GB2377501A

GB2377501A - A liquid crystal display device using disclination lines to accelerate a change of state

Info

Publication number: GB2377501A
Application number: GB0116983A
Authority: GB
Inventors: Harry Garth Walton; Diana Cynthia Ulrich; Robert Winlow; Brian Henley
Original assignee: Sharp Corp
Current assignee: Sharp Corp
Priority date: 2001-07-12
Filing date: 2001-07-12
Publication date: 2003-01-15
Also published as: GB0116983D0

Abstract

A liquid crystal display (LCD) device comprises a liquid crystal (LC) layer 1 disposed between two substrates 2. The device includes means for generating a pre-determined pattern of one or more disclination lines 10 in a portion of the LC layer. The generating means may include applying an electric field distribution to the LC layer. Each disclination line (domain wall) separates two different, stable asymmetric H-states (splay state). When a desired LC state, such as a V- (bend) or T-state (twisted), is generated in a region 9 of the LC layer, a disclination line 11 bounding region 9 can propagate along the disclination line(s) 10. This increases the speed with wich the region 9 of the desired LC state grows in area. A nucleation region 8 may be provided. A method of making a LCD substrate is described comprising providing a substrate having a portion of increased thickness forming a step. The edge of the step is made non linear (fig. 10). An alignment layer is disposed on the upper surface of the substrate and rubbed along a rubbing direction. The non-linear edge of the step is the leading edge during rubbing. The non-linear edge modifies the rubbing strength.

Description

A liquid crystal display device, a method of operating a liquid crystal display device, and a method of manufacturing a substrate for a liquid crystal display device The present invention relates to a liquid crystal display device, and in particular to a surface mode LCD such as a pi-cell device or a splay-bend device (SBD). The invention also relates to a method of driving such a liquid crystal display device. The invention also relates to a method of manufacturing a substrate for a liquid crystal display device, in particular to a method of manufacturing a substrate having regions of different pre-tilt.

The term"surface mode LCD"as used herein means an LCD in which the optical change caused by varying the electric field across the liquid crystal layer occurs primarily in the surface layers of the liquid crystal. Examples of surface mode LCDs are the pi-cell and the splay-bend device, although other types of surface mode LCDs are known. Surface mode LCDs are disclosed in"Sov. J. Quantum Electronics", 1973, Vol 3, p78-79.

The pi-cell (otherwise known as an"optically compensated birefringent device"or ZD OCB) is described in"Mol. Cryst. Liq. Cyst.", 1984, Vol 113, p329-339, and in US patent 4 635 051. The structure of a pi-cell is schematically illustrated in Figure 1.

The device comprises first and second transparent substrates 2,2'. The substrates are spaced apart by means of spacers such as glass or polymer beads (only one spacer 3 is shown in Figure 1, but in practice more than one spacers will be provided) or by a walllike structure. A layer of nematic liquid crystal 1 is disposed between the substrates 2, 2'. The thickness of the liquid crystal layer is typically in the range I-I OLm.

First and second alignment layers 4,4'are disposed on the first and second substrates respectively. Each alignment layer is typically formed by disposing a thin (less than 1 lam thickness) layer of a polymer such as polyimide on the substrate and then treating the polymer layer, for example by a rubbing or buffing technique, so that the treated

layer induces a preferred orientation direction in liquid crystal molecules adjacent to the layer. The alignment layers on the two substrates of a pi-cell are generally treated so that the orientation direction induced at the upper substrate 2 is parallel to the orientation direction induced at the lower substrate 2'.

Transparent addressing electrodes 5,5', for example formed from indium tin oxide (ITO), may be provided on the substrates 2,2', so that an electric field can be applied to the liquid crystal layer by applying a voltage across the electrodes. One or both of the addressing electrodes may be patterned to define a plurality of individually addressable regions in the liquid crystal layer 1, and these are known as picture elements or pixels.

Electrical circuitry such as thin film transistors or plasma-addressing circuitry (the so- called PALC circuitry) may be provided on the substrates to enable the electrodes to be addressed.

To provide a display device the liquid crystal layer 1 is placed between linear polarisers (not shown), whose transmission axes are crossed with one another and are at 45 to the optic axis of the liquid crystal layer. The transmissivity of the device may be varied by varying the voltage applied across the liquid crystal layer 1. A typical display device may comprise further components such as colour filters to provide a full-colour display, backlights, and birefringent optical films to improve the viewing angle and the contrast ratio of the display.

