surface normal.
Second, the LAD-SiOx surface is very smooth. The cross section analysis (Figure
5(d)) of the sample shows that a typical topographic feature on the surface is around 5nm
in height but 200nm in width (note that the horizontal and vertical scaling in the figures
are very different). The measured RMS roughness is generally around 1nm. So, it’s more
close to reality to picture the LAD-SiOx surface as a smooth ground with pebbles on it,
rather than hills and valleys that are typically seen in glancing angle (such as 5º)
deposition. On the other hand, a typical liquid crystal molecule is only about 2.5nm in
length and 0.5nm in diameter. With this kind of geometry it is difficult to produce any
significant elastic distortion in LC director field.
We also studied the evolution of surface topography as we increased the LAD-SiOx
layer thickness. The results are shown in Figure 6. Other than a tiny decrease in the low
thickness region, little change has been seen in either surface roughness or anisotropy
(defined as the difference in RMS roughness when sample is scanned along evaporation
30
direction and perpendicular to evaporation direction) when the thickness increases from
~30nm to ~350nm.
3.3.2 Stoichiometry and Surface Properties
The stoichiometry of LAD-SiOx/SiO2 thin films was studied by XPS (X-ray
photoelectron spectroscopy ). From each spectrum, atomic ratio of each element in the
sample can be calculated. As shown in Fig. 5(a) e-beam evaporated LAD-SiO2 has an
O/Si atomic ratio very close to 2/1. But for thermally evaporated LAD-SiOx we have seen
ratios from 1.2 to 1.7, depending on the deposition conditions. The data from atomic ratio
shows that thermally evaporated SiOx has an oxygen deficient chemical structure. The
Angle-Resolved XPS also allows us to do a depth profile of the stoichiometry. Increasing
the photoelectron takeoff angle by rotating the sample in the energy dispersive plane of
the analyzer reduces the sampling depth. Using this technique we were able to measure
the atomic ratio from the top of the surface to ~1.5nm underneath. The results we
obtained showed no significant difference in atomic ratio of Si and O.
The analysis software of XPS has the capability to fit the Si2p peak with the
characteristic Si2p peak from crystal SiO2, SiO, and silicon. The results shown in Fig 5(b)
imply that e-beam evaporated LAD-SiO2 is more close to crystal SiO2 in its chemical
structure while thermally evaporated LAD-SiOx has a big contribution from SiO and even
a small contribution from silicon. As we all know in a crystal SiO2 each Si atom bonds
with 4 oxygen atoms to form a network of tetrahedrons. However, in the case of SiOx,
there will be many unoccupied silicon orbits due to the lack of oxygen. Since the depth
31
profile of atomic ratio shows no obvious difference between the top surface and
underneath, we believe that the LAD-SiOx surface also has many dangling bonds or
empty orbitals that may attract nearby dipoles.
As a summary, the e-beam evaporated LAD-SiO2 surface is more passive compared
to the oxygen-deficient thermally evaporated LAD-SiOx surface, which may have lots of
empty Si orbits and dangling bonds on the surface.
32
Figure 5: AFM images of LAD-SiOx thermally evaporated at a medium angle. (a): 10µm
x 10µm tapping mode 3D image (b): 5µm x 5µm tapping mode 3D image (c): 3µm x
3µm contact mode 2D image of friction (d): Cross-section analysis
(a)
(c) (d)
(b)
33
Figure 6: (a): RMS Roughness of LAD-SiOx surface as a function of layer thickness (b):
Anisotropy in surface roughness as a function of layer thickness
RMS Roughness Anisotropy(Difference between RMS along and perpendicular to evaporation direction)
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0 50 100 150 200 250 300 350 400
SiOx Layer Thickness/nm
Delta RMSDelta Ravg
Roughness vs. Thickness
0
0.25
0.5
0.75
1
1.25
1.5
0 50 100 150 200 250 300 350 400
SiOx Layer Thickness/nm
along evaporation
perpendicular toevaporation
(a)
(b)
34
Figure 7: XPS spectrum of thermally evaporated LAD-SiOx and e-beam evaporated LAD-SiO2,
measured at 45º take-off angle. Atomic ratio of Si and O of the sample can be calculated from the
corresponding area of the peak. Signal of carbon is from the residual of CO2 or hydrocarbon
contaminations on the sample surface.
SiO1.95
SiO1.50
(a)
(b)
35
Figure 8: XPS spectrum analysis of silicon (Si2p) in (a) e-beam evaporated LAD-SiO2
and (b) thermally evaporated LAD-SiOx. The blue line is the characteristic peak of Si in
SiO2; The cyanic line is the characteristic peak of Si in SiO; The magenta line is the
characteristic peak of Si in Si crystal; The black line is the measured Si peak; The red line
is the synthetic peak based on characteristic Si peak in SiO2, SiO and Si crystal.
(b)
36
3.4 Summary
AFM data reveals that SiOx thin films evaporated at a medium or large angle exhibit
densely packed columnar structures in the direction close to the surface normal. The
surface roughness and anisotropy are so small that we believe surface topography and
elastic distortion energy is unlikely to have a significant effect on the anchoring of the
liquid crystals on LAD-SiOx thin films. It is also extremely hard to use topography to
explain all the anchoring effects we observed in the experiments. A mechanism that
shows a closer relationship between the physical-chemical properties of SiOx and the
liquid crystal molecules must be considered.
The stoichiometry of SiOx also plays an important role in the liquid crystal anchoring.
Unoccupied orbits or dangling bonds on the LAD-SiOx surface tend to interact with the
dipole moment strongly. So, a saturated surface will be more stable and less interactive to
liquid crystal molecules compared to an unsaturated one. From XPS data we can see that
on thermally evaporated the LAD-SiOx surface, silicon atoms are not saturated with
oxygen, leaving many orbits accessible to liquid crystal dipoles. On the other hand, e-
beam evaporated LAD-SiO2 is more like a crystal structure with each Si bonded to 4
oxygen atoms. As a result we can expect a stronger short range surface interaction
between the alignment layer and the liquid crystal on LAD-SiOx, compared to on LAD-
SiO2.
- 37 -
Chapter 4
Anchoring Transitions on LAD-SiOx Due to the Change in Liquid Crystal Composition
4.1 Introduction
In the previous chapter we reviewed experimental data on LAD-SiOx alignment
layers and concluded that the topography is unlikely to produce significant effects on
liquid crystal alignment in our particular case of medium angle evaporation. Another
piece of evidence that topography should not be held responsible for the entire alignment
phenomenon on SiOx is the material dependence of the alignment. In other words,
different liquid crystals tend to align in different ways on LAD-SiOx. This cannot be
explained using the model of elastic energy minimization. Here in this chapter we will
report experimental observations of the material dependence of liquid crystal alignment
on SiOx. Further, we will demonstrate anchoring transition phenomena due to the change
of the relative ratio of two components in liquid crystal mixtures.
The effect will be explained using the theory we introduced in Chapter 2, by considering
the competition between the long range van der Waals interactions and the short range
dipolar interactions. A novel method that may produce improved liquid crystal mixtures
38
for vertical alignment applications will be proposed based on our discovery involving
anchoring transitions.
4.2 Experimental Methods
4.2.1 Materials
Silicon monoxide powder (EVMSIO-1065B, >99.99% purity) purchased from Kurt J.
Lesker was used for the evaporation.
Commercial liquid crystals from Merck were used in the experiment. LC1 and LC2
(part number intentionally omitted for the proprietary of the research sponsor) are liquid
crystal mixtures with negative dielectric anisotropy. Table 2 lists the refractive index and
dielectric constant of these two mixtures. Table 3 lists the general composition of these
two mixtures. Notice that LC1 has a very large negative value of dielectric anisotropy.
Another liquid crystal known as 5CB or K15 (4-cyano-4-n-pentylbiphenyl), also
purchased from Merck was used. 5CB, as shown in Figure 9(a) is a small linear molecule
with a strong polar group on one end. Therefore, it possesses a strong longitudinal dipole
and a positive dielectric anisotropy. All other materials used in the experiments were
synthesized in-house. Among them, 1 – ethoxy – 4 – (4’ – trans -
propylcyclohexylcarboxy) - 2, 3 - dicyanobenzene (hereinafter referred to as C3) is of
particular importance. As shown in Figure 9(b), each C3 molecule has 2 cyano groups on
one side producing a large dipole moment in the direction perpendicular to the molecular
long axis. Also because of the cyano groups and the conjugation with benzene rings, C3
and compounds that have similar structures have been reported to have huge negative ∆ε
39
and are used in commercial liquid crystal mixtures as dopants to increase the magnitude
of the negative dielectric anisotropy. 32,33
4.2.2 Sample Preparation
1350Ǻ-thick SiOx films were deposited onto clean glass substrates at 45º by thermal
evaporation. Residual pressure was controlled to be around 1.0x10-5 torr by back-
bleeding air through a needle valve. Coated substrates were assembled into 20µm-thick
cells with anti-parallel deposition directions on the top and bottom plates. Liquid crystal
was forced into the cell under vacuum by capillary force at room temperature. Following
filling, the cells were sealed.
4.2.3 General Examination Methods and Definition for Alignment Quality
After filling, the liquid crystal cells were first examined on a light table between
crossed polarizers. With vertically aligned liquid crystals, cells should always look dark
when rotated. For planar cells, bright-dark alternation will be observed when rotated. A
cell is defined as uniform if all of following criteria have been satisfied: 1) More than
80% of the cell area has uniform brightness or darkness observed by visually; 2) Choose
5 spots in the uniform area that are at the area’s center and 4 corners. For a planar cell,
measure the extinction angle on each spot. The maximum difference between two
extinction angles should be smaller than 2°. Or, for a vertically aligned cell, measure the
40
pretilt angle on each spot. The maximum difference between two pretilt angles should be
smaller than 1°. Otherwise a cell is defined as non-uniform.
4.2.4 Pretilt Measurement
The pretilt angle of liquid crystals confined in a cell was measured by one of two
methods: Conoscopy and Crystal Rotation.
The Conoscopy method was mainly used to measure a homeotropic cell with a small
pretilt in which case an off-centered uniaxial cross can be recognized under conoscopic
observation. There’s a simple relationship that determines the pretilt angle:
(r/R)/N.A. = no sinθ (4.1)
Here r is the distance between the conoscopic image center and the cross center. R is
the diameter of the conoscopy. Detailed discussion of this method can be found in
reference [34].
The Crystal Rotation method has been used in our experiments to measure larger
pretilt angles that the conoscopy method is not capable of measuring due to the limitation
of microscope numerical aperture. Details of this method are available in reference [35].
4.2.5 Dielectric Anisotropy Measurement Method
20µm-thick empty cells with 1.0 cm2 patterned ITO electrodes were made in our lab.
