This is a repository copy of Liquid crystal displays.

White Rose Research Online URL for this paper:

https://eprints.whiterose.ac.uk/120360/

Version: Accepted Version

Book Section:
Jones, JC orcid.org/0000-0002-2310-0800 (2018) Liquid crystal displays. In: Dakin, JP and
Brown, RGW, (eds.) Handbook of Optoelectronics: Enabling Technologies. Series in
Optics and Optoelectronics, 2 . CRC Press , Boca Raton, FL, USA , pp. 137-224. ISBN
9781482241808

© 2018 by Taylor & Francis Group, LLC. This is an Accepted Manuscript of a book chapter
published by CRC Press in Handbook of Optoelectronics: Enabling Technologies on 06
Oct 2017, available online: https://www.crcpress.com/9781138102262. Uploaded in
accordance with the publisher's self-archiving policy.

Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.

Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing [email protected] including the URL of the record and the reason for the withdrawal request.

[email protected]
https://eprints.whiterose.ac.uk/
Liquid Crystal Displays

J. Cliff Jones

Soft Matter Physics, School of Physics and Astronomy, University of Leeds, Leeds, UK

IN J.P. Dakin and R.G.W. Brown, (2017) Handbook of Optoelectronics, CRC Press.

Objective 2 4.5 Ferroelectric Liquid Crystal 59

1. Introduction 2 Displays
2. Liquid Crystals and their display 3 4.5 In-Plane Switching LCDs 63
relevant physical properties 4.5.1 Introduction 63
2.1 Orientational Order 3 4.5.2 Basics of IPS Operation 64
2.2 General Anisotropic Physical 6 4.5.3 Fringe Field Switching 67
Properties Mode
2.3 Formulating Liquid Crystal 10 4.6 Vertically Aligned Nematic Modes 67
Mixtures 5. LCDs for non-mainstream and Niche 70
2.4 Functional Liquid Crystal 12 Markets
Compounds 5.1 Introduction 70
3. Basics of Liquid Crystal Devices 15 5.2 Cholesteric LCDs 70
3.1 Basics of Display Construction 15 5.2.1 Temperature, Strain and 70
3.2 LCD Polarisation Optics 21 Gas Sensors
3.3 Basic Operation: Electrically 22 5.2.2 Bistable Cholesterics 70
Controlled Birefringence Mode 5.3 Bistable Nematic LCDs and ZBD 71
3.4 Passive Matric Addressing and 28 5.4 Polymer Dispersed Liquid 75
the Mulitplexing Limit Crystals
3.5 Thin Film Transistors and Active 32 5.5 Liquid Crystal on Silicon (LCoS) 79
Matrix Addressing 6. LCD Modes for potential future 80
4. Standard LCD Modes 38 applications
4.1 Overview 39 6.1 Introduction 80
4.2 Dynamic Scattering Mode 43 6.2 Pi-Cell
4.3 Twisted Nematic Mode 44 6.3 Analogue Ferroelectric Liquid 82
4.3.1 Introduction 44 Crystal Modes
4.3.2 The OFF State and TN 44 6.4 N*-SmC* 82
Design 6.5 Anti-ferroelectric liquid crystals
4.3.3 The ON State and Optical 49 6.6 Deformed helix mode ferroelectric 83
Compensation liquid crystal.
4.3.4 Single Polariser Reflective 51 6.7 The Electroclinic Effect 83
Mode TN 6.8 Blue Phase TFT Displays 83
4.4 Supertwist Nematic LCDs 54 6.9 Cholesteric Flexoelectricity and 86
 the Uniform Helix Mode References 89
6.10 Liquid Crystal Lasers 86 Suggested Further Reading 98
7. Conclusions 89
Acknowledgements 89
technology that would eventually lead to Kubrick’s
Objective vision becoming reality: the first practical liquid
Over the past half-century, Liquid Crystal Displays crystal display, or LCD [3, 4].
have grown to be one of the most successful Although the first device utilising liquid
optoelectronic technologies, becoming an integral part crystal electro-optical effects dates back to 1934 [5],
of communication devices and often an enabling the era of LCD research had begun in earnest with the
technology. This success has required many adaptions ground breaking inventions at RCA in the mid 1960s
to meet the requirements of ever-increasing [2, 6, 7]. This work was driven by RCA CEO David
complexity and performance. Indeed, it is the Sarnoff’s dream for “hang on the wall” television
adaptability of liquid crystal devices that grounds their displays [6] and inspired by the work of George Gray
success. After outlining the basic physics of liquid [8] of Hull University in the UK, Glenn Brown at Kent
crystals and device construction, the various modes State University in the US [9] and Richard Williams
used in commercial displays will be reviewed, both for [10, 11] at RCA’s Sarnoff Laboratories in Princeton.
mainstream and niche markets, together with a LCDs offered both reflective mode and backlit
summary of the important complimentary transmissive modes, in displays with a simple
technologies. Finally, liquid crystal devices that have construction, light weight and flat form factor. Display
promise for future applications in optoelectronics will power consumption and operating voltages were
be discussed. attractively low and suited for being driven by CMOS
circuitry.
1 Introduction Soon after Heilmeier’s invention of the
Dynamic Scattering Mode liquid crystal display [3, 4],
On May 28th 1968, the cinema audience at the Loews RCA terminated their LCD investment, but the seed
Capitol Movie Theatre on Broadway, New York, USA had been sown. Research groups from around the
would have been astounded by the vision of the future world, notably from Europe [12] and Japan [7, 13],
presented to them in Stanley Kubrick’s masterpiece entered the field, stimulated by the RCA Press
2001: A Space Odyssey [1]. Amongst the announcement. Having teamed up with Martin Schadt
technological advances envisaged, from an at Hoffman La Roche in Switzerland, ex-RCA
International Space Station, video telephony to the researcher Wolfgang Helfrich invented the twisted
omnipresence of computers, the viewers will have nematic LCD in 1970 [14], a device with a more
been no less captivated by the tablet computers that attractive optical appearance than its Dynamic
astronauts Dave Bowman (Keir Dullea) and Frank Scattering Mode predecessor, and a design that was to
Poole (Gary Lockwood) apprise themselves with from become the mainstay of flat panel displays for the
the BBC12 podcast, after being awoken from following three decades. In the UK, Gray’s group had
hibernation. Those devices had flat panel displays able formulated the first stable room temperature liquid
to present full colour, video information, something crystal compounds [15] allowing mixtures with
that would be as unfamiliar to the audience in 1968, as operating temperatures below zero degrees centigrade
it is familiar to us now. Just one block away, at the [16] to be formulated by Peter Raynes at the Royal
Headquarters of RCA in the Rockefeller Plaza on 6th Signal and Radar Establishment in Malvern, UK. In
Avenue, George Heilmeier was at a press conference Japan, Tomio Wada at Sharp led a joint collaboration
[2] to announce his recently published patent of the with Dainippon Ink and Chemicals that launched the

2
world’s first commercial LCD, incorporated in those shown in figure 1. There are a variety of liquid
Sharp’s electronic calculator EL-805 in early 1973. crystal phases that may appear between the isotropic
The promise of flat-screen, low power, high liquid and solid crystal of certain organic compounds.
information content displays that met Sarnoff’s goal Another, less common expression for liquid crystals is
for television displays seemed imminent [17]. the term mesogenic, with molecules capable of
Developments in the fabrication of LCD devices [7, forming liquid crystals being called mesogens, and the
12, 13] led to early success in applications such as various phases termed mesophases. The simplest of
wristwatches and pocket calculators. However, it the liquid crystal phase is the nematic (denoted N for
would take a further three decades of investment, shorthand), figure 1a), which has no positional order,
invention and development before the LCDs replaced as with a conventional isotropic liquid, but the liquid
the dominant cathode ray tube displays used in crystal phase has long-range orientational order of the
televisions. Today’s state-of-the-art ultra-high long molecular axes. The molecular axes tend to
resolution LCDs for UHDTV combines technologies orientate in a common direction called the director,
such as thin-film transistors behind each colour sub- represented by the unit vector n. This orientational
pixel, optical compensation layers to give the widest order leads to crystal-like anisotropy of physical
angle of view, new alignment modes and addressing properties such as the refractive indices, dielectric
methods to achieve 120Hz frame rates, liquid crystal constants and conductivities. However, the lack of
mixtures that can operate from -20°C to +80°C to positional order means that the nematic is fluid, and
produce displays from several mm to over 100” can flow when perturbed. The statistical distribution of
diagonal, at a price affordable for mass market the molecular axes is quantified by order parameters.
adoption. Although it took this period to achieve For a cylindrically symmetric phase, such as the
Sarnoff’s goal, the roadmap includes many consumer nematic, which is composed of cylindrical molecules,
products that have been enabled by the emerging the degree of order is defined using the S order
LCD: flat screen desktop monitors, laptop computers, parameter:
mobile phones and tablets. LCDs are the purveyors of
the Internet age. #
𝑆= 3 𝑐𝑜𝑠 $ 𝜉 − 1 , (1)
$
There is a plethora of different liquid crystal
display types, each with properties that have been
where the brackets < > represent the statistical average
optimized for different markets. After reviewing some
over the coherence volume of the material, and the
basics of liquid crystal science and display
deviation from the director of an individual long
construction, a selection of typical LCD modes is
molecular axis is 𝜉. The anisotropic physical
described in this chapter, together with some of the
properties of the liquid crystal are related to S, which
more esoteric devices that meet the needs of niche
varies with temperature with the approximate form
markets or are yet to achieve commercial success.
[21]:

2 Liquid Crystals and their display relevant - 0

physical properties 𝑆 = 1− , (2)
-./

2.1 Orientational Order where ν is a material constant, typically 0.15 ≤ ν ≤

Although discovered by Friedrich Reinizer in 1888 0.2, and TNI is the nematic to isotropic transition
[18], the true nature of liquid crystals, and the temperature and T is the temperature below TNI.
oxymoronic term liquid crystal itself, was described a Above TNI, S = 0 by definition. This expression
year later by Otto Lehmann [19, 20]. The type of neglects the weakly first order nature of the nematic to
liquid crystal used in all devices to date is formed from isotropic phase transition, which jumps immediately
rigid rod-like (calamitic) organic compounds, such as

3
Figure 1. Some liquid crystals basics: a) Schematic representation of: a) the nematic phase and the n director;
b) the smectic A phase; and, c) the smectic C phase. The principal axes for; d) the cylindrical nematic and smectic A
phases, and e) the monoclinic smectic C. The spontaneous twist of the chiral nematic or cholesteric and definition of
the helical pitch length P are shown in f), and the spontaneous twist / bend of the chiral smectic C and ferroelectric
spontaneous polarisation Ps are shown in g).

from 0 to about S = 0.43 at TNI in actuality. However, A phial containing a nematic liquid crystal is
equation (2) is satisfactory a couple of °C below TNI, milky in appearance. Unlike the common colloidal
predicting typical nematic materials to have S ≈ 0.60 fluids with this type of behaviour (such as milk itself),
to 0.80 for typical nematic material at ambient this appearance occurs for a pure, single compound. It
temperatures. arises because, although the liquid crystal molecules
tend to point in the same direction locally, this

4
direction changes over micron length scales due to several hundreds of microns in a nematic material
thermal fluctuations, causing strong Tyndall scattering doped with a small concentration of a cholesteric
of the light. When heated to the isotropic liquid, the dopant. In the chiral smectic C phase (SmC*), figure
scattering disappears suddenly and the liquid is clear; 1g), the loss of mirror symmetry induced through the
for this reason, the temperature of the liquid crystal to addition of chirality has two notable effects. Firstly,
isotropic transition (for example, TNI) is often termed the n director undergoes a spontaneous twist and bend
the clearing point. deformation (so that the c director forms a twist in a
Almost all LCDs sold are based on the simple similar way to the cholesteric), again with pitch P.
nematic type of liquid crystal, because it is the easiest Secondly, the combination of the loss of the mirror-
to handle and most well understood. However, there plane with the monoclinic symmetry of the smectic C
are many other phases that can form that are leads to a spontaneous ferroelectric polarisation Ps in
intermediate between the liquid and crystal states, the plane of the layers and normal to the tilt plane [23].
from those with higher degrees of order (layered As the c-director precesses about the layer normal in a
smectics, helical cholesterics, and columnar phases), helical fashion, so the net ferroelectric polarisation
those using different shaped molecules (discs, banana- cancels out throughout the bulk of the phase.
shaped, bowl-shaped, main-chain and side-chain Liquid crystal phases are formed for
polymeric) to those systems where the liquid crystal molecules that comprise both a rigid core structure and
nature is due to changes in concentration, rather than, pendant flexible hydrocarbon chains. The occurrence,
or in addition to, temperature, (lyotropic and temperatures and phase sequences of liquid crystals
chromonic phases). The interested reader is referred to depend on the relative proportion of the flexible chains
reference [22] for a comprehensive review on all and rigid core: the flexible groups are needed to space
aspects of liquid crystal science. For the present the cores sufficiently to prevent crystallisation but
chapter, only the smectic and cholesteric phases are sufficiently small to allow orientational ordering of the
considered because of their (limited) application to cores. Predicting the temperature range over which
displays. Smectic liquid crystal phases have some liquid crystal phases form is a complex subject,
degree of positional order in addition to orientational requiring skill and experience for the synthetic
order. The simplest smectics are termed smectic A chemist. For the simple rod-shaped calamitic
and C (for which the shorthand SmA and SmC is used, molecules used in commercial displays, the transition
respectively) and they exhibit one-dimensional from isotropic to liquid crystal phase is related to the
positional order, to form layers of nematic-like length of the molecule. However, viscosity is also
material. The layers of such phases are described by dependent on molecular size, and so most practical
the unit vector a. Smectic A have n parallel to a, liquid crystal compounds are formed from two or three
figure 1b), whereas the director n is at an angle θC to a ring structures to keep viscosity low and switching
in the smectic C phase, figure 1c). Both the nematic speeds high. The ring structures in the core may be
and smectic A (SmA) phases have cylindrical benzene, cyclohexane, pyrimidine, thiophene, etc.,
symmetry about n, figure 1d), whereas the SmC has joined together with para- substitution (e.g. 1,4
monoclinic symmetry. Nematics comprised, at least in benzene) or with near-linear linking groups (such as
part, of chiral molecules undergo a spontaneous twist esters, ethyl groups), and terminated on one or both
of the director, to form a helical structure figure 1f). sides with alkyl chains. In each case, the molecular
Chiral nematics, represented by N*, are often called core retains its near-cylindrical symmetry with each of
cholesterics after the first liquid crystal behaviour was any low energy conformers allowed by the core, such
observed for this phase in a compound derived from as rotations about the ester or ethyl linkages. Table 1
cholesteryl benzoate [18]. A full 2π rotation of the n lists a range of different compounds that are typical for
director is defined as the pitch length, P, which may nematic liquid crystal mixtures, together with some of
vary from sub-micron in some pure compounds, to

5
the key physical parameters [24] related to typical
device behaviour. 𝜌𝑁D ℎ𝐹 1
𝑛@$ = 1 + 𝛼 − Δ𝛼𝑆
𝑀. 𝜀; 3
2.2 General Anisotropic Physical Properties
All liquid crystal displays operate because of (4)
the combination of the anisotropic physical properties 𝜌𝑁D ℎ𝐹 2
𝑛?$ = 1 + 𝛼 + Δ𝛼𝑆
combined with fluidity exhibited by liquid crystals. 𝑀. 𝜀; 3
Tensorial anisotropic physical properties such as the
refractive indices and electric permittivities can be where ρ is the density, NA Avogadro number, M the
diagonalised to give principal values parallel (𝒏|| ) and molecular weight, 𝛼 is the mean molecular
perpendicular (𝒏5 ) to the director, [25]. Most polarisability, and Δ𝛼 is the difference between the
commercially available devices use reorientation of molecular polarisability parallel to the long 𝛼J and
the director induced by an applied electric field E short 𝛼K axes. The cavity field factor h, and the
coupling to the dielectric tensor ε, to reduce the Onsager reaction field F relate to the internal field
electro-static free energy GE, given by: experienced by an average molecule, approximated for
a spherical cavity by:
# #
𝐺7 = − 𝑫. 𝑬 = − 𝜀; 𝑬. 𝜺. 𝑬 =
$ $
# $ 3𝜀 1
− 𝜀; Δ𝜀 𝒏. 𝑬 (3) ℎ= ≈1 ; 𝐹= 2𝜌𝑁 𝛼 𝜀−1 ≈1
$ 2𝜀+1 1−3 𝑀.𝜀𝐴 2𝜀+1
0

where the dielectric anisotropy Δε is the difference of (5)

the principal components 𝜺|| − 𝜺5 . Equation (3)
where mean anisotropic properties are expressed using
predicts that the director tends to reorient parallel to
a bar above the symbol, such as for the mean
the applied field if the material has a positive Δε, and
polarisability on the molecular scale:
perpendicular to the field if Δε is negative. Similarly,
many devices rely on changes of optical retardation #
𝛼= 2𝛼K + 𝛼J , (6)
from the material birefringence Δn (= 𝑛|| − 𝑛5 = 𝑛? − Q

𝑛@ ), where e and o refer to the extra-ordinary and

ordinary rays respectively) to give the perceived and permittivity and refractive index on the
optical modulation. For example, if viewed between macroscopic scale:
crossed polarisers, a device will appear isotropic, and
# #
therefore black, when viewed along the optic axis 𝜀= 2𝜀5 + 𝜀|| ;𝑛= 2𝑛@ + 𝑛?
Q Q
(parallel to ne) and will exhibit some white light (7)
transmission when viewed in a direction at an angle to
the optic axis due to the effect of the optical The birefringence is given by:
retardation.
The physical parameters depend on both STU VW [\
∆𝑛 = 𝑆 , (8)
chemical structure and temperature. Anisotropic X.YZ Q]^]_

properties, such as Δn and Δε are also related to the

order parameter S, whereas the fluid properties, such and is largely dictated by the polarisability anisotropy
as density, bulk viscosity and the mean refractive Δα of the constituent molecules. The density ρ is
indices and permittivities, are primarily related to usually close to that of water at 20°C, with a selection
absolute temperature. The uniaxial refractive indices of nematic compounds ρ = 990 ± 50 kgm-3 [26], and
are given by [25]: with a near linear temperature dependence through the

6
Table 1: Typical Nematic Liquid Crystal Compounds and their physical properties.

LC phases that is typically -1kgm-3K-1. Thus, the 𝜌𝑁D ℎ𝐹 $ $ 1

𝜀5 = 𝑛@$ + 𝜇 1 + (1 − 3 cos $ 𝛽)𝑆
temperature dependence of the birefringence is 3𝑀. 𝜀; 𝑘𝑇 2
dominated by the order parameter S. A calamitic
liquid crystal has a positive birefringence because Δα (9)
is positive for the cylindrical molecular core $
𝜌𝑁D ℎ𝐹
structures. If the rigid core is short and comprises 𝜀|| = 𝑛?$ + 𝜇 $ 1 − (1 − 3 cos $ 𝛽)𝑆
3𝑀. 𝜀; 𝑘𝑇
weakly polarisable saturated moieties such as
cyclohexanes, the birefringence will be low. It will be where 𝛽 is the angle between the molecular dipole 𝜇
higher for longer, unsaturated groups such as phenyls and the long molecular axis. The dielectric anisotropy
and ethynes, reaching Δn ≈ 0.4, but more typically is:
being maximum at Δn ≈ 0.26 for light stable
compounds. STU VW g
Δ𝜀 = 𝑛?$ − 𝑛@$ − (1 − 3 cos $ 𝛽)𝜇 $ 𝑆
Similarly, the uniaxial electric permittivities hX.YZ i-

are given by [25]: (10)

and

7
 STU VW jg # $ 𝟐𝝅 $
𝜀 =1+ 𝛼+ GK = 𝑘## 𝛁. 𝒏 + 𝑘$$ 𝐧. 𝛁×𝒏 − +
XYZ Qi- $ 𝑷
(11) $
𝑘QQ 𝐧×𝛁×𝒏 , (12)
There is more freedom for controlling the dielectric
anisotropy than for birefringence by placement of where P = ∞ for the usual achiral nematic case, and kii
strong dipole moments onto the core structure. (i = 1, 2, 3) are the splay, twist and bend elastic
Strongly positive materials result from polar moieties constants, respectively.
such as cyano- terminal groups, or 3,4,5 substitution of The elastic constants are important to the
fluoro- groups onto a terminal phenyl group. Such display engineer because they dictate the amount of
placement ensures that β is kept low, with Δε ≈ + 50 deformation induced by the applied electric field. Both
being readily achievable. It is harder to make strongly splay and bend elastic constants contribute to
negative materials for a number of reasons. Firstly, the distortions in the plane of the director and deforming
𝑛?$ − 𝑛@$ term of equation (10) is always positive and torque, whereas twist occurs where the distortion is
leads to Δε ≈ +2 even if there is a negligible dipole perpendicular to the director and torque. There have
moment. Equation (10) predicts that transverse been extensive studies of the relationship between
dipoles, where β ≈ 90º, are only half as efficient at molecular structure and the elastic constants, because
of this important role they play in the operation of
contributing to a negative Δε, due to the cylindrical
liquid crystal displays. To a first approximation, the
symmetry. Moreover, bulky polar side groups detract
temperature dependence of the elastic constants
from the rod-like shape of the molecule that leads to
follows S2, a relationship that holds well for k11 and
liquid crystallinity. Nevertheless, mixtures with Δε ≈ -
k22, where the ratio k22/k11 is relatively insensitive to
6 have been achieved for modern LCD TV
temperature and chemical structure with 0.5 ≤ k22/k11 ≤
applications. Note, equation (10) also predicts that a
0.8, and usually k22/k11 ≈ 0.5. The behaviour of the
dipole moment at β = 48.2º contributes equally to
bend elastic constant k33 is more complex [28], with
𝜀|| and 𝜀5 ; even a large dipole moment oriented at β ≈
k33/k11 being both temperature dependent and showing
52º gives Δε ≈ 0. a strong dependence on structure, with values ranging
Implicit in the definition of liquid crystal from: 1.0 ≤ k33/k11 ≤ 2.2. Approximating the
behaviour is the concept of orientational elasticity and constituent molecules of the liquid crystal to rigid,
the energetic cost associated with deforming the hard rods of length L and diameter W leads to the
director field. Whereas, a crystal solid has elasticity approximate relationship:
associated with the positional translation of the
constituent molecules, the liquid crystal has elasticity k33/k11 ~ L2 / W 2 (13)
associated with changing director orientation. The
curvature strain tensors of phases with cylindrical which provides a useful rule of thumb for the LC
symmetry contain terms in splay, twist and bend mixture designer. However, there is a strong influence
deformations, [27] as shown in figure 2a). The on this elastic ratio from short-range local ordering in
elasticity of the chiral nematic is equivalent to that of the nematic phase, due to intermolecular dipole
the achiral nematic, but includes 2π/P subtracted from correlations or local smectic ordering. For example,
the twist term to represent the spontaneous twist of smectic local ordering of positive Δε materials leads to
director. The elastic bulk free-energy density for a higher k33/k11.
nematics and cholesterics GK is given by:

8
Figure 2. Nematic elasticity and disclinations: a) Splay, Twist and Bend deformations; b) Line disclinations of
strength +1, +½, -½ and -1. In each case the disclination continues normal to the page, and the director remains
invariant in this direction (cylindrical symmetry). The defect core is indicated in grey; c) flexoelectric behaviour from
banana-like molecules in a bend deformation; and, d) flexoelectric behaviour from wedge-like molecules in a splay
deformation.

Liquid-like fluidity of a liquid crystal is case, there are five viscosity coefficients [29],
dictated by the viscosity, which determines the describing flow of the director in directions parallel
switching speed of LCDs. How quickly the director and perpendicular to the director and shear force
reorients with respect to a deforming torque is directions. The effect of fluid flow can play an
determined by the ratio of the viscosity and the elastic important part in device behaviour, such as the optical
constants. However, even for the simplest nematic bounce that occurs in twisted-nematic displays due to

9
backflow [30]. Full understanding of these effects which is typically of the order of 10nm in diameter.
requires numerical modelling [31]. For simplicity, For the nematic phase, disclinations of strengths +1, -
back-flow is often ignored, and the time dependence 1, +½ and -½ are common, with +½ and -½ appearing
of director reorientation without coupling to mass flow as pairs and usually form defect loops, or may
is represented by the single twist viscosity γ1. terminate at interfaces such as the containing
Viscosity has a strong temperature dependence related substrates of a device. Avoiding defects is a key part
to the change in order as well as the Arrhenius of the design of most liquid crystal displays, although
dependence common to liquids [32]: recently there has been an insurgence of interest in the
potential applications for liquid crystal devices with
t
controllable formation of topological defects, such as
𝛾# = 𝑎# 𝑆 + 𝑎$ 𝑆 $ 𝑒 uv
the zenithal bistable display described in section 5.3.
(14)
The constituent molecules for the liquid
crystal do not exhibit perfect cylindrical symmetry but
with a1 and a2 are material dependent constants, U is
have more complex shapes. For example, they may
the activation energy for molecular movement, k is the
not have a two-fold rotational symmetry about either
Boltzmann constant and T is the absolute temperature.
their short or long axes, leading to “wedged” or
Most commercial liquid crystal mixtures are based on
“banana” molecular shapes, respectively. Normally,
aromatic systems, where a1 >> a2 for compounds with
the nematic includes all possible orientations of the
low birefringence, a2 >> a1 for highly birefringent
molecules that maintain the symmetry of the phase.
compounds, and a1 ≈ a2 ≈ 10-14 s.m-2 and U ≈ 0.3eV for
However, with bend and splay deformations, the
Δn ≈ 0.18. At low temperatures, the viscosity often
distribution of asymmetric molecules is shifted slightly
deviates from this Arrhenius type of behaviour as it
as the molecules pack. For example, banana-like
diverges towards a low temperature glass transition.
molecules will tend to orient so that molecular
The standard Vogel-Fulcher-Tammann equation has
curvature follows that of the director field, figure 2c).
been applied successfully to a range of nematic
Similarly, the apex of wedge-like molecules points
compounds [26]:
into the splay, figure 2d). In either case, any
y
molecular dipole will then contribute to a macroscopic
log 𝛾# = 𝐴 + polarisation [33] and flexo-electric terms are required
-^-Z
(15) in the free energy. The degree of polarisation for splay
and bend deformations is represented by the
where A and B are material dependent constants, and flexoelectric coefficients e1 and e3 respectively, which
T0 is a temperature that is typically about 20°C below are dependent on the material shape, dipole moments
the liquid crystal glass transition. and density.
Continuum descriptions of the director field
are not always satisfactory, since the director field
often also includes topological discontinuities, called 2.3 Formulating Liquid Crystal Mixtures
defects or disclinations [27]. Indeed, it was the thread- Table 1 includes some of the important
like appearance of nematic samples containing such display-related physical properties for a number of
defects that originally led to the naming of the phase, common nematic compounds, including temperatures
from the Greek for thread (nematos). Both point for the nematic to isotropic transition TNI, as well as
defects and line defects are common; examples of the room temperature values for the optical, electrical,
nematic line disclinations with different strengths are elastic and viscous properties. Each of these
shown in figure 2b). At each discontinuity in the compounds is stable at operating temperatures and
director field, there is local melting of the phase and when exposed to light, allowing device lifetimes of
the liquid crystal order is zero at the disclination core, scores of years to be achieved. As shall be shown

10
through the following sections, the properties of each the Arrhenius behaviour of conventional liquids then
of the compounds might be suited to different display the viscosity of an n component mixture is given by:
technologies. For example, devices using optical
]
scattering effects require a high birefringence to log 𝛾# = { 𝐶{ . log (𝛾# ){
become strongly scattering whereas polarised light (16)
displays using the retardation effect of the liquid
crystal, often require a lower birefringence to enable where i = 1, 2, 3, …n, and Ci is the concentration of
higher device spacing (and hence higher yields). component i such that ]{ 𝐶{ = 1. This logarithmic
Modes such as the supertwist nematic require k33/k11 to concentration dependence means that highly viscous
be high, whereas the simpler twisted nematic display additives can be used at low concentration without
requires k33/k11 to be kept low. Such requirements are increasing the mixture viscosity significantly. For
rarely isolated, being combined with a range of target example, it is common practice to increase the clearing
physical properties, such as appropriate temperature point through the addition of three or four ringed
range of the required liquid crystal phase, low compound where the linear increase in TNI is
viscosity, to more complex needs, such as low accompanied with a disproportionately small viscosity
temperature dependence of elastic constant ratios, etc. increase.
This necessitates mixtures to be used, frequently None of the compounds in Table 1 show a
involving scores of components. room temperature liquid crystal phase, rather freezing
Properties such as the birefringence Δn, the to a crystal form. The liquid crystal phases usually
isotropic to nematic transition temperature TNI, and supercool below the melting point, because of the
splay elastic constant k11 depend roughly linearly with strong first order nature of crystallisation; this is
concentration. The dielectric anisotropy Δε, and twist particularly true in a thin container such as an LCD,
and bend elastic constants k33 and k22 are slightly more where crystallisation is suppressed by the surfaces.
complex, because of the effects of dipole correlations However, LCDs require operation typically between -
and smectic ordering. For example, the epoch-making 20ºC to +70ºC, and to be stored for months down to -
compound pentyl-cyano-biphenyl (5CB, compound 1 40ºC. Such temperature ranges require the formation
in Table 1) has a measured ∆ε of about 15 close to of eutectic mixtures to suppress melting, at
room temperature can be treated as ∆ε of about 24 concentrations approximated by the Schroeder-Van
when adding into mixtures. This is because the anti- Laar equation:
parallel dipole correlations of the pure compound that
[}~ # #
effectively reduce the parallel dipole moment are ln 𝐶{ = − −
• - -~
disrupted in a multi-component mixture. Care still has
(17)
to be taken when formulating mixtures, since the
breaking of the dipole correlations can lead to the
where R is the Rydberg constant, Ci, ∆Hi and Ti are
unmasking of smectic behavior: it is quite common for
the molar concentration, enthalpy of freezing and
a mixture of polar and apolar nematic compounds to
melting point of the ith component, respectively. This
exhibit an unwanted smectic phase despite neither
provides a guide for the material scientist to calculate
component having smectic behaviour. Such a phase is
the eutectic composition, given that, at the eutectic
called an “injected” smectic and arises because the
temperature, the concentrations sum to 100%. In
dipole correlations of the polar compound were
practice, more thorough empirical methods [34] are
preventing the formation of the smectic layers.
required to formulate commercial mixtures. Table 2
Ignoring the order parameter related terms in
includes the composition of three typical positive Δε
equation (14) and considering the viscosity to follow
eutectic mixtures, E7, ZLI2293 and ZLI 4792 together
with their important physical properties.