Figure 1 shows a transmissive LCD. A pi-cell can also be embodied as a reflective device by providing a reflector below the liquid crystal layer. The lower substrate 2' and the addressing electrode (s) 5'on the lower substrate 2'are not required to be transparent in a reflective display device, and indeed the addressing electrode 5'on the lower substrate can be made a reflective electrode. The lower polariser is not required with a reflective pi-cell.

The principle of operation of the pi-cell device is illustrated in Figures 2 (a) to 2 (c), which show how the director configuration of the liquid crystal layer varies as the voltage across the liquid crystal layer is applied. The electrodes, alignment films and

spacer have been omitted from Figures 2 (a) to 2 (c) for clarity. The orientation direction of the upper substrate is parallel to the orientation direction on the lower substrate.

As noted above, the alignment films on the substrates of a pi-cell are treated to ensure that liquid crystal molecules adjacent the substrates are oriented in a preferred orientation direction in the plane of the substrates. The alignment films of a pi-cell are also treated to ensure that liquid crystal molecules adjacent the upper substrate 2 are tilted with respect to the upper substrate at a so-called pre-tilt angle 81 and liquid crystal molecules adjacent the lower substrate 2'are tilted with respect to the lower substrate at a pre-tilt angle 82. The pre-tilt angle in a pi-cell is typically less than 45 and more typically is in the range 1 to 10 . The pre-tilt angles at the two substrates are preferably made equal to one another, since this has been found to provide the best optical properties.

When no electric field is applied across the liquid crystal layer, the liquid crystal layer of a pi-cell is in an H-state (homogenous state or splay state), in which the liquid crystal molecules in the centre of the liquid crystal layer are substantially parallel to the substrates. This is shown in Figure 2 (a). The short lines in the figures represent the director of the liquid crystal molecules. If the pretilt angles at the upper and lower substrates 01, 62 have equal magnitude, then the zero-field splay state will be a uniform splay state.

Since the pre-tilt angle at the upper substrate is opposite in direction to the pre-tilt angle at the lower substrate, there must be a point in the liquid crystal layer where the director of the liquid crystal molecules is parallel to the substrates 2,2'. This director is denoted by a in Figure 2 (a). In a uniform splay state, under no applied field, the director a is situated equidistant between the upper and lower substrates, so that the liquid crystal cell is symmetric about the midplane.

When a voltage is applied across the liquid crystal layer, the uniform splay state becomes unstable above a certain applied voltage (typically around IV). The director a that is parallel to the substrates 2, 2' is no longer in the cell midplane, but is displaced

vertically from the cell midplane leading to an asymmetric splay state. The director a can be displaced either above or below the cell-midplane, leading to the two asymmetric splay states Hup and Hdown shown in Figure 2 (b). In the state Hup the director a that is parallel to the substrates is displaced above the cell-midplane, and in the state Hdown the director a is displaced below the cell-midplane. If the pre-tilt angles at the upper and lower substrates are equal, that is 81 = 82, then the energy of the Hup state is equal to the energy of the Hdown state so that neither of the Hup or Hdown states is preferred over the other.

Thus, in a pi-cell in which the pre-tilt angles at the upper and lower substrates are equal in magnitude, application of a voltage (typically of around 1 volt) across the liquid crystal layer will result in the random formations of regions of the Hup and Hdown states.

Which state forms in any particular region of the liquid crystal layer will be dictated by minute random differences in the pre-tilt angles 81, 82, by random thermal fluctuations or by similar effects. A region in which the Hup state is stable is separated from a region in which the Hdown state is stable by a defect in the liquid crystal alignment, and this defect is shown schematically as"6"in Figure 2 (b). Such a defect is variously termed a "singularity", a"disclination", or a"domain wall"in the literature but, for convenience, it will also be referred to herein as a"Hup-Hdown line". In a conventional pi-cell, since the regions of Hup-state and the Hdown-state form in a random spatial fashion, the associated Hup-Hdown Imes are also randomly distributed.