Accurate cell gap thickness was measured from the interference patterns formed by the
41
reflection from top and bottom surfaces of the gap. The cell uniformity was also carefully
examined by measuring the gap thickness at the center and at four corners of the
patterned electrode. Only those cells with less than 2% thickness variation were used in
the experiments. Spin-coated polyimides were used as alignment layers using the
standard soft bake-hard bake procedure. For the homeotropic cell, SE-7511 was used, and
for the planar cell, SE-2555 was used. Cells were filled with liquid crystal and then
examined for uniformity. Pretilt angle of each cell was measured on a center-plus-four-
corners basis, as described before. The results show pretilt angle to be less than 1° for
planar cells and greater than 89° for homeotropic cells, all angles measured from the
surface.
For each material, the impedances (real and imaginary parts) of both planar cell and
vertical cell were measured on a Hewlett Packard 4284A 20Hz-1MHz precision LCR
meter as a function of frequency, ranging from 1 kHz to 1 MHz. ∆ε was calculated using
the equations of (2), (3), (4), and (5):
;1
0εεωω Aid
CiiZZZ ir −=−=+= (4.2)
;)(00 ir iZZAi
dZAi
d+
−=−=εωεω
ε (4.3)
)( ;
)( 220
220 ir
ri
ir
ir ZZA
dZZZA
dZ+
=+
=εω
εεω
ε ; (4.4)
)()(|| planarvertical rr εεεεε −=−=∆ ⊥ , (4.5)
42
Here Z is the impedance, ω is the angular frequency, C is the capacitance, A is the
area of the electrode, d is the cell gap, ε is the dielectric constant, and ε0 is the dielectric
permittivity of the free space. Subscript r and i stand for real and imaginary part
respectively. And subscript || and ⊥ stand for parallel and perpendicular to molecular
long axis respectively.
4.2.6 Birefringence Measurement Method
Birefringence of the liquid crystal mixtures were obtained from the optical retardation
measurements on the planar cells. The same center-plus-four-corners examination on cell
thickness uniformity and pretilt was performed and only cells with less than 2% thickness
variation and less than 1° pretilt (from the surface) were allowed. The optical retardation
of a cell was measured by the standard Senarmont Technique. Birefringence was
calculated from the optical retardation using equation (4.6):
dn δλ ⋅
=∆ (4.6)
Here λ is the wavelength of light, which is 632.8nm in our case, δ is the optical
retardation and d is the cell gap.
4.2.7 Electro-Optical Curve and Response Time Measurement Methods
Electro-optical curves and response times of tested cells were measured using a
home-built setup and software. The test cell is placed between crossed polarizers with its
43
surface projection of the easy axis making a 45º angle with the polarization axis. Light
coming out from a 632.8nm He-Ne laser passed through the setup and passed to a
detector. For the E-O curve, 1 kHz AC was applied to the cell with rms voltage ramping
from 0 to 10V. The transmitted light intensity was detected as a function of the ramping
voltage. For response time, the tested cell was switched between 0 and 5V at 1 kHz. The
detector recorded the transmitted light intensity as a function of time. All measurements
were done at 50°C.
Table 2: The refractive index and dielectric constant data of LC1 and LC2
ne no ∆n ε|| ε┴ ∆ ε
LC1 1.6567 1.4920 0.1647 4.5 10.2 -5.7
LC2 1.6560 1.4920 0.1640 3.7 6.4 -2.7
44
Table 3 General Composition of LC1 and LC2. Column 2 and 3 show the gas chromatography retain time of LC1 and LC2. Void indicates the missing of this
component. Column 4 shows the molecular weight of the component.
45
Figure 9: Chemical structure of (a) 5CB and (b) C3
C
(a)
N
O
C
O
O
C C
N N
(b)
46
4.3 Experimental Results
4.3.1 The Effect of Large Longitudinal Dipole
A commercial mixture LC1, which has a large negative dielectric anisotropy (∆ε = -
5.7) was filled into LAD-SiOx cells. While a liquid crystal with a moderate negative
dielectric anisotropy will typically align vertically on LAD-SiOx, LC1 on the contrary
assumes an orientation parallel to it.
5CB was filled into identical cells and was found to align parallel to the SiOx surface
as well. However in this case 5CB has a positive dielectric anisotropy.
Next we mixed 5CB into LC1 and filled the mixtures into LAD-SiOx cells. At room
temperature, when the mixture contains less than 3% 5CB (by weight, the same in the
following) it aligns parallel to the LAD-SiOx surface. When the concentration of 5CB
reaches about 3% an anchoring transition takes place that brings the mixture into
homeotropic alignment. Not until we increase the concentration of 5CB to about 55%
does another transition happen and switch the LC anchoring to parallel again. Figure 10
shows a photo of cells filled with liquid crystal mixtures of 5CB and LC1, observed
between crossed polarizers on a light table. Figure 11 plots the tilt angle of the LC
director (w.r.t. substrate surface) as a function of 5CB concentration.
47
Figure 10: Anchoring transitions from parallel to homeotropic to parallel again as the
concentration of 5CB in the mixture with LC1 decreases. From top left to bottom right:
pure 5CB, 50% 5CB, 25% 5CB, 10% 5CB, 5% 5CB, and pure LC1. Photo taken with
cells placed between crossed polarizers on a light table.
48
Figure 11: Anchoring transitions of liquid crystal mixtures (5CB/LC1) on LAD-SiOx due
to the change of the ratio of two components
0 10 20 30 40 50 60 70 80 90 100-10
0
10
20
30
40
50
60
70
80
90
100
Tilt
angl
e fro
m s
urfa
ce
C oncentra tion o f 5C B (w eight% ) in the LC 1/5C B m ix ture
49
4.3.2 The Effect of Large Lateral Dipole
As described in 4.2.1 C3 has two cyano groups on one side of the molecule. It also
has a huge negative dielectric anisotropy, which makes it desirable as an additive used in
making negative ∆ε commercial liquid crystal mixtures. A commercial liquid crystal LC2
from Merck (∆ε = - 2.7) was used to mix with C3. The reason we chose LC2 is that C3
has a good solubility in LC2 so the anchoring transitions can be more clearly
demonstrated.
LC2 by itself chooses homeotropic orientation on LAD-SiOx. We were not able to
know how C3 aligns on LAD-SiOx because it doesn’t have a nematic phase by itself. We
increasingly added C3 into LC2 and filled the mixtures into LAD-SiOx cells. When the
concentration of C3 is equal to or less than 5% the mixture still aligns perpendicular to
the surface. However, starting from 6% a transition takes place and finally changes the
LC anchoring into planar when the concentration of C3 is equal to or larger than 8%, as
can be seen in Figure 12. The LC anchoring can also be swung back to homeotropic with
the addition of small amount of 5CB to the mixture composed of LC2 and more than 8%
of C3. Figure 13 shows the transitions indicated by the change of LC tilt angle. Also seen
from the Figure 14 is for mixtures with higher concentration of C3, larger amount of 5CB
is needed to trigger the transition.
For the same experiment we have also used e-beam evaporated SiO2 as the alignment
layer. The SiO2 layer was produced with identical thickness and deposition angle to the
thermally evaporated SiOx. Figure 15 shows that on e-beam evaporated SiO2, more C3 is
50
needed than on thermally evaporated SiOx to cause its mixture with LC2 to change from
homeotropic alignment to planar alignment.
51
Figure 12: The addition of C3 into LC2 leads to an anchoring transition of liquid crystal
on LAD-SiOx from homeotropic to planar
0 5 10 15 20
0
20
40
60
80
100
Tilt
angl
e fo
rm s
ubst
rate
sur
face
Concentration of C3 (weight%) in the mixture of C3 and LC2
52
Figure 13: The addition of 5CB into the mixture of C3 and LC2 causes an anchoring
transition from planar to homeotropic on LAD-SiOx
-2 0 2 4 6 8 10 12 14 16
0
20
40
60
80
100
Concentration of 5CB (weight%) in the mixture of C3, 5CB and LC2
Pre
tilt a
ngle
from
sub
stra
te s
urfa
ce/d
egre
e
5.0% C3 7.5% C3 10.0% C3 12.5% C3 15.0% C3
53
-2 0 2 4 6 8 10 12 14 16
0
1
2
3
4
5C
ritic
al A
mou
nt o
f 5C
B (w
eigh
t%)
Concentration of C3 (weight%)
Figure 14: The correlation between the concentration of C3 and the critical amount of
5CB that is needed to maintain homeotropic alignment of C3/5CB/LC2 mixture on LAD-
SiOx
54
Figure 15: On E-beam evaporated SiO2, more C3 is needed than on thermally evaporated
SiOx to cause its mixture with LC2 to change from homeotropic alignment to planar
alignment
0 5 10 15 20
0
20
40
60
80
100
Tilt
angl
e fro
m s
urfa
ce/d
egre
e
Concentration of C3 (weight% )
Thermal Evaporated SiOx E-Beam Evaporated SiO2
55
4.3.3 The effect of varying the molecular structure of the additives
Through our experiments previously described in 4.3.1 and 4.3.2 we found that C3
and 5CB play a keen role in determining the alignment of liquid crystal mixtures on
LAD-SiOx. To understand how the chemical structure and physical properties of the
additives affect the anchoring, several compounds were carefully chosen to use in the
experiments. Their chemical structures are shown in column 2 of Table 4. They have
similarities as well as differences in many ways to 5CB and C3. Specifically, the first 5
compounds have a cyano group along the molecular long axis and thus strong
longitudinal dipole moments; but have different molecular length, shape, number of rings,
and types of linkage groups. Compound 6 has similar structure to compound 5 except that
the cyano groups are in the direction perpendicular to the molecular long axis, so it has a
strong lateral dipole moment. Compound 7 (C3) is the same as compound 6 except that it
has a cyclohexyl ring instead of benzene. Compound 8 to 13 are all similar to compound
1 but their functional groups along the molecular long axis are less polar or even non-
polar. Compound 14 has not only less longitudinal polarity, but also a cyclohexyl ring
instead of benzene. Except 5CB and 8CB (compound 1 and 2), all compounds are
synthesized in our institute. With these dopants, comparisons become possible between
their polarity, molecular shape, dielectric anisotropy, electronic conjugation, and other
properties.
Additives are mixed with LC1 to observe the effect on anchoring. The experimental
results are listed in Table 4.
56
Two conclusions can be made from the results. First, a small amount of a material
that has a cyano group at the end of its molecular long axis tends to promote homeotropic
alignment. This is clearly demonstrated in Figure 16 where a bunch of additives that have
similar structure to 5CB were tested for the effects on LC1 alignment. Second, a small
amount of a material that has cyano groups on the side of the molecular axis tends to
promote planar alignment.
Figure 17 shows a photo of two cells observed between cross polarizers. The left one
was filled with a mixture of LC1 and a material that has a longitudinal cyano end-group.