11
Table 2 Physical properties of selected nematic LCD mixtures. Data collected from references [35 - 37].

ZLI – 4792
Mixture E7 ZLI 1132
Typical Use Passive Matrix TN Active Matrix TN

Composition

Nematic
Temperature S < -30 N 58 I S < -6 N 71 I S < -40 N 92 I
range (°C)
Δn ; no
(589nm, 0.2253 ; 1.5211 0.1396 ; 1.4830 0.0969 ; 1.4794
20°C)
Δε ; ε ⊥

(1kHz, 13.8 ; 5.2 13.1 ; 4.6 5.2 ; 3.1

20°C)
k11 ; k22 ;
k33 11.7 ; 8.8 ; 19.5 1.95 13.2 ; 6.5 ; 18.3
(pN, 20°C)
Dynamic
bulk
465 200 150
viscosity η
(cP, 20°C)
Twist
viscosity γ1
180 250 109
(mPa.s,
20°C)
Flexo-
electric
coefficients
e1 + e3 ; +15 ; 12.2 |10| ; -15
e1 - e3
(pCcm-2,
20°C)

2.4 Functional Liquid Crystal compounds compounds, which are used for important display-
related purposes, as summarised below.
As shall be shown, it is not just the nematic phase that 1. The inclusion of one or more chiral centres
is used for LCDs. A variety of mesogenic compounds (denoted *) in the flexible end chain of a
have been used to impart some new functionality to mesogen imparts a tendency for spontaneous
the system. Table 3 lists examples of mesogenic twist of the director. The material may exhibit
inherent cholesteric (N*) or chiral SmC*

12
 phases (such as CE3 in Table 3) or may reversibly [41], or irreversibly by combining
induce chirality through doping into a the azo-mesogen into polymerisable groups
nematic or SmC host material. The important [42].
characteristic of the chiral compound as a 5. Reactive mesogens have become a major tool
dopant is its helical twisting power and for the displays field [43]. These compounds
handedness. These are related to molecular exhibit conventional liquid crystal behaviour,
structure through various empirical rules, either individually or when used in mixtures.
such as those of reference [38]. The inclusion of a photo-reactive group in
2. Smectic phases occur with longer terminal one or both of the terminal end-chains allows
chain groups on the mesogenic core. The the material to be polymerised by exposure to
formation of tilted phases, such as the SmC UV light. Hence, the reactive mesogen may
occurs with suitable placement of transverse be aligned appropriately (through interaction
dipole moments, though precise control of with the surface, electric fields or polarised
phase transition temperatures and sequences light) and that liquid crystal structure
remains somewhat an art for the chemist [e.g., stabilised by polymerising the reactive
reference 39]. mesogen. The solubility of the reactive
3. Pleochroic and fluorescent dyes can be added mesogens is high and so a variety of systems
to liquid crystals to give appropriate optical are possible, from linear elastomers, to
functionality. Pleochroic dyes such as the polymer stabilised networks and gels, and
anthraquinone shown in table 3 have liquid crystal polymers. The photo-reactive
reasonably good miscibility in the cyano- groups are often acrylates, and so may also be
biphenyls, leading to anisotropic absorption used in conjunction with standard acrylate
that is much greater parallel to the director, cross-linkers and photo-initiators. Reactive
than perpendicular to it [40]. This is used in mesogens have been used widely, from
Guest-Host displays, where the polariser is creating patterned optical retarders, functional
replaced by the constituent dye mixed into the alignment layers, to creating templates for
liquid crystal, allowing switching of the extending the temperature range of narrow
appropriate wave bands between absorbing liquid crystal phases.
and transmitting states. Performance is 6. Discotic liquid crystals have a symmetry axis
dictated by the combined order parameter of that is parallel to the short molecular axis
the system and the direction of the dye [44]. This leads to distinctive physical
transition moment with respect to the liquid properties, such as negative birefringence,
crystal director. one-dimensional conductivity and
4. Including a central azo- moiety into the core semiconductivity, and the formation of
of a mesogen allows photo-induced columnar rather than smectic phases. Many of
realignment of the director due to anisotropic the other anisotropic physical properties, such
absorption of the dye. Irradiating the as the order parameter, permittivities and
molecule with polarised light causes elastic constants are similar to those of their
reorientation of the dye molecule to the calamitic (i.e. rod –like) counterparts. A
direction normal to the polarisation, usually particularly important application for LCDs is
via transition between the trans- and- cis the use of polymerisable discotic liquid
isomers. This may be used to effect photo- crystals to form optical compensation plates.
reorientation of the nematic director, either

13
Table 3 Examples of Functional Mesogenic Compounds

# Compound Use Key properties

(+)-4-n-Hexyloxyphenyl- 4-(2-methylbutyl)biphenyl-4’-
carboxylate (CE3)
O N* ; Helical twisting
Chiral Dopant; power;
1 C2H5 C* CH2
O O-C6H13 Ferroelectric Induced Ferroelectric
compound polarisation.
CH3
K 65°C SC* 79°C N* 162°C I
2,3-difluoro-1-ethyloxy-4-[trans-4-propyl cyclohexyl]
biphenyl Low γ1
VAN
2 F F S C; High ∂ε
N phase;
FLC Host
C3H7 OC2H5 Wide SC

Example Anthraquinone dye

H
C4H9 N O OH High Order
Parameter;
3 Pleochroic Dye
High dichroism;
High miscibility
O N C4H9
H
1,2-Bis(4-hex-5-enyloxyphenyl)diazene
Solubility;
Optically induced
4 S;
reorientation
Absorption efficiency

RM257:
1,4-Bis[4-(3-acryloxyproyl benzoxyl]-2-methylbenzene.

Bifunctional Solubility;
5
Reactive Mesogen Photo-reactivity

DB126:
triphenylene hexa-(2-methyl-4-n-decyloxy)benzoate
R R
Discotic Nematic.
R= O Acrylate version
6 OC10H21 used for optical Negative Δn
R R O compensator films

R R
K 109 °C (Col 75°C) ND 164°C I

14
3 Basics of Liquid Crystal Devices ultra-thin (<100µm) glass have been used for
high-end displays, particularly for use in
3.1 Basic Display Construction portable display applications and curved
Much of the basic scientific understanding of liquid screen HDTV.
crystals was available in the first half of the twentieth • Indium Tin Oxide. Following the initial work
century. Mauguin [45] had found that, contrary to the done in the mid 1960s at Philips [50, 51], and
optical rotation caused by a quarter-wave plate, in the Far East at the Japanese Government
polarised white light could be guided by a liquid Research Institute [52], ITO has become the
crystal with a twisted structure, allowing light to be dominant transparent conductor for the display
transmitted between crossed polarisers by samples industry, as well as for many other
with a twist of 90°. Fréedericksz had shown field applications. The layer is sputtered onto the
induced switching and the existence of a critical glass surface at between 50nm to 200nm,
threshold for magnetic [46] and electric [47] fields depending on the display requirements. The
applied to a uniform sample. Châtelain [48] achieved thin layer causes loss of light through
uniform alignment of the liquid crystal director using interference and reflection, but these
thinly spaced samples and rubbing to define the thicknesses usually lead to about 88% and
orientation. Such scientific advances prompted the 85% transmissivity of the glass overall, and
first liquid crystal device to be patented by the
correspond to sheet resistances of 80 and
Marconi Wireless Company in the UK in 1936 [49].
However, the first LCD commercialisation had to wait 15 respectively. It is important to
until the early 1970s, and the availability of various understand how sheet resistance works: a
other components, many of which were invented and square of ITO gives a resistance of 15Ω
developed independently for other applications. These regardless of its area. Calculation of the
associated technologies include: resistance of any electrode requires the
• Glass. Typically 0.7mm thick, polished glass number of squares to be multiplied by the
is remarkably optically uniform (n = 1.520 sheet resistance. So, for example, the
±0.005), transparent (91% transmission from resistance of a display electrode that is 10cm
320nm to 700nm) and flat (±0.1m m). long and 200m m wide is typically about 15
Together with its excellent mechanical, x L/W = 7.5kΩ.
thermal and economic properties it is hard to • CMOS circuitry. Complimentary Metal-
imagine a better substrate material. Today’s Oxide-Semiconductor integrated circuits were
40” TV displays are made on $2B production invented by Wanlass at Fairchild in 1963 [53],
lines that handle Generation 10 glass whereby p-type and n-type transistors are used
(2850mm x 3050mm), though most small in parallel to provide low power logic circuitry
displays are still made using Generation 4 (and hence suitable for use with battery
production lines (550mm x 650mm). Two operated equipment) and logic output voltages
types are common: soda-lime glass for low of between 3.5 and 5V, which could be used
cost displays and boro-silicate glass for high- directly to address the liquid crystal. The first
end TFT (thin-film-transistor) displays. The LCD product was launched by Sharp in April
latter has a low alkali ion content and so adds 1973: the EL-805 electronic calculator. The
far fewer ionic impurities to the contacting LC CMOS provided both the logic for the
that would otherwise prevent the TFT calculator operation and the driving circuitry
maintaining its charge across the pixel. for the dynamic scattering mode LCD. This
Recently, ion-implanted reinforced glass and was followed in October 1973 by the 06LC

15
Figure 3. Construction and operation of a twisted nematic LCD: a) Schematic diagram of a 4 pixel TN in the
off-state; b) cross section of a modern chip-on glass LCD; c) Typical electro-optic response for a monostable LCD,
such as the TN; d) Schematic diagram of the 4 pixel TN with voltage Vs applied to one row, and data ±Vd applied to
the columns, such that one pixel (bottom left) is switched dark.

digital watch from Seiko, which used a twisted controlling the polarisation state. Today’s
nematic LCD. polarisers use aligned silver nano-particles,
• Polarising film. Although prism and wire-grid and comprise protective films, adhesives,
polarisers were established technologies, release liners and other filters required for
Land’s invention of iodine doped stretched different display purposes [56].
polymer film polarisers in 1929 [54] and the • The thin-film transistor [57] was invented at
subsequent use of polymerised dichroic RCA by Paul Weimar in the early 1960s. The
nematics by Dreyer in 1946 [55] enabled low original demonstration used Tellurium as the
cost flat-panel displays to operate by semiconductor but other materials researched

16
 at that time included Silicon, Germanium, polymer coating to impart either homeotropic (i.e.
Cadmium Selenide and Cadmium Sulphide. normal to the surface), planar homogeneous or tilted
Most displays today use either amorphous or homogeneous alignment of the director, figure 4. For
poly-crystalline silicon. example, the twisted nematic device includes
polyimide coatings rubbed in a direction to impart a
There are various different types of liquid crystal homogeneous alignment with a surface pre-tilt θS of
display but many of the design principles and modes typically 1° to 2°, though a supertwist TN would
of operation are common throughout. Consider the require a somewhat higher pre-tilt of 2° < θS < 6°. For
common twisted nematic LCD [14] shown in figure 3. the simple TN, a 90° twist from the top to bottom
The display is formed from two containing plates, the surfaces is induced in the device by arranging the
front one at least being transparent, and coated with rubbing directions on the opposing surfaces to be
transparent conducting electrodes, such as indium tin normal to each other. The prevention of tilt and twist
oxide (ITO). The plates are spaced a few microns apart disclinations is ensured by including a small helical
by glass beads, adhered together and the laminate twisting power to the nematic through the addition of
filled with the liquid crystal material and sealed. The very low quantities of a chiral nematic, and by
most common substrates in use are 0.5mm or 0.7mm matching the sign of the helicity to the two pretilts to
glass, though optically isotropic plastic substrates may minimise bend and splay distortion [58]. Ignoring
also be used for low weight and flexibility. Reflecting biaxial surface terms, the orientation of the director at
metal foil layer has been used as the rear-substrate each surface is governed by the surface energy GS:
where flexibility and high temperature thermal
processes are required for a reflective display. The # $
𝐺€ = 𝑊‚ 𝐧. 𝐬 − 𝛉𝐬 +
ITO is etched to form the appropriate electrode $ €
$ $
pattern. This is often a series of rows and column 𝑊… 𝐧. 𝐫 +𝑊‡ 𝐏. 𝐬 𝑑𝑆 , (18)
electrodes on the opposing internal faces of the device.
In such an arrangement, the electrodes form a parallel where s is the surface normal and r is the preferred
plate capacitor, wherein the individual pixels are alignment direction in the plane of the surface, figure
formed in the regions of overlap of the rows and 4a). Changes to the surface tilt are related to the
columns. Of course, other designs are possible, such zenithal anchoring energy, Wθ. Typically, 10-7 ≤ W θ ≤
as polar co-ordinates (formed from axial and radial
10-3 Jm-2 and 10-9 ≤ W ≤ 10-5 Jm-2 [59] though for
β
electrodes), alphanumeric characters and icons, or
typical commercial devices the anchoring energies are
inter-digitated electrodes on one of the substrates to
greater than 10-5 Jm-2 and are usually considered as
provide an in-plane electric field. The most
fixed boundaries. The polar surface term Wp is
sophisticated displays, used from mobile phones to
insignificant for nematics but is important for
HDTV use a thin-film transistor (TFT) on one of the
ferroelectric liquid crystal devices, where there is often
plates, addressed using copper bus lines to provide the
a preferred orientation of the spontaneous polarisation
signals to a pixel electrode. In these instances, the
with respect to the surface normal.
opposite plate is a single electrode, held at earth.
Each device is constructed in two stages: the
Creating an electro-optic effect requires some
back-end process is done under strict clean-room
optical property, such as the reflectance or
conditions, and the front-end, wherein the devices are
transmittance for a backlit display, to be changed with
filled with liquid crystal and the associated optical
an applied field. Today’s commercial LCDs almost
layer and driving electronics added and testing is done.
exclusively use re-orientation of the director from
The back-end processes are done in a clean room and
some initial pre-aligned state, dictated by alignment
consist of producing empty laminates each with the
layers on the inner surfaces of the display. Various
potential to form multiple displays. The factory
alignment layers are possible, but most devices use a

17
 e)

Figure 4. Nematic surface alignment leading to uniform director profile: a) Definition of in-plane azimuthal
angle β and out-of-plane zenithal tilt angle θ; b) Homeotropic; c) Planar homogeneous; and, d) Uniform Tilted from
anti-parallel surfaces with pre-tilt θs; e) Basic structure of the dedicated polyimide AL 1051 from JSR.

equipment dictates the size of the plates used; it is The processes for constructing a passive matrix
important for high yield and efficiency that the glass is display are shown in figure 5. Many of these processes
large enough to allow at least six or eight displays to are common to either passive matrix or TFT driven
be made on a single laminate. Even 47” diagonal active matrix devices. Common back end processes
television displays are produced eight per laminate include:
using generation 8 sized glass (2,200 × 2,500 mm), 1. Glass Cleaning: Glass is supplied with the
and large scale associated manufacturing equipment. appropriate thickness of ITO deposited onto one
The plates are made on a production line side. The plate is washed with de-ionized water,
preventing the use of many standard laboratory based ultra-sound, soaps and sometimes solvents such as
practices, such as oven baking and spin-coating. Each IPA (though this is not preferred due to
step occurs within the TAKT time, which is the time flammability).
that one plate moves from one step to the next; the 2. Electrode patterning: This is done using
TAKT time, by definition, must be the same for all of photolithography. A thin layer of positive photo-
the steps on a given line. Following a brief inspection, resist is printed onto the substrate, soft baked (to
the laminate is then sent to the front-end, where it is harden) and a mask is placed in contact with the
singulated, filled and the remaining components layer. Where exposed to UV, the photoresist is
attached. washed-off by developer, but the photoresist
remains in contact wit

18
 Figure 5 Construction steps for a passive matrix display

the ITO everywhere else. The exposed ITO is speed (rpm), r the roller radius (mm) and ν the
then removed by acid and finally the remaining speed of the plate (mm.s-1). Each of these steps
photoresist is stripped from the patterned ITO needs to be controlled to minimise display
using an aggressive solvent. It is common for the variation.
electrode patterning to be done separately from 4. Spacers. Polystyrene spheres of the required
the standard production process. spacing are either deposited in air or solvent onto
3. Alignment layer. Following another clean step, one of the glass plates that form the laminate. The
the alignment layer is pad-printed onto the density of the spacers depends on the cell gap, but
patterned glass in the areas that form the viewing is typically 50 – 100 per mm2.
area of each display. The alignment layer is 5. Glue seal. A thermal epoxy is syringe deposited
patterned to avoid the part of the electrodes where onto the other substrate to form the boundaries for
bonding (electrical access) is required, and where each individual display. The seal is designed to be
the glue-seal will be deposited (to ensure a strong as thin as can be reliable, with four edges
seal). The polyimide is then baked at high comprising the sealant but with one edge having a
temperature (e.g. 180°C) to harden it, and then ≈5mm wide hole to allow filling. After deposition,
rubbed using a rayon cloth mounted onto a roller. the glue solvent is removed in a soft bake and the
The resulting pre-tilt is not only dependent on the plate brought into contact with its opposing,
choice of the alignment polymer, but also on the spacer-coated plate. The laminate is then sealed
rubbing strength L: under pressure using a heated press. Usually, the
glue seal includes glass rod spacers to ensure the
$Υ] most uniform spacing to the edge of the panel.
𝐿 = 𝑁. 𝑙. 1 + , (19)
h;Ž
Often, the glue seal will also include a low density
of gold beads, each with a diameter 20% or so
where N is the number of passes of the roller, l is
higher than the spacers. These gold beads are
its contact length (mm), n is the roller rotation
distorted to near-cylinders under the pressure used

19
 to seal the device. In this fashion, they form elements that act to remove unwanted reflections,
multitudinous one-dimensional conductors and compensation films for improving viewing
randomly distributed throughout the sealant, angle.
linking the top and bottom substrates. These allow 9. Driver bonding. Connections are made from the
connection from one plate to the other, and exposed ITO on the bonding ledge to the driving
thereby allow a single driver to be used to supply circuitry by one of four methods:
the row and column signals. Careful patterning of a. Surface Mount Technology: Connectors are
the ITO electrodes ensures that no unwanted made directly from the bonding ledge to the
shorts occur, within the alignment tolerance of the Printed Circuit Board (PCB);
plates. b. Chip-on-Board: The silicon die is mounted
onto the PCB and connected to the panel
The completed laminate then leaves the high-end clean using gold wires, and protected by adhesive;
room for the front-end processes. Common front-end c. TAB mounted drivers. The driver
processes are: manufacturers often supply tape-automated-
bonded chips. The chip is mounted onto a
6. Singulation: Each display is cut from the laminate tape with gold I/O lines etched into it. The
using a diamond scribe. Each display is cut to give tape is adhered to the bonding ledge using a
at least one bonding ledge to allow electrical 1D conducting adhesive;
access to the electrodes. The panels are inspected d. Chip-on-Glass. COG uses gold-bump
for any non-uniform cell gaps, and tested for soldering to attach the silicon chip directly to
shorts. the bonding ledge. The chip must be
7. Filling and sealing: For small area LCDs, each protected from light and handling by
device is filled in a vacuum chamber, where the embedding in a black epoxy adhesive.
cell is initially evacuated, the liquid crystal 10. Inspection and test. Obviously, the degree of
brought into contact with the sealant’s filling hole testing is minimised to reduce costs, but
where it begins to fill by capillary action. Whilst manufacturers will operate a quality system to
still in contact the vacuum is removed and the air- determine and minimize the occurrence of optical
pressure used to force the liquid crystal into the and electrical defects.
remainder of the display. Once complete, each 11. Provision of Colour. Full-colour LCDs are made
display is sealed with a low-ion content UV glue in the same way as described above, but the glass
whilst pressure is applied to the cell to ensure that plates are pre-patterned with colour filters (for the
the correct cell gap is obtained on the sealed cell. front plate) and thin-film-transistors (TFT, for the
Large display panels are filled before singulation rear plate). Usually, RGB filters are used and
using a one-drop-fill method [60], which is aligned over sub-pixels in the column electrodes.
combined with sealant deposition and vacuum A black matrix surrounds each filter to maximise
assembly steps. One-drop filling is a major contrast, minimise colour leakage and to optically
enabling technology that helped LCDs meet the isolate the TFT (thereby preventing problems with
cost requirements for large area applications. photo-generated charge). The resulting structure
8. Polarisers: After cleaning, each cell has polarisers has a lower pixel fill factor, typically <50%. Given
and associated optical films mounted front and the losses associated with the low fill factor and
back. The films are supplied with an index- absorption of 67% of the light by the colour filters,
matching adhesive to reduce extra reflections from reflective full-colour LCDs are yet to be
the layers. If a reflective display is required, the successful commercially: the images are too dim
rear polariser also comprises a diffusive mirror. in all conditions but the brightest sunshine.
Front polarisers too may include lightly scattering Instead of using a reflective rear polariser, colour

20
 panels are combined with a backlight unit (BLU) depend on the orientation of the incident light with
to provide illumination. respect to the optic axis, and the phase difference
between the x and y components. The maximum
retardation due to the birefringence is Γ, given by:
3.2 LCD Polarisation Optics
$Œ[ ¡
Γ= . (22)
š
Although liquid crystals can be used to emit light [61],
all liquid crystal displays use the liquid crystal
Transformation of the Jones Vector requires the
medium to modulate light incident on the panel,
operators to be 2 x 2 matrices called a Jones Matrices
whether from ambient light or from a built-in source.
[62].
The anisotropic nature of the liquid crystal presents a
number of means through which contrast can be
Consider a simple uniform planar sample with parallel
produced, whether by scattering, absorption, selective
rubbing directions and a uniform cell spacing d.
reflection or changes of optical retardation and hence
Ignoring the reflections and absorption in this fashion
polarisation state. The great majority of LCDs are
is reasonable for thin (d < 25µm), well-aligned
sandwiched between film polarisers and use changes
samples, and it means that the polarisation state may
of optical retardation to give the required appearance.
be calculated using the Jones Matrix for a birefringent
The state of polarisation of light, as it travels through
retarder:
an optically transparent medium, can be linear,
elliptical or circular, depending on the relative ¢
magnitudes of two orthogonal polarisation 𝑒 ^{ g 0 . (23)
¢
components. As the light goes from one transparent {
0 𝑒 g
medium to another, light of each of the polarisation
components will be refracted, reflected, and the
polarisation reoriented. Finding the optical state Placing this retarder at an arbitrary angle φ between
requires solution of Maxwell’s equations at each crossed polarisers gives the Jones vector for light
interface for each polarisation. However, if we ignore transmitted by the device:
the reflections and refractions and just consider the
polarisation state, then we can consider the light to be ¦
𝐸£• 1 0 𝑐𝑜𝑠𝜙 −𝑠𝑖𝑛𝜙 𝑒 ^{ $ 0
given by a 2x1 Jones vector: =
𝐸£‘ 0 0 𝑠𝑖𝑛𝜙 𝑐𝑜𝑠𝜙 ¦
0 𝑒 {$
𝐸• 𝐴• 𝑒 ^{ ’K“”• 𝑐𝑜𝑠𝜙 𝑠𝑖𝑛𝜙 0 0 0
= , (20) .
𝐸‘ −𝑠𝑖𝑛𝜙 𝑐𝑜𝑠𝜙 0 1 1
𝐴‘ 𝑒 ^{ ’K“”–

¦
−2𝑖𝑠𝑖𝑛𝜙𝑐𝑜𝑠𝜙𝑠𝑖𝑛
where, A is the amplitude, x,y orthogonal components = $ . (24)
for light traversing in the z-direction and δ represents 0
the phase related to the wavelength λ through the
The transmission T is then given by:
refractive index by:
∗ $ $
$Œ]™ K 𝐸£ . 𝐸£ 𝐸£• + 𝐸£‘
𝛿˜ = ; 𝑗 = 𝑥, 𝑦 . (21) 𝑇= ∗ =
š 𝐸 .𝐸 𝐸• $ + 𝐸‘ $

If the light is incident on a birefringent layer which has ¦ Œ[ ¨

= 𝑠𝑖𝑛$ 2𝜙𝑠𝑖𝑛$ = 𝑠𝑖𝑛$ 2𝜙 𝑠𝑖𝑛$ (25)
refractive indices ne and no and thickness t, then the $ š
state of the polarisation of the transmitted light will

21
 ª‚ $
*
𝑘## cos $ 𝜃 + 𝑘QQ sin$ 𝜃 + 𝜀; ∆𝜀𝐸 $ sin$ 𝜃 = 𝐶
where E is the conjugate if E (which may be complex) ª®
(28)
and t = d is used for the spacing of the liquid crystal
cell.
where the constant C is found from realising that the
Equation (25) suggests that the transmitted
light is maximum if the cell gap and birefringence are maximum tilt θm must occur at the cell centre z = d/2
at the quarter wave plate condition Δn.d = λ/4, with because of symmetry, at which point dθ /dz = 0, and
the polarisers oriented at ±45° to the alignment hence:
direction. The eye is most sensitive to green
wavelengths, and setting λ = 550nm gives broad 𝐶 = 𝜀; ∆𝜀𝐸 $ sin$ 𝜃¯ . (29)
transmission across the wavelength range such that the
device will appear white; for a typical liquid crystal Substituting back into the Euler Lagrange equation
this condition occurs for cells spaced at about 2 – 3 (28) gives:
µm.
ª‚ ° ±² g ‚³ ^±² g ‚
= 𝜀; ∆𝜀 . (30)
ª® ª i´´ ^iµµ ±² g ‚“iµµ
3.3 Basic operation: Electrically Controlled
At low field strengths, the small angle limits to the tilt
Birefringence Mode
angles allow equation (30) to be re-expressed as:
Applying an electric field to the aligned liquid crystal
# iµµ i´´ ^iµµ ‚ g
will tend to align the liquid crystal director parallel to 𝑉. dz = 𝑑 . + 𝑑𝜃
YZ ∆Y ‚³ g ^‚ g ‚³ g ^‚ g
the field direction, if the material has a positive Δε, or
(31)
perpendicular to the field if negative. Considering the
pixel in one-dimension only and the director at the
which tends towards:
containing surfaces anchored strongly, the total free
energy of the liquid crystal F is given by combining µ
iµµ $ ^
equations (3) and (12) as: 𝑉. 𝑑𝑧 = 𝑑 . 𝜃¯ − 𝜃 $ g 𝑑𝜃 (32)
YZ ∆Y

# ª $ $
𝐹= 𝑘## ∇. 𝒏 + 𝑘$$ 𝒏. ∇×𝒏 + 𝑘QQ 𝒏×∇× as 𝜃 $ → 0. Integrating (24) to find the threshold field
$ ;
$ $
𝒏 − 𝜀; ∆𝜀 𝒏. 𝑬 𝑑𝑧 . (26) gives:

For the simple case shown in figure 6a), the electric ª

‚³ #
$ 𝑑 𝑘## ^
field is applied normal to the surfaces such that E = 𝑉º 𝑑𝑧 = . 𝜃¯ $ − 𝜃 $ $ 𝑑𝜃
; 2 𝜀; ∆𝜀 ;
(0,0,V/d) and initially n = (1,0,0). As the field is
increased, there is a torque on the director acting to
‚³
increase the tilt angle θ:
ª
$
𝑘## 𝜃
𝑉º 𝑧 ; = 𝑑. . tan^#
# ª ªg‚ 𝜀; ∆𝜀
𝐹= 𝑘## sin$ 𝜃 + 𝑘QQ cos $ 𝜃 − 𝜃¯ $ − 𝜃 $
$ ; ª® g
;
𝜀; ∆𝜀𝐸 $ sin$ 𝜃 𝑑𝑧 . (27)

𝑑 𝑘## 𝜃¯ 0
This integral may be solved using the Euler Lagrange 𝑉º =𝑑 . tan^# − tan^#
2 𝜀; ∆𝜀 0 𝜃¯
equation

22
Figure 6. Fréedericksz transitions in a) the planar homogeneous geometry for a positive Δε nematic, and b) the
homeotropic geometry with a negative Δε nematic. The dominant elastic constants at various field strengths is
indicated in red. c) Calculated tilt profile versus V/VC for a material with ε∥ = 20.25, ε5 =5.36, k11=11pN, k33=17pN in
a cell of gap d = 10µm, [63].

23
 iµµ
𝑉º = 𝜋 . (33)
YZ ∆Y ° ª Á ±² g ‚“#
. 𝑑𝑧 = 𝑑𝜃 , (36)
°À Œ ±² g ‚³ ^±² g ‚
Hence, there is a critical voltage that is independent of
cell gap at which the electric field induced distortion where κ = (𝑘QQ − 𝑘## )/𝑘## . Immediately above the
begins, a threshold that depends on the root of the ratio transition, a reasonable solution is found by integrating
of the relevant elastic constant and the dielectric (28):
anisotropy. By analogy, the threshold for a negative Δε
liquid crystal in the homeotropic geometry is: ° $ ‚³ #“Á ±² g ‚
= 𝑑𝜃 , (37)
°À Œ ; ±² g ‚³ ^±² g ‚

i´´
𝑉º = 𝜋 . (34)
YZ |∆Y| using a Fourier analysis to solve the complete elliptical
integral of the third kind:
Figure 6b) shows the situation where a positive Δε
material is used, but the field is applied in the plane of ° #
= 1+ 𝜅 + 1 sin$ 𝜃¯ + ⋯ (38)
°À Ã
the cell; for example using inter-digitated electrodes.
Similarly, a planar sample with the preferred
which, has the first order solution:
alignment direction normal to the applied field will
cause the director to twist at the electrode surface. In ® # ‚
such instances, the electric field is not uniform, though = sin^# −
ª Œ ‚³
it may be approximated at the electrode surface to be #“QÁ“⋯
V/l, where l is the separation between the electrodes of 𝜃 𝜃¯ $ − 𝜃 $ . µ (39)
#$Œ #“ Á“# ‚³ g “⋯
Æ
opposing voltage. In such cases, the cell gap does not
cancel in equation (33) and the threshold depends on where:
both cell gap and electrode spacing:
à °
J i´´ J igg 𝜃¯ = sin^# −1 (40)
Á“# °À
𝑉º ≈ 𝜋 ; 𝑉º ≈ 𝜋 (35)
ª YZ ∆Y ª YZ ∆Y

This indicates that, immediately above the threshold,

for the homeotropic and planar cases respectively. the tilt of the director increases linearly, with a
This latter geometry is utilised with in-plane switching gradient that is inversely related to the elastic ratio
devices, as will be described in section 4.6. If the field k33/k11. This is indicated in figure 6a), where the bend
is applied normal to a uniform director, the threshold elastic constant becomes increasingly pertinent at the
is well defined. In practice, the alignment layers centre of the device, and splay is increasingly pushed
induce a small pretilt, and the director profile of the to the alignment surfaces.
quiescent state is not uniform, neither is the electric Above the electric field induced Fréedericksz
field applied normal to the director. Thus, the transition, the situation is complicated by the effect of
threshold is rather second-order in nature, and begins the non-uniform dielectric properties of the distorted
to occur somewhat below the voltage predicted by director profile on the applied electric field. The
equations (33) to (35). electric torque is related to the electric displacement
Above the threshold, the solution to equation vector D and above the transition the director profile is
(30) must include the higher angle terms. Again no longer uniform such that E has a direction that is
considering the planar aligned case shown in figure dependent on the distortion. That is, the effect of the
6a), dividing equation through by VC gives:

24
field depends on the elastic energy, but conversely, the From equation (25), the transmission of polarised light
elastic energy depends on the field. Ignoring the (i.e. after the first polariser) becomes:
effects of charge, ∇.D = 0 and assuming D is a
function of z only, solutions to the free energy Œ[ £¨
𝑇 = 𝑠𝑖𝑛$ . (45)
š
expression require E to be calculated self-consistently
with the director profile:
With sufficient field, Δn’d → 0 and equation (45)
° predicts T → 0. Thus, above VC the cell changes from
𝐷® = Ï µ (41) white towards black. Figure 7a) shows this decrease in
Z ÈZ (ÉÈÊ~Ëg ÌÍÈÎ )
ª®
transmission for red, green and blue wavelengths,
calculated using equations (40) and (45) for a 0.67mm
and thus
cell filled with a liquid crystal material with ne = 1.7,
° no = 1.5 and k33/k11 = 2.0.
=
°À Figure 7b) shows the behaviour for a sample
$ [Y ‚³ #“Á ±² g ‚
with a higher initial retardation; the same material but
𝑠𝑖𝑛$ 𝜃¯ + 1 . ; ÉÈ 𝑑𝜃 with a 2.5µm cell gap. Each wavelength has
Œ YÎ Ð{]g ‚“# ±² g ‚³ ^±² g ‚
ÈÎ
transmission peaks corresponding to odd multiples of
(42)
the quarter wave plate condition (i.e. Δn’d/λ = ¾ and
¼); because these peaks occur at different voltages for
Solutions to this integral are done numerically. Indeed,
different wavelengths, the cell appears to change
display manufacturers use commercial packages, such
colour as the voltage is increased.
as “LCD Master” from Shintech, “TechWiz” from
There are a number of reasons why such a
Sanayi or “DIMOS” from Autronic Melchers [64], to
simple device is not used as a display effect,
calculate the director profile as a function of applied
illustrating the thinking that the display engineer needs
field for their devices.
to follow.
Consider the case of figure 6a) where the cell
• The optimum cell gap is very small, typically
retardation is set to the quarter-wave-plate condition at
less than 1µm, and always much lower than
550nm (the peak of the eye’s response, Δn.d=550nm/4
the cell gap required for a high production
= 137.5nm) and the alignment direction is set at φ =
yield (typically 4µm or higher).
45°. As the voltage is then applied, the overall
• The device cannot be passive-matrix
birefringence of the cell decreases as the optic axis of
multiplexed, since the saturation voltage is
the director at the centre rotates to towards the field
typically several times VC. This will be
and viewing direction, and the effective birefringence
explained in the following section.
is reduced to Δn’. Simplistically, we can consider the
• The optical configuration is very sensitive to
director to be uniformly at some average tilt angle 𝜃,
changes of cell gap. For example, a typical
the new refractive indices are given by:
LCD tolerance is ±0.15µm; such a change
# Ð{]g ‚ Ó@Ð g ‚
would cause a 12% reduction in the
𝑛′@ = 𝑛@ ; = + , (43) transmission and colouration for the ECB
]£Ò g ]_ g ]Ò g
device.
such that the effective birefringence Δn’ is: • The viewing properties of the device are very
poor in both high and low voltage states, as
]Ò ]_
Δ𝑛£ = 𝑛′? − 𝑛@ = − 𝑛@ . (44) illustrated in figure 8. In the 0V state, the
]Ò g Ð{]g ‚“]_ g Ó@Ð g ‚
effective birefringence decreases off-axis, the
quarter-wave plate condition is lost and the

25
Figure 7 Theoretical transmission characteristic for an electrically-controlled birefringence LCD. a) The
retardation at the quarter wave plate condition for green light (d = 0.67µm, Δn = 0.2) ; b) a thicker cell with
retardation at the full wave condition for λ = 500nm (d = 0.67mm, Δn = 0.2). Other fitting parameters used: ne =
1.7, no = 1.5 and k33/k11 =2. c) Experimentally determined transmission curve for a cell close to the full plate
condition. The results are taken for a white light source imaged through an eye-response filter.

