FEATURE ARTICLE
`
`www.rsc.org/materials | Journal of Materials Chemistry
`
`LCD alignment layers. Controlling nematic domain properties{
`
`Johan Hoogboom,{* Theo Rasing, Alan E. Rowan and Roeland J. M. Nolte*
`
`Received 25th July 2005, Accepted 2nd November 2005
`First published as an Advance Article on the web 24th November 2005
`DOI: 10.1039/b510579j
`
`From simple pocket calculators, to mobile telephones and LCD TV, over the past few decades
`devices based on liquid crystal display technology have proliferated into just about all conceivable
`aspects of everyday life. Although used in cutting edge technology, it is surprising that a vital part
`in the construction of such displays relies essentially on a process invented almost 100 years ago.
`This essential part, the alignment layer, dictates the macroscopic uniform alignment of liquid
`crystalline molecules (mesogens) near its surface. The current method for manufacturing such
`layers is the mechanical rubbing of spin coated polymers with a piece of velvet cloth. This very
`successful method is still at the basis of the production process of even the largest displays
`currently manufactured in industry. Unfortunately, the construction of ever larger displays with
`this technique is becoming a technological nightmare for engineers. Therefore, over the past
`decades, many alternatives to rubbing have been explored. This review will focus on advances
`towards achieving one of the most important goals in LCD technology: attaining rational control
`over the properties of nematic liquid crystal domains.
`
`Introduction
`
`From liquid crystals to LCDs
`
`The first observations of liquid crystalline behaviour, although
`not recognized as such at the time, were reported as far
`back as the 1850’s. Around that time, biologists Virchow,
`Mettenheimer and Valentin were studying nerve fibres. They
`found that a fluid substance from the nerve core, when left in
`water, exhibited strange behaviour if viewed in polarized light.
`Not recognizing the discovery of a new phase of matter, the
`
`Organic Chemistry, Institute for Molecules and Materials, Radboud
`University Nijmegen, Toernooiveld 1, NL-6525ED, Nijmegen,
`The Netherlands. E-mail: J.Hoogboom@science.ru.nl;
`R.Nolte@science.ru.nl; Fax: (+31)-24-2652929; Tel: (+31)-24-3652143
`{ This paper is partly based on a presentation given by R.N. at the
`ACS National Meeting, San Diego, CA, March 2005.
`{ Current address: Department of Organic Chemistry, Massachusetts
`Institute of Technology, 77 Massachusetts Avenue, Bldg 18-143,
`Cambridge, MA, 02139, USA. E-mail: hoogboom@mit.edu
`
`substance was labelled a ‘‘living crystal’’. Years later, the
`official discovery of liquid crystals was credited to Otto
`Lehmann and Friedrich Reinitzer. In 1888, the Austrian
`chemist Reinitzer investigated the phase transitions of various
`compounds using a polarizing microscope fitted with a heating
`stage. Over the course of his study, Reinitzer discovered that
`cholesteryl benzoate changed from a clear to a cloudy liquid
`before crystallizing. Unfortunately, he attributed the apparent
`occurrence of
`two melting points to an imperfect phase
`transition. Still puzzled, a year later Reinitzer wrote a letter
`with his findings to the German physicist Otto Lehmann,
`who was an expert in crystal optics. After conducting similar
`experiments, Lehmann suggested that the cloudy fluid was a
`new phase of matter. Subsequently,
`in a landmark paper
`entitled ‘‘U¨ ber fliessende Krystalle’’, he described the dis
`covery of a class of crystals which were so soft that one could
`call them liquid, coining the term ‘‘liquid crystal’’.1
`For several decades, liquid crystals were considered to be
`worthy of only scientific study, as they seemed to have little
`
`Johan Hoogboom completed his
`PhD in supramolecular chemis
`try and solid state physics in
`2004 at the Radboud University
`Nijmegen under the supervision
`of Roeland Nolte, Theo Rasing
`and Alan Rowan. He is cur
`rently a postdoctoral researcher
`at the Massachusetts Institute
`of Technology, Cambridge,
`USA, in the group of Timothy
`M. Swager. His
`research
`interests include the use of self
`assembling systems to induce
`liquid crystal alignment and
`LCD based biosensors.