The right hand side one was filled with a mixture of LC1 and a material that contains two
lateral cyano groups. As a result, the left cell is uniform homeotropic and the right cell is
uniform planar.
57
Table 4: Additives and their effects in determining the anchoring of their mixtures with
LC2 on LAD-SiOx. Here NUP, UVA, UP stand for non-uniform planar, uniform vertical
alignment (homeotropic) and uniform planar respectively.
Compound Host Liquid Crystal Weight % of dopant Alignment
LC1 0 NUP
Dopant Weight % of dopant Alignment
1 (5CB) CN
From 2.5% to 50% UVA
2 (8CB) CN 5%, 10% UVA
3 CN
20% UVA
4 N CN
O 10% UVA
5 O
O
O CN 10% UVA
6 O C
O
O C
O
O
NC CN
5% UP
7 (C3) O
C
O
O
C C
N N
5%,10% UP
8 Br 5%, 25% NUP
9 O 4%, 10% NUP
10 10%, 33.3% NUP
11 10%, 25% NUP
12 O O
4%, 10% NUP
13 O 5%, 10% NUP
14 O
4%, 10%, 33.3% NUP
58
Figure 16: Alignment of mixtures with different additives of LC1 on LAD-SiOx,
photographed between crossed polarizers on a light table. From top left to bottom right
cells are filled with: LC1; 10%C5-Ph-Ph-CN (5CB); 10%C5-Ph-Ph-O-C2; 5% C5-Ph-
Ph-Br, 10% C3-Cyclohexyl-Ph-O-C2 (PCH302); 10% C5-Ph-Ph; 10%C6-Ph-Ph-C5.
Here Ph represents a phenyl (benzene) ring; C stands for carbon; O stands for oxygen and
Br stands for bromine.
59
Figure 17: The effect of cyano groups on the liquid crystal anchoring on LAD-SiOx. Left:
20% C7-Cyclohexyl-Ph-CN; Right: 5% C3 (C3- Cyclohexyl-COO-Ph(-2CN)-O-C2)
60
4.3.4 A Method to Make Improved Liquid Crystal Mixtures for Vertical Alignment
Applications.
Electro-optical devices using vertically aligned liquid crystals with a negative
dielectric anisotropy (VAN) have been widely used in many applications because of their
high contrast ratio. To achieve lower driving voltage and faster response, liquid crystals
with large ∆ε are preferred. Unfortunately, it is well known that these types of liquid
crystals are very difficult to align vertically on SiOx, sometimes even on polyimides.
However, the experimental discovery we discussed in previous sections points out a
potential way to solve this problem. If an appropriate amount of a positive dielectric
material, such as 5CB is added to the host material that has a large negative dielectric
anisotropy, uniform vertical alignment can be easily achieved.
Commercial liquid crystal mixtures that have negative ∆ε are generally made by
mixing highly negative dopants into a neutral or slightly positive base liquid crystal
mixture. As an example, we will use a mixture of LC2 and C3 to explain how the method
we propose may help to improve the liquid crystal properties for VAN applications.
In section 4.3.2 we discussed the effect of introducing C3 into base material LC2. We
found that if greater than 4% of C3 was added, the alignment of the mixture deviated
from homeotropic toward planar. This effect practically prohibited us from producing
useful liquid crystal mixtures with LC2 and C3 with a larger negative ∆ε. We also
reported that the addition of 5CB allowed mixtures of C3 and LC2 to form vertical
alignment that they couldn’t do originally. Though 5CB exhibits a positive ∆ε itself the
overall effect still shows improved ability to produce vertical alignment with a larger
61
negative ∆ε. Figure 18 shows that improvement can be achieved in the magnitude of the
negative value of ∆ε vs. the amount of added 5CB. Also found was an improvement in
the birefringence as shown in Figure 19.
We made two 1.3µm-thick reflective liquid crystal cells using LAD-SiOx as the
alignment layer. One was filled in with LC2. Another was filled with the mixture of 88%
LC2, 10% C3 and 2% 5CB. The electro-optical and time response curves were measured.
As shown in Figure 20, the device with the improved liquid crystal mixture has a lower
threshold voltage and a higher optical retardation. The response times shown in Figure 21
are roughly the same, but since the improved LC has higher birefringence, the device
could be made thinner to achieve a faster response time.
Another interesting disclosure is that the addition of a small amount of 5CB or similar
materials into a liquid crystal with a large negative dielectric anisotropy also helps to
produce uniform vertical alignment on polyimide alignment layers. We used two
identical empty cells from EHC with homeotropic polyimide coatings inside. One was
filled with LC1 and another one was filled with the mixture of 10% 5CB and 90% LC1.
As shown in Figure 22, both cells look like vertical alignment. But closer examination
reveals that the one filled with LC1 has higher pretilt (bright) region around the gasket,
the ITO pattern and at the filling port. But the one filled with the 5CB mixture shows
almost perfect uniform vertical alignment.
62
Figure 18: The addition of 5CB enables the mixture of LC2 and C3 to obtain uniform
vertical alignment on LAD-SiOx with a greater negative dielectric anisotropy.
-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 162.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
4.2
4.4
4.6
4.8
5.0
5.2
|∆ε|
Concentration (weight%) of C3 in the mixture of C3, 5CB and LC2
Without 5CB, uniform vertical alignment cannot be achieved for C3 > 4%
With critical amount of 5CB
63
-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0.12
0.14
0.16
0.18
Without 5CB, uniform vertical alignment cannot be achieved for C3 > 4%
With critical amount of 5CB
Concentration (weight%) of C3 in the mixture of C3, 5CB and LC2
∆n
Figure 19: The addition of 5CB also allows higher birefringence of the LC2/C3 mixture
to be used for vertical alignment applications on LAD-SiOx
64
Figure 20: E-O curves of two identical LCoS devices filled with LC2 and improved
mixtures (88% LC2, 10% C3 and 2% 5CB) respectively.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
0
20
40
60
80
100
Nor
mal
ized
Lig
ht In
tens
ity
Voltage/V
LC2 Improved mixture of
10% C3, 2% 5CB and 88% LC2
65
Figure 21: Time response curves of two identical LCoS devices that used LAD-SiOx as
alignment layers and were filled with LC2 and improved mixtures (88% LC2, 10% C3
and 2% 5CB) respectively.
-1 0 1 2 3 4 5 6 7 8 9 10 11
0
10
20
30
40
50
60
70
80
90
100
Nor
mal
ized
ligh
t int
ensi
ty
Time/ms
LC2 switch on LC2 switch off Improved mixture switch on Improved mixture switch off
66
Figure 22: The addition of small amount of 5CB into a LC that has a large negative
dielectric anisotropy also helps to produce uniform vertical alignment on polyimide
alignment layers. Photo of SE-7511 coated cells purchased from EHC with ITO patterns.
Left cell was filled with LC1. Right cell was filled with 10% 5CB +90% LC1.
67
4.4 Discussions
4.4.1 The Effect of Large Longitudinal Dipole
In 4.3.1, we showed that the addition of small amount of 5CB into LC1 (which has a
large negative ∆ε) changes the LC anchoring on SiOx from parallel to vertical. But,
further increasing the amount of 5CB changes the anchoring back to parallel again. We
have discussed the competition between long range van der Waals forces and short range
dipolar forces in the chapter on Theory. Here we continue the discussion to explain the
experimental data.
Liquid crystal mixtures with a negative ∆ε are usually obtained by mixing a base LC
with materials that have very high negative ∆ε, such as C3. Since LC1 has a large
negative ∆ε we expect it to contain a relatively large amount of negative additives. The
effects of doing so on LC anchoring are two-fold. On one hand an increased ∆ε will
increase the anisotropy in van der Waals potential, making vertical alignment a more
preferred LC anchoring on LAD-SiOx. On the other hand, the short range interaction
between LAD-SiOx surface and negative additives such as C3 also increases, but with
parallel orientation as its more favorable anchoring direction. In the case of LC1, short
range dipolar interaction exceeds the long range van der Waals interaction so liquid
crystal aligns parallel to the SiOx surface.
5CB is a small molecule with a large longitudinal dipole moment and a positive ∆ε. It
favors parallel anchoring by van der Waals forces but vertical anchoring by short range
dipolar forces. When a small amount of 5CB is added to LC1 5CB molecules may bind
68
with the SiOx surface preferentially, which substantially changes the anchoring
preference of short range forces towards favoring homeotropic alignment. Due to the
limited amount of 5CB in the mixture, bulk properties are unlikely to be noticeably
altered so the long range force still favors homeotropic anchoring. Hence both long range
and short range interactions agree on the anchoring direction homeotropic alignment is
obtained. Further increasing the amount of 5CB makes the ∆ε of the bulk LC more
positive and turns the van der Waals potential preference towards parallel anchoring. This
can be seen in Figure 23, in which dielectric anisotropy of the mixture was measured as a
function of 5CB concentration. However since the SiOx surface becomes more or less
saturated with 5CB molecules the short range interaction doesn’t increase. As a result
long range van der Waals forces eventually prevail and the LC changes back to parallel
anchoring.
This process of anchoring transition is illustrated in the cartoons of Figure 24.
4.4.2 The Effect of a Large Lateral Dipole
In the experiment described in 3.3.1.2, C3 is mixed with LC2 causing an anchoring
transition of the liquid crystal from homeotropic to planar. Adding 5CB into the mixture
shifts the anchoring back towards vertical. This effect can also be explained by the theory
we proposed in Chapter 2.
C3 has a two cyano groups hence a very large dipole moment perpendicular to its
molecular long axis. The addition of C3 into LC2 leads to the increase of short range
69
dipolar interaction between the liquid crystal and the LAD-SiOx. It also contributes to
the van der Waals potential due to its negative ∆ε. But the increase in short range
interactions must be more profound to cause the anchoring transition towards planar
alignment.
The effect of adding 5CB afterwards is the same as discussed in the previous section
when 5CB is mixed with LC1. 5CB molecules bind with SiOx surface preferentially and
make the short range dipolar interactions favor an anchoring direction perpendicular to
the LAD-SiOx surface, while in the bulk, C3 still dominates the long range van der Waals
interaction so a homeotropic alignment is preferred. Since both long range and short
range interactions favor homeotropic alignment the anchoring is switched back to
homeotropic.
The reason why critical concentration of 5CB is quasi-proportional to the
concentration of C3 lies in the short range forces. For C3 short range force prefers
parallel anchoring while for 5CB it prefers vertical anchoring. The result is that to keep
the surface short range interaction in favor of homeotropic anchoring the favorable
interaction between 5CB and LAD-SiOx must exceed the unfavorable interaction between
C3 and LAD-SiOx. Therefore, more 5CB is required in a system that contains more C3 to
maintain homeotropic alignment.
In the comparison of SiOx and SiO2 we showed that the critical concentration of C3
needed to trigger the anchoring transition is much higher on LAD-SiO2 than LAD-SiOx.