26
 device looks dark instead of bright (in the For simplicity, the elastic anisotropy is discarded, and
direction orthogonal to this, there is little a single elastic constant k is assumed:
Ôg ‚ Ô‚
change and the device still looks white). In the 𝜀; ∆𝜀𝐸 $ sin 𝜃 cos 𝜃 + 𝑘 = 𝛾# (47)
Ô® g ÔK
high voltage state, the off axis transmission is
even darker than the full ON state when Applying a small angle approximation [65], the
viewed in the direction of the director tilt. characteristic ON and OFF times are:
When viewed in the other quadrant, the
retardation increases rather than decreasing ×µ ×µ ª g
𝜏ÖT = Ø g
= =
YZ ∆Y7 g ^ YZ ∆Y ° g ^°À g
as desired on switching causing contrast Ï
i

inversion, and the device begins to appear ×µ ª g

g ; (48)
Ù
coloured. Œg i ^#
ÙÀ

×µ ª g
𝜏ÖWW = .
The response time for even such a simple Œg i
device is complex, due to the effects of field induced These times are made short by using materials with
flow and the need for the anisotropic viscosities to be low viscoelastic ratios γ1/k, but most effectively by
considered. If flow is ignored, the response time keeping the cell gap d low. Examples of mixtures
depends solely on the twist viscosity γ1 that describes providing response times as low as 30ms at -20°C
rotation of the director. The Euler-Lagrange equation and10ms at +20°C in a 3µm cell have been produced
for the time dependent free energy of equation (28) [66]. Given the 3- 4µm lower limit on cell gap set by
then becomes: clean-room quality and device uniformity, it would
Ôg ‚ seem advantageous for high Δn materials to be used to
𝑘## 𝑐𝑜𝑠 $ 𝜃 + 𝑘QQ 𝑠𝑖𝑛$ 𝜃 + 𝑘QQ −
Ô® g enable the required optical effect with a lower cell gap.
Ô‚ $ Ô‚
𝑘## 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 + 𝜀; ∆𝜀𝐸 $ sin 𝜃 cos 𝜃 = 𝛾# In practice, however, γ1 tends to increase with Δn and
Ô® ÔK
(46)

Figure 8 Explanation for the poor viewing angle for the electrically controlled birefringence effect.

27
so the efficacy of this is limited. Ensuring that the minimise overlap with the common electrode. It is
voltage is made high can quicken the ON time essential for such a design that the common plate has
significantly. However, it is the sum of the ON and only a single point of overlap with ITO on the
OFF times that is relevant for display applications. electrode plate and conducting seal, to prevent
Moreover, if the display requires intermediate grey unwanted shorting of pixels. A typical transmission
levels or some degree of passive matrix addressing versus voltage characteristic for an LCD is shown in
(see section 3.4), the ON voltage is limited, and even figure 9b). With a direct drive scheme, the OFF
the ON time can be slow. For example, switching to voltage must be below the threshold T10 (usually 0V),
the grey level closest to VC is inherently slow. and the ON voltage should be above the saturation
In the case of a passive matrix addressed level T90.
device, even the OFF pixels have a voltage Vd applied A direct drive approach is satisfactory for
and the response time is given by [67]: devices with a small number of pixels. More complex
images require some degree of matrix addressing (or
×µ ª g multiplexing), where appropriate signals are applied to
𝜏ÖWW = ÙÏ g
; (49)
Œg i #^
ÙÀ
electrodes on both top and bottom plates. The
electrodes may also be shaped into alphanumeric
characters and designed for a low level of
In practice, the response is complicated by induced
flow of the liquid crystal, and the aligning effect of the multiplexing, or indeed into a more complex r(θ)
director in response to such flows. For example, the arrangement for the display of polar co-ordinates. The
OFF response is often slowed considerably by back- most common design, however, is for the electrodes to
flow that tends to initially reorient the director at the comprise N rows and M columns to form a rectangular
centre of the cell in the opposite direction to that M x N display. Time division multiplexing is used to
which gives the eventual lowest energy state, [68]. apply appropriate signals to the rows and columns,
Avoiding this “optical bounce” effect is considered when driven from only M + N connections.
further in section 6.2. Consider the simple 7 x 6 matrix display
illustrated in figure 10, where each pixel is addressed
3.4 Passive Matrix Addressing and the by sequentially scanning through the 6 rows, whilst
Multiplexing Limit the appropriate data is synchronously applied to the
For low information contents, it is satisfactory to form columns to discriminate between the ON and OFF
a display where each pixel is driven directly from the states. Figure 10 shows an example instant within the
driver. Usually, one of the electrodes is shaped into image frame, where the data voltages are being
the desired pattern, such as a pixel for a seven-segment applied to 7 columns and are those for the third row
or alphanumeric image, and the other electrode is a electrode. The data signal on the columns are ±D
common electrode. A display of N pixels then requires volts, where it is the sign of the signal that
N+1 electrodes. An example arrangement is shown in discriminates between the ON and OFF states of the
figure 9a), where a single seven-segment number is pixel on the third row. At this instant, +S volts is being
displayed using just eight electrodes. Rather than have
applied to the third row whilst all other rows are kept
separate connections to the two plates of the LCD,
at ground 0V. The data being applied at this instance
access to all of the electrodes is through a single
is (++--+-+)D. The potential difference at the pixels is
bonding ledge: the opposing common electrode is
defined as Row – Column, such that the third row
connected through via-electrodes that are connected
through the one-dimensional conducting seal. The experiences the voltage:
access electrodes to each segment is kept small, to

28
Figure 9 Design of a single element seven-segment LCD, showing front and rear plates face-up, and the
completed module. b) A typical Transmission Voltage characteristic for an LCD.

29
Row 3: +S-D ; +S-D ; +S+D ; +S+D ; +S-D ; +S+D ; the fact that each row is only addressed for a 1/N
+S-D , fraction of the frame time, and for the remainder of
the frame (N-1)/N each pixel experiences the data
and each of the other rows experiences: signal ±D.
Nematic displays respond to the root-mean-
square voltage applied over the whole frame. As soon
Row 1,2,4,5,6: -D ; -D ; +D ; +D ; -D ; as the signal has been removed, each pixel starts to
+D ; -D relax back to the off state, and so constant refreshing
at a frame rate sufficiently fast to prevent this
relaxation (or frame response) and maintain the image.
For a simple scheme, the frame time will equal the slot
Once the third row is addressed, the scan signal
time for each row τ multiplied by the number of rows
moves to the fourth row and the data signals change
N. It is independent of the number of columns, and
sign appropriately to supply the signal for that row.
hence it is common for the rows to be chosen as the
The aim of the multiplexing scheme is to ensure that
lower number in the matrix to ensure the faster frame
the high voltages |(S+D)| are sufficient to
rate.
discriminate from the low voltages |(S-D)|, despite

Figure 10 Example of passive matrix multiplexing for a 7x6 pixel graphic display. The row and column signals
are for the 1/6 proportion of the frame that corresponds to the 3rd row being addressed.

30
 The root mean square of the voltage applied The signal voltages S and D that are required for the N
over N rows for the ON and OFF signals are: x M display are found by substituting the optimum
multiplexing ratio 𝑆 = 𝐷 𝑁 into the RMS voltages of
(€“Ú)g “(T^#)Ú g equation (50):
𝑉ÖT = ; 𝑉ÖWW =
T

(€^Ú)g “(T^#)Ú g (Ú T^Ú)g “(T^#)Ú g $( T^#)

(50) 𝑉ÖWW = =𝐷
T T T
(55)
and D and S should be chosen to ensure that VON is
Ú g T“$Ú g T“Ú g “Ú g T^Ú g $( T“#)
above the Transmission saturation (i.e. > T90%) and 𝑉ÖT = =𝐷
T T
VOFF below the threshold (i.e. < T10%). The steepness
of the electro-optic response is related to R = which leads to the following data D and strobe S
VON/VOFF: voltages:

°Ü. $ (€“Ú)g “(T^#)Ú g Þ g “$Þ“T

𝑅$ = = = 𝐷 = 𝑉ÖWW
T
= 𝑉ÖT
T
(56)
°ÜÝÝ (€^Ú)g “(T^#)Ú g Þ g ^$Þ“T
$ T^# $ T“#
(51) and:
T T
where b = S/D is called the bias ratio. The maximum 𝑆 = 𝑉ÖWW (57)
$ T^#
number of lines that can be driven for a given S and D
is found by differentiating: Equations (56) and (57) are plotted in figure 11b); they
µ T
Ô
ßg ÍgßÍ. indicate that, at high N, 𝐷 ≃ 𝑉º and 𝑆 ≃ 𝑉º so
Ô •g ßg àgßÍ. Ã Þ g “T g $
= = =0 (52)
ÔÞ ÔÞ Þ g ^$Þ“T g that the maximum voltage that needs to be delivered
by the driver chip (S+D) increases with the level of
This has solutions when either b = ∞ or 𝑏 = 𝑆/𝐷 = multiplexing N, whilst the discriminating voltage D
𝑁. Substituting the latter into equation (52) gives the decreases. For this reason, STN displays with N=240
Alt-Pleshko [69] relationships: typically require drivers capable of delivering S+D =
20V.
°Ü. T“$ T“T T“# During the addressing frame, the liquid crystal
= = director for the ON pixels in the addressed row will
°ÜÝÝ ¯â• T^$ T“T T^#
(53) begin to decay as soon as the strobe voltage moves on
or: to the subsequent rows. Thus, it is important that the
g $ slot time of the addressing scheme is arranged to
ÙÜ.
“#
𝑁¯â• =
ÙÜÝÝ
(54) minimise any flicker whilst at the same time the pixels
g
ÙÜ.
ÙÜÝÝ
^# are ready to display new information in the following
frame; this occurs approximately where:
This relationship suggests that, as N increases, the
äÜ. “äÜÝÝ
°Ü. 𝜏≈ (58)
maximum must tend towards unity; that is, the $T
°ÜÝÝ
threshold characteristic must increase in steepness to
In practice, the scheme shown above is unsuitable for
allow more lines to be addressed, as shown in figure
long-term use, because there is no DC balance to the
11a). The typical TN characteristic shown in figure 9c)
waveform: this leads to eventual electrical breakdown
has a response suited to up to six levels of time-
of the liquid crystal and, if the image does not change,
division multiplexing.

31
Figure 11 a) The Alt-Pleshko Multiplexing Limit and b) addressing voltages.

can result in image sticking issues associated with active elements in displays had been recognised since
ionic conductivity. In practice, therefore, the their invention at RCA in the mid 1960s [57]. By 1971,
waveforms will be inverted periodically to maintain a the RCA team believed [70] that active components
net zero DC voltage. Given the data changes from were required to overcome the slow speed and poor
frame to frame, DC balancing can only be guaranteed electro-optic properties of the dynamic scattering
if the voltages are inverted twice per frame. That is, if mode device, proposing solutions using dual-diodes,
the slot time is halved, and the frame divided into a field effect transistors and storage capacitors to ensure
positive field (+S, ±𝐷) and a negative field with (-S, that the charge across the pixel remained constant
∓𝐷). The slot time should not be made too short, since across the frame. It was the team at Westinghouse [71]
the power dissipation of charging and discharging the that were the first to implement this in practice,
capacitances each field Pf is approximately related to: producing a TN 6x6 matrix driven by CdSe TFT. In
the UK, the team led by Hilsum at RSRE understood
Yéê D YD Ú g the handling, reliability and lifetime issues presented
𝑃ç ∝ . 𝑉 $. 𝑓 ≈ . (59)
ª ª K by CdSe and Te, and in 1976 approached Spear and
LeComber at the University of Dundee, who were
Given that power should be kept low, inversion is
working on the use of hydrogenated amorphous silicon
usually done only once per frame, often randomly
(α-Si:H) for use in photovoltaic cells. The Dundee
within the frame time to help ensure DC balance
team produced α-Si:H with electron mobility µe of
regardless of image pattern.
0.4cm2/Vs, and proved that the material was suited to
address LCDs [72]. The Dundee and RSRE teams
3.5 Thin Film Transistors and Active Matrix
fabricated insulated-gate field effect TFTs onto a glass
Addressing
substrate using photolithography, and fabricated the
world’s first α-Si active matrix LCD in 1980 [73].
Contrary to the approach taken by Sharp who targeted
the new market of portable electronic calculators, RCA This was followed in 1982 by a 240 x 240 α-Si TFT
concentrated on replacing the cathode-ray tube for TN from Canon in Japan [74], and the immense
televisual display with an LCD from the outset [2]. ensuing effort and investment in Asia to bring TFT
The potential application of thin-film transistors as

32
LCDs to the dominance of the display market that it and drain electrodes patterned on top of this. The
enjoys today. electrodes are usually made from Aluminium,
Very high levels of multiplexibility are Chromium or Molybdenum. With a negative-voltage
possible using thin-film transistors fabricated onto the applied between the gate electrode and the drain -VGD,
rear plate of the LCD behind each pixel (or sub pixel electrons are depleted from the semiconductor at the
in a colour display). A 4k UHD television, for interface with the insulator, preventing current
example, has 2160 x 3840 x 3 = 24.9M TFT per panel, between Source and Drain electrodes, figure 12a) ii).
and is 2160 ways multiplexed. Each TFT supplies When a positive-bias VGD is applied to the gate,
charge to a (sub-) pixel to switch the liquid crystal to however, electrons accumulate at this interface
any of 256 grey levels. The TFT is a field effect allowing current flow from source to drain ISD when
transistor (FET) formed from thin metal-insulator- the drain is positive, and from drain to source when
semiconductor layers, as shown in figure 12a). Most the drain is negative, figure 12a) iii). Figure 12b)
LCDs use a bottom-gate configuration, where the shows the transfer characteristic originally produced
amorphous silicon is deposited onto the pre-patterned by the Dundee / RSRE team in 1981.
gate electrodes, n+ doped with hydrogen, and source

Figure 12 a) Basic construction and operation of a TFT. b) Current-Voltage transfer characteristics for the
original Dundee / RSRE α-Si:H TFT [72], together with contemporary results for a-Si:H [74], low-temperature
polycrystalline silicon (LTPS) [75] and Indium-Gallium-Zinc Oxide (IGZO) TFT [76].

33
 A schematic of how a TFT is used in a typical time for the pixel to charge to the new level τ, the gate
LCD pixel is shown in figure 13. The LCD has an is reclosed (with -5VGD) and the gate pulse (+15VGD)
active matrix back-plane and a front plane with a scans to the next row and new data on the source
single common electrode. For backlit colour displays, columns. The frame is completed when all N rows
the front plate will also comprise the colour filters, have been scanned, after the frame time τN. It is
with a black matrix to ensure sufficiently high important to ensure that the liquid crystal material is
contrast, prevent colour leakage, and shield the TFT exposed to the minimum net DC over several frames.
from incident light and unwanted photo-induced To help ensure this, the polarity of the data signal is
charge generation. The TFT plate has row and column reversed, using frame-inversion, scan-line inversion,
metal bus lines that allow the TFT to be addressed, column inversion or pixel inversion (shown in figure
with the row electrode connected to the transistor gate 14a).
and the columns the transistor source. The ITO To understand some of the important design
electrodes that form the pixel are connected to the rules for a TFT LCD, consider the example of a 16.3”
drain electrode. The TFT is also protected using a QSXGA monitor. It has 2048 x 2560 pixels at 202dpi,
SiOx barrier layer, onto which the liquid crystal with the RGB sub-pixels on the columns (to maximise
alignment layer is deposited. The presence of bus- the time required to address each row). Such a panel
lines, shielded transistors and storage capacitor typically has an aperture ratio of 27% and operates at
reduces the active area of each pixel; aperture ratios as 60Hz (τ = 8.1µs). That is, the sub-pixel dimensions are
low as 30% to 40% are typical. This reduces the about 22µm x 65µm and a pixel capacitance CLC of
transmissivity of the panel, and necessitates brighter 0.16pF (assuming a cell gap of d = 4um and
backlighting to provide an attractive appearance. It permittivity εLC = 5).
also negates the use of the colour TFT LCDs in Firstly, the pixel should not fully discharge
reflective mode, because of the light losses associated within the 16.7ms frame time due to the off current of
with the polarisers, colour filters and the low aperture the transistor. Assuming an exponential decay, and a
ratio. 1% voltage tolerance gives the condition [78]:
Addressing a TFT matrix is done line-by-line
in a similar fashion to that described previously for °íî
𝜏ÖWW = 𝐶ìº . 𝑅ÖWW ≈ 𝐶 > 200𝜏𝑁 .
passive matrix displays, but now the row and column ïíî

signals are applied to the same substrate, as shown in (60)

figure 14a). From the α-Si:H TFT characteristic of
figure 12b) it is clear that applying +15VGD to the gate The α-Si characteristic from figure 12b) shows ISD is
electrode allows current ISD to flow between the source 20pA when the TFT is OFF and 10µA when ON.
and drain due to the signal voltage VSD applied to the Hence, equation (60) suggests that the TFT had a
column, whereas -5V on the gate turns the TFT off sufficiently low OFF current to drive a pixel of
and prevents current from flowing. With the scheme of capacitance 6pF, easily sufficient for the 40pF pixels
figure 14a), the transistors in each of the un-addressed used in the original work of reference [74], but is far
rows (rows n-1 and n+1) are turned OFF by the too leaky to drive the 0.16pF of a modern QSXGA
negative signal applied to the gate bus lines, whilst the monitor. Simply employing a storage capacitance CS
transistors in the addressed row (row n) are opened in parallel with the pixel, as shown in figures 13 and
using +15VGD to the gate electrodes and switched ON. 14, increases the capacitance to the desired level and
Synchronously, data voltages are applied to the source ensures that the pixel remains charged throughout the
bus; there is no current for any of the OFF rows, but frame. Similarly, ISD in the ON condition for the
the pixels on the ON row are charged to the highest signal level VSD should be sufficiently high to
appropriate voltage by the current ISD. After sufficient charge the pixel to at least 99% of that voltage level

34
Figure 13 Cross-section and Plan views of a typical TFT driven sub-pixel on a backlit colour LCD.

35
Figure 14. a) Active Matrix addressing of an LCD panel. b) Equivalent circuit for the pixel including the TFT
parasitic losses, line losses and the conductance of the liquid crystal.

36
 YZ Yöí~
within the addressing time τ, which occurs when τON is 𝐶ó€ = (65)
ªõ
10% of the addressing time τ:
and dG is the thickness of the SiNx insulator layer,
°íî
𝜏ÖT = 𝐶. 𝑅ÖT ≈ 𝐶€ + 𝐶ìº < 0.1𝜏 (61) figure 13. The α-Si leakage resistance dominates the
ïíî
off current:
Equations (60) and (61) can be combined to find the
ô °íî
maximum number of lines that can be addressed by a 𝐼€Ú = . . (66)
ì •îí
TFT, Nmax, with a given ration of ON to OFF currents:
Substituting equations (63) and (66) into (62) gives the
# ïíî (ÖT)
𝑁¯â• = . . (62) relationship:
$;;; ïíî (ÖWW)

The characteristic of reference [73] shown in figure 𝑁¯â• ∼ 𝜇? 𝐶ó€ 𝑅Ú€ . (67)
12b) suggests that the original TFT could address up
to 250 lines (as was claimed) but required That is, achieving the low OFF current is achieved by
improvement to address the 2048 rows of the monitor reducing the transistor width to length ratio W/L, but
display. Some of the improvements to TFT the high ON to OFF current ratio, and hence maximum
performance made to achieve such high levels of number of lines that can be addressed, is achieved by
multiplexing, 256 grey levels and high contrast ratios reducing the thickness of the insulating and
are outlined in the following discussion. semiconducting layers, leading to a typical CGS of
The equivalent circuit for a more realistic about 0.1µF/cm2. A low insulator thickness has the
pixel is shown in figure 14b). The TFT includes an additional benefit of reducing the area of the storage
inherent resistance for the semiconductor RDS and the capacitor and correspondingly increases the aperture
capacitance per unit area of the gate CGS. Also ratio. Figure 12b) also includes the characteristic of a
important is the conductivity of the liquid crystal itself TFT used for modern LCD panels, such as the
GLC and the resistance of the bus lines RBUS. The QSXGA monitor describe earlier (where W/L = 10µm
simplest model for field effect transistors predicts that / 6µm).
for sufficiently high gate voltages, well above the Decay of the voltage across the pixel is not
transistor threshold Vth and drain voltage VSD , the just related to the leakage current of the transistor and
current ISD is given by [78]: equation (60): loss of charge across pixel occurs if the
conductivity of the liquid crystal GLC is too high. This
ô # is quantified by the voltage holding ratio (VHR),
𝐼€Ú = 𝜇? 𝐶ó€ . 𝑉óÚ − 𝑉KV − 𝑉€Ú 𝑉€Ú
ì $
which represents the time it takes for the pixel voltage
(63)
to decay to 50% [79]:

where µe is the electron mobility, W is the width of the

gù.
TFT channel and L its length, as defined in figure 13. •Ü. º ^
úÜ. À
𝑉𝐻𝑅 = 1−𝑒 . (68)
The TFT threshold is typically about +3V, and is $äT

directly related to the charge density of free electrons

n 0: Ionic impurities in the liquid crystal must be
minimised to maximise VHR, and hence maintain a
ªõ
𝑉KV = −𝑒𝑛; (64) high display contrast. This cannot be done with
ºõí
nitrogenated compounds, preventing the use of highly
where the gate capacitance per unit area CGS: polar materials such as the cyano-biphenyls. Instead,
per-fluorinated compounds are essential. Although

37
mixtures produced from such compounds have a much Each of the four critical mask alignment steps needs
lower Δε, this is compensated by the low viscosity and very high resolution to prevent panel variability. These
hence fast switching speeds that can be achieved with steps are expensive, and various attempts to reduce the
these materials [79]. number of mask steps have been attempted, including
The pixel aperture ratio and hence the the use of back-to-back diodes. High tooling costs
transmission efficiency of the backlight is dictated by lowers the design flexibility, and so TFT panels tend
the target display resolution, the area of the panel (due to be available only in a range of standard sizes and
to the losses caused by RBUS) and the fabrication resolutions. If the requirement is for non-standard size
tolerances. A typical TFT is fabricated as follows [78]: in a niche market, the end-user may need to choose a
1. The gate metal, usually Cr or Mo, is sputtered passive matrix approach, which is why the market for
onto clean glass to a thickness of about passive matrix displays remains strong despite the
200nm. poorer performance. For a given display diagonal, the
2. The first set of electrodes are wet-etched: aperture ratio decreases linearly with increasing
photoresist is printed over the glass area and resolution. Various other pixel designs to that shown
exposed through a large area chrome mask in figure 13 have been suggested, often involving
placed with very high accuracy using a mask- alternative placement of the storage capacitor.
aligner. This is then developed to form the However, very high display resolutions require a
gate electrodes and bus line, and the bottom different approach, such as the use of low-temperature
electrode of the storage capacitor. poly Silicon (LTPS).
3. Plasma-enhanced Chemical Vapour Poly-crystalline silicon (p-Si) was amongst the
Deposition (PECVD) is used to deposit 400nm first semiconductors to be used for LCDs [81], and
of Si3N4, followed by130nm of intrinsic α-Si found in the first applications for TFT by Canon as the
and 50nm of n+ α-Si:H. watch used in the 1983 film Octopussy, and Sharp’s
4. The source, drain and storage capacitor 1991 hang-on-the-wall TV [7]. The material has a high
electrodes are then formed by sputtering Cr mobility of 200-400 cm2/Vs, which is intermediate
over the surface and wet etching the between the 1.5 cm2/Vs of amorphous silicon and
appropriate patterns, again using a wet etching 1400 cm2/Vs for crystalline. Such high mobilities
process and mask aligner. allow far smaller transistors, higher ON currents
5. The chrome electrode then acts as a self- (particularly important for Organic Light Emitting
aligned mask for plasma etching of the n+ α- Diode OLED displays), and potentially integrating the
Si:H to complete the source and drain display drivers onto the glass itself. This latter
electrodes (often using an etchant stopper). advantage potentially leads to significant overall cost
6. The remaining intrinsic α-Si:H is plasma savings, because the drivers would be produced in the
etched away using a third mask and mask same process steps as the pixel TFT. The problem with
aligner step, thereby forming the TFTs. producing p-Si TFTs were the very high processing
7. The ITO is sputtered and plasma etched to temperatures, requiring those early demonstrators to
form the pixels using a fourth mask and be produced on quartz substrates. In the mid-1980s
alignment step. [81], low-temperature poly silicon (LTPS) TFTs were
8. The last step uses PECVD to deposit the fabricated using excimer laser annealing of the α-Si to
350nm of SiNx or SiOx barrier layer, and uses form the polycrystalline structure whilst keeping the
a fifth mask to provide access to the bonding processing temperature to 260°C, equivalent to that
pads. A mask aligner is not usually required used for α-.Si. Today, many smart-phones benefit
for this step that is far less critical than the from the excellent properties of LTPS, which allows
previous photolithographic steps. resolutions above 400dpi and better battery life due to
the reduction in backlight power that the high aperture

38
ratio allows. However, the cost of LTPS is high, increase the complexity of the display to allow
because the fabrication of the top-gate transistors hundreds of lines of information to operate at video-
required uses 9 to 11 critical mask steps: this typically frame rate, with grey-scale and colour. By the mid
adds about 20% cost to the panels over equivalent α-Si 1990s, the technological steps to achieve this had
LCDs. largely been satisfied through the adoption of the TFT
Together with other disadvantages such as active backplane. At that point, the race moved from
high leakage current, the high production cost of LTPS complexity to appearance and in particular improving
has a driven research into other semiconductors, poor viewing angles. Several new modes were
including various metal oxides. ZnO is particularly developed, each finding market success for different
interesting, since it retains a high mobility, and manufacturers, first allowing LCDs to replace cathode
combines a very high ON to OFF state current ratio ray tubes (CRT) in computer montiors and in 2007,
with optical transparency over the visible region. surpass CRT sales for TV. However, the war shifted to
Recently, the 1:1:1 combination of Indium Gallium new battle grounds and yet further improvement of
and Zinc Oxide (IGZO) [82, 77] to form active appearance, including ultra-high resolution and wide-
element has been put into production by Sharp. A colour gamut drive the competition in the mobile
50nm amorphous IGZO layer is nitrogenated to form telecommunications marketplace. Before describing
the n+ doped semiconductor with a mobility of µe = the individual modes, it is apposite to discuss the
10cm2/Vs. Figure 12b) includes a comparison of the technological and market forces that dictated each
TFT transfer characteristics for α-IGZO against both shift in stage.
LTPS and α-Si. Although IGZO has a lower mobility Following the discovery of the Alt-Pleshko
than LTPS, it combines a very high ON/OFF current multiplexing limit for passive matrix addressing [69]
ratio (>109) with the same low-cost fabrication designs maximising display content was the key driver for the
of α-Si. It is likely that IGZO will play a leading roll LCD industry throughout the 1980s and early 1990s.
in future display devices, as displays continue to move Several approaches were taken:
to ever higher pixel contents, resolutions and power 1. Increasing the gradient of the LC
efficiencies. transmission-voltage response, either through
material improvement (controlling the elastic
4 Standard LCD Modes constant ratio k33/k11) or new LCD modes
4.1 Overview (supertwist nematic, STN).
There is a wealth of different liquid crystals modes 2. Designing bistable LCD modes, where pixels
possible, some of the important ones being shown in no longer responds to the RMS signal over the
figure 15. Some modes have niche applications, such frame, but are rapidly latched between the ON
as the cholesteric temperature sensor or the zenithal and OFF states and remains in the desired
bistable display (ZBD) and these will be described in state after the signals applied.
the section 5. Modes that have contributed to the 3. Introducing a non-linear element, such as a
mainstream LCD markets, from the early watches and TFT, that isolates and delivers the appropriate
calculators to today’s full colour video rate displays signal to each pixel.
are covered in this section. The first of these approaches was successful for low-
The evolution of the modern LCDs used in cost, black and white STN displays, but was limited by
mobile phones, computer monitors and HDTV has the slow frame time, low number of grey-levels and a
been undertaken in several stages, exemplified by maximum of 480 multiplexed lines. Many novel
listing the major advances, as in table 4. After the methods for making an LCD inherently bistable were
basic principles had been evaluated, as described in invented at the end of the 1970s and beginning of the
the previous sections, the most important step was to 1980s [101]. The most promising bistable technology
was the surface stabilised ferroelectric LCD, invented

39
Table 4 Significant advances on route to LCD market dominance.

Year Invention Protagonists Ref. Stage

1967 Dynamic Scattering Mode Heilmeier et al., [3, 4]

RCA; US
1971 Active Matrix Addressing Lechner et al., RCA; US [70]
1971 Twisted Nematic Mode (TN) Schadt and Helfrich, BBC; CH. [14]
Fergason, U. Kent; US
[83]
1971 Vertically Aligned Nematic Mode Kahn, Bell Labs; US [84]
(VAN) Schiekel and Fahrenschon AEG- [85]
Telejunken; W. Germany
1973 Cyano-biphenyl Nematics Gray et al. [15] Basics
Hull, RSRE, BDH; UK
1973 Formulation of Wide temperature Raynes [16]
range eutectic LCD for TN RSRE, UK
1973 In-plane switching proposed Kobayashi, U. Tokyo; JP [86]
Soref, Sperry; US [87]
1973 First commercial LCD products in Wada et al., Sharp; JP [7]
calculator and wristwatch Seiko; JP
1974 Defect-free TN Raynes and Waters [58]
RSRE; UK
1979 First amorphous silicon TFT used Spear and Le Coomber, [72]
to address LCD U. Dundee; UK [73]
Hilsum, Hughes, RSRE; UK
1980 Bistable Surface Stabilised Clark and Lagerwall, [88]
Increased
Ferroelectric (SSFLC) U. Göteborg, SE
Complexity
1982 Supertwist Mode (STN) Raynes RSRE; UK [89]
1984 Supertwist Birefringence Effect Scheffer and Nehring [90]
(SBE) BBC, CH
1987 Foil compensation for STN Katoh, Asahi Glass; JP [91]
1992 In Plane Switching Mode (IPS) Baur et al., Merck; DE [92]
Kondo et al., Hitachi; JP [93]
1997 Multi-domain Vertically Aligned Koike and Okamoto, [94,
Mode (MVA) Fujitsu; JP 95]
1998 Fringe Field Switching Mode Lee et al., [96, Wide
(FFS) Hyundai; KR 97] Viewing
Angle
2000 Patterned Vertical Alignment Kim et al., [98]
Mode Samsung; KR
2001 Axially Symmetric Mode (ASM) Yamada, Ishii et al., [99]
Sharp; JP
2004 IGZO TFT invented Nomura et al., Tokyo Institute of [82]
Improved
Technology, JP
resolution
2010 QD enhanced colour backlights Jang, et al., [100]
and colour
Samsung, KR

40
in 1980 [88], launching an immense worldwide effort SSFLC development effort targeting mainstream
for the next fifteen years in a two-way battle between displays was minimal.
the FLC and TFT approaches. The complexity of TFT Buoyed by early commercial success from
fabrication seemed to promise higher costs over the laptops, the TFT TN would attract sufficient
passive matrix bistable approach. However, despite the investment to explore new modes, targeting shifting
launch of a colour SSFLC monitor by Canon in 1992 the incumbent CRT from the monitor, and eventually
(figure 16d), the TFT-driven TN display was already TV, markets. The use of TFTs brought with it a new
beginning to enable the image quality required for a and very important advantage: the flexibility of LC
new market: the laptop. Early laptop screens were mode design. Modes that had previously been
black and white only (such as Apples’s MacIntosh impossible to multiplex became potentially useable. In
Powerbook 100, released in October 1991) but the the late 1990s, modes such as In-Plane Switching
sensitivity of SSFLC to shock made the technology (IPS) [92, 93] and Multi-domain Vertically Aligned
unsuited to portable products. Whilst there was no (MVA) nematic [94, 95] offered sufficiently wide
superior alternative to the TFT-TN LCD for laptop viewing angles to displace the CRT. By the beginning
displays, the competition for the SSFLC monitor came of the new millennium, the LCD was dominant from
from the superior performance and much lower cost of watches and calculators, to large area monitors [102],
the CRT, and well before the end of the 1990s, the as shown in figure 16.