`
`Theo Rasing received his PhD
`from Nijmegen in 1982 on
`incommensurate crystals and
`worked at
`the University of
`California at Berkeley between
`1982 1988. He is full professor
`in experimental physics at the
`Radboud University Nijmegen,
`working on the understanding
`and control of
`the relation
`between structure and pro
`perties
`of
`nanoscopic
`and
`molecular materials, with an
`emphasis on very small length
`scales (nm) and short
`time
`scales (fs).
`
`Theo Rasing
`
`Johan Hoogboom
`
`This journal is ß The Royal Society of Chemistry 2006
`
`J.Mater.Chem., 2006, 16, 1305 1314 | 1305
`
`Page 1 of 10
`
`Tianma Exhibit 1014
`
`

`
`merit with respect to industrial applicability. In fact, already
`before World War II, most scientists thought that the field of
`liquid crystals had matured and even been completed when
`a mathematical basis for the study of liquid crystals was
`provided by Oseen and Zocher.2 This all changed, however, in
`the sixties, when the liquid crystal display was invented.
`A crude display based on liquid crystals was first suggested
`by Richard Williams of RCA labs (Radio Company of
`America) in 1963 and developed into a device several years
`later.3 Unfortunately, his device required operating tempera
`tures well above 100 uC. Because of this, its power consump
`tion was too high and its display quality too low. The company
`soon developed mixtures that operated at room temperatures
`and continued research well
`into the seventies. A much
`improved version of the liquid crystal display, dubbed the
`twisted nematic liquid crystal display, was developed by James
`Fergason of Kent State University, Ohio. It was patented in
`1969 by his company ILIXCO, with the patent being applied in
`1971.4 Its principle was described by Schadt and Helfrich,5a
`who applied their subsequent patent5b several months earlier
`than Fergason.5c
`
`Twisted-nematic liquid crystal displays
`
`The twisted nematic liquid crystal display is still the most
`abundantly proliferated form of liquid crystal based device.
`Vital to its operating principle were two independent observa
`tions on liquid crystalline behaviour. When combined, these
`two curiosities, as they were deemed at the time, formed the
`basis of liquid crystal display technology still employed today.
`The first discovery was reported as far back as 1911, when
`Mauguin had made several observations that liquid crystal
`domains could be aligned by placing them in contact with a
`crystal surface.6 Most brilliant, however, was his observation
`that if a liquid crystalline compound was placed between two
`aligning surfaces of different orientations, the director of the
`liquid crystalline compound (i.e. the average direction of the
`molecular axis of the mesogenic molecules in a domain)
`smoothly followed the transition, rotating from one surface to
`the other.6,7 Also, during World War II, Chatelain showed
`
`that liquid crystal compounds could be uniformly aligned
`by placing them between two glass plates and subsequently
`rubbing the plates along each other.8
`The other discovery which contributed to the invention of
`the liquid crystal display was made by Freedericksz in the
`thirties. In elegant experiments, he showed that the orientation
`of liquid crystal molecules could be influenced by the applica
`tion of an electric field.9 If the field strength was higher
`than a certain threshold value, the liquid crystal molecules
`would orient themselves with their dipoles parallel to the field,
`abandoning their previous orientation. In honour of
`its
`discoverer, such a transition is now called a Freedericksz
`transition.