We believe that this is also because e-beam evaporated LAD-SiO2 has less surface
70
polarity than thermally evaporated LAD-SiOx, meaning the short range surface dipolar
interactions (that prefer parallel anchoring) between LAD-SiO2 and C3 is smaller than
that between LAD-SiOx and C3. Therefore, more C3 is needed for LAD-SiO2 to achieve
the same magnitude of short range torque on SiOx to compete with long range van der
Waals torque and change the anchoring direction.
4.4.3 The effect of molecular structure on liquid crystal anchoring on SiOx
In section 4.3.3 we tested several compounds for their ability to promote homeotropic
alignment on SiOx. The experimental results show that a molecule with a cyano group at
the end of its molecular axis helps to generate homeotropic alignment, while a molecule
with cyano groups on its side helps to obtain planar alignment. We have already
explained that this effect is due to the short range interactions between liquid crystal
molecules (especially cyano groups) and the LAD-SiOx surface. Though we believe that
any large dipole moment in general will cause similar effect, we give an explanation of
why cyano group looks particularly effective in our experiments. Let us review the
stoichiometry of LAD-SiOx described in the Chapter 2. LAD-SiOx is an oxygen-deficient
structure. According to the random-binding model, an Si atom forms a tetrahedron with 4
randomly selected atoms (Si or O). At the surface, Si has nothing to bind hence has a
vacant orbital. These orbitals are electron acceptors, which makes SiOx a Lewis acid. On
the other hand, a cyano group has a pair of spare electrons, which makes it a strong
electron donor, i.e., a Lewis base. The strong interaction between a Lewis acid and base
makes the cyano-group orientated along surface normal as shown in Figure 25.
71
Dielectric Anisotropy of LC1/5CB mixtures
-8
-6
-4
-2
0
2
4
6
8
10
12
0% 20% 40% 60% 80% 100%
5CB concentration (wt%)
die
lectr
ic a
nis
otr
op
y
Figure 23: Dielectric anisotropy of 5CB/LCI mixtures as a function of 5CB concentration
72
Figure 24: A cartoon showing the effect of adding 5CB into LC1. Green and orange rods
represent LC1 and 5CB molecules respectively. The blue surface represents the LAD-SiOx.
(a)
(b)
(c)
73
Figure 25: A cartoon that shows the interaction between the LAD-SiOx and the cyano
groups.
74
4.5 Summary
In this chapter we have shown our experimental results and explanations on the
material dependence of liquid crystal alignment on LAD-SiOx and the anchoring
transitions caused by the material dependence. The experimental results can be concluded
as follow:
• A liquid crystal with a positive ∆ε aligns parallel to LAD-SiOx surface while
that with a moderate negative ∆ε aligns vertical to LAD-SiOx surface.
However, a liquid crystal with a large negative ∆ε aligns parallel to LAD-
SiOx surface.
• The addition of small amount of a material that has a large longitudinal dipole
to a liquid crystal that has a large negative ∆ε promotes perpendicular
anchoring on LAD-SiOx. But further increasing the amount of additive leads
to planar alignment.
• The addition of a material that has a large lateral dipole to a liquid crystal
with a moderate ∆ε leads to an anchoring transition from homeotropic to
planar alignment on LAD-SiOx. Further introduction of a small amount of a
material with a large longitudinal dipole can switch the anchoring back to
vertical.
Results were explained by the following points:
• Dipole moment (or cyano group) tends to align perpendicular to the LAD-
SiOx surface. Therefore, a material with a longitudinal dipole prefers the
75
homeotropic boundary condition while a material with a lateral dipole prefers
the planar boundary condition.
• Assuming that the small additive molecules with large longitudinal dipole
moments will bind with the LAD-SiOx surface preferentially, they will
favored in covering the LAD-SiOx surface hence change the overall
orientational preference of surface short range dipolar interactions toward
homeotropic alignment.
• Long range van der Waals interaction between LAD-SiOx and the positive ∆ε
additive favors planar anchoring. Therefore, the addition of the additive with a
large longitudinal dipole also shifts the long range forces towards favoring
planar alignment.
• The final alignment depends on the relative strength of these two opposite
effects.
- 76 -
Chapter 5
Temperature Dependence of the Anchoring Transitions on LAD-SiOx
5.1 Introduction
Liquid crystal alignment on SiOx has been found to be temperature dependent. But
most commonly pretilt angle of liquid crystal has been mentioned in the literature without
much regard to the influence of temperature. Some work has been published on the
temperature behavior of liquid crystals 36, 37, 38, 39, 40, 41, 42 , most of which proposed to use
the temperature dependence of the order parameter (S) to explain the temperature
dependence of liquid crystal orientation. The temperature-correlated term can come into
the free energy either through the S-dependence of the van der Waals interaction, the S2-
dependent elastic adaptation to the surface topology term, or through the order electricity
term that depends on the gradient of S on the SiOx surface.
Most of the reported experimental results show that anchoring transitions become
obvious only when temperature is about 1°C below the TNI. However, Vithana et al. have
reported that there is a gradual increase in the pretilt angle of homeotropic alignment on
SiOx
77
starting from about 30°C below the TNI. 43 We observed similar effects in our
experiments on LAD-SiOx. We found that a smooth anchoring transition from
homeotropic to planar alignment can start at least 20°C below the clearing temperature.
More interestingly, the transition can be in the opposite directions for different nematics.
While the temperature dependence of the order parameter has been successful in
explaining many orientational effects of liquid crystals we propose in this chapter another
possible explanation that we found to be useful in discussing the observed phenomena in
our particular case.
5.2 Experimental Methods
5.2.1 Cell Preparation and Characterization
SiOx was evaporated onto substrates at a large angle of incidence. Cells were
assembled with anti-parallel evaporation directions on the top and bottom substrates. Cell
thickness was ~20µm. Liquid crystals were filled by capillary force under vacuum. Pretilt
angle was measured by conoscopy and crystal rotation.
5.2.2 Surface Adsorption and Thermal Desorption
The combination of Thermal Desorption analysis and Mass Spectroscopy makes an
effective tool for studying the surface interaction between LAD-SiOx and liquid crystal
molecules. The equipment we used was a Thermo Electron Polaris Q GC-MS with Direct
Exposure Probe (DEP). The principle of this method is illustrated in Figure 26. The
78
sample is placed on the probe and inserted into a vacuum chamber where the probe is
heated up with a pre-set temperature ramping profile. Molecules that are originally
absorbed on the sample surface are excited by the heat and leave the surface. Then the
free molecules are bombarded by an electron beam and get ionized. Ions fly into the mass
spectrometer and are analyzed. From the measurement, a time dependent mass spectrum
is obtained. This can be translated into the relative abundance of certain chemicals with
evolving temperature, from which we should infer the basic information of the binding
properties between these chemical and the surface.
In the experiments, LAD-SiOx was deposited onto both sides of aluminum foils and
soaked into diluted liquid crystal solutions (0.3% by weight in isopropyl alcohol). After
more than 8 hours the foils were taken out and gently dried. Dried foils were cut into
small pieces around 1mm x 6mm size to be compatible with the crucible in the probe.
This also helps with a good thermal conductivity between sample and the probe hence
accurate temperature control on the samples can be achieved. Sample was heated up at
10°C/min. Real-time temperature measurement was done by a thermal coupler on the
probe. The mass spectra of desorbed materials were recorded as a function of time.
It is true that the LAD-SiOx surface will absorb many things other than the target
liquid crystal molecules during the handling. When heated, all the absorbents tend to be
set free from the surface and will be all recorded by the mass spectrometer. So before the
measurement we had first obtained the standard mass spectrum of each target molecule
so we can look only at their spectral signatures.
79
Figure 26: The working principle of a TDMS (thermal desorption mass spectroscopy)
80
5.3 Results
5.3.1 Thermal Induced Anchoring Transitions
Several liquid crystals have been found to align on LAD-SiOx according to the
temperature. For example the mixture composed of 1/3 LC1 and 2/3LC2 aligns vertically
on LAD-SiOx at room temperature. But as the temperature increases the liquid crystal
director tilted down towards planar alignment, which can be seen from the microscopic
photos shown in Figure 27.
Cells filled with different liquid crystals were placed between crossed polarizers with
the evaporation direction 45º to the polarizer axis. A He-Ne laser passes through the
polarizers and the sample in the normal direction. We measured the transmitted light
intensity as a function of temperature. For the cells filled with LC1 mixtures we found
that when the cell was heated the intensity of transmitted light went up. This effect
becomes more abrupt when the temperature is beyond ~50°C. Figure 28 shows the
measured results of several mixtures. This data shows that the optical retardation of the
cell increases with the temperature, indicating a deviation of LC director from
perpendicular towards planar orientation.
5.3.2 The Effect of Temperature on the Critical Concentration of 5CB
In the anchoring transitions described in 4.3.1 (Figure 11) we observed a shift of
anchoring transition point due to this temperature dependence effect. As shown in Figure
31 the first anchoring transition (happens at lower concentration of 5CB) is obviously
81
shifted due to the temperature change. The concentration of 5CB required to obtain
homeotropic alignment (defined as the critical concentration) ascends when the
temperature increases. On the other side of the plot, since 5CB has a nematic-isotropic
transition temperature as low as ~37°C the anchoring transition points of some mixtures
become unable to be measured when the concentration of 5CB is high. However, judging
from the available data we don’t see any indication of noticeable thermal induced
anchoring shift on the high concentration side.
82
Figure 27: Microscopic images of a LAD-SiOx cell filled with 1/3 LC1 and 2/3 LC2 at
different temperatures. Left side photos were taken with crossed polarizers. Right side
photos were taken with parallel polarizers.
Figure 1 : SiO 1/3-2/3, T=84.0 C Figure 2 : SiO 1/3-2/3, Very Big Pretilt T=84.9 C
45deg with crossed polarizers 0deg with crossed polarizersFigure 3 : SiO 1/3-2/3, Planar T=85.0 C
Figure 4 : SiO 1/3-2/3, Planar to Isotropic T=89.0 C Figure 5 : SiO 1/3-2/3, Planar to Isotropic T=90.0 C
Hom
eotropic Planar Isotropic
Figure 1 : SiO 1/3-2/3, T=84.0 C Figure 2 : SiO 1/3-2/3, Very Big Pretilt T=84.9 C
45deg with crossed polarizers 0deg with crossed polarizersFigure 3 : SiO 1/3-2/3, Planar T=85.0 C
Figure 4 : SiO 1/3-2/3, Planar to Isotropic T=89.0 C Figure 5 : SiO 1/3-2/3, Planar to Isotropic T=90.0 C
Figure 1 : SiO 1/3-2/3, T=84.0 C Figure 2 : SiO 1/3-2/3, Very Big Pretilt T=84.9 C
45deg with crossed polarizers 0deg with crossed polarizersFigure 3 : SiO 1/3-2/3, Planar T=85.0 C
Figure 4 : SiO 1/3-2/3, Planar to Isotropic T=89.0 C Figure 5 : SiO 1/3-2/3, Planar to Isotropic T=90.0 C
Figure 1 : SiO 1/3-2/3, T=84.0 C Figure 2 : SiO 1/3-2/3, Very Big Pretilt T=84.9 C
45deg with crossed polarizers 0deg with crossed polarizersFigure 3 : SiO 1/3-2/3, Planar T=85.0 C
Figure 4 : SiO 1/3-2/3, Planar to Isotropic T=89.0 C Figure 5 : SiO 1/3-2/3, Planar to Isotropic T=90.0 C
Hom
eotropic Planar Isotropic
83
Figure 28: Intensity of transmitted light as a function of temperature. Samples were held
between crossed polarizers with evaporation direction 45º to the polarizer axis. All cells
have the same cell gap ~20µm.