Figure 15 Dendrogram of the important LC modes. The LCD mainstream described in section 4 is indicated by
the blue colouration. Abbreviations used are: PDLC: Polymer dispersed liquid crystal; ECB: Electrically controlled
birefringence; VAN: Vertically Aligned Nematic; IPS: In-plane switching; MVA: multi-domain vertically aligned;
PVA Photo-aligned vertical alignment; ASM: Axially symmetric mode; FFS: Fringe field switching; STN: Supertwist
nematic; SSFLC: surface stabilised ferroelectric liquid crystal; SBE: Supertwisted birefringence Effect; F-STN: film /
foil compensated STN; ABD: Azimuthal bistable display; ZBD: Zenithal Bistable Display; OCB: Optically
compensated bend-mode.

41
Figure 16: State of the Art LCD at the turn of the Millennium. a) Watches and calculators using the reflective
TN; b) A dual display phone and PDA using foil compensated STN; c) a laptop computer with TFT driven TN; d), a
14” monitor display based on bistable SSFLCD; and e), one of the first IPS mode TFT monitors. [100].

The battle for the television market was currents than was possible, thereby meeting the
complicated by another new technology: the plasma requirement for another emissive display technology:
display panel (PDP). As an emissive display, the Organic Light Emitting Diodes (OLED). Not only do
viewing angle and contrast of PDP could not be OLED surpass LCD for contrast and viewing angle,
surpassed by LCDs. This gave PDP an early lead in but also for colour saturation. Again, the adaptability
the flat-panel television market. However, on-going of LCD technology provides a solution: replacing the
LCD improvements, such as the invention of the cold-cathode backlights with a blue LED, and adding a
fringe-field-switching (FFS) mode [96, 97], and the film containing red and green quantum dots (QD),
inability to fabricate high resolution PDP below 40”, provides sharper colours than is possible using colour
led to the eventual triumph of LCDs in this most filters alone [100]. Once more, LCD performance
important of markets too. increases to meet the market challenge. Mid-way
Coincident with LCD domination of the TV through the 2010s, OLED is making grounds for
display market, the first i-Phone was launched in June portable displays, where features such as form factor
2007, heralding the age of the smart-phone. The goal and weight also play an important role, but LCDs
of providing resolution at the limit of visual perception continues to dominate in all other markets, figure 17.
helped drive LTPS into the marketplace. However, Will OLED eventually replace the LCD altogether? At
LTPS also had the advantage of providing higher ON each stage of LCD evolution, new markets have been

42
enabled by particular LCD modes. Many of these the supertwist nematic for instrumentation, or TFT
modes have retained their market share despite more driven TN for low-cost monitors, the successful
advanced options becoming available. Whether it is technologies become difficult to supplant once
the simple twisted nematic in watches and calculators, established.

Figure 17: History of the mainstream of LCD, from the calculator to QD 4k FHD TV.

4.2 Dynamic Scattering mode nematic with a DC or low frequency AC field. This
The first operating liquid crystal device was an turbulence was induced by ionic flow in the liquid
electro-optic shutter, devised by Heilmeier’s crystal disrupting the liquid crystal, thereby causing
predecessor at RCA, Richard Williams, in 1962 [103, strong optical scattering in the birefringent medium.
104]. Williams showed that strong turbulence could be Heilmeier used this switching between scattering and
induced in a roughly planar sample of a negative Δε non-scattering states in RCA’s first display
demonstrators in 1968. Unlike most of the other

43
modes described, Dynamic Scattering Mode devices causes sufficient reorientation of the director in the
did not need polarisers and operating in a 20 -100µm bulk of the cell to cause the NW cell to appear dark,
spaced cell at typically 30 – 50V. The onset voltage and the NB cell bright. TN may be used in
for scattering is approximately [105]: transmission, usually driven by TFT and including
colour filters, or may be used as a two-polariser
üj reflective mode. In this latter device, the rear polariser
𝑉º ~𝑇 , (69)
Y
incorporates a diffusive reflector so that light incident
from the front of the panel is reflected back through
where η is the bulk flow viscosity, µ is the ion
the device for a second pass. This leads to a much
mobility, and T is a constant typically about 100. For
higher contrast ratio, since any light leakage in the
η = 0.5 Pa.s and µ = 10-4 cm2/Vs then VC is about 5 -
dark state from the first pass is dramatically reduced
10V. The resistivity of the liquid crystal could be
on the second. However, it also means that the
made lower than 10 GΩ.m through deliberate addition
colouration of the white state is more important. Also
of mobile ionic impurities. The original displays of
possible are single-polariser reflective modes, as
Heilmeier used DC fields to induce the ion flow,
described at the end of this section.
presumably because of the desire to address the
A key part of any display design is to ensure
devices using transistors. However, the lifetime was
uniformity of texture for the ON and OFF states. For a
too short due to electrostatic breakdown effects, and so
90° TN this means ensuring that degeneracy of both
the first commercial devices used low frequency AC
twist and tilt are removed [58]. For twist alone this is
instead.
simply done using slightly uncrossed alignment
directions or, preferably, by inducing a natural sign of
4.3 The Twisted Nematic LCDs
twist with the addition of a small amount of
cholesteric to give a pitch of a few hundred micron.
4.3.1 Introduction
Removing tilt degeneracy requires that there is pre-tilt
By 1980, practically all commercial LCDs were based
on both alignment surfaces. However, it remains
on the twisted nematic. This device set the benchmark
essential that the pre-tilt and sign of twist are matched,
for future developments, operating using alignment
so that there is minimal splay from one surface to the
layers to obtain a uniformly aligned director profile, a
other when undergoing the correct twist. Otherwise,
cell gap of a few microns sandwiched between
reverse tilt and twist domains may still form as the
polarisers either side. The electro-optic effect is the
director relaxes to the quiescent state after switching.
basic Fréedericksz transition described in section 3.3,
wherein the director responds to the RMS voltage
4.3.2 The OFF State and TN Design
coupling to the dielectric anisotropy affecting a change
The polarisation optics from a uniformly
to the birefringence profile through the device, and
aligned birefringent liquid crystal cell was calculated
hence altering the polarisation state of the transmitted
in section 3.2. It was shown that the polarisation state
light. Each of the remaining devices described in this
after transmission through multiple elements is
section uses these principles (although the FLC is a
calculated simply by multiplying the appropriate Jones
field effect device, with a polar coupling of the field to
matrices together. The twisted nematic has a director
the ferroelectric spontaneous polarisation). The
twist angle that varies linearly from one surface to the
twisted nematic uses a positive Δε material, and
other. Thus, it can be described as a series of very thin
usually a 90° twist between top and bottom surfaces,
birefringent retarders, each with a slightly different
with either crossed or parallel polarisers, to operate in
orientation angle. Slicing the device into N layers, the
normally white (NW) or normally black (NB) modes,
retardation and orientation of each is given by:
respectively. Applying a voltage three or four times
greater than the threshold (i.e. applying typically 5V)

44
 ¦ $Œ[ ¨ ý overall twist Φ =0, and crossed polarisers are used (φ =
= ; 𝜙˜“# = 𝜙˜ + , 𝑗 = 0,1,2, … 𝑁 − 2
T šT T
(70) φ1 and φ2 = φ1 +90°).
For the Normally White mode twisted nematic
where Φ is the total twist angle. The overall Jones TN with Φ =90°, the polarisers are crossed (φ2 = φ1
matrix M is the multiplicative sum of each of these +90°) and oriented with either φ1 = 0° or 90°. The
elements Mj: transmittance from equation (73) then simplifies to the
Gooch – Tarry expression [107]:
T

𝑀= 𝑅 𝜙˜ . 𝑀˜ . 𝑅 𝜙ÿ Ø gÉ&.' g
Ð{]g #“
g (
˜!# 𝑇 =1− gÉ&.' g
, (74)
T #“
(
≡ 𝑅 −𝜙˜ . 𝑀˜ . 𝑅 𝜙˜
˜!# and the value of Δn.d is chosen to ensure that the right-
hand term is zero for λ ≈ 550nm, so that the TN
¢ T
appears white. If the polarisers are kept parallel
= 𝑅 −Φ 𝑒 ^{g. 0 𝑅
ý
=
{
¢ T instead, then the transmittance becomes:
0 𝑒 g.
¢ ¢ T
ý ^{ ý
𝑐𝑜𝑠 𝑒 g. 𝑠𝑖𝑛 𝑒 ^{g. Ð{]g
Ø
#“
gÉ&.' g
T T
𝑅 −Φ (71) g (
ý ¢
ý ¢ 𝑇= , (75)
−𝑠𝑖𝑛 𝑒 {g. 𝑐𝑜𝑠 𝑒 {g. #“
gÉ&.' g
T T (

As N tends to infinity, then each retarder becomes and the TN is Normally Black at the same retardation
infinitesimally thin, and [106]: conditions. Equations (74) and (75) are plotted versus
𝑀= retardation in figure 18. There is minimum with NB,
¦ Ð{]% Ð{]% maximum with NW, where the twisted structure of the
𝑐𝑜𝑠Φ −𝑠𝑖𝑛Φ 𝑐𝑜𝑠𝑋 − 𝑖 Φ
$ % %
Ð{]% ¦ Ð{]% birefringent material transmits linearly polarised light
𝑠𝑖𝑛Φ 𝑐𝑜𝑠Φ −Φ 𝑐𝑜𝑠𝑋 + 𝑖
% $ % that is orthogonal to the input polarisation and
(72) perpendicular (for NB, parallel if NW) to the output
where: polariser. This occurs at a series of conditions, given
$ by:
Γ
𝑋= Φ$ +
2
∆].ª #
Solving for the transmission: = 𝑚 $ − , 𝑚 = 1, 2, 3 … , (76)
š Ã

ý
𝑇 = 𝑐𝑜𝑠𝑋𝑐𝑜𝑠 Φ + 𝜙# − 𝜙$ + 𝑠𝑖𝑛𝑋𝑠𝑖𝑛 Φ + 𝜙# − including the first minimum at Δ𝑛. 𝑑 = µg Q 𝜆 and
%
µ
$ ý $ second at Δ𝑛. 𝑑 = #+ 𝜆. At any one of these minima
𝜙$ + 1− 𝑠𝑖𝑛$ 𝑋𝑐𝑜𝑠 $ Φ − 𝜙$ − 𝜙$ g
% (or maxima for the NW, although the term minima is
(73) used conventionally regardless of the polariser
orientations), the transmission is least sensitive to
where the input and output polariser directions with changes in retardation (including viewing angle and
respect to the input director are φ1 and φ2, cell non-uniformity) and polariser orientations. As the
respectively. As expected, equation (73) reduces to the retardation is increased (i.e., for high cell gaps)
transmission of a retardation plate, equation (25) if the successive minima are increasingly less sensitive to
retardation changes until the Mauguin condition is

45
Figure 18 The Gooch-Tarry curve for the quiescent transmission of a 90° Twisted Nematic Display versus
retardation, for normally black (NB) mode (parallel polarisers) and normally white (NW) mode (crossed polarisers).

approached at the high retardation limit [45], where because this cell spacing is conducive to a high
polarised light transmitted by the cell is rotated manufacturing yield and the switching speed is
through 90° regardless of cell gap, wavelength, satisfactory. If switching speed is the priority, rather
temperature or polarisation angle. than use higher birefringence materials with the
The choice of retardation is dictated by concomitant increases of viscosity, an alternative
fabrication limitations on cell gap, the operating speed approach is to use a 75° twist, [67]. The first minimum
that is required, the colouration of the white state and peaks at a lower transmissivity than for a 90° cell
the required contrast. Response times depend on the (98.4%), but this can be corrected by reorienting the
square of the cell gap, through equation (48). This polarisers through -7.5°, as shown for the results in
usually limits practical operation to either the first or figure 19. Typically, the cell gap for the 75° TN mode
second minimum condition. The wavelength is 80% that of the standard 90° TN, potentially leading
dependences for 1st min. and 2nd min. devices are to a 50% speed increase, provided the reduced cell gap
shown in figure 19. Again, choosing the wavelength remains suitable for manufacture.
to occur at the peak eye response of λ=550nm, the first For each of the modes, it is important to
and second Gooch – Tarry minima for a TN filled with consider the colouration of the white state. For a
5CB are 2.6µm and 5.9µm, respectively. Most TFT backlit colour TN, this is less important than for a
LCDs operate at the first minimum, using lower black and white reflective device, because any loss of
birefringence materials, Δn ≈ 0.1 and d ≈ 4.5µm, white colour balance can be compensated in the

46
Figure 19 Choices of white state and cell gap for the TN. a) The Gooch Tarry curve for the 90° and 75° twist
angle TN. b) Theoretical wavelength dependences of the transmission for a material with a typical birefringence (Δn =
0.15) for 90° TN cells operating at the first minimum, first maximum and second minimum and for a 75° TN with
uncrossed polarisers at 85°. The cell gaps are also indicated.

addressing signals, back light spectrum and colour Figure 19b) shows the theoretical wavelength
filters. For a dual-polariser reflective display, the dependences in transmission for four TN modes.
second pass of the light through the cell exaggerates Converting these spectra to CIE 1931 x and y colour
any unwanted colouration. co-ordinates, as shown in figure 20, gives a locus of

47
white hues depending on the retardation and the slightly blue tinged polariser. The optimum cell gap
colouration of the polarisers [108]. Two loci are for this is:
plotted, corresponding to a standard polariser centred
on green and a slightly bluer variety. These polarisers #
𝑑≈ 0.45 15 + 0.68 3 , (77)
Ã∆]
represent the range of colours that can be targeted
whilst retaining a black OFF state; narrower spectrum
where the blue light is close to the second minimum
polarisers can also be used to modify the ON state
and red wavelengths close to the first minimum, and
colouration, but with a noticeable colouration of the
the green transmittance is decreased somewhat, figure
dark state too. Also shown on the chart are the co-
19b).
ordinates for the D65 standard, representing a target
The other important consideration for the OFF
for the ideal white state. Operating at the first
state is viewing angle. Figure 21a) shows a polar plot
minimum tends to give a greenish hue to the white,
of transmissivity for the azimuthal and zenithal
state whereas, the second minimum gives an improved
directions for a normally white 1st minimum TN
white, due to the increased blue transmittance shown
operating in transmission. The OFF-state viewing
in figure 19b). Thus, a second minimum TN is
angle is good: light at any azimuthal angle experiences
preferred for applications where the response time is
the same retardation profile due to the twisted
unimportant (being five times slower than the
structure, and deviations of zenithal angle away from
equivalent first minimum mode) but the attractive
the display normal (shown as the central point)
bluish white is preferred. Alternatively, an
represent reduction of the effective retardation, to
intermediate retardation close to the point of the first
which devices operating at the Gooch-Tarry minima
maximum (i.e. at Δn.d/λ = 2/7√22) has been used [108] are insensitive.
to produce a neutral white as close to the D65 standard
as possible, particularly when combined with the

Figure 20 Central region of the CIE colour chart (shown in full on the right) indicating the region of white
transmission for a dual-polariser reflective mode TN. Two loci are shown, corresponding to polarisers with a slightly
green tint (continuous line) and a slightly blue tint (dashed). The D65 white state reference is shown as a target for
pure white. The location on the colour chart of the 1st min, 2nd min and 1st maximum are indicated.

48
4.3.3 The ON State and Optical Compensation dependence of the normal incidence transmittance can
be estimated by considering the mid-plane tilt of the
The ON state transmission is very viewing director θm [109, 110]. Immediately, above VC this is:
angle dependent. This is evident from the contrast
ratio of OFF / ON transmission shown in figure 21b) 0 g u´´ g
° i´´ ^ “igg ^i´´ [Y
$ Ø ugg
since contrast is dominated by the dark ON state 𝜃¯ = 4 −1 g +
°À iµµ “
0
i´´ ^$igg YÎ
transmission. The blackest ON state occurs when the Ø

light is transmitted parallel to the liquid crystal optic (80)

axis, the direction where the liquid crystal behaves as
an isotropic medium and any light transmitted is due which becomes, for the 90° TN cell:
to leakage of the polarisers. Contrast is highest for on-
axis light, since this is the direction of the applied field 𝜃¯ $ =
u g
towards which the director reorients, although it ° +i´´ ^ ´´ “igg
ugg [Y
remains affected by the distorted regions close to the 4 −1 +
°À Ãiµµ “ i´´ ^$igg YÎ
surface. Off-axis light experiences a retardation with a ° +i´´ [Y
direction dependence that is related to the director ≈4 −1 +
°À 1iµµ YÎ
profile through the cell, which in turn depends on the (81)
applied voltage. for the reasonable approximation k33 ≈ 2k22. Given the
The Fréedericksz threshold voltage for a TN is surface pre-tilt is small, this expression represents the
given by: linear part of the transmission - voltage characteristic
Œg iµµ “ýg i´´ ^$igg above VC shown in figure 9b). The multiplexibility of
𝑉º =
YZ ∆Y the TN is related to the steepness of this characteristic
(78) through the Alt-Pleshko expression, equation (54).
which simplifies to: Thus, to ensure the highest number of lines that a
passive-matrix TN can be multiplexed, the
iµµ “ i´´ ^$igg à denominator of equation (81) should be kept low. This
𝑉º = 𝜋
Y_ ∆Y is done in practice using hybrid mixtures of polar and
(79) non-polar compounds [111], in which k33/k11 is
for a standard Φ= 90° TN. Often, k22 ≈ k33/2 in lowered due to short-range order effects, offered a
practice, and so the Fréedericksz threshold is similar to route to multiplexing of up to 20 lines.
that of the ECB mode, and typically around 1V. Off-axis calculations of the transmissivity for
Figure 21c) shows schematically what happens as the the ON state certainly require numerical modelling.
voltage is increased above VC, and numerical However, the form of the contrast ratio curve shown in
calculations of the response for typical elastic figure 21b) is anticipated by considering the director
constants are given in figure 21d). As the voltage profile of the high voltage state as a vertically aligned
increases, the initially linear twist from one surface to nematic in the bulk, with two orthogonal hybrid
the other becomes increasingly concentrated into the aligned states in the surface regions. The director in
centre of the cell where the director approaches the these two regions are oriented parallel to the
vertical condition, whereas regions of splay and bend polarisers, at azimuthal directions 45°and 135° in
are increasingly pushed towards the surfaces. Thus, figure 21b), and hence appears dark at all angles. This
the director profile approaches a vertically aligned understanding led to pronounced improvements of
central region, with twist-free splay-bend regions transmissive mode TN viewing angles through optical
oriented at 90° to each other close to the two surfaces. compensation. Discotic liquid crystals are similar to
The optical transmission through the cell is complex the standard rod-like calamitic mesogens used in
and requires numerical modelling. The voltage displays but exhibit negative birefringence.

49
Figure 21 Measured viewing angle characteristics of a first minimum NW TN operating in transmission,
showing polar plots of a) ON state transmissivity and b) ON / OFF contrast ratio. c) Schematic of the TN with low,
intermediate and high voltages. d) Twist and tilt director profiles for the TN versus voltage.

50
 Combining discotic and calamitic layers with 4.3.4 Single Polariser Reflective Mode TN
equal but opposite retardations leads to a net optically
isotropic medium, black at all angles when between Conventional two-polariser reflective TN and
crossed polarisers. Figure 22 shows how this works in STN LCDs have two further optical limitations.
practice [112, 113]. Polymer film fabricated from a Parallax caused by the separation of the rear reflecting
hybrid aligned discotic are laminated either side of the polariser from the image plane by the thickness of the
LCD with the optic axes parallel to the fully ON state rear plate, leads to shadowing of the image when
director at the adjacent surface, and the splay anti- viewed off axis. This can be distracting for black and
symmetric. This leads to a greatly improved off-axis white devices, but is severely detrimental to reflective
contrast ratio, as shown in figure 22b), and helped lead colour devices due to colour leakage between sub-
to sufficiently good performance for TN displays to be pixels. Secondly, the optical efficiency of the white
used in laptops and monitors. state is relatively low because the light passes four
Two-polariser reflective mode TN devices are times through the polarisers. The transmission of the
usually used in low cost applications such as watches highest quality LCD polariser is 43%. However, a
and calculators where the cost of extra optical further 5% is absorbed on each pass so that the
compensation layers is prohibitive. However, the maximum reflectance possible with a two-polariser
viewing angle characteristic of such devices is far reflective display is less than 28%, typically 23%. This
more symmetrical than that of the transmissive device is also a severe limitation for pixelated reflective
shown in figure 21b), due to a self-compensating colour LCDs, where the colour filters and decreased
effect. Off axis light that experiences a lower Δn from aperture ratio then lead to prohibitively low 7%
the director tilted towards it on the first pass through reflectivity.
the device, experiences a correspondingly higher Δn Single polariser reflective TN reduces these
when traversing in the other direction on the second optical losses by using a front polariser only and using
pass after reflection. an internal reflector on the inside surface of the rear
With all reflective displays, it is important to substrate of the display to remove the parallax. The
consider the illumination conditions; in particular the reflector is made slightly diffusive to scatter the
colour balance of the incident light and its degree of reflected light in different directions, so that the OFF
diffusivity. Indeed, the viewing angle characteristic state has an attractive appearance and good viewing
can be considered both as an optical output for diffuse angle even when lit by a point source. The reflector
illumination, or as means for ensuring the maximum can form the rear electrode, as shown in figure 23a).
light input. In practice, the LCD reflectivity is Single polariser operation requires that the liquid
controlled to some extent by the diffusivity of the crystal profile is designed to rotate the input polarised
reflector and / or front polariser. Indeed, adaptions of light through 90° after both passes through the liquid
the reflective layers can be used to deliberate trade-off crystal, thereby being absorbed by the single front
viewing angle and reflectivity. The appearance of the polariser and appearing black. That is, the polarisation
display can be remarkably different when viewed by is elliptically polarised after the first pass, the
the diffuse light of a cloudy day or by the highly indicatrix is rotated on reflection and the light
directional light on a sunny day or in a dark room with becomes linearly polarised after the second pass
a single light source. In the former case, the orthogonal to the input polariser: the display appears
appearance of a scattering mode display will easily dark. For a positive Δε material, this will always
surpass that of the polarised LCD, but the situation is correspond to the OFF state, since in the ON state the
reversed for directional lighting and viewing closer to director approaches the vertical condition and appears
the specular angle. optically isotropic: no change to the polarisation
occurs on either pass and the light is transmitted. The
theoretical curves shown in figure 23 ignore losses: in

51
 a)

b)

Figure 22 Foil Compensated TN. a) Use of discotic foils to compensate the viewing angle of the OFF state for a
transmissive TN LCD; b) and experimental measurements for the resulting viewing angle.

52
 practice the reflectivity may reach up to 35% where α = Γ/Φ and 𝛾 = Φ 1 + 𝛼 $ . This gives dark
in the single polariser modes. The image may be state solutions (R = 0) when the following two
inverted and Normally White operation achieved by conditions are met simultaneously:
the addition of a quarter-wave plate between the
polariser and the liquid crystal, as shown in figure
23a). 1 1 + 𝛼$
Φ= sin^#
Consider a single polariser display using the 1 + 𝛼$ 2𝛼 $
simplest LCD geometry, the ECB mode of section 3.3.
Here, the liquid crystal acts as a switchable wave- (85)
plate. Multiplying the Jones matrices for a polariser, 1
𝜑# = tan^# α$ − 1
wave-plate with optic axis at angle φ1 to the polariser, 2
and the same wave-plate and polariser after reflection
gives: These solutions represent the cases where the liquid
𝑅 = 1 − 4𝑠𝑖𝑛Ã Γ𝑠𝑖𝑛$ 2𝜑# , (82) crystal is at the quarter-wave condition, rotating the
polarisation through 90° over the two passes, and tend
where the retardation is approximately: to those of equation (82) as Φ tends to zero (α⟶ ∞),
with ϕ1 = ±π/4 and Δn.d/λ = (2π+1)/4 for m =
Œ∆].ª 0,1,2…. Figures 23b) and c) show the solutions to
Γ= 𝑐𝑜𝑠 $ 𝜃 , (83)
š equation (85) for the first three orders.
Consider the solution for ϕ1 = 0, shown in
and 𝜃 is the average director tilt and ϕ1 is the polariser
figures 23d) and e): equations (85) predicts R=0
orientation measured from the input director. The OFF
occurs when α = ±1, for which Φ = π / (2√2) = 63.64°
state appears dark (R = 0) if Γ = ϕ1 = mπ/4 (m = 1, 3,
and Δn.d/λ = 0.353. This is a rather low retardation;
5…). That is, at the quarter wave-plate condition light
even if a commercial liquid crystal with the lowest
is circularly polarised with one handedness, which is
birefringence available (Δn ≈ 0.09) is used, this LCD
swapped to the other handedness after reflection and
mode requires a 2.2µm cell gap, potentially lowering
no longer transmitted by the retarder. This mode
production yield and increasing cost. Second and third
suffers from poor dark state leakage and colouration,
order examples are also listed in table 5. Although the
due to wavelength and angular dispersion away from
cell gap becomes more typical, the colouration of the
the quarter-wave condition. As for transmissive state
dark state is far too high for use in a conventional
devices, improved performance is achieved using the
display. Such modes, however, are useful for liquid-
twisted nematic geometry.
crystal-on-silicon (LCOS) projection systems, where
There are several options for single polariser
three panels are tuned to operate at the RGB
reflective TN operation [67], including twist angles
wavelengths.
ranging from 60° to 90°. Replacing the wave-plate
The 63.6° single-polariser TN [114] was
with the Jones matrix for a TN in the derivation above
successfully deployed in Nintendo’s Color Game Boy
gives [114]:
games console, from 1998 – 2003. The display was
$ manufactured by Sharp, who included achromatic
$\ g
𝑅 = 1− 𝑠𝑖𝑛$ 𝛾 + retarders to invert the display to Normally White
#“\ g
$\.Ð{]× $ $ operation. This minimised dark state leakage, which is
1 + 𝛼 $ 𝑐𝑜𝑠𝛾. 𝑐𝑜𝑠2𝜑# − 𝑠𝑖𝑛𝛾. 𝑠𝑖𝑛2𝜑#
#“\ g essential for any colour display, whereas the resulting
(84) wavelength dependence of the white state was readily
compensated through the colour filters. Although the
design gives the highest reflectivity possible with a

53
polarised light mode LCDs, the introduction of colour the natural pitch of the chiral nematic P is much lower
filters still leads to a maximum reflectance of less than allows twists of greater than 90° to be reached [115].
10%. The approach taken by Sharp was to introduce The range of conditions for such “supertwist” states is
prismatic elements into the back reflector, thereby [89]:
directing off-axis light into the viewing direction; the
reduced viewing angle display was considered suitable Φ 1 𝑑 Φ 1
− ≤ ≤ + (86)
for this single viewer application. The adoption of 2𝜋 4 𝑃 2𝜋 4
better performing backlights and high aperture ratio
LTPS back-planes eventually led to transmissive or:
𝜋 𝜋
displays only becoming acceptable for colour portable 1− ≤𝛽 ≤ 1+ (87)
applications. Interestingly, full colour reflective and 2Φ 2Φ
transflective displays remain an area where no
where β = 2π.d/PΦ. The STN range of twist angles is
technology, liquid crystal or otherwise, has yet met the
90° < Φ ≤ 270°, which corresponds to differences in
performance required for market success.
the rubbing direction ϕ1-ϕ2 of:
φ# − φ$ = Φ − π (88)
4.4 Supertwist Nematic LCDs where it is important to ensure that the sign of the
liquid crystal helix matches the pre-tilt of the two
The TN device usually includes a trace amount of surfaces, in the same fashion as the TN. In practice,
cholesteric dopant, simply to impart a natural higher pre-tilts are needed for STN, typically 2° ≤ θS ≤
handedness and prevent domains impairing the 8°.
appearance. Increasing the cholesteric content so that

Table 5 Examples of normally Black Solutions for Single Polariser TN

Cell gap
Order of Director Twist Polariser Angle
Δn.d / λ d / µm
Minimum Φ/° ϕ1 / °
(Δn = 0.09)
1st 63.6 0 (or 90) 0.353 2.2

45 56.4 (or -33.6) 0.683 4.2

nd
2
60 63.6 (or -26.4) 0.606 3.7

3rd 90 32.7 (or -57.3) 1.175 7.2

54
 a) b)

b) d)

e)

Figure 23 Options for single polariser TN modes. a) Schematic of single polariser TN with internal diffusive
reflector. b) Solutions for the Normally Black mode single polariser TN. Continuous lines show the twist angle Φ for
the first three solution sets to equations (85), and the dashed lines the corresponding polariser angles ϕ1. c) The data of
b) re-plot as twist and polariser orientations. The numbers indicated represent the retardation with 90° twist, whereas
the untwisted cells give Δn.d/λ = 0.25, 0.75 and 1.25, for each mode respectively, each with ϕ1 = 45°. d) Calculated
reflectivity for Φ = 64°, 75° and 90° twists, with polariser angles set to give R = 0 for m =1, 2 and 3, respectively. e)
Dark state transmission for Φ = 64°, 75° and 90° for a low Δn materials (0.09).