If the voltage applied across an asymmetric display liquid crystal state is steadily increased, the arrangement of the liquid crystal director across the liquid crystal layer becomes increasingly distorted and asymmetric. Above some threshold voltage (typically around 2V), the asymmetric display state is replaced by a bend state, or a Vstate, as shown in Figure 2 (c). The energies of the Hup state and the Hdown state remain equal as the voltage is increased, so that an Hup state will not spontaneously transform to an Hdown state, or vice versa) as the voltage across the liquid crystal layer is increased.

The bend state shown in Figure 2 (c) is the most suitable state for operating the pi-cell as a display device. In general, a pi-cell is operated between a low-voltage bend state and

a high-voltage bend state. Thus, it is desirable for the transition from the splay state to the bend state to be effected as easily as possible. The transition from an undesired symmetric or asymmetric splay state to the desired bend state is not straightforward, however, since applying a voltage to a pi-cell cannot continuously transform a splay state into a bend state. As a result, even if the applied voltage across the liquid crystal layer is increased to a voltage at which a bend state is the most energetically favourable liquid crystal state, the liquid crystal may nevertheless remain in the splay state if there is no suitable means for transforming the liquid crystal from the splay state to the bend state. This problem is sometimes referred to as the"nucleation problem".

A further state, in addition to those shown in Figure 2 (a) to 2 (c) can exist in a pi-cell.

This state is known as the 1800 twist state, or T-state. This state forms transiently upon removal of applied voltage from a pi-cell which is in a bend-state. Whereas a splay state may not be continuously transformed into a bend state, a 180 twist state can be easily and continuously distorted into a bend state by application of a suitable voltage and, for the purpose of this application, the 180 twist state and the bend state of a pi- cell can be considered equivalent.

If a region in which a splay state is stable and a region in which a bend state is stable are present in a pi-cell, the two regions are separated by a defect in the liquid crystal alignment, and this defect is variously known as a"singularity", a"disclination", or a "domain wall". For the purposes of this application, a defect line that separates a splaystate region from a bend-state region will be known as a"H-V disclination line".

The action of creating a region in which a bend-state is stable within a region of a picell in which a splay-state was previously the stable state (or vice versa), or equivalently the creation of the associated H-V disclination line, is termed"nucleation". Applying a sufficient voltage (typically around or greater than 2V) will cause the area in which the bend-state is stable to increase at the expense of the area in which the splay state is stable so that the bend state region will grow into the splay state region. This growth of the bend state into the splay state is equivalent to a movement of the H-V disclination line separating the bend state region from the splay state region.

In a conventional pi-cell it generally takes a considerable time for the bend state to be nucleated and to grow so as to cover the liquid crystal layer (or a selected portion thereof, such as the portion that corresponds to a particular picture element). This increases the time required to put a pi-cell into its operating states.

It is known that adding a chiral dopant to a pi-cell can assist in the nucleation of the bend state. For example, adding chiral dopant to the liquid crystal layer such that the ratio of the liquid crystal layer thickness d and the pitch p of the twist of the doped liquid crystal layer satisfy the relationship 0 < d/p < 0.25 will cause the energy of the 180 twist state to become more nearly equal to the energy of the splay state, and this will aid in the nucleation of the 180 twist state (if sufficient chiral dopant is added to produce d/p > 0.25, the splay state ceases to appear and the nucleation problem is avoided, although other disadvantages occur). However, adding a chiral dopant has the disadvantage that this may reduce the brightness of the display.

A number of approaches have been proposed to aid the initial nucleation of the bend state in a pi-cell. In a conventional pi-cell, the positions within the liquid crystal layer where nucleation of the bend state occurs are essentially randomly located. It is found that nucleation of the bend state often occurs at inhomogeneities in the liquid crystal layer such as, for example, scratches in the alignment film or dust particles within the liquid crystal panel. Some techniques for systematically introducing nucleation sites into a panel have been suggested, as means for aiding the initial nucleation of the bend state. For example, JP-A-9 90 432 suggests that the glass or polymer spacer beads that are normally provided in a pi-cell to control the spacing of the substrates can, under suitable circumstances, act as sites that promote the nucleation of a bend state.

Co-pending UK Patent Application Nos. 0002733.4 and 0024636.3 suggest that regions of increased thickness of the liquid crystal layer can act as nucleation sites for aiding the nucleation of a bend state.

JP-A-11 7 018 describes a further nucleation site, which consists of a region of high pre-tilt provided within a liquid crystal panel having a predominantly lower pre-tilt.