20 30 40 50 60 70 800.1
1
10
100
1000
Temperature/oC
Inte
nsity
/mV
1/3 LC1 and 2/3 LC2 5% 5CB in LC1 MLC-6609
84
Figure 29: Temperature dependence of the anchoring transitions of 5CB/LC1 mixtures on
LAD-SiOx
0 20 40 60 80 100
0
20
40
60
80
100
Tilt
Ang
le fr
om S
urfa
ce
Weight Percentage of 5CB
RT degC40 degC50 degC60
85
5.3.3 Thermal Desorption
To understand the temperature dependence of anchoring transitions we have studied
the surface adsorption of LC on LAD-SiOx. In an attempt to measure the binding energy
of different liquid crystals with the LAD-SiOx surface, we tried to measure the thermal
desorption curve of monolayer liquid crystal on LAD-SiOx. 5CB is dissolved into IPA to
make a series of solutions with different concentration ranging from 1% to 0.01%.
Solutions were spin-coated onto LAD-SiOx substrates to make a thin film. Substrates
were then heated at a few degree below the N-I temperature and monitored by measuring
the optical retardation. We believe a monolayer is achieved when the optical retardation
stops decreasing dramatically. Substrates with monolayer liquid crystal were then baked
in a vacuum oven with controlled temperature for 30min. After that, XPS was used to
detect the nitrogen signal from the surface. As shown in Figure 30 we found that the
thermal desorption of 5CB happens between 49.5ºC and 100ºC.
However, the continuation of our study using the XPS has been interrupted by
technical difficulties. First of all, the signal to noise ratio is too small (because it’s a
monolayer), and it becomes difficult to be convinced that we are measuring monolayer
desorption instead of reaching the instrumental limit of signal to noise ratio. Second, the
10-8~9 torr UHV (ultra high vacuum) pulls liquid crystal molecules from the surface very
quickly. We found the signal to be very weak even if we use a very thick liquid crystal
layer. So in principle we should measure as fast as we can once the sample is inside the
chamber. However to get a good S/N ratio we have to repeat the measurements many
times in a period of time around 20 minutes.
86
Thermal Desorption Mass Spectroscopy (TDMS) provided us with another solution.
Figure 31(a) shows the desorption curve of 5CB measured from 3 samples of LAD-SiOx
that had been soaked in 0.3% 5CB solution of IPA. As can be seen, the thermal
desorption peak is at 55±2.5°C. The error probably comes from the difficulty in
reproducing 5CB monolayer on each sample. We also tried to measure the thermal
desorption curve of C3. Unfortunately, even if we increased the temperature to about
165°C we didn’t see any significant desorption of C3 on any sample, as shown in Figure
31(b). Nevertheless using XPS we have successfully detected nitrogen atoms (in the
cyano groups) on the LAD-SiOx samples prepared by the same method. This indicates
that C3 should be absorbed on the TDMS samples, too. Considering the two lateral cyano
groups on each C3 molecule and its relatively large molecular weight we conclude that
C3 has a very strong surface interaction with LAD-SiOx which prevents it from
desorption at temperatures lower than 165°C.
87
Figure 30: XPS spectrum showing nitrogen atoms of 5CB on LAD-SiOx. On the spectrum
of the original sample and the sample that has been baked at 49.5ºC, a peak of Nitrogen
has been observed. This implies the existence of 5CB on the SiOx surface. However on
the spectrum of the sample that has been baked at 100ºC the nitrogen peak no longer
exists, indicating that the thermal deposption temperature of 5CB is between 49.5ºC and
100ºC.
385 390 395 400 405 410 4153800
3850
3900
3950
4000
4050
4100
4150
Bonding energy/ev
Cou
nts
N1s original 49.5OC 30min 100OC 30min
88
Figure 31: (a): Thermal desorption curve of 5CB (3 samples of 5CB absorbed on LAD-
SiOx were prepared by the same methods). (b): Thermal desorption curve of C3
0
10
20
30
40
50
60
70
80
90
100
40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120
Temperature / Degree C
Rel
ativ
e A
bund
ance
0
1
2
3
4
5
6
7
8
9
10
40 50 60 70 80 90 100 110 120 130 140 150 160 170
Temperature / Degree C
Rel
ativ
e A
bund
ance
89
5.4 Discussions
5.4.1 Thermal Induced Anchoring Transitions
According to our hypothesis, the two major origins of LC alignment are long range
van der Waals interactions and short range dipolar interactions. Thermal induced
anchoring transitions can be viewed as the result of change in their relative strengths.
In a liquid crystal mixture with a negative dielectric anisotropy such as the mixture
of 1/3LC1 and 2/3LC2 we expect there exist two kinds of LC molecules: molecules with
strong lateral dipole moment (negative material) and molecules with longitudinal dipole
moment (positive material). On the LAD-SiOx surface both types of molecules can be
absorbed and contribute to the short range interactions. Nevertheless the binding
strengths could be very different. Assuming positive materials have a weaker binding
with LAD-SiOx and can be broken at lower temperatures compared to negative materials;
the number of positive material molecules absorbed on LAD-SiOx will decrease more
substantially. Hence the short range interactions will shift towards favoring planar
alignment at a higher temperature, as shown in Figure 28. This effect will be especially
obvious when the temperature is close to the characteristic binding temperature between
positive materials and LAD-SiOx.
Some liquid crystal mixtures do not have obvious temperature dependence in the
anchoring on LAD-SiOx like is shown in Figure 28. There could be a few reasons. Some
mixtures that have a negative ∆ε do not contain (or contain less) materials that have
strong lateral permanent dipoles. So the short range interaction that favors planar
alignment will be minimized. Also if the binding energy between liquid crystal molecules
90
and the surface is too strong to break during our measuring temperature range the
anchoring transition will not be obvious either.
5.4.2 The Effect of Temperature on the Critical Concentration of 5CB
In 5.3.2 (Figure 29) we showed the effect of temperature on the critical concentration
of 5CB to sustain homeotropic alignment of 5CB/LC1 mixtures on LAD-SiOx. To
explain this effect let’s look at the surface adsorption and desorption of LC on LAD-SiOx
first. Normally, surface adsorption can be described by the surface coverage ratio Θ ,
which satisfies the following equation44:
TkH
TkH
B
B
eAx
eAx∆
−
∆−
⋅+
⋅=Θ
1
(5.1)
Here A is a constant; x is the concentration of adsorbent; H∆ is the enthalpy of the
adsorption process; Bk is Boltzman constant; and T is temperature.
Let 0TkH B−=∆ (5.2)
and 0T is the characteristic temperature of the thermal desorption process.
TT
TT
eAx
eAx0
0
1 ⋅+
⋅=Θ (5.3)
Known from the thermal desorption experiment data (Figure 31) the characteristic
temperature is ~55°C for 5CB on LAD-SiOx but higher than 165°C for C3.
91
For a mixture of 5CB and LC1 contacting with LAD-SiOx, the surface coverage ratio
of 5CB is:
T
T
CB
TT
CBCB
CB
CB
exA
exA0
5
05
5
55
'1
'
⋅+
⋅=Θ (5.4)
Suppose the original concentration of the C3-like material in LC1 is 03Cx . Its
concentration in the mixture should be )1( 50
33 CBCC xxx −= . The surface coverage ratio of
C3 is therefore:
TT
CBC
TT
CBCC
C
C
exxA
exxA0
3
03
)1(''1
)1(''
50
3
50
33
⋅−+
⋅−=Θ (5.5)
From Equation (2.12) at the critical point where the anchoring transition happens,
CB
TT
criticalCB
TT
criticalCB
CBcriticalCBC
TT
criticalCBC
TT
criticalCBC
CcriticalCBC W
exA
exAUxW
exxA
exxAUxx
CB
CB
C
C
5
5
5553
50
3
50
335
03 0
5
05
03
03
'1
'
)1(''1
)1('')1(
⋅+
⋅−=
⋅−+
⋅−−−
(5.6)
When the concentration of 5CB is relatively small, the coverage ratio of 5CB should
also be small. So,
TT
CB
TT
CB
TT
CBCB
CB
CB
CB
exA
exA
exA0
5
05
05
5
5
55 '
'1
'⋅≈
⋅+
⋅=Θ (5.7)
92
Since 5CB concentration is small the concentration of C3 can be treated as constant.
Due to the relatively high thermal desorption temperature of C3 (>165°C), it is
reasonable to assume that the surface coverage ratio is close to 1 and the change in
surface coverage ratio of C3 due to temperature or 5CB concentration is negligible.
So the critical point relies on:
Constant)(' 30
330
35555
05
CWUxWexAUx CCCCCBT
TcriticalCBCB
criticalCB
CB
=Θ−=⋅− (5.8)
)'/(0
5
555T
T
CBCBcriticalCB
CB
eWAUCx −= (5.9)
Since C3 prefers parallel alignment by itself, we know
030
330
3 <Θ−= CCCC WUxC (5.10)
Let CUCWA CBCB /,/' 55 −=−= βα , (5.11)
)/(1 /5
05 βα −⋅= TTcritical
CBCBex (5.12)
where 0, >βα . Now let’s consider the effect of temperature at anchoring transition
point. When T increases, TT CBe /05⋅α decreases, and )/(1 /0
5 βα −⋅ TT CBe increases.
So as a result, the critical concentration of 5CB increases with temperature. Using
Equation (5.12) we were able to fit the critical concentration as a function of temperature
as plotted in Figure 32. This explains why the rising edge (left edge) in Figure 29 has
been shifted by temperature.
93
The falling edge (right edge) of Figure 29 does not have complete data showing the
effect of temperature because the clearing point of the mixture drops with the increasing
concentration of 5CB. Nonetheless available data shows no indication of temperature
dependence. We propose to explain this as follows. When the concentration of 5CB is
large from Equation (5.4) we know that the adsorption on SiOx may be saturated
( 15 ≈Θ CB ) even at relatively high temperatures. Hence, there will be no obvious
temperature dependence.