`foresee any
`time, however, did not
`that
`Engineers at
`immediate applicability of these discoveries. This certainly
`changed when they were combined to construct the twisted
`nematic liquid crystal display.4,5 Its operating principle is as
`brilliant as it is simple (Fig. 1). A liquid crystalline compound,
`usually a calamitic nematic liquid crystal,
`is sandwiched
`between two conductive surfaces mostly indium tin oxide,
`ITO covered with an alignment layer, which induces uniform
`planar alignment of the liquid crystal near the surface. Instead
`of orienting the alignment directions of the two surfaces in a
`parallel fashion, they are rotated 90u with respect to each
`other. As Mauguin showed, this prompts the director of the
`liquid crystal between the aligning surfaces to also rotate 90u
`going from one surface to the other (Fig. 1a). If the distance
`between the two plates is much larger than the wavelength of
`light transmitted through the cell, any incident light will also
`have its polarization direction rotated by 90u while traversing
`the cell.10 By adding polarizers to the cell, parallel to the liquid
`crystal alignment direction on their respective surfaces, a
`twisted nematic liquid crystal cell is made which is capable of
`allowing light to pass crossed polarizers (off state, Fig. 1a).
`This property can be switched by applying an electric field
`between the two conductive alignment layers. If the field
`strength exceeds the Freedericksz transition, the liquid crystal
`molecules will align parallel to the field direction (Fig. 1b). As
`the director of the liquid crystal no longer smoothly changes
`
`Alan Rowan completed his
`PhD in physical organic chemi
`stry in 1991 at the University
`of Liverpool, England. After a
`period of postdoctoral research
`at
`the University of Otago,
`Dunedin, New Zealand, he
`returned to Europe and became
`an assistant professor at the
`Radboud University Nijmegen
`in the group of Roeland Nolte.
`In 2004, he became a full
`professor in molecular mate
`rials. His scientific interests
`are the design and construction
`of supramolecular assemblages
`possessing catalytic and electronic properties.
`
`Alan Rowan
`
`Roeland Nolte received his PhD in physical organic chemistry
`
`from the University of Utrecht
`(1973), where he stayed and
`became assistant professor and
`then associate professor. In
`1981, he was a visiting scientist
`at UCLA in the group of
`Donald J. Cram. In 1987 he
`m o v e d t o t h e R a d b o u d
`University Nijmegen and
`became full professor of
`organic chemistry. Since 1994
`he has also been a professor of
`supramolecular chemistry at
`the Eindhoven University of
`Technology.
`In 2003,
`the
`Royal Dutch Academy of
`Sciences awarded him an Academy Professorship. His principal
`research interest is supramolecular chemistry, focusing on the
`design of catalysts and molecular materials.
`
`Roeland Nolte
`
`1306 | J.Mater.Chem., 2006, 16, 1305 1314
`
`This journal is ß The Royal Society of Chemistry 2006
`
`Page 2 of 10
`
`

`
`Fig. 1 Operating principle of the twisted nematic LCD; a) off state;
`b) on state.
`
`between the two alignment surfaces, any incident light will
`not have its polarization direction altered. As such, it is no
`longer able to exit the cell through the second polarizer, which
`causes the cell to appear dark (on state, Fig. 1b). When the
`electric field is switched off, the liquid crystal reverts back to
`its original structure and the cell is once again capable of
`transmitting light.
`The almost simultaneous development of micro electronics
`in the sixties and seventies provided small switching elements,
`which could regulate the electric field strength in a single cell.
`When individual
`twisted nematic cells, also called pixels,
`were connected in two dimensional arrays, images could be
`produced by their
`coordinated switching,
`enabling the
`construction of liquid crystal displays.11
`
`Liquid crystal alignment: rubbing
`
`One crucial element of the twisted nematic liquid crystal
`display is the alignment layer, which is used to induce uniform
`and planar alignment of the liquid crystal molecules near the
`surface. Until the invention of the twisted nematic cell, not
`much research had been done on the subject. Needless to say
`that the industrial production of this type of display spurred
`research into alignment surfaces.11a,12
`Surprisingly, the process first applied to induce uniform
`alignment in the late sixties is very similar to the process used
`today. Simply called ‘‘rubbing’’, it had already been developed
`in the twenties.13 The method entails the unidirectional
`rubbing of a spin coated polymer surface with a piece of cloth
`(Fig. 2a). As polymers, usually polyimides are used (Fig. 2b).