The above prior art approaches address the initial nucleation of a bend-state in the liquid crystal layer of a pi-cell. As noted above, however, once an initial region in which a bend state is stable has occurred, it is necessary for this region to grow into the entire liquid crystal layer, or into an entire pixel, before an image can be displayed. The above prior art techniques do not address the issue of minimising the time taken for the bend state to grow into the entire liquid crystal layer, or into an entire pixel, after it has initially nucleated.

A first aspect of the present invention provides a liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; and means for generating a pre- determined pattern of one or more disclination lines in a selected portion of the liquid crystal layer, the or each disclination line separating a first region in which a first asymmetric H-state is stable from a second region in which a second asymmetric H- state different from the first asymmetric H-state is stable, the pre-determined pattern extending over a substantial part of the selected portion of the liquid crystal layer.

The invention makes use of the finding that the speed of movement of an H-V disclination line through a liquid crystal layer is increased significantly if the H-V disclination line moves along a pre-existing Hup-Hdown disclination line. The present invention sets out to control systematically the formation of disclination lines in the liquid crystal layer, to increase the speed with which the desired operating state grows into the liquid crystal layer. According to the invention, a pre-determined pattern of one or more disclination lines is generated in a selected portion of the liquid crystal layerthat is, one or more disclination lines are formed at pre-determined positions. When the desired operating state is nucleated, the disclination line surrounding the region in which the desired operating state is stable will encounter, as it moves, a disclination line between the regions of first and second asymmetric H-states, and this will increase the

speed of movement of the disclination line that surrounds the region in which the desired operating state is stable.

As noted above, disclination lines between a region in which a first asymmetric H-state is stable and a region in which a second asymmetric H-state is stable exist in conventional pi-cells. As also noted above, which one of the asymmetric H-states is stable in any particular region of a liquid crystal layer will be dictated by minute random differences, so that the regions of the different asymmetric H-states are randomly distributed. The disclination lines between the regions will therefore also be randomly distributed. Adjacent pixels of a display device will normally have quite different distributions of disclination lines, and the arrangement of disclination lines away from the boundaries of a pixel is random. In contrast, the present invention provides a pre-determined disclination line pattern.

Japanese Patent No. 8-328 045 describes a method of eliminating Hup-Hdown disclination lines in a pi-cell. This patent teaches that the alignment layers on the upper and lower substrates of a pi-cell should have uniform pre-tilt angles, but with the pre-tilt angle of the upper alignment film being different from the pre-tilt angle of the lower alignment film. This forces the liquid crystal layer within the pi-cell to adopt either the Hup or the Hdown state, so preventing the formation of disclination lines. In contrast, the present invention makes use of disclination lines to reduce the time need for a pi-cell to be put into the operating state.

The liquid crystal display device may further comprise a nucleation region for nucleating a desired liquid crystal state, and a part of at least one disclination line of the pattern may be in the vicinity of the nucleation region. This ensures that, once the initial region in which the desired liquid crystal state exists has been nucleated, a disclination line that bounds the region of the desired liquid crystal state does not have far to move before it encounters a disclination line between the first and second regions.

At least one disclination line of the pattern may extend into the nucleation region.

Doing this ensures that, as soon as the desired liquid crystal state has been nucleated in

the nucleation region, the disclination line surrounding the region of the desired liquid crystal state is able to move along the disclination line between the first and second regions.

The nucleation region may be for nucleating a V-state. The first asymmetric H-state may be a Hup-state and the second asymmetric H-state may be a Hdown-state.

The means for generating the pattern of disclination lines may be adapted to generate a plurality of disclination lines in the selected portion of the liquid crystal region, each disclination line extending over a substantial part of the selected portion of the liquid crystal layer. This further decreases the time taken for the desired liquid crystal state to grow into the selected portion of the liquid crystal layer, since the disclination line bounding the desired liquid crystal state can move quickly in many directions.

At least one disclination line of the pattern may extend to a boundary side of the selected portion of the liquid crystal layer.

The means for generating the pattern of one or more disclination lines may comprise means for inducing a first pre-tilt angle in liquid crystal molecules adjacent a first region of the first substrate and for inducing a second pre-tilt angle different from the first pretilt angle in liquid crystal molecules adjacent a second region of the first substrate.