94
Figure 32: The critical concentration of 5CB in the planar-to-homeotropic anchoring
transition of 5CB/LC1 mixtures as a function of temperature.
15 20 25 30 35 40 45 50 55 60 65 70
0
2
4
6
8
10
12
14
16
x=1/(αeT0/T-β)
Chi^2/DoF = 0.00009R^2 = 0.96831 T0 55 ±0α 7.21501 ±1.7361β 13.04596 ±5.26203
Crit
ical
Con
cent
ratio
n of
5C
B (w
eigh
t%)
Temperature/ oC
95
5.5 Summary
In this chapter we have reported our experimental observations of thermally induced
anchoring transitions on LAD-SiOx. Temperature has also been found to be capable of
changing the transition point in an anchoring transition on LAD-SiOx caused by the
liquid crystal composition change. We propose to explain the observed effects using our
theory of competition between long range and short range interactions. Assuming that the
number of absorbed molecules on the LAD-SiOx surface is proportional to the
contribution of short range interactions to the total free energy we expect the heating will
change the liquid crystal anchoring toward the opposite way short range interactions
prefer. This change will be especially obvious when the binding energy between liquid
crystal and LAD-SiOx is close to the thermal energy. Different liquid crystals may have
very different binding energy with LAD-SiOx. By increasing temperature, the thermal
desorption of each component in a liquid crystal mixture can be very different, too. The
result may be a change in overall short range interaction preference of liquid crystal
orientation, hence an anchoring transition.
- 96 -
Chapter 6
The Effect of LAD-SiOx Thickness on Liquid Crystal Anchoring
6.1 Introduction
The dependence of liquid crystal alignment on the thickness of inorganic alignment
layer has been discovered by many previous researchers.45,46,47,48 In most of the cases,
authors found a critical thickness of the alignment layer at which the orientational
transition would be observed. In particular reference [Error! Bookmark not defined.]
and [48] reported that pretilt angle (w.r.t. surface normal) of slightly tilted homeotropic
alignment tended to decrease with the increase of alignment layer thickness. This kind of
effect is by all means important but so far only ambiguously explained.
In this chapter we will report our experimental results on the LAD-SiOx thickness-
induced anchoring transitions. We will also show the long range nature of the LAD-SiOx
thickness dependence by the screening effect experimental results. Later we will use the
competition between long range van der Waals interactions and short range dipolar
interactions to explain the results.
97
6.2 Experimental Methods
6.2.1 LAD-SiOx Sample Preparation
LAD-SiOx thin films were prepared by thermally evaporating silicon monoxide (SiO)
powders onto substrates at a large angle of incidence. The thickness of coating is
measured in-situ by an oscillating quartz crystal thickness monitor. The reading of the
thickness monitor has been calibrated with ellipsometry measurement data. Coated
substrates were made into anti-parallel cells with ~20µm cell gaps.
6.2.2 Polyimide Sample Preparation
SE-7511L polyamic acid was used to produce polyimide (PI) alignment layers that
gave liquid crystal homeotropic alignment. We diluted SE-7511L into a series of
solutions with different concentrations. These solutions were spin-coated onto substrates
at 3000rpm for 30 seconds followed by the standard procedure of 60 seconds soft bake at
90°C and 1 hour hard bake at 180°C. After bake, no rubbing was performed. Since it’s
difficult to measure the PI thickness on substrates with multiple coatings we used an
identical procedure to coat PI onto blank silicon substrates and measured the PI thickness
by a J.A. Woolam spectroscopic ellipsometer. The thickness values measured on silicon
were used for PI coatings on LAD-SiOx and glass substrates using the same PI solutions.
However, we would rather look at these numbers as a relative values instead of absolute
PI thickness because the wetting properties of silicon could be different from LAD-SiOx
or glass, resulting in coatings with different thicknesses for the same polyamic acid
solution and the same coating process. Cells were made with ~20µm cell gaps.
98
6.2.3 Pretilt Measurement
Pretilt angle was calculated from the optical retardation data measured using a
tuneable compensator.
The optical retardation can be expressed as
dnnnd oeff )( −=∆=Λ (6.1)
Here the effective extraordinary refractive index of liquid crystal
θθ 2222 sincos/ oeoeeff nnnnn += (6.2)
θ is the pretilt angle from the surface normal.
From equations (6.1) and (6.2) we get
( )( ) ( )
−+Λ−= 222222 ///arcsin oeooee nnndnnnθ (6.3)
The refractive index ne, no and cell gap d were measured separately.
6.3 Experimental Results
6.3.1 The Effect of LAD-SiOx Thickness on Liquid Crystal Alignment
Cells were made from a series of LAD-SiOx coatings that were identical except for
the thickness. A commercial liquid crystal mixture that has a moderate negative dielectric
anisotropy was used to fill the cells. As shown in Figure 33 the liquid crystal forms a
quasi-homeotropic alignment with a large pretilt angle w.r.t. the surface normal. But the
99
pretilt angle decreases as the LAD-SiOx thickness increases until it stabilizes after a
certain thickness.
6.3.2 The Effect of LAD-SiOx Thickness on the Critical Concentration of 5CB
We discovered that anchoring transitions depend on the thickness of the LAD-SiOx
layer. In our experiments LAD-SiOx was thermally evaporated onto 4 identical glass
substrates with identical deposition parameters but increasing layer thickness. Cells were
made from these coated substrates and were filled with mixtures of LC1 and 5CB. As we
described in 3.3.1 we observed planar-to-homeotropic-to-planar anchoring transitions as
the concentration of 5CB increased. But we also found that the anchoring transitions have
different starting and ending points on LAD-SiOx versus thickness. On the first transition
where LC anchoring switches from planar to homeotropic (on the left edge in Figure 34)
we didn’t notice any obvious change of 5CB critical concentration due to LAD-SiOx
thickness. However, on the second transition where LC anchoring switches from
homeotropic to planar (on the right edge of Figure 34) we found that with a thicker LAD-
SiOx layer on the substrate, lower concentration of 5CB is needed to make the transition
happen.
100
Figure 33: The effect of LAD-SiOx thickness on the alignment of liquid crystal. A
commercial liquid crystal mixture with a negative dielectric anisotropy was used in the
experiment.
0 20 40 60 80 100 120 140
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
40o evaporation angle 50o evaporation angle
Pre
tilt A
ngle
w.r.
t. S
urfa
ce N
orm
al/d
egre
e
SiOx Layer Thickness/nm
101
Figure 34: The anchoring transitions in a 5CB/LC1 mixture depend on the underlying
LAD-SiOx layer thickness.
-10 0 10 20 30 40 50 60 70 80 90 100 110-10
0
10
20
30
40
50
60
70
80
90
100
Tilt
angl
e fro
m s
urfa
ce/d
egre
e
Concentration of 5CB (Weight%)
SiOx thickness 35nm 65nm 130nm 350nm
102
6.3.3 Screening Effect
The thickness dependence remains even if the LAD-SiOx layer is screened by other
films. To show this effect we put a thin layer of polyimide on top of LAD-SiOx. SE-
7511L polyamic acid is widely used to produce PI alignment layers that give LC
perpendicular anchoring.
In one experiment we fixed the top PI layer thickness and increased the LAD-SiOx
thickness from 0 to 130nm. The cells were filled with 5CB. We found that the pretilt
angle of 5CB (w.r.t. surface) decreased when we increased the LAD-SiOx thickness, even
though LAD-SiOx has been separated from 5CB by the PI layer. The results are shown in
Figure 35.
In another experiment PI 7511 layers with different thickness were coated onto
substrates with fixed LAD-SiOx thickness. 5CB was filled into the cell and tested for its
alignment. The results show that 5CB forms homeotropic alignment when PI 7511 is
thick enough. When the PI layer becomes thinner the pretilt (w.r.t. surface) decreases
accordingly. At a certain point, the 5CB director jumps down and quickly changes into
planar alignment. The experiment has been repeated on bare ITO glass, and ITO glass
coated with 22nm, 65nm, and 130nm LAD-SiOx. Similar results were observed, as shown
in Figure 36.
Significantly, the same experiments using MLC-6609 (which has a negative ∆ε)
instead of 5CB produce no anchoring transition.
103
We also noticed that the LC on the bare ITO glass (0 nm SiOx) showed tilt angles
very different from the LC on other LAD-SiOx substrates and doesn’t fit into the
transition curve in Figure 35. We believe that this is due to the very different surface
wetting between glass and LAD-SiOx. So the produced PI films are not directly
comparable in thickness and other properties.
104
Figure 35: The effect of LAD-SiOx layer thickness on the alignment of 5CB screened by
polyimide that prefers homeotropic anchoring.
0 20 40 60 80 100 120 140
20
30
40
50
60
70
80
90
Tilt
Ang
le fr
om S
urfa
ce /
Deg
ree
SiOx Layer Thickness / nm
38.2nm PI 28.1nm PI 16.4nm PI
105
Figure 36: Anchoring Transitions induced by the screening effect of polyimide on top of
LAD-SiOx surface.
0 10 20 30 40 50 60 70-10
0
10
20
30
40
50
60
70
80
90
100
PI 7511 thickness / nm
Pre
tilt f
rom
sur
face
/ de
gree
5CB/PI 5CB/PI/130nm SiOx 5CB/PI/65nm SiOx 5CB/PI/22nm SiOx MLC-6609/PI/130nm SiOx
106
6.4 Discussions
6.4.1 The Effect of LAD-SiOx Thickness on the Alignment of Liquid Crystal
We have already shown in Chapter 2 that the LAD-SiOx surface roughness and
anisotropy doesn’t change obviously when the deposition thickness increases from 0 to
130nm. This excludes the topography from causing the LAD-SiOx thickness dependence
effects we have shown. To explain this effect we propose to look at the van der Waals
interactions between LAD-SiOx and liquid crystals.
The van der Waals potential between two infinite flat surfaces (Figure 37) is given as
212/ DAW π−= (6.4)
In ref. [15],[16] de Gennes used this approach by assuming there was a small
distance δ between the alignment layer and liquid crystal layer. However, to understand
the anchoring transition dependence on LAD-SiOx thickness this is inadequate. In this
paper we follow the method described by Israelachvilli49 to deduce the influence of LAD-
SiOx thickness on the van der Waals potential.
Let’s start with the basic van der Waals interaction potential between two molecules:
6)(rCrw −= (6.5)
Consider a half slab of a liquid crystal cell as shown in Figure 38
For molecules in a circular ring of cross-sectional area dxdz and radius x, the ring
volume is 2πxdxdz. The number of molecules in the ring will be 2πρxdxdz.