`Not only do they exhibit the thermal stability needed for
`application purposes, but they are also relatively cheap and
`easy to apply. In addition, their structure offers multiple
`possibilities for molecular interactions with the liquid crystal
`molecules in the display. The contact between the cloth and
`polymer layer creates microscopic grooves on the surface of
`the polymer (Fig. 2c) and aligns (substituents of) individual
`
`rubbing; b) polyimide
`Fig. 2 a) Simplified representation of
`1
`PI 2555; c) AFM image of a rubbed polyimide surface.
`Pyralin
`Reused with permission from ref. 18c. Copyright 2001, American
`Institute of Physics; d) simplified representation of the alignment of
`liquid crystals on a rubbed surface.
`
`polymer chains. When a liquid crystalline compound is
`interacted with such a surface,
`the mesogen adopts a
`macroscopic alignment (Fig. 2d).
`The alignment mechanism associated with rubbing has been
`debated for decades and can be roughly divided into two
`contributing components. The first component is a purely
`physical model, which holds that only the formation of
`microgrooves on the surface is responsible for liquid crystal
`alignment. The foundations of
`this theory were laid by
`Berreman, in his elastic continuum theory of 1972.14 In it,
`molecular interactions between the alignment layer and the
`liquid crystal molecules are assumed not to play a role in the
`alignment mechanism. Instead, the liquid crystal molecules
`orient themselves parallel to the microgrooves, which are
`formed as a result of the rubbing. The driving force for this is
`the minimization of the so called elastic distortion energy.
`Simply put, this means that the alignment of liquid crystal
`molecules in and around the microgrooves, with their director
`parallel to the direction of the groove, results in the lowest
`surface energy and hence the most stable conformation
`(Fig. 3a).15
`The other component is a molecular model, which holds that
`the factors responsible for alignment are mainly based on
`
`Fig. 3 a) Berreman model of alignment; b) molecular model of
`alignment.
`
`This journal is ß The Royal Society of Chemistry 2006
`
`J.Mater.Chem., 2006, 16, 1305 1314 | 1307
`
`Page 3 of 10
`
`

`
`molecular interactions. First suggested by Geary,16 this theory
`has recently won ground with other experimental confirma
`tions.17 The theory hinges on the notion that
`individual
`polymer chains and/or their substituents are aligned by the
`unidirectional rubbing of the sample. This results in an ordered
`surface which is capable of engaging in anisotropic molecular
`interactions with liquid crystal molecules in the vicinity
`(Fig. 3b). This causes the unidirectional alignment of the first
`layers of liquid crystal molecules, after which the liquid crystal
`bulk follows epitaxially.
`Research into the mechanism of liquid crystal alignment has
`intensified over the past decades and has led to a significant
`increase in the understanding of
`the factors underlying
`liquid crystal alignment on ordered surfaces. For polymer
`coated surfaces, the current consensus is that the predominant
`alignment mechanism can be ascribed to the generation of a
`statistically important number of similarly oriented bonds in
`the top layer of the polymer,18a,c as described by the molecular
`model of alignment (vide supra).
`Another vital property of the alignment layer is the ability to
`induce a tilt in the liquid crystal molecules near the surface.
`Also called the pretilt angle, this quantity is responsible for
`the uniform texture of
`the liquid crystal display during
`operation.11b A detailed discussion, however, is beyond the
`scope of this review.
`The rubbing technique has several important advantages. It
`is relatively straightforward, easy and cheap. In addition, the
`resulting alignment surfaces are very stable and the techniques
`are already widely applied in industry.18a,b There are, however,
`also some serious drawbacks associated with this technique,
`which are all the result of the direct contact between the
`alignment
`layer and the velvet cloth during manufactur
`ing.18a,19 For instance,
`the rubbing creates debris, which
`remains on the alignment layer and interferes with uniform
`liquid crystal alignment. In addition, as parts of the layer
`are rubbed off, the contrast ratio differs from site to site.