They may further comprise drive means for applying an electric field across the selected portion of the liquid crystal layer.

The means for generating the pattern of one or more disclination lines may alternatively comprise drive means for applying an electric field across the selected portion of the liquid crystal layer, the drive means being adapted to produce an electric field distribution so shaped as to define the pattern of one or more disclination lines in the selected portion of the liquid crystal layer.

The drive means may comprise a first electrode disposed on the first substrate and a second electrode disposed on the second substrate, one of the first and second electrodes

being shaped so as to produce the electric field distribution. One or more apertures may be provided in the one of the first and second electrodes. Alternatively, the drive means may comprise: a first electrode disposed on the first substrate; a second electrode disposed on the second substrate; and a dielectric material disposed on one of the first and second electrodes so as to produce the electric field distribution.

The disclination lines may be substantially parallel to one another. They may be substantially parallel to the rubbing direction of one of the substrates (for example as determined by the rubbing direction of the alignment film on that substrate). Arranging the disclination lines to be substantially parallel to the rubbing direction has been found to provide smooth, continuous disclination lines that extend across the selected region of the liquid crystal layer.

The disclination lines may alternatively be at an angle of up to 30 to the rubbing direction of one of the substrates The selected portion of the liquid crystal layer may be a picture element.

The liquid crystal display device may comprise an array of picture elements and means for generating, in each of the picture elements, a pre-determined pattern of one or more disclination lines, the or each disclination line separating a first region in which a first asymmetric H-state is stable from a second region in which a second asymmetric Hstate different from the first asymmetric H-state is stable. Each picture element of the array of picture elements may comprise comprises means for generating a predetermined pattern of one or more disclination lines in the respective picture element.

A second aspect of the invention provides a method of operating a liquid crystal display device as defined in any preceding claim, the method comprising generating a predetermined pattern of one or more disclination lines in the selected region of the liquid crystal layer including at least a first disclination line separating a first region in which the first asymmetric H-state is stable and a second region in which the second asymmetric H-state is stable; generating a desired liquid crystal state in a region of the

liquid crystal layer; and propagating a second disclination line bounding the desired liquid crystal state along the first disclination line.

A third aspect of the present invention provides a liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; a nucleation region defined in a selected portion of the liquid crystal layer; and means for generating a non-random pattern of disclination lines in the selected portion of the liquid crystal layer between first and second liquid crystal states, a part of the pattern of disclination lines being in the vicinity of the nucleation region.

A fourth aspect of the present invention provides a liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; drive means for applying an electric field across a selected portion of the liquid crystal layer; wherein the drive means is adapted to produce an electric field distribution so shaped as to define a pre-determined pattern of one or more disclination lines in the selected portion of the liquid crystal layer.

The drive means may comprise a first electrode disposed on the first substrate and a second electrode disposed on the second substrate, one of the first and second electrodes being shaped so as to produce the electric field distribution. One or more apertures are provided in the one of the first and second electrodes.

The drive means may alternatively comprise: a first electrode disposed on the first substrate; a second electrode disposed on the second substrate; and a dielectric material disposed on one of the first and second electrodes so as to produce the electric field distribution.

The selected portion of the liquid crystal layer may be a picture element.

The liquid crystal display device may be a surface mode liquid crystal display device, and may be a pi-cell.

A fifth aspect of the present invention provides a method of manufacturing a substrate for a liquid crystal display device, the method comprising the steps of : providing a substrate having a first portion having a first thickness and a second portion having a second thickness greater than the first thickness whereby an upper surface of the substrate includes a step portion, the edge of the step adjacent the second portion of the substrate being non-linear; disposing an alignment layer over the upper surface of the substrate; and rubbing the alignment layer along a rubbing direction; wherein the substrate is arranged during the rubbing step such that the non-linear edge of the step is the leading edge during the rubbing step.

The rubbing cloth or other rubbing device used in the rubbing process will initially make contact with the non-linear edge, since the substrate is arranged so that the non- linear edge is the leading edge. The strength of the rubbing on the alignment layer on the upper surface of the greater thickness portion is modified by the non-linear edge. As the rubbing cloth is brought into contact with the non-linear edge, the areas of the alignment film that the rubbing cloth first makes contact with are rubbed more strongly than areas of the alignment film that the cloth subsequently makes contact with. Thus, the alignment layer is rubbed to produce areas of high rubbing strength and areas of low rubbing strength.