107
22 zxr += (6.6)
The net interaction for a liquid crystal molecule at a distance h from the LAD-SiOx
surface will therefore be
∫ ∫+=
=
∞=
= +−=
Dhz
hz
x
x xz
xdxdzCDw0
622
2)( ρπ
| 0222 )(1
2∞=
=
+=
= += ∫
x
x
Dhz
hz xzdzCρπ
|34
162
Dh
h
Dhz
hz zC
zdzC +
+=
=
−== ∫ρπρπ
])(
11[6 33 DhhC
+−−=
ρπ (6.7)
Here h is the thickness of distance from the liquid crystal molecule to the SiOx-liquid
crystal interface and D is the thickness of the SiOx layer, as shown in Figure 38.
To calculate the van der Waals potential between the liquid crystal and LAD-SiOx
per unit area, we need to integrate this expression over the thickness of liquid crystal
layer. As de Gennes pointed out, we have to assume a small distance δ between the
alignment layer and liquid crystal layer to prevent the integration from diverging. δ
should be of the magnitude of absorbed monolayer thickness, a good estimation would be
~1nm. Therefore we have
dhDhh
CHDw
Hh
h∫=
= +−−=
δ
ρρπ]
)(11[
6),( 33
21
108
])(
1211
21[
6 || 2221 HH
DhhC
δδ
ρρπ+
+−−=
]11)(
1)(
1[12 2222
21
HDHDC
+−+
−+
−=δδ
ρρπ (6.8)
In our experiments, H(half cell gap) is about 10 µm, D(LAD-SiOx thickness) is from
0.035 µm to 0.35 µm, δ is about 0.001 µm. Therefore H>>D>> δ and the equation can be
approximated to
∆+−=+−−−≈ 221
222221
12]1111[
12)(
DC
HHDCrw ρρπ
δρρπ (6.9)
Here ∆ is a constant.
From eq. (6.9) we can see that when SiOx thickness increases the van der Waals
potential also increases. For a liquid crystal that has a negative dielectric anisotropy, van
der Waals potential prefers homeotropic alignment. An increase in van der Waals
potential will lead the liquid crystal alignment towards homeotropic as we have seen in
Figure 33.
6.4.2 The Effect of LAD-SiOx Thickness on the Critical Concentration of 5CB
The critical concentration of 5CB in Figure 34 can be regarded as a balance point at
which the relative strength of long range van der Waals torque is the same as short range
109
dipolar torques. As we increase the LAD-SiOx thickness we have increased the strength
of the van der Waals interactions, which will cause the balance point to be shifted.
As described in eq. (2.12), at critical point
CBCBCBcriticalCBCCC
criticalCBC WUxWUxx 55553335
03 )1( Θ−=Θ−− (6.10)
Since C3 has very strong bonding (high desorption temperature) we assume that the
coverage ratio of C3 is always 1. When the concentration of 5CB is high, the coverage
ratio of 5CB is also close to 1. As a result we get
CBCBcriticalCBCC
criticalCBC WUxWUxx 555335
03 )1( −=−− (6.11)
CBCC
CBCCB
CBCC
CCBCCcriticalCB UUx
UWWUUx
WWUxx
530
3
535
530
3
3530
35 1
+−−
+=+
−+= (6.12)
Since U is in the form of 1/D2, let 323
3,525
5 CC
CCBCB
CB DU
DU ∆+
Ψ=∆+
Ψ= (6.13)
x5CBcritical =
xC 30 UC 3 + W5CB −WC 3
xC 30 UC 3 + U5CB
=xC 3
0 (ΨC 3
D2 + ∆C 3) + W5CB −WC 3
xC 30 (ΨC 3
D2 + ∆C 3) + (Ψ5CB
D2 + ∆ 5CB )
253
0353
03
2353
033
03
)()()(
DxxDWWxx
CBCCCBCC
CCBCCCC
∆+∆+Ψ+Ψ−+∆+Ψ
= (6.14)
It’s easy to see that x is in the form of
110
ϕγξ
++
= 251
Dxcritical
CB (6.15)
Here γ, φ and ξ are positive invariants to SiOx thickness.
From equation 6.15 we know when the thickness of SiOx increases, the critical
concentration of 5CB should decrease. The physical meaning of this effect can be
understood in the following way. At the point of the transition, 5CB has a concentration
of over 50% and we assume that the surface is saturated with absorbed 5CB molecules.
Further we have confirmed in our experiments that the dielectric anisotropy of the
mixture becomes positive when the concentration of 5CB is higher than 33%. So, van der
Waals potential would prefer planar alignment but surface short range interaction prefers
vertical alignment. With the increase of LAD-SiOx thickness more van der Waals
potential is gained to compete with the same amount of short range interactions.
Therefore less 5CB (smaller ∆ε) is needed to cause the transition due to the greater
polarization of the surface for thicker LAD-SiOx. Experimental data has been fitted using
Equation (6.15) as shown in Figure 39.
111
Figure 37: Two infinite surfaces separated by distance D
Figure 38. The cross section of a half slab of a liquid crystal cell.
D
D
LC
x
z
H
SiOx
112
Figure 39: The critical concentration of 5CB in the homeotropic-to-planar anchoring
transition of 5CB/LC1 mixtures (shown in Figure 34) depends on the thickness of
underlying LAD-SiOx layer
0 50 100 150 200 250 300 350 400
50
52
54
56
58
60
62
64
66
68
70
x=ϕ+1/(ζ+γD2) Chi^2/DoF = 0.00026R^2 = 0.98346 ϕ 0.49864 ±0.01858γ 0.00090 ±0.00049ζ 4.76581 ±0.75567
Crit
ical
Con
cent
ratio
n of
5C
B (w
eigh
t%)
SiOx Layer Thickness / nm
113
6.4.3 Screening Effect
Because of its “long range” nature, van der Waals forces originating from SiOx can
be “felt” by liquid crystal molecules even if they are separated by intermediate polyimide
layers. While PI 7511L is known to promote homeotropic anchoring on its surface the
preference of SiOx underneath depends on the dielectric anisotropy of the liquid crystal.
For 5CB, which has a positive ∆ε, the long range van der Waals interactions would prefer
planar alignment on LAD-SiOx, which is different from the homeotropic alignment
favored by PI 7511L. Therefore, anchoring transitions between homeotropic and planar
alignment can take place when the relative strength of long range and short range
interactions changes. In Figure 35 we have shown that with a fixed PI thickness
increasing the LAD-SiOx thickness shifted the anchoring towards planar. We interpret
this as a result of van der Waals potential (that prefers planar alignment) being
strengthened. While in Figure 36 we observed anchoring transitions due to the increased
short range surface interactions resulted from the increased PI thickness. The
experimental data therefore fits with our theory.
On another hand MLC-6609 has a negative ∆ε so the long range van der Waals
interaction between SiOx and liquid crystal prefers homeotropic alignment, which agrees
with short range forces between PI 7511L and the liquid crystal materials. Hence no
anchoring transition was observed in the experiments.
114
6.5 Summary
We have observed in our experiments anchoring transitions of liquid crystal
materials on LAD-SiOx alignment layers due to the change of SiOx thickness. For a liquid
crystal with a negative ∆ε that aligns perpendicularly on LAD-SiOx with a small pretilt,
the pretilt angle decreases when the SiOx thickness increases. The thickness of LAD-SiOx
also shifts the critical point of anchoring transitions caused by the addition of additives
that have large longitudinal dipole or lateral dipole, which we have discussed in previous
Chapters. We found that the effect of LAD-SiOx thickness is long range. It can affect the
liquid crystal that has been separated from it by a thin layer of polymer. More specifically,
when the polymer has opposite preference of the liquid crystal orientation, anchoring
transitions can be observed by changing the SiOx thickness; while if the polymer has the
same preference of the liquid crystal orientation no anchoring transition will be observed.
We proposed to explain the observed LAD-SiOx thickness dependence by
considering the van der Waals interactions between the liquid crystal and LAD-SiOx. The
van der Waals potential depends on the LAD-SiOx thickness and is long range in nature.
Our particular experimental results of anchoring transitions can be interpreted as a change
in the relative strength of long range and short range interactions, which, we believe,
have different preference in the liquid crystal orientation.
- 115 -
Chapter 7
Conclusions and Suggestions for Future Work
7.1 Summary of Dissertation Work
In this dissertation work we have shown an experimental study of the anchoring
transitions of liquid crystal materials on silicon oxide alignment layers coated at a
medium angle of evaporation.
• We have characterized LAD-SiOx thin films and discovered that LAD-SiOx
used in our experiments has little surface topography and is unlikely to have a
significant effect on the liquid crystal anchoring. We also discovered that
thermally evaporated LAD-SiOx has an unsaturated chemical structure with
the atomic ratio of Si and O changing with deposition pressure. On the other
hand, e-beam evaporated LAD-SiO2 shows a saturated structure that is close
to crystalline silica, with a stable atomic ratio of Si to O equal to ½.
• We found that a liquid crystal with a moderate negative ∆ε aligns
perpendicularly on our LAD-SiOx substrates. But, a liquid crystal with a large
negative ∆ε forms planar alignment.
• We found that the addition of a small amount of a positive ∆ε material that
has a large longitudinal dipole into a liquid crystal that has a large negative
∆ε changes the liquid crystal alignment from planar to homeotropic. But
further addition of the positive material will change the alignment back to
planar
116
The anchoring transitions we observed were found to be dependent on
temperature and LAD-SiOx thickness.
• We have studied the effect of additive chemical structures on the alignment
of a liquid crystal that has a large negative ∆ε. We found that an additive with
a large lateral dipole moment tended to promote homeotropic alignment
while an additive that had a large lateral dipole moment tended to promote
planar alignment.
• We reported experimental results showing the effect of temperature and SiOx
thickness on the alignment of liquid crystal on LAD-SiOx. The SiOx
thickness dependence was observed even in the case where LAD-SiOx was
screened by another polymer thin film.
Some theoretical work has also been done as part of the dissertation:
• We have reviewed important theoretic works on this topic.
• We adopted the de Gennes model that described competition between long
range van der Waals interactions and short range dipolar interactions that
have opposite preference in liquid crystal orientation, and expanded it to use
in mixtures that contain liquid crystals with both positive and negative ∆ε.
• We have theoretically derived the temperature dependence of anchoring
transitions based on the idea of surface adsorption and thermal desorption
using classic Langmuir isotherm model and the Van’t Hoff equation.
• We also did a theoretical calculation of the SiOx thickness dependence of van
der Waals potential and its effect on anchoring transitions.
7.2 Conclusions
We have seen interesting anchoring transitions of liquid crystal materials on the
LAD-SiOx substrates we produced in our experiments. Based on the experimental data
117
we reported in this dissertation, we propose that the liquid crystal anchoring phenomenon
we observed on LAD-SiOx can be explained as a result of competition between long
range van der Waals interactions and short range dipolar interactions. The LAD-SiOx we
studied is birefringent. We believe it’s more polarizable in-plane than perpendicular to
the plane. The more polarizable directions of liquid crystals tend to be parallel to those of
SiOx because of van der Waals interactions while the short range dipolar interactions
prefer dipoles to be perpendicular to the surface. As a result, long range and short range
interactions have completely opposite preference in the anchoring direction. The final
alignment of liquid crystal depends on their relative strengths. Any changes in the
balance between long range and short range forces may lead to anchoring transitions.