`Furthermore,
`the rubbing of
`the two materials creates
`electrostatic charges, which interfere with the switching of
`the display and destroy electronic circuits. All these drawbacks
`result in the creation of faulty pixels. In order to minimize the
`harmful effects of these drawbacks, the entire process is
`performed in a clean room, devoid of dust and charges,
`making this a very labour intensive process. Also, as the need
`for larger displays is increasing, e.g. with the onset of large
`LCD TV, the process requires ever more scaling up. Although
`factories are currently capable of processing plates of more
`than 4 m2, building a rubbing machine for even larger
`substrates, as demanded by market pressure, would be ‘‘a
`nightmare for engineers to build and operate’’.19 As such,
`research into novel ways of aligning liquid crystals has become
`increasingly important over the past decades.
`
`Alternatives to rubbing
`
`Photoalignment
`
`One of the most studied alternatives for rubbing is the
`photoalignment technique, which provides directional control
`over mesogens using polarized light.20 As such, this is a
`completely non contact technique, eliminating most of the
`
`Chart 1
`
`problems associated with rubbing (vide supra). The molecules
`which are employed in this technique invariably contain
`double bonds, e.g. azobenzenes (Chart 1), which can be
`isomerized with light.
`Early work on photoalignment conducted in the seventies
`mainly focused on the disruption of an already present
`(aligned) mesophase when the chromophore dissolved in it in
`small quantities underwent a shape change going from the
`trans to the cis conformation during irradiation.21 In effect,
`the photoisomerization caused a phase transition of the liquid
`crystal medium in the LCD, which enabled the switching
`between two states with different mesogenic orientations,
`resembling the operating principle of twisted nematic displays
`(Fig. 4). Upon irradiation with the proper wavelength(s) of
`polarized light, azobenzenes dissolved in a liquid crystal
`medium undergo a trans cis isomerization, followed by (fast)
`thermal relaxation, until the dipole of the double bond is
`perpendicular to the polarization direction of the incident
`light, which occurs after about 10 minutes. This results in a
`cooperative, uniform alignment of both the chromophores
`and mesogens. At higher irradiation times, another process
`becomes dominant: optical pumping. As
`irradiation is
`sustained, the concentration of the less stable cis form will
`increase with time. The shape change upon going from trans to
`
`Fig. 4 Operating principle of one of the first photoaligned displays
`(see text).
`
`1308 | J.Mater.Chem., 2006, 16, 1305 1314
`
`This journal is ß The Royal Society of Chemistry 2006
`
`Page 4 of 10
`
`

`
`after which cooling the phase to below the Tg (the temperature
`below which the system is thermally stable) would preserve
`the image.20a,25 The technique, however, was deemed too
`laborious and prone to errors. This left most azobenzene based
`systems ill suited for practical applications, which spurred
`research into other types of photoalignment
`(vide infra).
`Recently though, several azobenzene based systems have been
`reported which can be switched between two stable photo
`induced states, which might be used in future devices.26
`One of the first alternatives to azobenzenes were molecules
`based on cinnamic acid. In contrast to azobenzenes, these
`molecules
`exhibit
`two distinct photo induced processes
`(Scheme 1). Besides the photo induced trans cis and cis trans
`isomerizations and thermal relaxations, these molecules are
`capable of undergoing [2+2] photocycloaddition reactions
`upon irradiation with UV light.
`By irradiating a spin coated layer of poly(vinyl p methoxy
`cinnamate) with polarized (UV) light, Schadt et al. were able
`layer for LCD purposes.27 This
`to create an alignment
`landmark paper led to a boom in photoalignment research,
`which resulted in numerous papers and patents on the
`subject.28 Although the actual mechanism behind the align
`ment of liquid crystals in photoaligned systems is still under
`some debate, increasingly more insight has been attained into
`the two photochemical processes occurring during photo
`induced ordering and their influence on mesogenic ordering.