The substrate may be positioned during the rubbing step such that the non-linear edge of the step is generally perpendicular to the rubbing direction.

The non-linear edge of the step may contain lateral variations. Since the edge is generally perpendicular to the rubbing direction during the rubbing step, the lateral variations in the edge will be parallel to the rubbing direction.

The height, above the first portion of the substrate, of the non-linear edge of the step may vary over the edge.

The step of providing the substrate may comprise: disposing a layer of photoresist over a substrate; masking a selected portion of the layer of photoresist ; and reducing the thickness of the unmasked portion of the layer of photoresist.

Preferred embodiments of the present invention will now be described by way of illustrative example with reference to the accompanying figures in which: Figure 1 is a schematic sectional view of a pi-cell; Figures 2 (a) to 2 (b) show the variation of the director configuration of liquid crystal molecules in a pi-cell with applied voltage; Figure 3 (a) is a sectional view of a pi-cell according to a first embodiment of the present invention; Figure 3 (b) is a plan view of the pi-cell of Figure 3 (a); Figure 4 illustrates the speed of growth of the nucleation region of the pi-cell of Figures 3 (a) and 3 (b); Figure 5 is a plan view of a pi-cell according to a further embodiment of the present invention; Figure 6 is a sectional view of a pi-cell according to a further embodiment of the present invention; Figure 7 (a) is a sectional view of a conventional liquid crystal device; Figure 7 (b) is a sectional view of a liquid crystal device according to a further embodiment of the present invention;

Figure 7 (c) is a sectional view of a liquid crystal device according to a further embodiment of the present invention; Figure 8 (a) is a plan-view of a pi-cell according to another embodiment of the invention; Figure 8 (b) is a plan-view of a pi-cell according to a further embodiment of the invention; Figure 9 is a plan-view of a pi-cell according to a further embodiment of the invention; Figures 10 (a) and 10 (b) are a cross-sectional view and a plan view of a substrate suitable for use in a display device of the present invention; Figure 10 (c) is a schematic view of the rubbing strength for the substrate of Figures 10 (a) and 10 (b); Figure 10 (d) is a schematic plan view of a liquid crystal display device incorporating the substrate of Figure 10 (c); Figure 11 (a) is plan view of another substrate suitable for use in a display device of the present invention; Figures 11 (b) and 1 I (c) are schematic sectional views of the substrate of Figure 11 (a); Figure 1 I (d) is a schematic view of the rubbing strength for the substrate of Figure 11 (a); Figure 12 (a) is plan view of another substrate suitable for use in a display device of the present invention; Figures 12 (b) and 12 (c) are schematic sectional views of the substrate of Figure 12 (a);

Figure 12 (d) is a schematic view of the rubbing strength for the substrate of Figure 12 (a).

Like reference numerals refer to like components throughout the figures.

The principle of a liquid crystal display device according to the present invention is illustrated in Figures 3 (a) and 3 (b). In these figures the liquid crystal display device is a pi-cell.

The liquid crystal display device shown in Figures 3 (a) and 3 (b) has a layer of nematic liquid crystal 1 disposed between an upper substrate 2 and a lower substrate 2'. The upper surface of the lower substrate 2'is preferably parallel to the lower surface of the upper substrate 2 so that the thickness of the liquid crystal layer is uniform. The upper and lower substrates can be spaced apart by any suitable spacer means (not illustrated).

The substrates are provided with aligning films 4,4'for defining the orientation direction of liquid crystal molecules adjacent to the substrates, and also for imparting a pre-tilt angle to liquid crystal molecules adjacent to the substrates. The device further

comprises drive electrodes 5, 5', for example ITO electrode, for applying a voltage across the liquid crystal layer. The drive electrodes 5, 5' are connected to suitable driving circuits (omitted from Figure 3 (a) and 3 (b) for clarity), and the combination of the drive electrodes 5,5'and the driving circuits forms a drive means for the display device. The orientation direction 7 induced by the upper alignment film 4 and the orientation direction 7'induced by the lower alignment film 4'are parallel.