55
 Figure 24 shows the transmission YÎ iµµ
Φ= ≈ 𝜋 + (92)
characteristics for a set of devices with twist angles ∆Y i´´

operating across the supertwist range [110]. As the

twist increases, both the Fréedericksz threshold Low VC requires highly positive Δε, for which
voltage and the steepness of the curve increase, the typically 0.3 ≤ 𝜀5 ∆𝜀 ≤ 0.5. Ensuring that Φ= is
latter allowing an increase of the number of lines that maintained at an attainable twist then necessitates
can be passive-matrix addressed [89]. 𝑘QQ 𝑘## is as low as possible, contrary to the
The inherent twist of the chiral nematic requirement for the standard TN. For example, a
effectively reduces the twist elastic constant by the typical mixture suitable for STN may have 𝜀5 ∆𝜀 ≈
term (1-2π/P), and the Fréedericksz threshold 0.4 and 𝑘QQ 𝑘## ≈ 0.85 [79] such that the gradient is
becomes: highest for Φ= ≈ 200°. Equation (92) only acts as a
guide for material design, but it shows that keeping
# ý $ both k11/k33 and Δε/ 𝜀5 low is likely to give the highest
𝑉º = 𝜋 𝑘## + 𝑘QQ − 2𝑘$$ (1 − 𝛽)
YZ ∆Y Œ degree of STN multiplexing. Therefore, the material
(89) scientist targets a high k33, Δε and 𝜀5 whilst keeping
k33 low, to combine a low threshold voltage with a
high multiplexibility. Where the gradient exceeds Φ=
Hence, VC increases with twist angle Φ. Similarly, the the voltage response becomes hysteretic, as shown for
voltage dependence of the mid-plane tilt angle is also the 270° STN in figure 24. This is unusable in RMS
affected, with equation (80) now given by [110]: addressed displays, and so typically twists of Φ = 4/3π
= 240° are used in practice. Even then, switching can
‚³ g be disrupted through the formation of stripe domains,
Ù
=
à ^# electro-hydrodynamic instabilities that cause the
ÙÀ
0 g u´´ g helical axis to rotate into the plane of the cell [116].
i´´ ^ “igg (#^Ã…“… g )^i´´ ($…^#) [Y
Ø ugg
1 0 g
+ These require a lower d/P to be used than the β = 1
iµµ “ i´´ ^$igg (#^…) YÎ
Ø condition, together with ensuring the pre-tilt is high.
(90) + ##
For a 240° STN, ≤ 𝛽 ≤ and the range of
1 1
As for the TN case, it may be assumed that k22 ≈ ½k33 Fréedericksz thresholds is:
such that the denominator becomes:
# #;
𝜋 𝑘## + 𝑘QQ < 𝑉º <
µ 0 g YZ ∆Y >
i´´ ^ i´´ (<^1…“… g ) [Y
g Ø
0 g
+ (91) # $$
iµµ “ i´´ … YÎ 𝜋 𝑘## + 𝑘QQ (93)
Ø YZ ∆Y >

Assuming that the chiral doping is chosen to be in the For the typical STN mixture with k33 /k11 ≈ 1.2, VC is
central range for the given twist (β=1), then the 20% to 70% higher than for the equivalent TN
denominator of (82) becomes zero, and the 𝜃¯ 𝑉 operation, with the higher threshold more practical if β
gradient infinite when that twist is set to Φ= : is kept low and the stripe voltage is to be avoided.

56
Figure 24 a) Schematic of a typical Two-polariser Reflective Supertwist Nematic, STN. b) STN transmission
characteristics for increasing twist angle.

Also, the elastic term that appears in the square # Œ

(Φ − ). The retardation for the first minimum is
$ $
brackets of equation (89) is applicable to the response
double that of the TN, potentially decreasing the speed
times: inserting into equations ( (48) show that the
of the device by a factor of four. Even when high
STN response compared is inherently quicker than the
birefringence materials are chosen, the effect is too
TN. However, this is rarely found in practice, since the
slow for animation, with >100ms response times
STN is usually highly multiplexed, and the response
typical. However, the most significant drawback from
much slower due to the much smaller changes of RMS
the original technology were the optical properties: the
voltage.
OFF state is a prominent shade of yellowy-green, or
The first STN demonstrators used a Guest-
blue if the NB polariser orientations are chosen, the
Host mode, in which anisotropic pleochroic dyes give
viewing angle is poor, and the contrast is much lower
contrast between the ON and OFF states [89].
than that of the TN, because the twist remains more
However, optical contrast could also be achieved using
evenly distributed through the cell in the STN case as
optical retardation effects, in what is called the
the field is applied [117]. Various attempts were made
Supertwist Birefringence Effect (SBE), [90]. STN also
to improve the appearance, such as the use of blue
have a Gooch-Tarry type dependence of the
polarisers to give a White on Blue display with high
polarisation optics, with the first minimum mode
transmissivity, but the most successful were in the
occurring at increasing Δn.d/λ with twist, as shown in late-1980s, when optical compensators were added.
figure 25, and the crossed polariser angles set to 𝜑# =

57
Figure 25 Optics of the Supertwist Birefringence Effect mode.

The first method was to mount an inactive dummy best colour compensation is achieved through the use
STN panel onto the front of the LCD, equivalent to the of two polymer films [7], as indicated in figure 24.
active panel but with the opposite handedness [91]. Even where the transmission characteristic is
This corrected both the colouration and the viewing made infinitely steep, the effect of non-uniformities,
angle, and would do so across the whole temperature including electrode resistive losses and temperature
range. However, the additional cost and weight variations, prevents multiplexing much beyond 240
prohibited this approach commercially, and polymer lines. Driving alternate rows from opposite sides of the
film retardation plates are now used instead [118]. The panel doubles the maximum number of rows. Such

58
high degrees of multiplexibility represents a have two optic axes and three principal permittivities.
considerable improvement to the passive matrix TN, The difference between the two directions orthogonal
and for that reason the STN continues to find a market to the n director is called the biaxiality. The refractive
where there is demand for a high Black and White index biaxiality is negligible and FLC are best treated
image content combined with low cost. However, the as optically uniaxial materials, with a single optical
lack of grey-scale and slow response speed prohibits axis along the director. Thus, the polarisation optics
video applications, and it was these factors that fed on- can be treated in the same way as a conventional
going research efforts for highly multiplexed displays. nematic. However, the dielectric biaxiality ∂ε = ε2-
ε2 is significant [120, 121] and, in essence, acts in a
similar fashion to the nematic Δε, dictating the RMS
4.5 Ferroelectric Liquid Crystal Displays response of the c-director. Reflections of the SmC
molecules in the tilt plane do not disrupt the phase
Although making only a small commercial symmetry: molecular dipoles aligned along the C2 axis
impact, Ferroelectric Liquid Crystals commanded have equal probability of being parallel or antiparallel.
major R&D efforts through the 1980s and 1990s, If the phase is chiral, this mirror symmetry is
immediately after the publication by Clark and lost and there is a net dipole in the direction of ±C2:
Lagerwall of the Surface Stabilised Bistable FLC the ferroelectric spontaneous polarisation PS. This
mode [88], SSFLC. Companies across Europe and the unique feature provides a polar switching torque many
Far East each produced demonstrators to rival the times greater than that possible in nematic liquid
nascent TFT technology [119]. However, the only crystals, resulting in switching times that can be faster
panel to receive notable sales in the mainstream than 10µs.
displays market at that time was the 15” 1280 x 1024, For simplicity, a single elastic constant B and
16-colour monitor produced by Canon, figure 16. the flow-free viscosity γ1 are used, although the effect
Unlike the other mainstream LCDs described of the elastic anisotropy and the uniaxial dielectric
in this section, FLC do not use nematic liquid crystals
anisotropy Δε are important too [122, 123]. The
but rather a particular type of smectic, the tilted
switching torque of the FLC is then given by:
smectic C phase, denoted SmC. Unlike the nematic
case, reorientation of the director for smectics is Ô?À Ô g ?À
constrained by the presence of the smectic layers. The 𝛾# sin$ 𝜃º = 𝐵 cos $ 𝛿º +
ÔK Ô® g
smectic C phase is constrained in this fashion too, but 𝑃€ 𝐸® cos 𝛿º sin 𝜙º − 𝜀; 𝜕𝜀. 𝐸®$ sin 𝜙º cos 𝜙º cos $ 𝛿º
the director is relatively free to reorient about the layer (94)
normal in an imaginary cone of possible orientations.
The component of the director in the layer plane is The ferroelectric torque differs from the dielectric
described by the unit director c, which acts as a two- having a linear dependence on the applied field E,
dimensional nematic. It is reorientation of the c- rather than the RMS E2. This means that the torque
director in response to applied electric fields that depends on the field polarity as well as its strength. A
yields the potential for electro-optic effects suited for typical FLC may have a PS = 50nCcm-2 and dielectric
display operation. However, what makes the SmC so biaxiality ∂ε ≈ +0.5. For a typical ±10V signal
interesting is the inherent ferroelectricity allowed by applied across a 2µm spaced cell, the ferroelectric
the reduced symmetry of the chiral smectic C phase torque is fifty times greater than the dielectric,
(SmC*). Whereas the nematic phase has cylindrical dominating the electro-optic behaviour. Assuming the
symmetry, the SmC is monoclinic: there is a single C2 material has a positive PS, then +E will tend to reorient
symmetry axis corresponding to the direction the c-director on one side of the cone (towards φ =0)
orthogonal to the layer normal a and director n. Such and –E to the other (φ = π). Simplistically, if these two
symmetries are inherently biaxial, meaning that they

59
conditions equate to the n-director being oriented π/4 occurs. After the pulse, the director remains in that
apart, and the device has a spacing set to give the state, with the director relaxing back to one of the two
quarter wave-plate condition and crossed polarisers quiescent states.
parallel and perpendicular to the director in one of the The surface pre-tilt and anchoring energies
states, equation (25) predicts that the device will play fundamental roles in SSFLC devices.
switch between minimum and maximum Conventional polyimides for nematic alignment are
transmissivity. used. These have azimuthal and zenithal anchoring
A second important aspect of the SSFLC is energies in the range of 0.3 to 1mJcm-2; high enough
bistability. As for any LCD, the starting point is to to be considered infinite for nematic LCDs, but
achieve the desired alignment, uniformly over the insufficient to cause changes to the layer orientation or
whole sample [124]. There are several steps to SmC* cone angle in FLC. Immediately below the
consider for ferroelectric LCDs. The usual SSFLC SmA to SmC* transition, the layers always tend to tilt
device geometry relies on the N* - SmA – SmC* in the direction favoured by the surface pre-tilts. This
sequence, as shown in figure 26a). The device requires is the C1 state, shown in figure 26a). As the
parallel alignment with the required pre-tilt θS. The temperature is cooled further, θC and δC increase,
cell gap must be sufficiently low to unwind the forcing the surface director to lie further from the
cholesteric helix and provide an almost uniform rubbing direction and pre-tilt angle. Close to the
nematic texture, as shown: devoid of twist and with temperature where the difference between the cone
only a slight splay and bend associated with the and layer tilt angles (θC-δC) approaches the surface
surface pre-tilt. Furthermore, the helicity of the SmC pre-tilt θS, the tilt of the layers swaps direction to form
n- director must also be suppressed in a similar the C2 layer structure. For low surface tilt angles (0.1°
fashion. This uniform nematic texture should be ≤ θS ≤ 2°) samples usually form a mixture of both C1
retained on cooling into the SmA phase, where the and C2 states. The lack of uniformity and zigzag
layers will align uniformly perpendicular to the cell defects that separate the regions of opposite layer tilt
walls in what is termed “bookshelf” geometry. On severely damages the device appearance. However,
cooling into the SmC* phase, the director tilts from the with intermediate pre-tilts (2° ≤ θS ≤ 8°) the layers
layer normal by the cone angle θC, which grows form the C2 state completely and uniformly, whereas
continuously from 0° at the second-order SmA to for higher pre-tilts still (12° ≤ θS ≤ 35°) the layers do
SmC* phase transition to typically 22° to 25° at not undergo the alignment transition at all and the
ambient temperatures, figure 26b). The smectic layer sample is uniformly C1. Both layer geometries have
spacing contracts as the director tilts from the layer been used in SSFLC devices [119, 101], though the
normal, causing the layers to tilt by an angle δC and Canon monitor used the high pre-tilt C1 approach.
form a symmetric chevron-like structure. The layer The out-of-plane tilt of the director θ, and in-
tilt remains a constant fraction of the cone angle, plane tilt angle of the projection of the n-director into
typically δC ≈ 0.85θC, so that the director remains the cell plane are shown in figure 26b). They are given
continuous across the sharp chevron interface at the by the expressions:
cell centre. It is this interface that gives the SSFLC its
bistability. The c director can be at either of two cos 𝜙𝐶 sin 𝜃𝐶
𝛽 = tan^#
orientations, φi and π− φi in one arm, and -φi and π+φi sin 𝜙𝐶 sin 𝛿𝐶 sin 𝜃𝐶 +cos 𝛿𝐶 cos 𝜃𝐶

in the other. These correspond to just two orientations ;

^#
of the n director, as shown in figure 26b). Application 𝜃 = sin sin 𝜙𝐶 cos 𝛿𝐶 sin 𝜃𝐶
of a DC field couples to the ferroelectric polarisation, + sin 𝛿𝐶 cos 𝜃𝐶 (95)
eventually causing such a high torque at the chevron
interface that latching from one state to the other

60
 a)

b)

Figure 26. a) Ferroelectric Liquid Crystal alignment on cooling through the sequence N* - SA – SC* for parallel
aligned surfaces. C1 and C2 chevron layer textures with low and high tilt and triangular director profiles are shown. b)
Definition of angles for SC* devices, and the source of bistability from the chevron interface.

At the chevron interface, there is no out-of-plane tilt which is typically about 8°. For the C1 layer geometry,
and the director has an in-plane twist angle βm given the surface tilt is chosen to be high, typically 25°. For
by: typical values of θC = 25° and δC = 22°, equations (95)
give an in-plane tilt of βS ≈ 27°. Thus, the director
CD± ‚À twists from 27° at one surface to 8° in the cell centre
𝛽¯ = ± cos ^# , (96)
CD± ”À
and back out to 27° at the other surface. Assuming that

61
the twist is approximately linear then the director FLCD panels are addressed in a similar
profile is triangular [125], as shown in figure 26a). fashion to most passive matrix displays, and in the
The transmission of polarised light through such a cell same line-scanning method described in section 2.4.
is similar to that of the reflective TN mode given by However, the response is no longer to the RMS over
equation (84), except the opposite polariser is crossed the frame, but rather the signal applied to each row
to the input polariser in the FLC case, rather than must be sufficient to latch the pixels into a new state
being parallel. This has the simple solution for the within the line-time. In this fashion, the information is
angle to align the polarisers with respect to the rubbing built up line-by-line. The row waveform has one
direction, βext: particular sign of operation, allowing only one set of
states to be selected appropriately. Both states are
¡F …³ ^…í
µ
#“ \ g
addressed, either using two sub-frames of apposite
Æ
𝑡𝑎𝑛2 𝛽?•K − 𝛽€ = , (97) polarities (with a cost of increasing overall frame time)
µ
#“ \ g or by preceding each addressing pulse by a blanking
Æ

where; pulse that selects the black state regardless of the data
being applied to the previous lines, and then latching
𝛼=𝜋
∆].ª
, (98) selectively or not in the addressing line (with the cost
š …Ê ^…³
of reduced brightness for pixels that should remain
white in consecutive frames). Various addressing
as before. Equation (97) suggests that the device will schemes are possible [126], influencing speed,
appear highly coloured in its quiescent state for operating window and appearance.
retardations close to the full wave-plate condition. A second approach to operating the SSFLC
However, if the cell spacing is reduced to the half was also attempted jointly by RSRE (then DERA) and
wave-plate condition, the device appears black with Sharp Corporation. This maximised device speed by
the polarisers aligned at the angle: using the C2 geometry, and multiplexibility using a
lower PS (≈10nCcm-2) and higher dielectric biaxiality
…³ “…í
𝛽?•K = ± . (99) ∂ε (≥ +1), [120, 122]. Equation (94) predicts that the
$
dielectric and ferroelectric terms become equal at
The example above gives βext ≈ ±18°. Thus, if the about 32.5V for such high biaxiality, low PS materials.
polarisers are placed at +18° and+108° to the rubbing If the field has the correct polarity to reorient the
direction, the cell will appear black for one state, and director from one side of the cone to the other, the
transmit most of the light when in the other domain, dielectric biaxiality supresses switching and, above the
where the optic axis is about 36° from the polarisers. voltage where the torques balance (≈ 32.5V for these
Approximating the structure to a uniform retarder, and values), the director will remain unswitched
setting the retardation to be at the half-wave plate indefinitely. In fact, electrical pulses at about 70% of
condition for green light, then the FLC switches this voltage [121, 122] start to slow the response
between states with the optic axis either parallel or at rapidly, creating a minimum in the switching
2βext to the polariser. Equation (25) suggests that the characteristic (τVMIN). Operating close to this voltage
transmittance should be 91%. This could be (i.e. about 23V, for the high biaxiality SmC* material
maximised so that βext = 22.5° (and the director in this example) gives a highly non-linear response,
reorients through 45°), for example by using a material thereby enabling thousands of lines to be addressed
with a higher SmC* cone angle θC. However, this also with a high degree of insensitivity to temperature
causes a decrease in switching speed, and so Canon variations and line losses. Using this τVMIN mode,
used the lower optical efficiency to help achieve a fast, Sharp created a prototype colour ¼ HDTV, operating
flicker-free frame for their monitor.

62
with a 12µs line address time, to give 256 grey levels achieve microsecond pulses across a passive matrix
and a 60Hz frame rate [101]. required metal bus lines to be prepared. Moreover,
By the mid-1990s, the key advantages of the smectics are fundamentally sensitive to shock, since
SSFLC over TFT TN were its perceived lower cost any flow in the panel disrupts the carefully aligned
and excellent viewing angle. The good viewing angle layers irreparably. To prevent mechanical damage to
was inherent to the SSFLC mode due the fact that the the LC alignment, polymer walls were defined photo-
director remains in the same plane for both switched lithographically, again introducing a critical mask step.
states, as shown in figure 27. Such in-plane switching Together with the poorer yield associated with
gives excellent viewing properties, surpassing that of achieving perfect alignment, any cost advantage was
the foil compensated TN of figure 27. However, in the severely eroded. The final battle came with the
mid-1990s, SSFLC lost the war with TFT driven introduction of in-plane switched TFT nematics that
nematics, because of two reasons. Firstly, the number matched, and eventually surpassed, the viewing angle
of critical mask steps grew with FLC complexity. To of even the SSFLCD, figure 16.

Figure 27 Schematic representation of a switchable quarter wave-plate with in-plane switching. One direction of
viewing self compensates the change reduction of birefringence with the increased optical path. The increase in
retardation in the other direction (usually set to be the horizontal direction for the white state) is relatively small, and
easily compensated. A polar plot for the contrast is sketched on the right.

4.6 In-plane Switching LCDs LCD modes to be chosen that meet the more stringent
optical requirements for large area monitor and
4.6.1 Introduction televisual displays. The crucial weakness of TN LCDs
was viewing angle, even with optical compensation.
Operating with TFT removes steepness of the electro- Viewing angle is particularly important for large area
optic transition as a critical design issue. This enables displays, where images must appear uniform from the

63
centre of viewing to the corners. The viewing angle cylindrical symmetry of the director profile in both of
target is harsher still if the image is to satisfy multiple the ON and the OFF states. Two approaches were
viewers. The television market also requires fast developed during the late 1990s and early 2000s: in-
response times, not just for black to white transitions plane switching (IPS) and vertical aligned (VA)
but also between adjacent grey levels. Extremely high modes, as shown in figure (28). Different
contrast ratios are needed to compete with emissive manufacturers championed each mode, and numerous
technologies such as CRT, PDP and, most recently, modifications were tried. This section will concentrate
OLED. on the IPS mode, and its derivative fringe-field-
Achieving pixel contrasts in excess of 10,000:1 switching, whereas VA modes are dealt with in the
and >160° horizontal viewing angles requires following section.

Figure 28. Improvement of viewing angle using IPS or VAN modes.

4.6.2 Basics of IPS Operation plate (in addition to the Source and Gate lines), in-
plane electrodes are etched onto the active back-plane
In-plane fields are impractical for passive matrix in either a comb-like structure, as shown in figure
displays but are readily achieved with Active Matrix 29a), or more commonly a zigzag configuration. The
LCDs by moving the common electrode to the active field produced by comb-like electrodes is complex,

64
varying both in the plane of the cell in the direction 45°, thereby causing the transmission to decrease, as
normal to the electrodes. Ignoring the effect of the shown in figure 29b).
liquid crystal permittivity, and taking the first Fourier The IPS mode can operate with either positive
component only, the field has components [127]: or negative Δε materials, figure 29c), depending on the
orientation of the rubbing direction with respect to the
° Œ‘ Œ®
𝐸‘ ∼ 𝑠𝑖𝑛 𝑒𝑥𝑝 − ; 𝐸® ∼ electrodes φ0. In this diagram, it is assumed that the
J J“G J“G
° Œ‘ Œ® anchoring on the surfaces is strong and director
𝑐𝑜𝑠 𝑒𝑥𝑝 − (100)
J J“G J“G reorientation occurs in the bulk of the cell, although
somewhat closer to the electrode plate due to the
That is, the electric field components have a periodic transverse decay of the field. Clearly, the director in
form in the plane of the cell, but decay exponentially the positive mode will tend to orient in the field
across the bulk of the cell. Rather unsatisfactorily, direction. For the negative mode, the director reorients
most treatments ignore the complex field behaviour to be orthogonal to the applied field, either remaining
and assume that, if the cell gap and electrode width are in the cell plane or tilting out of that plane. In-plane
sufficiently low, the field is uniform and restricted to switching is always favoured in this case, since the
the cell plane. If the quiescent state of the nematic is elastic energy associated with the twist elastic constant
aligned with the rubbing direction orthogonal to that of is lower than the splay-bend that would be induced by
the applied field, the IPS mode has a Fréedericksz the tilt (i.e. k22 < k11 < k33). Although the early
threshold field that is approximately given by equation demonstrators used negative Δε, positive mode is now
(34), although the errors for this can be severe in more common because the materials combine higher
practice [127]. Assuming a uniform in-plane field, the Δε with lower viscosity, and hence give lower
Euler-Lagrange equation for the elastic distortion operating voltages and faster response times.
above the threshold is [128]: With the TN, STN and SSFLC modes,
analytical calculations for the on-axis optical and
Ôg ?
𝑘$$ + −𝜀; Δ𝜀. 𝐸®$ sin 𝜙 cos 𝜙 = 0 , (101) electro-optic behaviour are reasonably accurate.
Ô® g
Optimisation of viewing angle required numerical
which has the simple solution: solutions, due to the importance of off-axis refractive
and reflection effects. However, with the adoption of
Œ.® the IPS mode and the variety of multi-domain VAN
𝜙 = 𝜙¯ 𝑠𝑖𝑛 , (102)
ª modes covered in the following section, both the
electric field and the director profile vary in two or
for fixed boundary conditions. Aligning the polarisers even three dimensions. This necessitates numerical
parallel and crossed to the rubbing direction, the first simulation to optimise the LCD [31, 64]. For example,
order solution for the optical transmission is given by figure 29b) shows the calculated transmission for a 2D
that of a wave-plate, equation (25). Approximating the simulation of the IPS mode [129]. Above the
sinusoidal dependence in equation (102) to the electrodes, the field is almost vertical and the director
triangular form of equation (97) [125] and setting the either reorients vertically for positive Δε or remains
device thickness at the half-wave retardation gives an unaffected by the field if negative Δε; in either
effective in-plane tilt of the director of < 𝜙 >≈ 𝜙¯ / instance the electrodes appear as unwanted dark bands
2, and the transmission approaches 50% with in the pixel. This necessitates that the gap should be
increasing voltage. At fields higher than that which much larger than the width of the electrodes l >> w. If
gives 𝜙¯ = 90°, the twist deformation becomes too high, the field is reduced and the switching
increasingly concentrated towards the surfaces. Thus, voltages become higher. Typically, the electrodes are
the effective twist angle <φ> exceeds the optimum w = 3µm wide, and have a gap l = 6 - 8µm. Therefore,

65
the banding represents a major reduction in the optical viewing cone. However, the stringent requirement to
efficiency of IPS mode. maintain very high contrast well off-axis still
Ideally, the transmission–voltage necessitates the use of optical compensators. For
characteristic should approach linearity across the example, the crossed polarisers themselves leak at the
voltage range, readily giving grey levels. The response high angles in the four quadrants centred at ±45° and
of the IPS mode is strongly dependent on the angle of ±135°. A typical IPS mode LCD used in television
the rubbing direction with respect to the electrodes, φ0 achieves pixel contrast ratios in excess of 2000:1 for
[129]. For positive mode IPS, φ0 ≈ 30° is used direct viewing, and contrast in excess of 1000:1 over
typically [130], not only leading to suitable 175° horizontal and vertical viewing. The lowest pixel
transmission characteristic and near linearity of the contrasts still surpass 200:1 at 140° viewing in the 45°
grey scale response times, but also significantly quadrants. Such impressive viewing angle figures are
improving the response time, which decreases linearly achieved using a front uniaxial wave-plate with its axis
with increasing φ0. oriented crossed to the rubbing direction (and input
The viewing angle characteristics of the IPS polariser) combined with a negative uniaxial wave
mode exceed even that of the SSFLCD, shown in plate formed from a homeotropic discotic liquid
figure 27. The first improvement was to orient the crystal polymer [131].
electrodes in a small-angled zigzag, to help widen the

Figure 29 IPS Principals. a) Schematic of the TFT in-plane electrode arrangement; b) Theoretical transmission
calculated for low and high switching voltages for one electrode period [125]. c) Mode choices associated with
positive Δε and negative Δε materials.

66
4.6.3 Fringe-Field Switching mode. 4.7 Vertically Aligned Nematic Modes

Despite the impressive performance of IPS, As for IPS, the invention of vertically aligned nematic
the mode suffers from poor optical efficiency due to (VAN) occurred early in the history of LCDs [85], but
the banding structure caused by lack of director was reinvigorated with the widespread adoption of
switching above the electrodes. For high-resolution TFT and the need for wide-viewing angle technologies
portable display applications, this is very important in the mid-1990s. Homeotropic alignment gives a near
because increasing the output from the backlight to perfect black state at normal incidence, and a viewing
compensate for optical inefficiency is too costly for angle easily compensated using a negative uniaxial
battery life. Many mobile phones and tablets deploy retardation plate. Being dominated by the dark state,
fringe-field-switching mode LCD [96, 97]. This is a the contrast of VAN mode devices is generally
modification to IPS mode, where a uniform counter exceptional at all angles and is independent of
electrode is placed beneath a dielectric layer with the temperature or cell gap variations. The director is
interdigitated electrodes on top. The pitch of the switched into the plane of the cell when a negative Δε
electrodes is much smaller than the cell gap, figure 30. liquid crystal is used. The direction of tilt will be
There is little electric field directly above the central degenerate and form scattering domains unless some
line of the data electrodes, but immediately either side preferred orientation is imparted to the cell. For
of this there is an in-plane field component that causes example, if one of the homeotropic surfaces is rubbed,
the director to reorient. This arrangement greatly it gives a pre-tilt of typically 89.8°, and the director
reduce the transmission loss caused by the banding, will tilt uniformly in this direction with increasing
and can be operated with electrodes that are closer field. However, this will give a poorer viewing angle
together, therefore allowing lower operating voltages for the white state, as indicated in figure (8). To
to be used. The fringing fields break the symmetry overcome this limitation, Fujitsu invented the
sufficiently to allow operation without alignment pre- multidomain vertically aligned mode in 1997 [94, 95],
tilt, again helping to improve viewing angle towards termed MVA mode. The aim of the invention was to
the limit. This is difficult with conventional rubbed maintain the vertical cylindrical symmetry as the
polymers but can be achieved using photo-alignment. director reorients with applied voltage, using domains
Typically, the dielectric layer is 500nm thick, of opposing tilt. Typically each pixel is subdivided
providing much higher fields close to the electrode into two or four areas with orthogonal tilt directions
surface than typical for the IPS mode [132]. Together for each. Different area ratios may be used for the
with the reduced electrode spacing, this lower horizontal and vertical directions, provided that the
operating voltages, and hence power. The dielectric area for opposing pairs is equivalent.
layer acts as the storage capacitor for the pixel, thereby Initial attempts to produce multi-domain
increasing the aperture ratio and concomitantly alignment used the rather impractical approach of dual
increasing optical efficiency still further. Typically, rubbing, where a rubbed surface was protected during
FFS mode has double the optical efficiency of the IPS a second antiparallel rubbing by a photo-
mode, with the only disadvantage being that two lithographically defined mask that was subsequently
transparent electrodes need to be deposited and etched removed. The approach that Fujitsu took was to
onto the rear plate during fabrication process. The arrange dielectric protrusions onto the electrodes and
success of FFS mode can be measured by its adoption underneath the homeotropic alignment layer, as shown
for many portable products, notably the Apple iPad in figure 31a). Each protrusion has a convex shape
from 2011. with sidewalls angled to the surface, but is sufficiently
small and rarefied to have negligible effect on the
overall pixel alignment. When the voltage is applied,
the slight field fringing around the protrusion causes

67
Figure 30 Fringe-Field Switching mode. a) Basic structure and operation, and b) typical transmission
distribution with varying voltage within a period of the data electrodes, shown in blue.

tilting in opposite directions on either side of the pitch to ensure a twisted orientation of 90° for the
protrusion, thereby breaking the symmetry and given cell gap. Unlike the MVA and PVA modes,
automatically causing domains of the opposite tilt disclinations are avoided altogether because the
sense. Early modules included protrusions on both director forms a mono-domain whilst retaining the
inner surfaces, as shown in figure 31, but including the axial symmetry that results in the exceptional viewing
protrusion on a single surface only saved additional angle.
photolithographic steps, and was found to give These VAN modes share the properties of
satisfactory performance. wide viewing angle, very high contrast, and fast
Samsung made a further advance for the VAN operation. High switching speeds are common to each
mode in 2000. Similar to the MVA mode, the of the vertically aligned modes because the field
patterned vertically aligned mode (PSA) replaced the induced distortion is dominated by the bend elastic
protrusions with slits in the electrodes to create the constant k33, which is usually significantly higher than
fringing fields, figure 31b) [98]. Sharp [99] designed both k11 and k22, equation (48). The mode also has the
the axially symmetric micro-cell (ASM) mode LCD, advantage of not requiring the rubbing step during
where polymer walls surround the nematic, forming a fabrication, which can give improved yield and costs.
microcell container for the liquid crystal, figure 31c). However, the mode is slower than modern IPS mode
The walls are formed by photo-induced phase panels, because full switching is needed to obtain the
separation of a monomer during the polymerisation bright state, whereas IPS switches the director through
that occurs on UV exposure of a grid like pattern. A a lower angle if φ0 >> 0.
chiral dopant is added to the nematic with the correct

68
Figure 31 Vertical Aligned Nematic Modes. A) Multi-domain Vertical Aligned (MVA) mode; b) Patterned
Vertical Aligned (PSA) Mode; c) Axially symmetric multi-cell (ASM) mode.