`Drevensˇek Olenik et al. showed that the interaction between
`liquid crystal molecules and photoaligned cinnamoyl contain
`ing alignment layers is mainly governed by the p p interactions
`between the mesogens and the trans cinnamoyl groups.28e
`In addition, photocycloaddition occurs preferentially between
`cis cinnamoyl groups, resulting in the truxilate structure
`depicted in Scheme 1.28e,g These results imply a decrease of
`the interaction energy (also called anchoring energy) between
`the liquid crystal molecules and the surface during irradiation,
`as the photoprocesses entail a depletion of the trans species. As
`such, the final surface ordering and its stability are provided by
`
`Fig. 5 Effect of irradiation of a chromophore containing surface with
`linearly polarized light
`(out of plane reorientation or command
`surface effect).
`
`cis is rather drastic, with the latter possessing a small cavity
`(Fig. 4). In addition, the cis form has more affinity for the
`surface of the cell than the trans form. The combination of
`these two effects causes the alignment of the mesogens in the
`direction parallel to the cell walls (Fig. 4). In 1991, Gibbons
`et al. used this principle to construct the first photoaligned
`LCD with a rewritable surface ordering.22
`Another type of photoaligned display based on azobenzenes
`was developed by Ichimura et al. in 1988.20a,23 Instead of
`using the photochromic molecule as a dopant, it was chemo
`or physisorbed to the surface of the display. The switching of
`these tethered azobenzenes by polarized light also induced
`a change in the ordering of
`the mesogens, going from
`homeotropic to planar alignment (out of plane reorientation,
`Fig. 5). Because of their ability to switch the orientation of
`mesogens, these surfaces were named ‘‘command surfaces’’.
`Dependent on the type of azobenzene used, the switching time
`and interaction between surface and mesogen could be
`controlled.24
`Azobenzene tethered command surfaces are capable of two
`types of photo reorientation of mesogenic molecules. Besides
`the out of plane reorientation shown in Fig. 5, the surfaces
`are also capable of
`inducing in plane reorientations of
`mesogens (Fig. 6). The double bond of an azobenzene only
`stops isomerizing when its dipole moment is perpendicular to
`the polarization of the incident light. This implies that a
`rotation of the polarization direction of the incident light will
`entail the reorientation of the azobenzenes on the surface. In
`turn, the mesogenic molecules will follow the newly written
`order, which causes controllable in plane movements of the
`mesogens (Fig. 6).20
`the systems based on azobenzenes
`Unfortunately, all
`showed one major drawback: they were not stable. Due to
`random cis trans relaxations and in plane thermal motions,
`the written surface ordering was lost over a period of time,
`varying from seconds to months. Several methods were devised
`to stabilize the ordering. For instance, the use of a smectic
`liquid crystal instead of a nematic one allowed imprinting of
`the surface ordering of the azobenzenes into the smectic phase,
`
`Fig. 6 Effect of changing the polarization direction of incident light
`on liquid crystal alignment (in plane reorientation).
`
`Scheme 1 Schematic representation of the photochemistry of cinna
`moyl derivatives.
`
`This journal is ß The Royal Society of Chemistry 2006
`
`J.Mater.Chem., 2006, 16, 1305 1314 | 1309
`
`Page 5 of 10
`
`

`
`Scheme 2 Photoisomerization of: a) a hydrazono b ketoesters; b) spiropyran merocyanine systems.