We also proposed to correlate the magnitude of short range surface interactions with
the number of molecules adsorbed on the surface (or coverage ratio of the first layer). As
a result, the addition of materials that have strong affiliation with the surface can
significantly change the anchoring preference of surface short range interactions hence
affect the bulk alignment on the surface. Certainly the addition of materials may also
change the bulk dielectric properties, thus change the anchoring preference of long range
van der Waals interactions. Therefore, the effects of adding an additive into a base LC
that has opposite dielectric anisotropy are always two-fold. The balance between the
long and short range forces may shift toward different directions during the course of
mixing, causing anchoring transitions.
Based on the surface adsorption and thermal desorption theory we can explain the
temperature dependence of the anchoring on our LAD-SiOx substrates. Thermal
118
desorption changes the coverage ratio of liquid crystal on the LAD-SiOx surface, which
leads to a change in the magnitude of short range interactions. If the change is big enough
to shift the balance between long range and short range forces, an anchoring transition
takes place. In a liquid crystal mixture, different components may have different
preference in boundary conditions. Due to the different characteristics of interaction with
LAD-SiOx they may respond to temperature in different ways. The result is that changes
in coverage ratio of each component may be different, which could possibly cause a
change in overall preference in liquid crystal anchoring.
The dependence of liquid crystal anchoring on the thickness of LAD-SiOx is
believed to be a direct result of the r-dependence of van der Waals potential. Varying the
LAD-SiOx thickness in our experiments changes the van der Waals interaction strength in
the competition with short range interactions and may cause anchoring transitions.
Because of its long range nature the effect of LAD-SiOx can be detected when a thin
layer of polymer comes between the LC and the LAD-SiOx.
In the system we studied, the surface properties of LAD-SiOx contribute a lot to the
short range interactions between LAD-SiOx and the liquid crystal. A clean passive
surface with saturated silicon bonds and quartz-like crystal structure tends to have fewer
interactions with dipoles on liquid crystals so the anchoring preference of van der Waals
potential will determine the overall anchoring direction.
The above conclusions are made based on the experimental results and assumptions
that are specific to our LAD-SiOx –liquid crystal system. In some other systems it may be
119
found that things such as topography are of great importance. So it may be incorrect to
generalize our conclusions to other cases where our assumptions don’t apply.
7.3 Suggestions for Future Work
Though we have successfully used our theory to explain some aspects of the liquid
crystal alignment and anchoring transitions on LAD-SiOx thin films, a comprehensive
understanding that in general explains the fundamentals of the liquid crystal alignment on
inorganic thin films is still not available. I would like to make a few suggestions that
hopefully will be helpful toward a better understanding of this topic.
Monolayer behavior of liquid crystals on LAD-SiOx would be very interesting to
understand. It might allow us to probe the anchoring preference of short range
interactions directly. As a matter of fact, we had an unsuccessful attempt to do this study
using FTIR and ATR. But I believe that with more sophisticated techniques this work
should be doable and will provide valuable information.
It will also be very helpful to understand the optical/dielectric properties of LAD-
SiOx thin films. The optical/dielectric axis direction will be very useful information for a
better understanding of the anisotropy of van der Waals potential.
The topography of SiOx changes a lot when the evaporation angle increases. It would
be very important to understand this evolution and find out how this changes the liquid
crystal alignment on SiOx.
120
Lastly, numerous inorganic materials have been found capable of producing uniform
alignment of liquid crystals. Using the same deposition angle, thickness and other process
parameters, different materials still produce different alignment of the liquid crystals. For
example, Janning reported that using 5º evaporation SiO produced planar alignment
while gold produced homeotropic alignment of 5CB.50 It would be very interesting to
understand this phenomenon and correlate the inorganic material properties with the
liquid crystal alignment.
121
1 W. R. Heffner, D. W. Berreman, M. Sammon, and S. Meiboom, Appl. Phys. Lett. 36, 144 (1980). 2 W. Urbach, M. Boix, and E. Guyon, Appl. Phys. Lett. 25, 479 (1974). 3 F. J. Kahn, Appl. Phys. Lett. 22, 386 (1973). 4 Chien-Hui Wen, Sebastian Gauza, Jun Li, Haying Wang, and Shin-Tson Wu, Liq.
Cryst., Vol. 32, No. 5, 643 (2005)
5 Atrushi Kubono, Hidetoshi Onoda, Kai Inoue, Katsufumi Tanaka, and Ryuichi Akiyama, Mol. Cryst. Liq. Cryst. 373, 127 (2002). 6 Tatsuo Uchida, Hideo Watanabe, and Masanobu Wada, Jpn. J. Appl. Phys. 11, 1559 (1972). 7 E. Perez, J. E. Proust, and L. Ter-Minassian-Saraga, Mol. Cryst. Liq. Cryst. 42, 167 (1977). 8 E. Dubois-Violette and P. G. de Gennes, Le Journal de Physique - Letters 36, 255
(1975).
9 E. Dubois-Violette and P. G. de Gennes, J. Coll. Int. Sci., 57, 403 (1976)
10 A. A.Sonin, The Surface Physics of Liquid Crystal, (Gordon and Breach Publishers
1995), p.118-127
11 H. Yokoyama, Mol. Cryst. Liq. Cryst. 165, 265 (1988).
12 E. R. Smith and B. W. Ninham, Physica 66, 111 (1973).
13 J. Bernasconi, S. Strassler, H. R. Zeller, Phys. Rev. A, 22, 276 (1980)
14 Minhua Lu, K. H. Yang, T. Nakasogi, and S. J. Chey, SID Intl. Symp. Dig. 29, 446 (2000). 15 Hemasiri Vithana, David Johnson and Philip Bos, Jpn. J. Appl. Phys. 35, 320 (1996)
16 Hyung-Ku Kang, Jeoung-Yeon Hwang, Chang-Joon Park, Dae-Shik Seo, Kyung-Chan Kim, Han-Jin Ahn, Jong-Bok Kim, and Hong-Koo Baik, Mol. Cryst. Liq. Cryst. 434, 135 (2005). 17 L. M. Blinov and A. A. Sonnin, Langmuir 3, 660 (1987).
18 L. M. Blinov and A. A. Sonnin, Mol. Cryst. Liq. Cryst. 179, 13 (1990).
122
19 E. Guyon, P. Pieranski, and M. Boix, Lett. In Appl. And Eng. Sci. 1, 19 (1973).
20 D. W. Berreman, Phys. Rev. Lett. 28, 1683 (1972);
21 T. Wilson, G. D. Boyd, E. H. Westerwick, and F. G. Storz, Mol. Cryst. Liq. Cryst. 94,
359 (1983).
22 Lawrence A. Goodman, J. T. Mcginn, and Charles H. Anderson IEEE Trans. Elect.
Dev. 24, 795 (1977).
23 J. Papanek, Ph. Martinot-Lagarde, J.Phys II France 6 (1996) p 205
24 Alfred Wagendristel and Yuming Wang, An introduction to physics and technology of
thin films, (World Scientific, Singapore ; New Jersey, 1994).
25 M. Monkade, Ph. Martinot-Lagarde, G. Durand, and C. Granjean, J. Phys. II France 2,
1577 (1997).
26 H.R. Philipp, J. Phys. Chem.. Solids 32 (1971) 1935
27 H.R. Philipp, J. Non-cryst. Solids 8-10 (1972) 627.
28 A. Hohl, T. Wieder, P.A. van Aken, T.E. Weirich, G. Denninger, M. Vidal, S. Oswald,
C. Deneke, J. Mayer, H. Fuess, J. Non-cryst. Solids 320 (2003) 255.
29 Yuri Reznikov, Oksana Ostroverkhova, Kenneth D. Singer, Jae-Hoon Kim, Satyendra
Kumar, Oleg Lavrentovich, Bin Wang, and John L. West, Phys. Rev. lett. 84, 1930
(2000).
30 H. Katsten and G. Strobl, J. Chem. Phys. 103, 6768 (1995).
31 S. Meiboom and M. Sammon, in The Physics and Chemistry of Liquid Crystal
Devices, edited by G. J. Sprokel (Plenum, NY, 1980), p.13.
32 V. Šeškauskas, P. Adomėnas, and E. Kutaitė, Mol. Cryst. Liq. Cryst. 103, 213 (1983).
123
33 T. Inukai, K. Furukawa, H. Inoue, and K. Terashima, Mol. Cryst. Liq. Cryst 94, 109
(1983).
34 Brett L. Van Hom and H. Henning Winter, Appl. Opt. 40, 2089 (2001).
35 Tatsuo Uchida and Hidehiro Seki, in Liquid Crystal: Applications and Uses, edited by
Birenda Bahadur (World Scientific, 1992), Vol. 3, Chap. 5, p.36.
36 A. Toda, H. Mada, and S. Kobayashi, Jpn. J. Appl. Phys. 17, 261 (1978)
37 S. Shimoda, H. Mada, and S. Kobayashi, Physics of Liquid Crystal Devices, edited by
G. J. Sprokel (Plenum, New York, 1980), pp47-60
38 K. Hiroshima and M. Mochizuki, Jpn. J. Appl. Phys. 19, 567 (1978)
39 D. Armitage, J. Appl. Phys. 51, 2552 (1980)
40 M. Schadt and F. Muller, Rev. Phys. Appl. 14, 256 (1979)
41 H.A. van Sprang and R.G. Aartsen, J. Appl. Phys. 56, pp.251-262 (1984)
42 H. Yokoyama, S. Kobayashi, and H. Kamei, J. Appl. Phys. 61, 4501 (1987)
43 H. Vithana, D. Johson and P. Bos, Jpn. J. Appl. Phys. 35, 320 (1996)
44 Karlheinz Graf, Hans-Jürgen Butt, Michael Kappl, Physics and Chemistry of
Interfaces, (Willey-VCH), Chap. 9, p.185.
45 Koki Hiroshima and Manabu Mochizuki, Jpn. J. Appl. Phys. 19, 567 (1980)
46 M. Monkade, M. Boix and G. Durand, Europhys. Lett., 5, 697 (1988)
47 Miller et al, U.S. Patent # 5011267, April.30 (1991)
48 Shao Shan Chen, SID 84 DIGEST, pp.219-222
49 Jacob Israelachvili, Intermolecular & Surface Forces, 2nd edition (Academic Press,
1991), Chap. 10, p.155.
124
50 J.L. Janning, Appl. Phys. Lett. 21, 15 (1972)