69
 blades [135]. Indeed, it was encapsulation that
5 LCDs for non-mainstream and Niche eventually led to mass production of liquid crystal
Markets thermometers that remain popular today.
The cholesteric has a natural helical structure
5.1 Introduction shown in figure 1. Light travelling along the helical
The wealth and diversity of different device modes axis undergoes Bragg reflection due to the repeating
that exploit liquid crystals is a measure of both the nature of the optic axis. This occurs for circular
variety of phases and behaviours exhibited by these polarised light with the same handedness as the
organic materials, and by the ingenuity of the cholesteric pitch P and at a band of wavelengths
scientists, engineers, inventors and innovators working centred on λ0 given by:
in this field over the past half century. The path to
providing flat screen monitors and FHD television
λ0 = n.P.cosθ , (103)
displays is littered with unsuccessful attempts.
However, many of those technologies found, or indeed
created, niche markets, offering unique selling points where θ is the angle of incidence with respect to the
in other applications. A selection of some of the more helical axis and n is the average refractive index (= 1/2
important LCDs used outside the conventional direct ne+1/2no). Strong colouration occurs when λ0 matches
view displays are summarised in this section. a visible wavelength, from 400nm to 700nm. The
width of the reflection band Δλ is approximately:
5.2 Cholesteric LCDs
Δλ ≈ Δn.P . (104)
5.2.1 Temperature, Strain and Gas Sensors
From the first discovery of liquid crystals by Reinitzer
The colour is therefore made more vivid by using
in 1888 [18], the bright colours of cholesteric liquid
weakly birefringent cholesteric materials. Light of
crystal textures have fascinated observers. Indeed, the
wavelengths outside the band of selective reflection is
first liquid crystal application was not an electronically
transmitted through the sample, as is all light of the
addressed display, but rather as sensors that deployed
opposite handedness. Optical contrast then requires the
this colouration. In the early 1960s, prolific inventor
liquid crystal layer to be mounted on to a dark backing
James Fergason observed that the peak wavelength of
material, to absorb the transmitted light. Tuning of the
cholesteric selective reflection depended strongly on
temperature range over which the pitch varies in the
both temperature and trace amounts of chemical
visible regime usually requires two or more
vapour [133], envisaging device applications. His
components with different pitches and temperature
work inspired many applications, from detecting
dependences. The pitch diverges as the second order
minute temperature changes on human skin due to the
cholesteric to smectic phase transition is approached,
influence of nicotine or underlying tumour, to
leading to rapidly changing colour from red to blue
visualisation of faulty electronic circuitry.
with increasing temperature. Varying the
Photographer Henry Groskinsky, inspired by the vivid
concentration of components with strong smectic local
colours of the cholesteric, recorded these applications
ordering in the cholesteric mixture provides control
for Life Magazine in 1968 [134]. The article also
over the operating temperature range for the
mentioned that the liquid crystal could be encapsulated
thermometer.
into a polymer, another of Fergason’s inventions that
will be described in section 5.5. That was used by
5.2.2 Bistable Cholesterics
NCR to detect the mechanical strain within a loaded
Selective reflection of coloured light from cholesteric
spring, an application that was later extended to
liquid crystals can be used for electronic displays too.
aeronautical testing of planes, rockets and turbine

70
Moreover, the cholesteric electro-optic characteristic reflection bands. Stacking three layers in series then
can be arranged to be bistable, allowing unlimited allows full reflective colour. This type of technology
multiplexibilty using low cost passive matrix was successfully applied to large area signage by the
addressing. Amongst the first optical switching modes company Magink, providing bright billboards for
to be studied at RCA was the bistable cholesteric advertising purposes [142]. As a reflective display, the
[136], the switching mechanism for which was panels were ideal for bright sunlit conditions, where
elucidated during the 1970s [137, 138]. However, the they could outperform LED electronic signage. KDI
success of the bistable cholesteric is largely due to the also used a triple stack to produce full colour reflective
team at Kent State University headed by Doane [139], displays for portable products. Parallax between the
and the many innovations made by the engineers at the layers was minimised through polymer stabilisation of
spin-out company Kent Displays Inc. (KDI), [140, the individual active layers mounted directly onto a
141]. backing foil [141], as shown in figure 32b). Not only
The basic operation of the display is to switch did this enable good optical performance, but also the
the cholesteric liquid crystal between the Grandjean resulting display was very flexible indeed, as is
texture, where the axis of the helix is largely normal to apparent from the demonstrator shown in figure 32c).
the display which therefore appears reflective, and the Although not successful commercially, this also
focal conic texture, where the helix lies in the plane of remains true of all other reflective colour display
the cell and light is forward scattered to be absorbed modes, and remains an important gap for future
by the rear (black) substrate. Latching between the developments.
states is done via an intermediate homeotropic state, The technical advances made by KDI for
where the field is unwound by a high electric field flexible colour plastic displays promised new
coupling to the positive Δε of the liquid crystal, as applications outside the display field, such as
shown in figure 32a). If the field is switched off electronic skins, figure 32d), and electronic writing
immediately, the helix forms with its axis vertical to tablets, figure 32e). The Boogie board is an electronic
the pixel, thereby appearing coloured due to selective writing pad that uses mechanical pressure to induce
reflection. If the field is reduced more gradually the reflective Grandjean texture, on a black
through an intermediate level, the helix forms in the background. When the image needs to be refreshed, an
plane of the cell, in the focal conic state, and the pixel electrical blanking pulse is applied to erase the page.
appears dark. The degree of hysteresis and the This product continues to be successful in a niche
sensitivity of the device to mechanically induced market that the technology has created.
damage is controlled using polymer stabilisation.
Most displays are designed to be 5.3 Bistable Nematic LCDs and ZBD
monochrome, using materials with high birefringence
to give the broadest reflectivity. The highest practical Bistable operation of an LCD allows many
levels of Δn give yellowish-green colouration against lines of information to be passive matrix addressed,
black, though some customers prefer the inverted where each frame is written line by line and each line
optics of white and blue when a blue background is retains its information until it receives the next
used. Together with the slow response speed, high addressing signal. Before the large-scale adoption of
voltage and sensitivity to shock, the poor appearance TFT back-planes and active matrix addressing, various
means that the devices are not commercially bistable nematic modes were invented, complementing
successful, and have been largely superseded by other the contemporary efforts to develop bistable
choices. However, bistable cholesterics remained of cholesteric and ferroelectric liquid crystal displays.
interest for full colour bistable reflective displays. Few of these modes made it even to demonstrator
Lower birefringent materials are used to give stage. However, in the late 1990s, there was a
individual layers with sharp red, green and blue resurgence of interest in bistable nematics [127],

71
Figure 32 Bistable Cholesterics. a) operating principle; b) tri-layer stack for full colour; c) Prototype of the film
backed triple N* stack; d) Switchable electronic skin using a single layer cholesteric; e) The successful Boogie-Board
product from KDI, using bistable cholesteric to form a writing tablet.

72
notably the work at Seiko-Epson on the Bistable However, these drivers, were also attractive to other,
Twisted Nematic (BTN) Mode, the 180° BTN or non-liquid crystal display modes, such as Janus
Binem™ developed by Nemoptic, and the Zenithal colloids from Gyricon, interference mode MEMS from
Bistable Device (ZBD) by ZBD Displays Ltd., (now Iridigm / Qualcomm, electrowetting from Liqua-vista /
Displaydata). Several factors drove this renaissance: Amazon and Gammadynamics, and electrophoretics
• Simple passive matrix displays based on TN from Bridgestone and E-ink. In 2005, Amazon selected
and STN did not lose market share with the E-ink displays for its Kindle e-Book readers, based
introduction of TFT; rather, the markets that largely on the appearance of its white state and despite
they had created, such as watches, calculators, the significantly higher cost due to the active matrix.
instrumentation, etc., continued to need low This created demand for scattering mode devices, and
cost, usually reflective displays. Whereas new, many new markets chose electrophoretics over LCDs.
high investment production plants in Japan, One bistable LCD technology that survived was the
Korea and Taiwan developed TFT LCDs for Zenithal Bistable Display, marketed by ZBD Displays
high end applications, a plethora of Ltd, now Displaydata. This spinout from R.S.R.E (then
manufacturers in China continued to serve DERA) in the UK targeted the retail signage market,
what remained a $1B passive matrix market and in particular electronic shelf-edge labelling. For a
well into the 2000s. retailer to replace tens of thousands of electronic labels
• Bistability promised to create new niche LCD per store, cost was paramount, and the bistable LCD
markets, particularly where ultra-low power has a significant advantage over electrophoretics,
was required. Markets included electronic whilst offering superior performance and higher image
shelf-edge labels with high information content than the incumbent directly addressed
content, electronic book readers, smart-card segmented TN LCD. The company’s success was not
displays, secondary displays for mobile only due to the bistable LCD, but also by development
phones, and an enormous range of indicator of a novel RF communications protocol [126], that
displays, from car key-fobs to razor blades. allowed small to mid-sector retailers to take advantage
Indeed, the image storage offered by of the labelling, whereas only hyper-stores had been
bistability continues to be a unique selling able to afford the infrastructure required for previous
point for many applications associated with labelling systems. The company has sold several
the Internet-of-things, where displays can be million labels worldwide, and now offers a
updated automatically using radio frequency combination of the ZBD LCDs alongside E-ink for
(RF) signals, rather than rely on costly higher-end application.
batteries and associated circuitry. Bistability results where a device has two
• Bistable displays often have good potential for stable states with similar free energies that are
plastic displays. TFT on plastic remained a separated by an energy barrier, wherein transitions
challenge to fabricate through the 2000s, due from one state to the other are discontinuous, or first
to the difficulties of registering high- order. An early approach was the Bistable Twisted
resolution patterns on a flexible backplane, Nematic (BTN) mode [143], which followed similar
and the low fabrication temperatures required principles to that of the STN, but set the d/P ratio of
for plastic substrates. Bistable nematics the chiral nematic to lie halfway between states of low
offered the possibility of high image content and high twist angles Φ. For example, setting d/P =
based on simple, low temperature, low cost 0.5 with parallel surface alignment should give a π
fabrication on plastic. twist state. However, if the pre-tilt on both surfaces is
sufficiently high, the cost of the induced splay energy
becomes greater than that for twist. Thus, the chiral

73
nematic may either unwind to a uniform 0π state to this instance, the director cannot change its twist from
match the surface condition, or may wind further to one state to the other without breaking the anchoring
form the Φ = 2π twist state, figure 33a). Switching at one of the surfaces, requiring that one of the
from one state to the other then relies on whether or surfaces be deliberately weakly anchored [145].
not flow is induced immediately after a high electrical Switching of the device again utilised back-flow
pulse coupling to a positive Δε. If the pulse returns to depending on the trailing shape of the addressing
0V via an intermediate voltage, there is little induced pulses. Marketing the device under the trade name
flow and the 0π state is formed, whereas a direct Binem™, Nemoptic produced various demonstrators
transition to 0V induces flow that encourages director [146], including full colour reflective displays, TFT
twist at the cell centre and the Φ = 2π state is formed. driven panels and, intriguingly, a pixelated switchable
The two states are metastable, so the texture relaxes quarter-wave plate mounted onto the front of an
OLED display to switch between high-power emissive
back to the intermediate π-state after a second or two
video frame rate display and ultra-low power Black
on removal of power. This means the device was not
and White E-reader mode [147]. Despite the excellent
suited to zero power applications. Rather, Seiko Epson
used it as the display for Hi-Fi Graphic Equalizer optical appearance, 0-π BTN required cell gaps below
displays, due to its very fast optical response [144]. 2µm and suffered from manufacturing tolerances that
A similar approach was taken by Nemoptic, were difficult to achieve. Perhaps the biggest cause of
who also used a BTN configuration, but with d/P = the company’s eventual demise in 2010 was that it
0.25 to give either 0 or π twist states, figure 33b). In failed to find the correct niche for its product.

Figure 33. Operation of Bistable Twisted Nematics. a) 0 – 2π mode ; b) 0 – π mode or Binem™.

74
 The Zenithal Bistable Display (ZBD) is rather Typical cell gaps are 7µm, operating at 15 – 20V using
unusual in several aspects [148, 149]. It uses a grating standard STN driver electronics, LC mixtures and
as a surface alignment layer, designed to impart polarisers. Operating temperatures from -25°C to 40°C
bistable pre-tilts of the contacting nematic regardless and -5°C to +60°C were achieved using low and high
of the overall geometry chosen, figure 34. This allows anchoring energies, respectively.
many different LCD designs to be used, including The main technical challenge for ZBD
VAN- HAN mode [148], HAN to TN mode [150, Displays Ltd., was to introduce a low cost and
148], multistable VAN – HAN – TN modes [151], reproducible manufacturing method for a 0.8µm pitch,
grey scale displays [152], single polariser mixed TN 1µm high blazed sinusoidal grating into a standard
modes [153], scattering modes [154] and ultra-fast Pi passive matrix LCD production line using Gen 2 glass.
cell modes [151]. The device is an early example of, This was done by copying a photo-lithographically
and the only LCD currently on sale, that utilises the defined master grating into a lacquer on Polyethylene
flexoelectric effect for latching between the two states. terephthalate (PET) film, via a nickel sputtering and
Moreover, the device deliberately uses disclinations at electro-forming replication technique [149, 156]. The
the deep, homeotropic-grating surface to stabilise the film was shipped from the UK to manufacturers in
low tilt or D state. The potential barrier between this China, where it is used to emboss the grating into a
state, and the high tilt defect-free or Continuous state homeotropic photopolymer deposited on the glass
(C state) is mediated by the creation and annihilation surface, thereby replacing the conventional rubbing
of these surface defects. The voltage for latching from step for that plate. This method allows the technology
one state to the other VCD is related to the pulse width τ to meet the same price point as conventional STN,
by [155]: where the costs of the compensation foils required for
the STN offsets that of the ZBD grating film. Having
°Àî ×µ JÊ $ôÌ
≈ + achieved such low price-points for its chosen niche
ª ?µ “?´ ä ?µ “?´ “ YZ [YI´´
×µ JÊ market, it remains to be seen if ZBD can replace
= + 𝑉KV , (105) conventional TN and STN displays in other markets.
?µ “?´ ä

where Wθ is the zenithal anchoring energy of the 5.4 Polymer Dispersed Liquid Crystals
homeotropic grating surface, ls the coefficient of slip
A weakness of LCDs that is particularly evident in
for defects moving across that surface, e1+e3 is the
reflective mode devices is the constraint for polarisers,
sum of the splay and bend flexoelectric coefficients for
which absorb more than half of the available light.
the liquid crystal material, and the dielectric effect of
This inefficiency was avoided in the first dynamic
the grating has been ignored. Controlling the
scattering displays, but the contrast and lifetime of
anchoring energy in the range 0.2mJcm-2 to 1mJcm-2
those devices was far inferior to retardation based
allows the threshold voltage Vth to be adjusted to
LCDs such as the TN, that their period of success was
typically about 1Vµm-1, typically 7V.
very short-lived. Hilsum [157] produced a scattering
The device is usually configured with the
device by mixing glass micro-spheres into a nematic,
grating opposite a standard parallel aligned rubbed
creating a scattering texture that could be switched to a
polymer surface to create a 90° TN when in the low
non-scattering state by an electric field. A more
tilt D state. This state is always formed first on cooling
practical device was the polymer-dispersed liquid
from the isotropic to nematic phase, because the
crystal (PDLC). This is another example of an LCD
defects are stabilised at the surface when the S order
hat initially aimed to produce bright displays, but
parameter is low. Thus, the interpixel gaps remain in
which found success when the technology was applied
the TN state, and good display reflectivity results
when operating in the normally white TN mode.

75
Figure 34 The Zenithal Bistable Display Mode, ZBD™. a) Latching from the C to D state occurs with a bi-polar
pulse with the trailing part positive with respect to the grating. The applied field nucleates defects at the bifurcation
point on the grating sidewall, and the positive polarity favours the D state with -1/2 disclinations at the convex surface
and +1/2 defects at the concave surface. The resulting low surface tilt creates a 90° TN state when placed opposite a
conventional monostable rubbed polymer surface as shown. Operating in the normally white mode, with the grating
aligned parallel to the front polariser and with n matched to the liquid crystal ordinary refractive index gives an
excellent white state. b) Applying pulses of the opposite polarity causes the defects to retrace their path along the
grating surface until they annihilate. The resulting C state is continuous and homeotropic, thereby creating HAN
alignment. This appears black when between the crossed polarisers as shown, and the viewing angle self-compensates
when in reflective mode.

to a novel product with a niche market; for PDLC this matrix in the early 1980s [158]. Originally called
was privacy glass. NCAP by its inventor, PDLC usually takes the form of
Following his successes with cholesteric a plastic layer that can switch between scattering and
sensors and the invention of the twisted nematic LCD non-scattering states with an applied electric field
soon after Schadt and Helfrich (claiming precedence [159, 160]. An ITO coated glass or transparent plastic
in the US), Fergason invented a method of cell is filled with a nematic liquid crystal mixed with a
encapsulating liquid crystal droplets into a polymer monomer. Curing of the monomer into a solid polymer

76
form is done either by evaporating the monomer light intensely due to the refractive index mismatch
solvent, applying a thermal treatment or, most between the polymer and the liquid crystal.
commonly, through the initiated polymerisation of Application of an electric field across the sample then
constituent photo-reactive groups. Phase separation of causes reorientation of the liquid crystal director,
the liquid crystal occurs as the polymer forms from the leaving the cross-linked polymer undisturbed. Usually,
monomer, creating droplets within the polymer matrix. the liquid crystal has a positive Δε and an ordinary
Often, a surfactant is included, both to help control the refractive index no matched to that of the polymer.
dispersion of droplet size and align the nematic into a Thus, the scattering is reduced as the liquid crystal
radially symmetric or bipolar state, as shown in figure aligns parallel to the field, and light incident close to
35. Droplets in the 0.5 to 5µm range scatter incident the normal direction is transmitted unchanged.

Figure 35 Polymer Dispersed Liquid Crystal for a positive Δε operating in the scattering to non-scattering
mode. a) Droplets of different sizes with random orientations of the director profile, forming either a radial or bipolar
configuration. Light is strongly scattered, including back scattering. b) Application of the field initially causes little
distortion of the director profile within each droplet, but the profile adjusts to align the disclinations within the plane
of the film and the director with a net direction parallel to the field. The scattering reduces somewhat. c) Increasing
the field causes distortion of the director field within each droplet towards the vertical orientation, where the
refractive index of the droplet and polymer matrix approach the matched condition, and little scattering occurs. d)
Conventional back-scattering mode of PDLC; e) Novel, low voltage mode using specular reflection as the black state;
f) Common mode for privacy windows, where the plastic film switches between translucent and transparent.

77
 Droplets with radial alignment have a single are typically used. Thus, the voltage of such PDLC
defect at the droplet centroid (called a “Boojum”) films can be high, typically much higher than 30V to
leading to spherical symmetry, whereas bipolar achieve saturation of the transparent, non-scattering
droplets have point defects on opposing surfaces and state.
cylindrical symmetry of the director field. Typically, If the droplets are sufficiently small, and the device
droplets are between 1 and 10 micron, and form the spacing high, then the OFF state can lead to significant
bipolar structure with randomly oriented symmetry back scattering. This can be used in a reflective mode
axes. The electrical field behaviour is complex. The display, by mounting the PDLC film onto a black
field required to align the director parallel to the field, background to provide optical contrast, figure 35d).
and hence approach the minimum deflection of However, the switching voltage inherently increases
incident light, is inversely proportional to the droplet both with smaller droplets and higher device spacing.
radius, a [161]. For a radial droplet, the critical field is Moreover, the shallow switching response necessitates
approximately related to [162]: the use of TFT to matrix address the pixels, limiting
the switching voltage to 15V and preventing any
à i useful degree of optical effect. Recently, Sharp has
𝐸Ó = , (106)
â YZ [Y produced low voltage TFT PDLC displays where the
layer is mounted instead onto a specular mirror [163].
where 𝑘 is a mean elastic constant. The symmetry of a Relatively large droplets keep the voltages sufficiently
perfectly spherical bipolar droplet prevents analytical low for TFT addressing (≈6V), and the forward
solutions being made, and needs the symmetry to be scattered light is scattered more strongly after being
broken. A practical approach is to assume that each reflected back towards the observer a second time by
droplet is slightly elongated with different semi-major the mirror. In the ON state, the forward scattering is
axis a and semi-minor axis b. For liquid crystals with removed and the viewer sees the specular reflected
positive Δε, weak applied fields reorient the symmetry light. In applications where the display is illuminated
axis parallel to the direction of the field. The critical by a point source, this will appear black in all
field is then given by [162]: directions except where the display is oriented directly
between the viewer and light source, figure 35e).
# JKÀ i´´ â g ^Þ g Although never successful for mainstream
𝐸º = +2 , (107) displays, one market that PDLC has found a niche is
QâÞ JL YZ ∆Y
for privacy windows. The PDLC can be made as a
where σLC and σp are the low frequency conductivities laminated plastic roll, which can be applied adhesively
of the liquid crystal and polymer respectively. The to existing infrastructure, and trimmed to the
response times are given by: appropriate size and shape. The window can then
switch between a translucent, "milky white"
×µ appearance in the OFF state, to being clear and fully
𝜏ÖT = u
YZ ∆Y7 g “ g´´g â g ^Þ g transparent when ON. Although this adds costs to the
é ß

window, it is increasingly popular in high-end

×µ â g Þ g architectural projects for both interior and exterior
𝜏ÖWW = . (108)
i´´ â g ^Þ g
applications. Uses include privacy control of
conference rooms, intensive-care areas or bathroom
Typically, droplets of a = 1µm, b = 0.9 µm, k33 = doors, for example, to providing architectural design
10pN, γ1 = 0.05 kgm-1s-2 and Δε = 15, and σLC/σp ≈ features. Moreover, the material can also be adjusted
28, giving EC ≈ 1.5V/µm, and τOFF ≈ 20ms. to act as an environmental smart window, switching
Substantial scattering needs twenty or so droplets in between heat retention and loss as required [164].
the direction of incidence, so film thicknesses of 20µm Other applications such as temporary projection

78
screens, active camouflage, and switchable diffusers be applied, with typical pixel pitches for an XGA
are all possible [165]. projection display being below 3µm.
Another potentially profitable application for LCoS spatial light modulators are the active
PDLC is switchable holographic media [166]. If a element in light projectors for conferences and home
coherent image is used to cure the polymer during the cinema systems. This is one of the largest niche
fabrication process, switchable diffractive optical markets for LCDs, despite strong competition from
elements can be fabricated with droplets ranging from micro-electro-mechanical-systems (MEMS) projection
100nm to 1µm forming holographic gratings from the displays (Digital Light Projectors, DLP). Recently,
interference pattern of the illumination. Rather than there has been great interest in near-eye displays and
cause scattering, the droplets locally modulate the pico-projectors, such as those used in Virtual and
refractive index to generate a diffraction grating. Augmented Reality headsets, and for Google glass.
Holographic optical elements (HOE) are used for a Such displays tend to use a small poly-silicon
variety of optoelectronic applications, including transmissive TN LCD, rather than reflective LCoS.
focusing, beam-steering, filtering and optical The projector systems either use a single
multiplexing. The use of holographic PDLC provides LCoS panel, or use three panels tuned for red, green
the means for switchable HOE, allowing dynamic and blue wavelengths and the final image combined
beam steering, tuneable filtering, and optical signal using a prism system. A 45° single polarise TN can be
processing. For Holographic PDLC the droplet size a used, but VAN mode provides the high contrast
is typically sub-micron and the OFF time is very fast: needed for home cinema, as utilised by the top of the
typically 50µs to 500µs. However, EC also increases range projectors from Sony and JVC, [168].
with decreasing droplet size, and so typically 100 to Considerations such as viewing angle and colour
200V is required. This high field dominates the ON balance become inconsequential in such projection
time, which again is fast, typically 50µs. systems, where speed and contrast are the dominant
factors, and performance can be optimised for
individual colour bands.
5.5 Liquid Crystal on Silicon (LCoS) A different approach to colour uses a
ferroelectric LCoS. Rather than using three separate
Rather than apply a semi-conductor onto glass to modulators, the fast response time of the ferroelectric
provide transistors, Liquid Crystal on Silicon (LCoS) liquid crystal is used to provide operation at 360Hz,
places the liquid crystal element directly onto a CMOS thereby allowing frame sequential colour when
integrated circuit [167, 168]. Of course, as the silicon illuminated by 120Hz alternating colour band
is opaque, the devices must operate in reflection; illumination [169, 170]. This is used for helmet
usually, silver electrodes patterned directly onto a mounted VR, binocular displays and viewfinders by
CMOS integrated circuit provide both the pixelated Forth Dimension Displays and Cinoptics.
electric field pattern and the highly reflective rear LCoS has also proven successful for non-
substrate. The chip not only provides the drive signals display applications, particularly for spatial light
for the contacting liquid crystal but can also deliver modulators in optoelectronic systems such as
extra functionality, such as grey-scale gamma wavelength selective switching, structured
correction, temperature compensation, edge detection, illumination and optical pulse shaping [171].
etc. The devices are typically only 1cm to 2cm in A related technology to LCoS is the Optically
diagonal, and are used for projectors and camera Addressed Spatial Light modulator [172]; rather than
viewfinders. Keeping such low dimensions enables the electrically addressing each pixel, the device is
ultra-high resolution of the silicon circuitry industry to addressed using incident light. The LCD includes un-
pixelated ITO electrodes sandwiching a photo-
conducting semiconductor and liquid crystal layers. In

79
the dark state, the dielectric permittivity of the Moreover, non-display applications of liquid
photoconductor is low, and most of the electric field crystals continue to grow in importance and variety. A
applied across the ITO electrodes is dropped across good example of this is the advent of liquid crystal
this layer, leaving the liquid crystal unswitched. lasers that have more far-reaching possibilities for
However, charge is transferred to the liquid crystal optoelectronics. Brief appraisals of each of these
interface where light is incident on the modes are included in this section.
photoconductor, which switches the liquid crystal
accordingly. Early devices used GaAs photoconductor 6.2 Pi- Cell
and a nematic. Recently [173, 174], an α-Si OASLM Nematic liquid crystals are simple to align, usually
with an FLC modulation layer has been used to maintain alignment quality after receiving mechanical
produce dynamic computer generated hologram with or thermal shock, and are well understood and
over 108 pixels. The speed and the bistability of the characterised. The down side is often a slow response
FLC allowed the image to be built successfully in a speed, particularly at low operating temperatures
projection system using a series of electrically where the Arrhenius form of the viscosity dominates,
addressed nematic SLM. The resulting system remains equation (14). An early example of a fast nematic
arguably the most complex display to be produced, mode is the Pi-cell [175], where ON and OFF times
and the nearest to providing full 3D holographic approaching 1ms are achieved.
dynamic images, which remains the ultimate display The pi-cell is the most well known example of
goal. a surface mode liquid crystal [176]. It relies on parallel
alignment of high pre-tilt surfaces, between which is a
positive Δε nematic. For low surface pre-tilts, the
6 LCD Modes for potential future lowest energy quiescent state is predominantly splayed
applications from one surface to the other. For high pre-tilts, a bend
state of the director becomes favoured energetically.
6.1 Introduction This occurs above the pre-tilt given by [177]:
The variety of different LCDs is huge and only a
handful has made it to commercial success, whether 4θ s − π k 33 − k11
mainstream or niche. Some of these modes are =
sin 2θ s k 33 + k11 , (109)
important to describe, since they include principles yet
to be utilised (e.g. the Pi-cell, V-shaped switching
FLC and Anti-ferroelectric Liquid Crystals); others are which predicts that the splay and bend states are
recent modes that are yet to find a market (such as the energetically equivalent states for 40° < θs ≤ 50° for
Blue-phase, or flexoelectric cholesteric modes). It was typical calamitic nematics (0.5 ≤ k33/k11 ≤ 2). Pre-tilts
suggested earlier that the principal motivation for LCD of this magnitude have been hard to achieve
developments has evolved from increasing image historically, with θs ≈ 25° typically being the
complexity, to achieving the widest viewing angle, maximum. With pre-tilts of this magnitude, the splay
high resolution and recently colour depth. Alongside state is energetically favoured but the bend state is
these developments has been a constant need for metastable, and can exist alongside the splayed state
increasingly fast LCD switching. Further speed separated by a π disclination. With an applied voltage
improvement continues to be important for gaming, coupling to the positive Δε, the central director of the
virtual and augmented reality (VR, AR), and splayed state can tilt in either a clockwise or
ultimately for glasses-free 3D displays with eye anticlockwise direction, concentrating the elastic
tracking for multiple viewers, and frame sequential distortion close to the top or bottom surface, figure
colour. 36b). If a sufficiently high field is applied, the elastic
distortion may become sufficiently high to break the

80
anchoring at one of the surfaces and allow a first order “optical bounce” effect slows the response for ECB
alignment transition to the bend or π-state [178]. On and TN devices significantly [68]. However, for the π-
removal of the field, the director relaxes back to the state, the final tilt remains vertical in the cell centre
meta-stable state where the director remains vertical in and the flow acts in both halves in the same direction
the cell centre. Whilst in this state, the director field as the relaxing director, thereby hastening the
close to the surfaces can be affected by the application relaxation process. Thus, viscous flow supports
of a lower field in a continuous fashion, thereby switching so that the device is very fast – typically
modulating the retardation and causing optical contrast switching in about 2 - 5ms for both ON and OFF
between ON and OFF states, with a fast response. switching.
To understand why the pi-cell gives this fast response, The device is oriented with the alignment
the effect of viscous back-flow needs to be considered. directions at 45° to crossed polarisers, and the cell gap
Figure 36a) shows the situation for a uniform director set to give the half-wave plate condition. This occurs
produced by anti-parallel surface alignment. When the for a higher cell spacing for the bend-state due to the
field is removed, the director is subject to both elastic higher overall tilt of the director. The viewing angle of
and viscous restoring forces. Lateral flow of the the bend state is better than the splayed state, because
material is set up in opposing directions in either half of the symmetry of the director arrangement. That is,
of the cell, acting to kick the director in one half in the the change in retardation for off-axis light is increased
opposite direction to that of the other, thereby slowing in one half of the cell and decreased in the other half,
the relaxation to the final state. This back-flow, or so that there is little retardation change. This self-

Figure 36. The effect of back-flow and the Pi-cell. a) With anti-parallel alignment, there is flow alignment in
opposing directions on removal of the field, hindering the relaxation back to the 0V uniform state. b) Parallel
alignment initially gives a splayed state, but application of sufficiently high field to break the surface anchoring
allows a bend, or p-state to be formed. When the field is removed, relaxation to the un-switched bend state is
supported by the flow throughout the cell. However, the bend state is metastable, so the splay state reforms after
sufficient time, usually nucleated from the pixel edges or LCD spacers.