`
`This ordering effect could be attributed to two competing
`processes, which occur simultaneously. The first process is
`related to the large p surface of 1, compared to similar
`cinnamates. As a result of this large p surface, the compound
`oligomerizes in solution,
`forming rigid cyclic and linear
`oligosiloxanes (Fig. 7a). As the oligomer grows, it becomes
`increasingly less soluble, resulting in its precipitation. The
`second process
`involves
`the covalent grafting of
`these
`
`Chart 3
`
`two separate effects. First, the photocycloaddition provides a
`rigid framework to stabilize the ordering induced by mainly
`the trans conformation of the cinnamoyl derivative. Second,
`the photocycloadduct itself is a source of surface ordering, as
`one isomer is preferentially formed.
`Besides derivatives of cinnamic acid, various other mole
`cules containing a double bond have been used in the
`e.g. a hydrazono
`construction of photoaligned LCDs,
`b ketoesters29 (Scheme 2a) which are thermally stable at
`low operating temperatures due to intramolecular hydrogen
`bonding, and spiropyran merocyanine systems30 (Scheme 2b),
`which might be used in devices.
`One other breakthrough in photoalignment was reported
`in 1996 by Schadt et al., who used a spin coated polymer
`containing a coumarin type dye (Chart 2).31 The irradiation of
`these polymers under an angle with respect to the surface
`resulted in the formation of an alignment layer which, when
`used in an LCD, offered a much improved field of view,
`compared to cinnamates.
`
`Hierarchical rubbing
`
`Another alternative to rubbing, based solely on hierarchical
`self assembly processes from solution, was reported by us
`in 2003.32 When an ITO plate was immersed into a solution
`of the naphthalene functionalized siloxane 1 (Chart 3), an
`alignment surface was generated without the use of external
`stimuli (e.g. rubbing or photoalignment).
`
`Fig. 7 Schematic representation of the construction of an LCD with
`a self assembled alignment layer; a) oligomerization of 1; b) covalent
`grafting of oligomers of 1 to the ITO surface containing nanogrooves,
`thereby amplifying their ordering over several orders of magnitude;
`c) interaction of the alignment layer with liquid crystals; d) LCD
`construction. Reproduced with permission from ref. 32. Copyright
`Wiley VCH 2003.
`
`Chart 2
`
`1310 | J.Mater.Chem., 2006, 16, 1305 1314
`
`This journal is ß The Royal Society of Chemistry 2006
`
`Page 6 of 10
`
`

`
`precipitating oligosiloxanes onto the ITO surface (Fig. 7b).
`By very carefully tuning the rates of both processes, i.e. by
`ensuring that oligomers with the right size precipitate and graft
`at the right time during layer formation, a highly ordered
`surface could be constructed, which was shown to be capable
`of inducing uniform surface alignment of liquid crystalline
`molecules (Fig. 7c). These self assembled alignment surfaces
`could be used in the construction of standard twisted nematic
`LCDs (Fig. 7d) and rivalled industrially manufactured displays
`in terms of interaction energy between the alignment layer and
`the liquid crystalline matrix. Surprisingly, the overall direction
`of the surface ordering and hence the liquid crystal align
`ment was determined by the ITO plate, which was shown to
`contain parallel groove like structures of nano sized dimen
`sions, which formed during its manufacturing. During the
`self assembly process, the size and ordering of these nano
`grooves was amplified over several orders of magnitude. The
`whole process could be performed under laboratory condi
`tions, eliminating the need for clean room facilities
`in
`alignment layer construction.
`
`Other techniques
`
`the most well studied
`remains
`Although photoalignment
`alternative to rubbing, during the past decades several other
`promising techniques have been developed. For instance, the
`group of Abbott reported a method for controlling the
`alignment of liquid crystals by using gold sputtered plates
`covered with a self assembled monolayer (SAM) of chemi
`sorbed thiols (Fig. 8).33 They elegantly showed that the
`alignment properties of the layer could be fine tuned by
`changing the deposition conditions and nature of the thiol.33a
`The latter was varied by using thiols with different alkyl tails.