81
compensation effect leads to the alternative term for 6.3.1 N*-SmC*
the Pi-cell as optically compensated bend-mode
(OCB) [179]. Further improvements to the viewing Conventional SSFLC uses an N*-SmA-SmC* to give
properties can be achieved using film compensators, a chevron layer arrangement and the implicit
[180]. bistability that brings. If the ferroelectric liquid crystal
The device is yet to be used commercially. It has a phase sequence where the SmC* cools directly
cannot achieve the high contrast ratio enjoyed by the from the unwound chiral nematic, the director remains
IPS and VAN modes, due to the remnant surface in the rubbing direction at the two surfaces and the
retardation; even with optical compensation, dark state layers form in a uniform layer bookshelf geometry but
leakage is too high to meet the high contrast ratio with the layers angled at the cone angle θC to the
requirement of modern TV displays. Moreover, the rubbing [182]. Cooling with an applied DC field
bend-state is meta-stable, and usually decays back to ensures that a single sign of orientation is achieved.
the splayed state after removal of the field. Unless a Thus, the device is mono-stable, with an analogue
constant bias voltage is maintained, domains of the response as the director switches about the cone in the
unwanted splay state spread from unswitched areas bulk of the cell with one polarity of field, and no-
such as the inter-pixel gaps and close to spacers. response with the opposite polarity. This “half V-
However, the principles deployed in this mode remain shaped” switching can be addressed by TFT to give a
of interest. For example, a transflective display has fast, analogue response.
recently been suggested that uses sub-pixellation for A modification is the twisted Ferroelectric
the bend state operating in reflection, and the splay LCD [183] which combines a material with a first-
state operating with in-plane electrodes in transmissive order N*-SmC* transition with rubbing directions
mode [181]. crossed to each other. The unwound N* forms a
conventional TN but, as the twist elastic constant
6.3 Analogue Ferroelectric Liquid diverges on approaching the smectic C phase, the
Crystal Modes smectic layers form uniformly at 45° to the rubbing
directions. The director still forms a 90° twist, but
The surface stabilised FLC mode described in section moves about the SmC* cone from one surface to the
4.5 is bistable: either the black or white states is other with the applied DC field, leading to a “V-
retained after the addressing pulse. Grey scale is shaped” switching response. This mode works best
provided using spatial or temporal dither [126]. with a strongly first order transition, wherein the
Bistability was integral to the original interest in SmC* cone angle is independent of temperature and is
FLCDs, since it provided an alternative to TFT. typically close to the optimum 45°; this allows the
However, once TFT had become sufficiently low cost, director to lie parallel to the crossed rubbing
interest was retained in utilising the fast optical speed directions.
of ferroelectric liquid crystals, but using an analogue
response combined with active matrix. 6.3.2 Anti-ferroelectric liquid crystals
Various analogue ferroelectric liquid crystal
modes were studied, as outlined below. Each of the Certain compounds that form chiral tilted smectic
modes described suffers from the need to achieve and phases exhibit higher ordered phases, where there is
maintain uniform smectic layer structures, which has correlation of the director orientation between adjacent
prevented display applications to date. However, the layers. With antiferroelectric liquid crystals (AFLC),
high speeds that are possible means that these modes the SmC* c-director and polarisation directions
may yet find utility in non-display and optoelectronic alternate by π from one layer to the next. Application
applications. of a DC field switches each alternate layer parallel to

82
the field, thereby forming a V-shaped switched Blue phases are a subset of cholesteric liquid
response [184]. crystals, which occur when the cholesteric natural
helicity is strong and the pitch is very short, close to
6.3.3 Deformed helix mode ferroelectric the transition to the isotropic phase [187]. Rather than
liquid crystal. spontaneously twist along a single axis perpendicular
to the local director, the director twists along two
Although formed from chiral liquid crystals, most FLC mutually orthogonal axes to form a double helix
modes use sufficiently low spaced devices to unwind cylindrical structure, as shown in figure 37a). The
the pitch. This is not the case for the deformed helix director at the centre of each cylinder lies parallel to
mode, where the pitch is made sufficiently low that the cylinder axis, and the diameter of the cylinder is
even devices at the quarter wave plate condition wind P/4 so that the director twists through 45° from one
continuously in a helical manner. Application of the side to the other. This means that the director remains
field then causes distortion of the sinusoidal director continuous across adjacent cylinders oriented with
variation and a shifting of the optic axis from that of their axes orthogonal to each other, as shown in figure
the helix towards ±θC [185]. 37b). These double twist cylinders pack into a cubic
array as shown in figure 37c), mediated by disclination
6.3.4 The Electroclinic Effect. lines throughout the structure, that occur at the
interstices of the cylinders. Of the three possible blue
Smectic A phases formed from chiral molecules phases, the simple cubic Blue phase I, figure 37f) and
undergo a field induced director tilt with an applied body-centred cubic Blue Phase II, figure 37d) have the
DC field, forming a structure similar to the SmC* widest temperature ranges. The occurrence of the
phase [186]. The strength of this electroclinic effect is disclinations means that the cubic structure is only
greatest immediately above an SmA to SmC* stable where the pitch and order parameter are very
transition, where the induced tilt is linear with the low, typically for about 1°C to 2°C below the isotropic
applied field. As with the AFLC and DH FLC modes, to cholesteric phase transition. Optically, the structure
the switching is within the cell plane, and hence it is appears an iridescent colour due to Bragg reflection
not only fast but has good viewing angle. However, from the defects arranged on the regular cubic lattice,
the limited temperature range and temperature figure 37e); hence, the phase became known as the
dependence have so far discounted electroclinics from “blue” phase, although other colours are possible
application. depending on the lattice. Otherwise, the phase is
optically isotropic due to the cubic symmetry of the
6.4 Blue Phase TFT displays phase, and it appears dark between crossed polarisers.
Blue-phases exhibit a Kerr effect with an applied field,
Attendees at the 2008 exhibition that occurs each year due to deformation of the local cholesteric structure
alongside the Society of Information Display annual coupling to the dielectric anisotropy of the liquid
conference were surprised to see an unheralded novel crystal [188, 189]. This induces a net birefringence
liquid crystal display mode based on the blue phase at ΔnE with the optic axis in the direction of the applied
the Samsung stand. The company demonstrated a 15” field [190]:
TFT monitor operating at a ground breaking 240Hz
frame rate. Liquid crystals were known for their slow Δ𝑛7 = 𝜆𝐾𝐸 $ ~ − 𝑛Q 𝐸 $ , (110)
response, and most attendees were unaware of the
Blue-phase. Those that were, most likely believed that where n is the isotropic refractive index of the blue
the blue phase would never receive such serious phase and the Kerr constant K is also wavelength
interest from a manufacturer, due to the notoriously dependent.
narrow temperature ranges exhibited by these phases.

83
Figure 37 Blue Phase Liquid Crystal. a) The Double helix structure; b) The intersection of three double twist
cylinders and the direction of the -1/2 disclination that occurs close to the cylinder intersections; c) Cubic packing of
the double twist cylinders; d) Packing of the cubic Blue Phase II, with associated defect lattice; e) Photomicrograph of
the transition between Blue Phase I and II; f) Body centred packing of the cubic Blue Phase I and the associated
defect lattice.

Although the Kerr effect provided means for liquid crystal at temperatures where the material
electro-optic modulation, the extremely limited formed the blue phase [189]. The resulting polymer
temperature range of the blue-phase seemed severely formed at the defect sites, stabilising the lattice
prohibitive to application. This was particularly true structure over a wide temperature range but without
for displays, where operation from -20°C to +70°C is affecting the electro-optic properties of the liquid
usually the minimum specification. In 2002, Kikuchi crystal significantly.
et al succeeded in greatly extending the temperature The Samsung display used in-plane electrodes
range by photo-polymerising a monomer added to the to switch the polymer stabilised blue phase from an

84
optically isotropic structure, to a birefringent medium showed that using angled electrode walls produced
with the optic axis approximately in the plane of the higher transmissivity, by reducing the angle of the
cell, figure 38. The device does not require alignment optical axis at the electrode edges. Using these
layers, and gives excellent high contrast and wide structures, operating voltages below the 15V target for
viewing angle due to the isotropic nature of the dark TFT addressing were achieved. Indeed, since that
state and in-plane switching effect. The most original work, a host of different device layouts and
important issue to resolve was achieving suitably low liquid crystal materials have been studied [192],
operating voltages. This was done using shaped indicative of the on-going interest in the Blue Phase
protrusions as the electrodes [191], with 2µm wide mode for applications.
electrodes and 2 to 4µm electrode gaps. Modelling

Figure 38 Principal of operation for In-plane switching mode Blue Phase LCD. The applied field induces
optical birefringence (shown by the ellipsoid shape of the refractive index indicatrix). Good performance is achieved
using shaped protrusions allowing 10V operation to be achieved [187].

85
 6.5 Cholesteric Flexoelectricity and the optimised for small pitch lengths P, promising speeds
Uniform Lying Helix (ULH) mode below 100µs and greater than 45° switching angles for
strongly flexoelectric materials [194], and in-plane
Recently, there has been much renewed interest in the switching for good viewing angle. The pitch is
flexoelectro-optic effect in chiral nematic liquid unwound by the RMS effect of the field coupling to
crystals, originally proposed by Patel and Meyer [193] Δε, so liquid crystal materials with strong flexoelectric
in 1987. This interest stems from the very fast in-plane effect but low Δε are required. This has been achieved
switching effect, typically 100µs. If an electric field is using nematic dimers with odd alkyl spacer groups
applied perpendicular to the helical axis of a [195].
cholesteric liquid crystal, coupling between the field Fabrication of devices based on this effect
and the flexoelectric polarization causes splay and requires a method for obtaining the required ULH
bend of the local director field, which in turn causes alignment uniformly and stably over the panel.
the local director to tilt away from the helical axis in Strongly planar anchoring at the cell walls leads to the
the direction normal to the field, figure 39a). Without Grandjean texture being formed, where the helical axis
the field, the uniaxial optic axis of the cholesteric lies lies parallel to the cell normal (sometimes called the
parallel to the helical axis. The field-induced local tilt Uniform Standing Helix). Homeotropic alignment
of the director causes the optic axis to tilt through gives degenerate focal conic domains, and a scattering
angle β in the plane orthogonal to the applied field. texture. Various methods have been investigated,
If the cholesteric can be uniformly aligned including using periodic planar and homeotropic
with the helical axis parallel to the plane of a standard surfaces with the periodicity matched to P/2 [196],
LCD with transverse electrodes (Uniform Lying Helix, surface relief gratings [196, 197] and polymer walls
ULH), then applying the electric field causes rotation [198]. The director profile for a cholesteric liquid
of the optic axis in the cell plane, giving contrast when crystal aligned on a homeotropic grating with a pitch
observed through polarisers parallel and crossed to the of P/2 and a cusped shape to give the lowest elastic
helical axis. This in-plane change of retardation axis distortion of the ULH state [197], as shown in Figure
provides a switchable half-wave plate. 39b). Although yet to be proven over large areas, these
Assuming negligible dielectric anisotropy, the methods show considerable promise for future
twist angle is approximately linear with the applied applications of the ULH mode LCDs.
field:

?µ ^?´ N 6.6 Liquid Crystal Lasers

𝑡𝑎𝑛𝛽 = 𝐸 , (111)
$Œ iµµ “i´´

With the successful commercialisation of high

and the response time τ is:
resolution, high frame rate LCDs for TV, interest in
non-display applications for liquid crystals has
×µ N g
𝜏= 𝐸 . (112) continued to grow at an amplified rate. One such
$Œg iµµ “i´´
application particularly relevant to optoelectronics is
Flexoelectric switching is polar, so the ULH can be the invention [199] and demonstration [200, 201] of
driven in either direction by swapping the field the liquid crystal laser. This allows straightforward
direction. Thus, equation (112) is appropriate for both tunability of laser light across the optical wavelength
ON and OFF switching, since the optic axis can be range combined with simple low cost fabrication, and
driven back to the OFF state using signals of the offers the potential for electric field tuning of the
opposite polarity to those for switching. Both lasing wavelength.
expressions show that the strength of the effect is

86
Figure 39 The cholesteric flexoelectric effect. a) Rotation of the optic axis; b) Alignment of N* into the ULH
texture using cusped grating surfaces [193].

87
 Selective reflection from the cholesteric liquid or changing to a dye with a longer wavelength
crystal is an example of a one-dimensional photonic emission swaps the lasing wavelength to the upper
band-gap, wherein circularly polarised light of the photonic band edge.
correct wavelength range and handedness cannot The stimulated emission occurs in a single
propagate through the material in the direction parallel direction, as shown in figure 40. Such devices offer
(or antiparallel) to the helical axis. This means that potential for forming laser arrays, competing against
spontaneous emission from a fluorophore within the III-V Semiconductor Vertical Cavity Surface Emitting
cholesteric is inhibited within the photonic band-gap, Lasers (VCSEL) but without the need for the multiple
leading to photons being emitted at the band edge fabrication steps. Indeed, the simplicity of fabrication
[202]. That is, optical pumping of a cholesteric with allows ink-jet printing of the liquid crystal to form an
either a fluorescent dye dissolved into it or as array of laser dots, each with a signature emission
constituent part of the mesogenic molecule causes [204]. The low cost and ability to print onto a plastic
emission of lower energy photons That overlap with backing layer has already earned the technology
the forbidden band of wavelengths corresponding to commercial application for anti-counterfeiting.
the range of selective reflection. Where emission Lasing is possible in other liquid crystal
occurs across the band-gap lasing will occur at both phases that include Bragg type structures with optical
edges. In the example system shown in figure 40, a length scales, including the ferroelectric chiral SmC*
tetra-aryl-pyrene derivative dye [203] pumped at and Blue-phases [205]. The former offers the potential
410nm gives emission peaking in the range 450nm – for fast electric field modulation, whereas the latter
500nm. Once dissolved into a cholesteric with a produces lasing in multiple directions due to the cubic
photonic band from 500 to 520nm lasing occurs on the structure of the Bragg lattice.
lower band edge. Either lowering the cholesteric pitch

Figure 40 Lasing in dyed cholesteric. a) Example of the optical properties for the fluorescent dye and
cholesteric liquid crystal, leading to lasing on the lower band edge of the selective reflection; b) schematic
representation of cholesteric lasing.

88
6 Conclusion operation and OLEDs emit and so cannot modulate
reflected light. Solutions to achieving the required
The market success of liquid crystal displays performance at a suitably low cost have yet to be
is predicated by scientific and technological commercialised; perhaps the developments of new
achievements from across the globe, advances that nematic modes [e.g. 180] or the application of a new
have needed collaborative efforts of mathematicians, LC phase, such as the Blue-phase III [206] will prove
chemists, physicists, engineers and entrepreneurs. The successful.
breadth of the subject is unusually broad and An area of enormous growth is the use of
involving. A major part of the on-going success of liquid crystals for non-display applications [207]. In
LCDs is the continual evolution and adaptation that addition to the conventional applications described in
these technological advances allow. Throughout the the previous sections, liquid crystal main-chain
history of the LCDs the naysayers have maintained a polymers such as Kevlar continue to prove one of the
mantra that LCDs cannot achieve the complexity, biggest LC applications, liquid crystal elastomers are
cannot achieve the viewing angle, cannot achieve the used as the active element in nano-machines and
resolution, cannot achieve the speed, cannot achieve molecular motors [208], liquid crystal semiconductors
the colour balance of the latest competing technology. are attracting interest for photo-voltaic and OLED
In each case they have proven incorrect. Often, this [209], and the great interest in liquid crystals in
has been through the invention of new modes, chemical and biological sensors continues with
sometimes associated with different arrangements, unabated enthusiasm [210]. One of the largest areas of
sometime with different materials and phases. Often, growth, however, is in the field of optoelectronics.
the improvements have come from the use of Whether this is for smart-windows, beam steering and
partnering technologies; the active backplanes, light guiding, printable lasers, switchable lenses,
polariser and optical films, back-light units, light optical computing, adaptable photonic structures, or
guides or manufacturing advances. Terahertz modulators, successful solutions to such
As it has always been, the competition is applications in the future will require both the
strong. OLED displays are already making inroads ingenuity and the multidisciplinary approach that
into small, high-resolution displays for mobile phones. made our predecessors so successful.
If performance, cost and lifetimes can all exceed those
of the LCD, then they will gain an ever increasing Acknowledgements
share of the market, from low to high diagonal The author wishes to thank Professors Peter Raynes,
displays. Low cost reflective passive matrix LCDs Cyril Hilsum, Phil Bos, Tim Wilkinson and Dr.
retain a large market presence too. However, as the Mamatha Nagaraj for invaluable discussions.
cost of the TFT backplane continues to reduce, the
competition from electrophoretic displays is strong
and growing. Winning these battles will take yet more References
ingenuity from the LCD scientists and engineers.
There remain several display markets where [1] S. Kubrick (Dir.) (1968) “2001: A Space
no display technology has yet delivered, whether LCD Odyssey”, Metro-Goldwyn Mayer, Released 2nd
or alternative. Although reflective colour and April 1968. Premiered in New York from 3rd
April 1968 at the Loews Capitol Theatre,
transflective displays have been marketed, solutions to
Broadway, NY.
date have not been adequate to gain market [2] B. J. Lechner, (2008). History crystallized: A
acceptance. LCDs have a natural advantage for first person account of the development of
transflective mode operation, being based on matrix-addressed LCDs for Television at RCA
transparent media that modulates ambient light. in the 1960s. Information Display, 08 (1), 26–
Electrophoretics absorb light preventing transmissive 30.

89
[3] G.H. Heilmeier and L.A. Zanoni (1967). [18] F. Reinitzer (1888) “Contributions to the
“Electro-optical Device” US patent 3499112 A. understanding of cholesterol,” (transl.),
[4] G.H. Heilmeier, L.A. Zanoni and L. A. Barton Monatshefte für Chemie (Wien) vol. 9, pp 421 –
(1968) “Dynamic Scattering: A new electro- 441.
optic effect in certain classes of nematic liquid [19] O. Lehmann, (1889) Zeitschrift für
crystals,” Proc. IEEE, 56, pp1162 – 1171 Physikalische Chemie, 4, p462
[5] B. Levin and N. Levin (1936) “Improvements in [20] D. Dunmur and T. Sluckin (2011) Soap,
or relating to Light Valves,“ UK patent 441, Science, and Flat-screen TVs: a history of liquid
274. crystals, Oxford University Press ISBN 978-0-
[6] J. A. Castellano, (2005) “Liquid Gold: The story 19-954940-5.
of Liquid Crystal Displays and the Creation of [21] I. Haller (1975) “Thermodynamic and static
an Industry,” World Scientific, Singapore properties of liquid crystals”, Prog. Sol. State
[7] H. Kawamoto (2001) “The History of Liquid Chem., 10 (2) pp. 103 – 118
Crystal Displays,” Proc. IEEE vol. 90, no. 4, [22] J. W. Goodby, P. J. Collings, T. Kato, C.
pp460 – 500 Tschierske, H. Gleeson, & P. Raynes (Eds.),
[8] G. W. Gray (1962) “Molecular structure and (2014) “Handbook of Liquid Crystals”, Wiley
properties of Liquid Crystals”, Academic Press, VCH, Weinheim. ISBN: 978-3-527-32773-7.
London. [23] R.B. Meyer, L. Liébert, L. Strzelecki and P.
[9] G. H. Brown and W.G. Shaw (1957) “The Keller (1975) “Ferroelectric liquid crystals,”
Mesomorphic State: Liquid Crystals”, Chem. Journal de Physique, vol. 26, pp L69–71.
Rev., 57, pp 1029 – 1157. [24] M. Klasen-Memmer and H. Hirschmann (2012)
[10] R. Williams (1963) "Domains in Liquid “Liquid Crystal Materials for Devices”, IN J.
Crystals," J. Chem. Phys. 39, No. 2, 384-388 Chen, W. Cranton and M Fihn (eds.)
[11] R. Williams (1962) U.S. Patent 3,332,485 A. “Handbook of Virtual Display Technology”,
Filed Nov. 9, 1962 and issued May 30, 1967 Springer-Verlag Berlin Heidelberg.
[12] C. Hilsum, (2010). “Flat-panel electronic [25] D. Dunmur and G. Luckhurst (2014) “Tensor
displays: a triumph of physics, chemistry and properties and Order Parameters of anisotropic
engineering”, Philosophical Transactions. materials” IN J.W. Goodby et al. “Handbook of
Series A, Mathematical, Physical, and Liquid Crystals: Volume 2”, Wiley VCH,
Engineering Sciences, 368(1914), 1027–1082. Weinheim, pp189 – 281.
doi.org/10.1098/rsta.2009.0247 [26] J. Constant and E.P. Raynes, (1985) RSRE
[13] N. Koide (2014) “The Liquid Crystal Display Internal Memorandum, unpublished
Story: 50 Years of Liquid Crystal R&D that lead [27] F. C. Frank, (1958). I. Liquid crystals. “On the
the Way to the Future”, Springer, Japan. theory of liquid crystals”, Discussions of the
[14] M. Schadt and W. Helfrich, (1971) “Voltage- Faraday Society, 25 (1), 19- 28.
dependent optical activity of a twisted nematic [28] J.C. Jones (2000) “Liquid Crystal Elastic
liquid crystal,” Appl. Phys. Lett., vol. 18, no. 4, Constants”, Encyclopaedia of Material Science
pp.127–128. [29] F. M. Leslie (1966) “Some constitutive
[15] G. W. Gray, K. J. Harrison, and J. A. Nash, equations for anisotropic fluids”, Q. J. Mech.
(1973) “New family of nematic liquid crystals Appl. Math, 19, pp 357-370; F.M. Leslie (1968)
for displays,” Electron. Lett., vol. 9, no. 6, pp. “Some constitutive equations for anisotropic
130–131. fluids”, Arch. Ration. Mech. Anal. 28, pp 265-
[16] A. Ashford, J. Constant, J. Kirton, and E. P. 283.
Raynes, (1973) “Electro-optic performance of a [30] M. G. Clark and F. M. Leslie, (1978) “A
new room-temperature nematic liquid crystal,” calculation of orientational relaxation in nematic
Electron. Lett., vol. 9, no. 5, pp. 118–120. liquid crystals, Proc. Royal Soc. Lond. A, 361,
[17] G. H. Heilmeier, “Liquid crystal displays: An pp 463–485.
experiment in interdisciplinary research that [31] D.W. Berreman (1983) “Numerical Modelling
worked,” IEEE Trans. Electron Devices, vol. of Twisted Nematic Devices”, Phil. Trans. R.
ED-23, July 1976. Soc. Ser. A: Math. And Phys. Sci. 309, (1507)
pp 203 -216.

90
[32] S.T. Wu and C.S.Wu, (1990). “Rotational Rigid Polymer Networks to Elastomers”, CRC
viscosity of nematic liquid crystals A critical Press, Florida. ISBN 978-1-4200-4622-9.
examination of existing models”, Liquid [44] R.J. Bushby and N. Boden (2014) “Responses
Crystals, 8(2), 171–182. of Discotic Liquid Crystals to Mechanical,
doi:10.1080/02678299008047339 Magnetic and Electrical Fields” IN J.W.
[33] R. B. Meyer (1969) “Piezoelectric effects in Goodby et al. “Handbook of Liquid Crystals:
liquid crystals” Phys. Rev. Lett. vol. 22, pp 918– Volume 3; Part III: Discotic, Biaxial and Chiral
921. Nematic Liquid Crystals”, Wiley VCH,
[34] E.P. Raynes, (2014) “Mixed systems: phase Weinheim, pp1 - 33.
diagrams and eutectic mixtures”, IN J.W. [45] C. V. Mauguin, (1911) “Sur les Cristaux
Goodby, P. J. Collings, T. Kato, C. Tschierske, Liquides de Lehmann”, Bull. Soc. Fr. Min, 34,
H. Gleeson, & P. Raynes (Eds.), “Handbook of 71-117.
Liquid Crystals: Volume 1”, Wiley VCH, [46] V. Fréedericksz and A. Repiewa (1927)
Weinheim. pp 351 – 363. “Theoretisches und Experimentelles zur Frage
[35] W. Martienssen and H. Warlimont (Eds.) (2005) nach der Natur der anisotropen Flüssigkeiten”,
Springer Handbook of Condensed Matter and Zeitschrift für Physik 42 (7): 532–546.
Materials Data, Springer Verlag; Heidelberg. [47] V. Fréedericksz and V. Zolina, (1933) "Forces
ISBN 3-540-44376-2 causing the orientation of an anisotropic liquid".
[36] K. Tarumi, B. Schuler, E. Bartmann and D. Trans. Faraday Soc. 29: 919–930.
Pauluth (1996) “Liquid Crystalline Medium”, [48] P. Châtelain, (1941) Sur l’orientation des
US Patent 5,744,060 cristaux liquides par les surfaces frottées,
[37] A. Buka and N. Éber (Eds.) (2013) Comptes rendus de l’Académie des Sciences
“Flexoelectricity in Liquid Crystals: Theory, 213, 875–76
Experiments and Applications”, Imperial [49] B. Levin and N. Levin (1936) “Improvements in
College Press, London. ISBN 978-1-84816-799- or relating to Light Valves”
5. British Patent 441, 274.
[38] G.W. Gray and D.G. McDonnell, (1976) “The [50] H.J.J. van Boort and R. Groth, (1968) “Low-
Relationship Between Helical Twist Sense, pressure sodium lamps with indium oxide filter”
Absolute Configuration and Molecular Structure Philips Tech. Rev., 29, pp17-18.
for Non-Sterol Cholesteric Liquid Crystals”, [51] M. Mizuhashi (1980) “Electrical properties of
Molecular Crystals and Liquid Crystals, 34, (9), vacuum-deposited indium oxide and indium-tin
211–217. oxide films”, Thin Solid Films, 70 (1980) 91-
[39] M. Hird (2011) “Ferroelectricity in liquid 100
crystals—materials, properties and applications, [52] Y. Katsube and S. Katsube (1966) “The Tin-
Liquid Crystals”, 38, (11-12), pp. 1467-1493. oxide Films Deposited by Vacuum Evaporation
[40] F.C. Saunders, K.J. Harrison, E.P. Raynes and (I)”, Shinku: J. Vacuum Soc. Jap., 9, (11),
D.J. Thompson, (1983) “New photostable pp443 – 450.
anthraquinone dyes with high order [53] F.M. Wanlass (1963) “Low stand-by power
parameters”, IEEE Trans. Electr. Dev., ED-30, complementary field effect circuitry” US patent
pp 499–502. 3, 356,858
[41] I. Jánossy, (1999) “Optical reorientation in dye [54] E. H. Land and J.S. Friedman (1933)
doped liquid crystals”, J. Nonlin. Opt. Phys. “Polarizing refracting bodies” US patent
Mat., 8, (3) pp 361 – 377. 1,918,848 A
[42] M. Schadt, K. Schmitt, V. Kozinkov and V. [55] J.F. Dreyer (1946) “Dichroic light-polarizing
Chigrinov (1992) “Surface-Induced Parallel sheet materials and the like and the formation
Alignment of Liquid Crystals by Linearly and use thereof”; US Patent 2544659.
Polymerized Photopolymers”, Jap. J. Appl. [56] T. Nagatsuka (2014) “Chapter 4.1 –
Phys., 31, (7) pp Development of polarizers to support the
[43] D. Broer, G. Crawford and S. Žumer, (2011) development of the liquid crystal industry” IN
“Cross-linked Liquid Crystalline Sytems: From N. Koide (ed.) The Liquid Crystal Display
Story: 50 years of liquid crystal R&D that led

91
 the way to the future, Springer, Japan. pp 81 – [69] P. M. Alt and P. Pleshko, (1974) “Scanning
95. limitations of liquid-crystal displays, IEEE
[57] P. K. Weimer, (1966) “Thin film transistors,” Trans. Electron Devices, vol. ED-21, pp. 146–
IN ” J. T. Wallmark and H. Johnson, (Eds.) 155.
Field Effect Transistors, Devices, Englewood [70] B.J. Lechner, F.J. Marlowe, E.O. Nester, and J.
Cliffs, N. J. Prentice-Hall. Tults, (1971) “Liquid Crystal Matrix Displays,”
[58] E.P.Raynes, (1974) “Improved contrast Proceedings of IEEE, 59, (11), pp. 1566–1579.
uniformity in twisted nematic liquid-crystal [71] T.P. Brody, J.A. Asars and G.D. Dixon, (1973)
electro-optic display devices”, Electron. Lett., “A 6x6 inch, 20 lines-per-inch liquid crystal
10, pp 141-142. display panel.” IEEE Trans. Electron Devices,
[59] A. A. Sonin (1995) “The surface physics of ED-20, 995-1001.
liquid crystals,” Gordon and Breach Publishers, [72] P.G. LeComber, W.E. Spear and A. Ghaith,
Luxembourg. (1979) “Amorphous silicon field effect device
[60] A. Hirai, I. Abe, M. Mitsumoto and S. Ishida and possible application,” Electron. Lett. vol.
(2008) “One Drop Filling for Liquid Crystal 15, pp. 179- 180.
Display Panel Produced from Larger-sized [73] A. J. Snell, K. D. Mackenzie, W. E. Spear, P. G.
Mother Glass” Hitachi Review 57, (3), pp 144 - LeComber and A. J. Hughes (1981)
147. “Application of amorphous silicon field effect
[61] M. O’Neill and S. Kelly (2012) Chapter 6: transistors in addressable Liquid Crystal Display
Optical Properties of Light-emitting Liquid Panels,” Appl. Phys. vol. 24, pp. 357 - 362.
Crystals IN R. J. Bushby, S. Kelly and M. [74] Y. Okubo, T. Nakagiri, Y. Osada, M. Sugata, N.
O’Neill, Liquid Crystalline Semiconductors: Kitahara, and K. Hatanaka, (1982) “Large-Scale
Materials, properties and applications, Springer LCDs Addressed by α-Si TFT Array,” SID
Science and Business Media, London, pp 163 – Digest, pp. 40–41.
195. [75] F. R. Libsch (2009) “Thin-film Transistors in
[62] R.C. Jones (1941) “New calculus for the Active Matrix Displays” IN C.R. Kagan and P.
treatment of optical systems”. I. J. Opt. Soc. Andry “Thin-film Transistors”, Dekker, New
Am., 31: p488 – 493. York. pp187 – 269.
[63] K. R. Welford,and J.R. Sambles, (1987). [76] M.-K., Kang, S. J., Kim, and H.J. Kim, (2014).
“Analysis of Electric Field Induced Fabrication of high performance thin-film
Deformations in a Nematic Liquid Crystal for transistors via pressure-induced nucleation.
any Applied Field.” Mol. Cryst. Liq. Cryst, 147 Scientific Reports, 4, 6858.
(1) pp25 -42. [77] X. Huang, C. Wu, H. Lu, F. Ren, D. Chen,
[64] “LCD Master”, Shintech; R.Zhang and Y. Zheng, (2013). Enhanced bias
http://shintech.jp/wordpress/en/software- stress stability of a-InGaZnO thin film
2/lcdm3d.html; “TechWiz” Sanayi transistors by inserting an ultra-thin interfacial
http://www.sanayisystem.com/eng/index.asp; InGaZnO:N layer. Applied Physics Letters, 102,
DIMOS, Autronic Melchers. 193505.
[65] E. Jakeman and E.P. Raynes (1972) “Electro- [78] E. Lueder. (2001) “Liquid Crystal Displays:
optical response times in Liquid Crystals”, Addressing Schemes and Electro-optic effects”,
Phys. Lett., 39A, (1), pp. 69 – 70. John Wiley and Sons, Chichester.
[66] H. Chen, M. Hu, F. Peng, J. Li, Z. An and S.T. [79] S.M. Kelly and M. O’Neill, (2000). “Chapter 1:
Wu, (2015). “Ultra-low viscosity liquid crystal Liquid Crystals for Electro-optic Applications”
materials”. Optical Materials Express, 5, (3), pp IN H.S. Nalwa, Ed., Handbook of Advanced
655 – 661. Electronic and Photonic Materials, Volume 7:
[67] S. T. Wu and D. K. Yang, (2001) Reflective Liquid Crystals, Displays and Laser Materials.
Liquid Crystal Displays Wiley, New York. Academic Press, UK.
[68] D.W. Berreman (1975) “Liquid‐crystal twist [80] S.W. Depp, A. Juliana and B.G. Huth, (1980)
cell dynamics with backflow”, J. Appl. Phys., “Polysilicon FET devices for Large Area
46, 3746 - 3751. Input/Output Applications”, Tech. Digest
International Electron Devices Meeting,

92
 Washington DC, December 1980, IEEE, Paper plane switching mode,” Appl. Phys. Lett. 67, pp.
27.2, pp703 – 706. 3895–3897.
[81] T. Sameshima, S. Usui and M. Sekiya, (1986) [94] A. Takeda, S. Kataoka, T. Sasaki, H. Chida, H.
“XeCl Excimer laser annealing used in the Tsuda, K. Ohmuro, T. Sasabayashi, Y. Koike,
fabrication of poly-Si TFT’s”, IEEE Electron and K. Okamoto, (1998). “A super-high image
Device Letters, 7 (5), pp276–278. quality multi-domain vertical alignment LCD
[82] K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. by new rubbing-less technology,” SID Digest
Hirano and H. Hosono, (2004). Room- 98, pp. 1077–1100
temperature fabrication of transparent flexible [95] Y. Koike and K. Okamoto, (1999) “Super high
thin-film transistors using amorphous oxide quality MVA-TFT liquid crystal displays,”
semiconductors. Nature, 432 (7016), 488–492. Fujitsu Sci. Tech. J., 35, pp. 221–228.
[83] J. Fergason (1971) “Liquid Crystal Non-linear [96] S.L. Lee and S.H. Lee, (1996). “Liquid Crystal
light modulators using electric and magnetic Display Device with Improved Transmittance
fields”, US Patent 3,918, 796. and method for manufacturing same:, KR Patent
[84] F.J. Kahn, (1971) “Nematic liquid crystal No. 19960059509, US Patent No. 5,914,762.
device”, US Patent 3694053 A [97] S.H.Lee, S.L.Lee and H.Y.Kim (1998) "High-
[85] M. F. Schiekel and K. Fahrenschon, (1971) Transmittance, Wide-Viewing-Angle Nematic
“Deformation of Nematic Liquid Crystals with Liquid Crystal Display Controlled by Fringe-
Vertical Orientation in Electrical Fields”, Appl. Field Switching," Proc. 18th International
Phys. Lett. 19, 391-393. Display Research Conference (Asia Display), p.
[86] S. Kobayashi, T. Shimojo, K. Kasano, and I. 371-374.
Tsunda, (1972) “Preparation of Alphanumeric [98] J.O. Kwag, K.C. Shin, J.S. Kim, S.G. Kim, and
Indicators with Liquid Crystals” SID Digest of S.S. Kim (2000) “Implementation of New wide
Technical Papers, p68-72. viewing angle mode for TFT-LCDs”, SID
[87] R.A. Soref (1973) “Transverse Field Effect in Technical Digest, 31, pp256 – 259.
Nematic Liquid Crystals” Phys. Lett., 22, pp165 [99] Y. Ishii, S. Mizushima, and M. Hijikigawa,
[88] N.A. Clark and S.T. Lagerwall, (1980). (2001) “High performance TFT-LCDs for AVC
“Submicrosecond bistable electro-optic applications,” SID Digest 01, pp 1090–1093.
switching in liquid crystals”, Applied Physics [100] E. Jang, S. Jun, H. Jang, J. Lim, B. Kim and Y.
Letters, 36, pp899–901. Kim, (2010) “White-light-emitting diodes with
[89] C. M. Waters and E.P.Raynes (1982) “Liquid quantum dot color converters for display
Crystal Devices with particular Cholesteric backlights”, Advanced Materials, 22 (28),
Pitch / Cell thickness ratio”, US Patent pp3076–3080.
4,596,446. [101] J.C. Jones (2012) “Bistable LCDs” IN J. C.
[90] T.J. Scheffer and J. Nehring (1984) Janglin, W. Cranton and M. Fihn (eds)
“A new, highly multiplexable liquid crystal Handbook of Visual Display Technology,
display” Appl. Phys. Lett. 45, 1021-1022. Springer Verlag, Berlin Heidelberg. pp 1507 –
[91] K. Katoh, Y. Endo, M. Akatsuka, M. Ohgawara 1543.
and K. Sawada (1987) “Application of [102] J.C. Jones (1999) “The evolution of liquid
Retardation Compensation; A New Highly crystal displays”, Opto and Laser Europe, 69,
Multiplexable Black-White Liquid Crystal (12), pp 41-42.
Display with Two Supertwisted Nematic [103] R. Williams, “Electro-optic elements utilising
Layers” Jpn J. Appl. Phys., 26, (11) L1784 – an organic nematic compound” US Patent
L1785. 3,322,485.
[92] R. Kiefer, B. Webber, F. Windscheid, and G. [104] R. Williams, “Domains in Liquid Crystals” J.
Baur, (1992) “In-plane switching of nematic Chem Phys., 39, (2) pp 384 – 388.
liquid crystals,” Proc. Japan Displays’92, pp. [105] L. M. Blinov (1998) “Behaviour of Liquid
547–550 Crystals in Electric and Magnetic Fields”, IN
[93] M. Oh-e and K. Kondo, (1995) “Electro-optical D. Demus, J.W. Goodby, G.W.Gray,H.W.
characteristics and switching behavior of the in- Spiess and V. Vill, (Eds.) Handbook of Liquid

93
 Crystals, Volume 1: Fundamentals”, Wiley Volume 1: Fundamentals”, Wiley VCH,
VCH, Weinham, Germany. pp 437 – 544. Weinham, Germany. pp 727 – 762.
[106] S. Chandrasekhar (1992) “Liquid Crystals”, [118] O. Okamura, M. Nagata and K. Wada, (1987)
Cambridge University Press, UK “Neutralized STN (NTN) – LCD full colour
[107] C.H. Gooch and H. A. Tarry (1975) “The picture image”, Inst. Television Eng. Japan,
Optical Properties of Twisted Nematic Tech. Rep. 11, (27) pp 79 -84.
Structures with Twist Angles ≤90°”, J. Phys. [119] S.T. Lagerwall (2014) “Ferroelectric Liquid
D: Appl. Phys. 8, 1575 – 1583. Crystal Displays and Devices”, IN J. W.
[108] J.C. Jones and G.P. Bryan-Brown, (2010), Goodby et al (Eds.), “Handbook of Liquid
“Low cost Zenithal bistable device with Crystals, Volume 8: Applications of Liquid
improved white state”, Proc. SID, Int. Symp. Crystals”, Wiley VCH, Weinheim, Germany.
Digest, Technical papers, vol. 41, pp. 207 – pp 213 -236.
211. [120] J.C. Jones, and E.P. Raynes, (1992)
[109] E. P. Raynes (1986) “The theory of supertwist "Measurement of the Biaxial Permittivities for
transitions,” Mol. Cryst. Liq. Cryst. Lett. vol. Several Smectic-C Host Materials used in
4, no. 1, pp1 – 8. Ferroelectric Liquid Crystals Devices." Liquid
[110] E.P. Raynes (2014) “ TN, STN and Guest Crystals, 11, (10) pp199-217.
Host Liquid Crystal Display Devices”, IN J. [121] J.C. Jones (2015) “On the biaxiality of smectic
W. Goodby et al (Eds.), “Handbook of Liquid C and ferroelectric liquid crystals”, Liquid
Crystals, Volume 8: Applications of Liquid Crystals 42, pp732-759.
Crystals”, Wiley VCH, Weinheim, Germany. [122] J.C. Jones, C.V. Brown and P.E. Dunn, (2000)
pp 3 -20. “The Physics of τVmin Ferroelectric Liquid
[111] M.J. Bradshaw and E.P. Raynes (1983) “The Crystal Displays”, Ferroelectrics, 246, pp 191
Elastic Constants and Electric Permittivities of – 201.
Mixtures Containing Terminally Cyano [123] J.C.Jones, M.J. Towler, and J.R. Hughes
Substituted Nematogens”, Mol. Cryst. Liq. (1993) "Fast, high contrast Ferroelectric
Cryst, 91, 145-148. Liquid Crystal Displays and the Role of
[112] T. Sergan, S. Jamal and J. Kelly, (1999). Dielectric Biaxiality" Displays, 14, (2) 86 -
“Polymer negative birefringence films for 93.
compensation of twisted nematic devices”, [124] M.J. Bradshaw, V. Brimmell and E.P. Raynes
Displays, 20, (5), pp 259–267. (1987) “A novel alignment technique for
[113] H.K. Bisoyi and S. Kumar, (2010) “Discotic ferroelectric smectics”, Liquid Crystals, 2,
nematic liquid crystals: science and (1), pp. 107 – 110
technology”, Chemical Society Reviews, 39, [125] M.H. Anderson, J.C. Jones, E.P. Raynes and
pp. 264–285. M.J. Towler (1991) "Optical Studies of Thin
[114] M.D. Tillin, M.J.Towler, K.A.Saynor and E.J. Layers of Smectic C Materials", J. Phys. D:
Beynon, (1998) “Reflective single polariser Appl. Phys., 24, pp 338 - 342.
low and high twist LCDs”, SID Tech. Digest, [126] J.C. Jones (2014) “Bistable Nematic Liquid
29, pp 311 – 314. Crystal Displays”, IN J.W. Goodby, P.J.
[115] S. Aftergut and H.S.Cole Jr., (1978) “Method Collins, T. Kato, C. Tschierske, H.F. Gleeson
for improving the response time of a display and P. Raynes, The Handbook of Liquid
device using the twisted nematic liquid crystal Crystals. Volume 8: Liquid Crystal Devices,
composition”, US Patent 4,143,947. Chapter 4, Wiley.
[116] V. G. Chigrinov, V. V. Belyaev, S. V. [127] V. Y. Reshetnyak, O.V. Shevchuk and M.
Belyaev, and M. F. Grebenkin, (1979) Osyptchu (2004) “Director profile in the in-
“Instabilities of Cholesteric Liquid Crystals in plane switching of nematic liquid crystal cell
an electric field”, Sov. Phys. JETP 50, (5), pp with strong director anchoring”,Proceedings
994-999. of SPIE, (July 2004), pp 283–292.
[117] I.C. Sage (1998) “Displays”, IN D. Demus, [128] M. Oh-E and K. Kondo (1997). The in-plane
J.W. Goodby, G.W.Gray, H.W. Spiess and V. switching of homogeneously aligned nematic
Vill, (Eds.) Handbook of Liquid Crystals,

94
 liquid crystals. Liquid Crystals, 22, (4), pp Doane (2005) “Reflective cholesteric displays:
379–390. From rigid to flexible,” J. SID, 13, (6), pp. 469
[129] H. Ki Hong, and H. Ho Shin (2008). Effects of – 474.
rubbing angle on maximum transmittance of [141] T. Schneider, G. Magyar, S. Barua, T. Ernst,
in‐plane switching liquid crystal display. N. Miller, S. Franklin, E. Montbach, D. J.
Liquid Crystals, 35, (2), 173–177. Davis, A. Khan and J. W. Doane (2008) ”A
[130] S. Jung, S. Jang, H. Park and W. Park, (2003). Flexible Touch-sensitive Writing Tablet,”
“Optimization of the Electro-Optical Proc. SID Int. Symp. Digest Tech. Papers, vol.
Characteristics of In-Plane-Switching Liquid 39, pp. 1840 – 1842.
Crystal Displays by Three-Dimensional [142] D. Coates (2009) “Low-power large-area
Simulation”, J. Korean Phys. Soc., 43 (5), cholesteric displays” Information Display, 25,
935–940. (3), pp. 16 – 19.
[131] D.K. Yang and S.T. Wu (2006) [143] D. W. Berreman and W. R. Heffner (1980)
“Fundamentals of Liquid Crystal Devices”, “New bistable liquid crystal twist cell,” Appl.
John Wiley and Sons, Chichester, UK. Phys. Lett. 25, pp 109-111.
[132] D.H. Kim, Y.J. Lim D.E. Kim, H. Ren, S.H. [144] H. Nomura, T. Obikawa, Y. Ozawa and T.
Ahn and S.H. Lee (2014) “Past, present, and Tanaka (2000) “Recent studies on multiplex
future of fringe-field switching-liquid crystal driving of BTN-LCDs,” J. SID, 8, (4) pp. 289
display”, Journal of Information Display, 15, – 294.
(2), pp99–106. [145] I. Dozov, M. Nobili and G. Durand (1997)
[133] J.L. Fergason (1964) “Liquid Crystals”, “Fast bistable nematic display using
Scientific American, 211, (Aug), pp 76 – 85. monostable surface switching,” Appl. Phys.
[134] H. Groskinsky (1968) “The Chameleon Lett. 70, (9), pp. 1179 – 1181.
Chemical” Life Magazine, 64, (2), pp 40 – 45. [146] S. Joly, P. Thomas, J. Osterman, A. Simon, S.
[135] P Bonnett, T. V. Jones and D. G. McDonnell, Lallemant, L. Faget, J.-D Laffitte, M. Irzyk, L.
(1989) "Shear Stress Measurements in Madsen, J. Angelé, F. Leblanc and P.
Aerodynamic Testing Using Cholesteric Martinot-Lagarde (2010) “Demonstration of a
Liquid Crystals", Liquid Crystals, 6, (3), pp technological prototype of an active-matrix
271-280. Binem liquid-crystal display,” J. SID., 18,
[136] G. H. Heilmeier and J. E. Goldmacher (1968) (12), pp. 1033 – 1039.
“A new electric-field-controlled reflective [147]
optical storage effect in mixed liquid crystal https://www.youtube.com/watch?v=j3Ob93X
systems,” Appl. Phys. Lett,. 13, (4), pp 132 – GM6A
133. [148] G. P. Bryan-Brown, C.V. Brown and J.C.
[137] W. Greubel, U. Wolf, H. Krüger, (1973) Jones (1995) "Bistable Nematic Liquid Crystal
“Electric field induced texture change in Device", Patent US 06249332.
certain nematic/cholesteric liquid crystal [149] J.C. Jones (2008) “The Zenithal Bistable
mixtures,” Mol. Cryst. Liq. Cryst. 24, pp 103 - Display: From concept to consumer,” J.
106. SID.16, (1) pp.143 - 154.
[138] W. Greubel (1974) “Bistability behavior of [150] E.L. Wood, G.P. Bryan-Brown, P. Brett, A.
texture in cholesteric liquid crystals in an Graham and J.C. Jones, (2000), “Zenithal
electric field,” Appl. Phys. Lett., 25, (1) pp. 5 – bistable device (ZBD) suitable for portable
7. applications”, Proc. SID, Int. Symp. Digest,
[139] D.-K. Yang, J. L. West, L.-C. Chien and J.W. Technical papers, 31, pp. 124 – 127.
Doane, (1994) “Control of Reflectivity and [151] J.C. Jones (2006), “Novel Geometries of the
Bistability in Displays using Cholesteric Zenithal Bistable Device,” Proc. SID Int.
Liquid Crystals,” J. Appl. Phys., 76, pp 1331 - Symp. Digest, Technical papers, vol. 37, pp.
1333. pp 1626 – 1629.
[140] A. Khan, I. Shiyanovskaya, T. Schneider, N. [152] J.C. Jones, S.M. Beldon and E.L. Wood
Miller, T. Ernst, D. Marhefka, F. Nicholson, (2003) “Gray scale in zenithal bistable
S. Green, G. Magyar, O. Pishnyak, J. W. displays: The route to ultra-low power color

95
 displays,” J. SID, vol. 11, no. 2, pp. 269 – PDLC)”, Annu. Rev. Mater. Sci. 30, pp 83 –
275. 115.
[153] J.C. Jones, P. Worthing, G Bryan-Brown and [167] D. Armitage, I. Underwood and S. –T. Wu
E. Wood (2003) “Transflective and single (2006) “Introduction to Microdisplays”, John
polariser reflective Zenithal Bistable Wiley and Sons Ltd., Chichester, UK.
Displays”, SID Int. Symp. Digest, Technical [168] M. S. Brennesholtz and E. H. Stupp (2008)
papers, 34, 14.1, pp 190 – 193. “Projection Displays”, John Wiley and Sons
[154] J.C. Jones (1999) “Bistable nematic liquid Ltd., Chichester, UK.
crystal device”, US Patent 7,371,362. [169] M. J. O’Callaghan and M.A. Handschy,
[155] T. J. Spencer, C. Care, R.M. Amos and J.C. (2001) “Ferroelectric liquid crystal SLMs:
Jones (2010) “A Zenithal bistable device: from prototypes to products", Proc. SPIE
Comparison of modelling and experiment”, 4457, Spatial Light Modulators: Technology
Phys. Rev. E., 82, 021702 pp1 – 13 and Applications, 31.
[156] G.P. Bryan-Brown, D.R.E. Walker and J.C. [170] E. E. Kriezis, L.A. Parry-Jones and S.J. Elston
Jones (2009) “Controlled Grating Replication (2003) Optical Properties and applications of
for the ZBD Technology”, SID Int. Symp. ferroelectric and antiferroelectric liquid
Digest, Technical papers 40, pp 1334 – 1337. crystals”, IN L. Vicari “Optical Applications
[157] C. Hilsum, (1976) “Electro-optic Device” UK of Liquid Crystals”, CRC Press, Cambridge,
Patent 1,442,360. UK.
[158] J.L. Fergason (1984) “Encapsulated liquid [171] N. Collings, T. Davey, J. Christmas, D. Chu
crystal and method”, US patent 4,435,047. and W. Crossland (2011) “The Applications
[159] D. Coates, (1995) “Polymer-dispersed liquid and Technology of Phase-Only Liquid Crystal
crystals”. J Mater. Chem. 5, pp 2063–2072. on Silicon Devices”, IEEE Journal of Display
[160] P. Dzraic (1995) “Liquid Crystal Technology, 7, (3), pp 112–119.
Dispersions”, World Scientific, Singapore. [172] D. Armitage, J.I. Thackara and W. D. Eades
[161] P. Dzraic (1986) “Polymer dispersed nematic (1989) “Photoaddressed Liquid Crystal Spatial
liquid crystal for large area displays and light Light Modulators”, Appl. Opt. 28, (22), pp
valves”, J. Appl. Phys., 60, (6) pp 2142 – 4763 - 4771
2148. [173] M. Stanley, P.B. Conway, S. Coomber, J.C.
[162] B.-G. Wu , J. H. Erdmann and J. W. Doane, Jones, D.C. Scattergood, C.W. Slinger, R.W.
(1989) “Response times and voltages for Bannister, C.V. Brown, W.A. Crossland, A.J.
PDLC light shutters”, Liquid Crystals, 5, (5), Travis, (2000) “A novel electro-optic
pp 1453-1465 modulator system for the production of
[163] Y. Asaoka, E. Satoh, K. Deguchi, T. Satoh, K. dynamic images from giga-pixel computer
Minoura, I. Ihara, S. Fujiwara, A. Miyata, Y. generated holograms”, Proceedings of SPIE,
Itoh, S. Gyoten, N. Matsuda and Y. Kubota 3956, 13 – 22.
(2009), “29.1: Polarizer-Free Reflective LCD [174] C. Slinger, C. Cameron and M. Stanley (2005)
Combined with Ultra Low-Power Driving “Computer-generated holography as a generic
Technology”, SID 09 Digest, pp 395–398. display technology”. Computer 38, 46–53.
[164] M. Kim, K.J. Park, S. Seok, J.M. Ok, H.T. [175] P.J. Bos and K.R. Koehler-Beran (1984) “The
Jung, J. Choe and D.H. Kim (2015) pi-cell: A fast liquid crystal optical switching
“Fabrication of Microcapsules for Dye-Doped device”, Mol. Cryst. Liq. Cryst. 113, pp 329-
Polymer-Dispersed Liquid Crystal-Based 339.
Smart Windows”, ACS Applied Materials and [176] J. L. Fergason and A. L. Berman (1989) “A
Interfaces, 7(32), 17904–17909. push/pull surface-mode liquid-crystal shutter:
[165] For example, Technology and applications”, Liquid
http://intelligentglass.net/categories/switchabl Crystals, 5, (5), 1397–1404.
e-glass/ [177] X. Y. Yu and H. S. Kwok, (2004) “Bistable
[166] T. J. Bunning, L.V. Natarajan, V.P. Tondiglia bend-splay liquid crystal device”, Appl. Phys.
and R.L. Sutherland, (2000) “Holographic Lett., vol. 85, (17), pp 3711 – 3713
Polymer Dispersed Liquid Crystals (H-

96
[178] E.P. Raynes (1975) “Optically Active [190] Z. Ge, S. Gauza, M. Jiao, H. Xianyu and S.T.
Additives in Twisted Nematic Devices”, Rev. Wu, (2009) “Electro-optics of polymer-
de Phys. Appl., 10, pp. 117 – 120. stabilized blue phase liquid crystal displays”,
[179] Y. Yamaguchi, T. Miyashita and T. Uchida, Appl. Phys. Lett., 94, (10), pp 9–12.
(1993) "Wide-viewing-angle display mode for [191] L. Rao, Z. Ge, S.-T Wu and S. H. Lee, (2009)
the active-matrix LCD Using bend-alignment “Low voltage blue-phase liquid crystal
liquid-crystal cell", SID 93 Digest, pp. 273- displays”, Appl. Phys. Lett., 95, (23), 231101.
276. [192] M.D.A. Rahman, S. Mohd Said and S.
[180] P. J. Bos and J. A. Rahman, (1993) “An Balamurugan (2015) “Blue phase liquid
optically ’self-compensating’ electro-optical crystal: strategies for phase stabilization and
effect with wide angle of view”, SID 93 device development”, Sci. Tech. Adv. Mat., 16,
Digest, pp 277-280. (3), 033501.
[181] C. G. Jhun, M. S. Park, P. K. Son, J. H. Kwon, [193] J. Patel and R.B. Meyer (1987) “Flexoelectric
J. Yi and J. S. Gwag, (2011) “Transflective electro-optics of a cholesteric liquid crystal”,
Liquid Crystal Display with Pi-cell”, Mol. Phys. Rev. Lett. 58, 1538
Cryst. Liq. Cryst., 544, 88 – 94. [194] P. Rudquist, M. Buivydas, L. Komitov and
[182] M.J. Bradshaw and E.P. Raynes (1986) S.T. Lagerwall, (1994) “Linear electro-optic
“Smectic Liquid Crystal Devices”, US Patent effect based on flexoelectricity in a cholesteric
4,969,719. with sign change of dielectric anisotropy”, J.
[183] J.S. Patel (1992) “Ferroelectric liquid crystal Appl. Phys., 76, (12), pp 7778–7783.
modulator using twisted smectic structure”, [195] K. L. Atkinson, S. M. Morris, M. M. Qasim,
Appl. Phys. Lett. 60, pp 280–282 F. Castles, D. J. Gardiner, P. J. W. Hands, S.
[184] A.D.L. Chandani, E. Gorecka, Y. Ouchi and S. Choi, W. –S. Kim and H. J. Coles (2012)
H. Takezoe H (1989) “Antiferroelectric chiral “Increasing the flexoelastic ratio of liquid
smectic phases responsible for the tristable crystals using highly fluorinated ester-linked
switching in MHPOBC”. Jpn. J. Appl. Phys. bimesogens” Phys. Chem. Chem. Phys. 14,
Lett., 28, pp L1265–L1268 (47) pp 16377-16385.
[185] B.I. Ostrovsky, A.Z. Rabinovich and V.G. [196] L. Komitov, G.P. Bryan-Brown, E.L. Wood
Chigrinov (1980) “Behavior of ferroelectric and A.B.J. Smout, (1999) “Alignment of
smectic liquid crystals in electric field” IN: L. cholesteric liquid crystals using periodic
Bata (ed.) “Advances in liquid crystal research anchoring”, J. Appl. Phys., 86, (7) pp 3508 –
and applications”, Pergamon Press, Oxford, 3511.
UK, p 469 – 482. [197] G. Carbone, D. Corbett, S.J. Elston, P.
[186] S. Garoff and R.B. Meyer (1977) Raynes, A. Jesacher, A. R. Simmonds and M.
“Electroclinic effect at the A-C phase change Booth, (2011) “Uniform Lying Helix
in a chiral smectic liquid crystal”, Phys. Rev. Alignment on Periodic Surface Relief
Lett. 38, pp 848–851. Structure Generated via Laser Scanning
[187] D. Coates and G.W. Gray (1973) “Optical Lithography”, Mol. Cryst. Liq. Cryst., 544,
studies of the amorphous liquid-cholesteric (1), pp 37-49.
liquid crystal transition: the blue phase”, Phys. [198] C.C. Tartan, P.S. Salter, M.J. Booth, S.J.
Lett. A 45, pp.115–116. Morris, and S. J. Elston, (2016) “Localised
[188] Y. Haseba and H. Kikuchi (2006) “Electro- polymer networks in chiral nematic liquid
optic effects of the optically isotropic state crystals for high speed photonic switching”, J.
induced by the incorporative effects of a Appl. Phys., 119, (18), 183106.
polymer network and the chirality of liquid [199] N.V. Kukhtarev, (1978) “Cholesteric liquid
crystal”, J. SID, 14, (6) pp. 551-556 crystal laser with distributed feedback”, Soviet
[189] H. Kikuchi, M.Yokota, Y. Hisakado, H. Yang Journal of Quantum Electronics, 8, (6), 774-
and T. Kajiyama, (2002) “Polymer-stabilized 776.
liquid crystal blue phases”, Nature Materials, [200] V. I. Kopp, B. Fan, H. K. M. Vithana and A.
1 (1), 64–68. Z. Genack, (1998) “Low-threshold lasing at

97
 the edge of a photonic stop band in cholesteric [205] H. Coles and S. Morris, (2010) “Liquid-crystal
liquid crystals. Opt. Lett. 23, 1707–1709. lasers”, Nature Photonics, 4, (10), pp 676–
[201] B. Taheri, A. Munoz, P. Palffy-Muhoray and 685.
R. Twieg, (2001) “Threshold lasing in [206] H.-Y Chen, S. –F. Lu, P.-H., Wu, and C.S.
cholesteric liquid crystals”, Mol. Cryst. Liq. Wang, C.-S. (2016) “Transflective BPIII mode
Cryst. 358, 73–82. with no internal reflector” Liquid Crystals,
[202] E. Yablonovitch (1987) “Inhibited 8292(August 2016), 1–6.
Spontaneous Emission in Solid-State Physics [207] Q. Li, (2012) “Liquid Crystals Beyond
and Electronics”, Phys. Rev. Lett. 58, (20) pp Displays: Chemistry, Physics, and
2059-2062. Applications”, John Wiley and Sons, New
[203] M. Uchimura, Y. Watanabe, F. Araoka,, J. Jersey. ISBN: 978-1-118-07861-7
Watanabe, H. Takezoe, and G.I. Konishi, [208] WH. De Jeu (Ed.) (2012) “Liquid Crystal
(2010) “Development of laser dyes to realize Elastomers: Materials and Applications”,
low threshold in dye-doped cholesteric liquid Springer, US. ISBN: 978-3-642-31581-7
crystal lasers”. Advanced Materials, 22, (40), [209] R.J. Bushby, S.M. Kelly and M. O’Neill (Eds)
pp 4473–4478. (2012) “Liquid Crystalline Semiconductors:
[204] D.J. Gardiner, W. -K.Hsiao, S. M. Morris, Materials, properties and application”,
P.J.W. Hands, T.D. Wilkinson, I.M. Hutchings Springer-Verlag Berlin Heidelberg. ISBN
and H.J. Coles, (2012) “Printed photonic 978-90-481-2873-0
arrays from self-organized chiral nematic [210] R.J. Carlton, J.T. Hunter, D. S. Miller, R.
liquid crystals”, Soft Matter, 8 (39), 9977 – Abbasi, P.C. Mushenheim, L. N. Tan and N.
9980. L. Abbott (2013) “Chemical and Biological
Sensing using Liquid crystals”, Liq. Cryst.
Rev., 1, (1) pp 29 – 51.

Recommended further reading

R. H. Chen (2011) “Liquid Crystal Displays”, John

Wiley and Sons, New Jersey.

N. Koide (2014) “The Liquid Crystal Display Story:

50 Years of Liquid Crystal R&D that lead the Way to
the Future”, Springer, Japan.

D. Dunmur and T. Sluckin (2011) Soap, Science, and

Flat-screen TVs: a history of liquid crystals, Oxford
University Press ISBN 978-0-19-954940-5.

98