`Experiments revealed an odd even effect of the length of the
`alkyl tail on the alignment properties. When odd numbered
`alkyl tails were used, the liquid crystal molecules would align
`parallel to the surface, but perpendicular to the direction of
`the deposition of the gold layer (Fig. 8a). In contrast, when
`even numbered alkyl tails were used, liquid crystal molecules
`would align parallel
`to both the surface and deposition
`direction of the gold layer (Fig. 8b). When a mixture of even
`and odd numbered tails was used, the liquid crystal molecules
`adopted a homeotropic orientation (Fig. 8c). Usually,
`
`alignment layers based on gold are not suited for application
`purposes, as they are opaque to visible light. Abbott et al. were
`able to overcome this drawback by using thin gold sputtered
`layers of around 100 A˚ thick.33a Although these layers were
`capable of transmitting incident light, they still displayed a
`poor maximum transmission, preventing their use in industrial
`manufacture.34 Bastiaansen et al. have recently overcome this
`problem by using an ultra thin gold layer of 1 100 A˚ thick,
`which decreased the light absorption and shifted the absorp
`tion band of the gold layer into the UV.34
`In 2001, a group led by Chaudhari of IBM showed that
`liquid crystal alignment could be achieved by the bombard
`ment of a carbonaceous surface with a low energy ion beam at
`a glancing angle.35 By linearly moving the surface in front of
`an argon sputtering gun, they were able to create an alignment
`layer without the use of a polymer. In pilot line manufacturing,
`this method has already yielded 15 and 22 inch displays for
`laptop and desktop computers. The same year, Sto¨ hr et al.
`were able to explain the results obtained by Chaudhari. They
`showed that the directional sputtering of argon ions caused the
`scission of only certain surface bonds, leaving a surface with a
`directional molecular bond order.36 In fact, they showed that
`any orientational bond order on a surface is capable of
`inducing liquid crystal alignment, which had been suggested
`before by Lee,37 and is in agreement with the proposed
`alignment mechanism on (rubbed) polymer
`substrates
`(vide supra).
`One other interesting technique is the modification of
`substrates with a tip from a scanning probe microscope,
`so called ‘‘nanorubbing’’. Already in 1992, Majumdar et al.
`reproduced the surface characteristics generated by conven
`tional rubbing, by scanning a gold coated AFM tip across a
`polymer layer.38 By applying a bias voltage between the
`surface and tip, the polymer could be burned away along
`the path of the tip. Several years later, Ruetschi et al. and
`Pidduck et al. showed that similar results could be obtained by
`mechanical scratching of a spin coated polymer layer with an
`AFM tip (Fig. 9).39 Consequently, these systems were shown
`to be able to align liquid crystals.39,40 In addition, it was shown
`that scratches made on a bare ITO surface were also capable
`of inducing liquid crystal alignment, eliminating the use of a
`polymer alignment layer.15a Although it is unlikely that the
`process could be scaled up to magnitudes which are of interest
`
`Fig. 8 Schematic depiction of the effect of different thiol containing
`SAMs on liquid crystal alignment; a) alkylthiols with an odd
`numbered aliphatic tail cause alignment perpendicular to the deposi
`tion direction of the gold layer; b) alkylthiols with an even numbered
`aliphatic tail cause alignment parallel to the deposition direction of the
`gold layer; c) a mixture of alkylthiols with odd and even numbered
`alkyl tails causes homeotropic alignment. The deposition direction of
`the gold layer is indicated by the arrows.
`
`Fig. 9 SEM micrograph of spin coated poly(methylmethacrylate)
`(Mw = 335 000) scratched with an AFM tip. Bar = 1 mm.
`
`This journal is ß The Royal Society of Chemistry 2006
`
`J.Mater.Chem, 2006, 16, 1305 1314 | 1311
`
`Page 7 of 10
`
`

`
`orientations of liquid crystalline molecules, e.g. planar or
`homeotropic, including controllable tilt angles.43
`
`Towards rational control over nematic domains
`
`the hierarchical rubbing
`Recently, we demonstrated that
`process described above can also