In accordance with the invention, the liquid crystal display device is provided with means for generating a pre-determined pattern of disclination lines, each disclination line separating a first region in which a first liquid crystal state is stable and a second region in which a second, different liquid crystal state is stable. The pattern of disclination lines extends over a substantial part of the liquid crystal region. In this embodiment the pre-determined pattern of disclination lines comprises a single disclination line, which is formed at a pre-determined position, but a display device of

the invention may alternatively comprise, in use, a plurality of disclination lines arranged in a pre-determined pattern. In the embodiment of Figures 3 (a) and 3 (b) the two liquid crystal states are two different asymmetric H-states, for example, an Hup-state and an Hwn-state. The upper and lower aligning films 4, 4'in one region of the liquid crystal device are treated (in a manner to be described below) such that application of a suitable voltage across that region of the liquid crystal layer stabilises an Hup-state. The aligning films in a second, adjacent region of the liquid crystal device are treated such that, upon application of a suitable voltage, an Hdown-state is the stable state in that region. Thus, the combination of the suitably-patterned aligning films and the drive means together forms a means for generating a disclination line 10 between a region in which an Hup-state is stable and a region in which an Hdown-state is stable.

The liquid crystal cell further contains a nucleation site 8 for generating a desired liquid crystal state. In the embodiment of Figures 3 (a) and 3 (b) the nucleation site 8 generates a V-state since, as noted above the desired operating states of a pi-cell are V-states.

The method of generating the desired operating state in the liquid crystal display device of Figures 3 (a) and 3 (b) is as follows. Initially, when no voltage is applied across the liquid crystal layer a symmetric H-state is stable over the entire liquid crystal layer, for the case where 81 = 82. As a voltage is applied across the liquid crystal layer, one or more regions in which the Hup state is stable are generated, and one or more regions in which the Hdown state is stable are also generated. As a result, one or more Hup-Hdown disclination line is set up. If the pre-tilt difference between the upper and lower substrates is small, the Hup state will be similar to the Hdown state, so that a disclination line will be relatively diffuse. The greater the difference between the pre-tilt angles on the upper and lower substrates, the sharper will be a disclination line.

In principle, it would be possible for one of the asymmetric H-states to be the stable state at zero applied field, depending on the pre-tilt angles of the upper and lower alignment films.

As the voltage is further increased, a region 9 in which the V-state is stable will be generated around the nucleation site 8. The V-state region 9 will be separated from the Hup-state region and the Hdown-state region by an H-V disclination line 11. The H-V disclination line 11 will tend to move away from the nucleation site 8, so that the V- state region 9 will grow into the Hup and Hdown regions until the V-state is the stable state throughout the entire liquid crystal 1. Figure 3 (b) illustrates the pi-cell after the V-state region 9 has been generated around the nucleation site 8 and the H-V disclination line 11 has started to move away from the nucleation site 8. Figure 4 illustrates the speed at which the H-V disclination line 11 moves through the liquid crystal layer. Figure 4 shows the speed of movement of the H-V disclination line 11, as a function of the voltage applied across the liquid crystal layer, in the direction along the Hup-Hdown disclination line 10 (direction"A"), and in a direction perpendicular to the Hup-Hdown disclination line (direction"B"). It will be seen from Figure 4 that the H-V disclination line 11 moves much faster in direction A, along the Hup-Hdown disclination line 10 than it does in the direction B. For example, when the voltage applied across the liquid crystal layer 1 is 4Vrms (the measurements in Figure 4 were taken using an AC voltage applied across the liquid crystal layer, as is conventionally done to avoid electrolysis of the liquid crystal material), the speed of movement of the H-V disclination 11 is over 2.5 times as fast in the direction A as in the direction B. As the applied voltage across the liquid crystal layer increases, the difference between the speed of movement of the H-V disclination in the direction A and its speed of movement in the direction B increases further. The results of Figure 4 were obtained with a liquid crystal layer at room temperature, which is the temperature at which most liquid crystal display devices are used in practice.

Thus, by providing a liquid crystal display device with means for generating one or more predetermined disclination lines between an Hup-state region and an Hdown-state region, the present invention is able to reduce the time taken for an initial region containing a desired operating state to grow into the entire liquid crystal layer. The present invention thus reduces the time taken before the liquid crystal layer is able to be addressed for displaying an image.

In a conventional pi-cell liquid crystal display device, the regions in which an Hup-state is stable and the regions in which an Hdown-state is stable form in a random spatial fashion. The associated Hup-Hdown disclination lines are also randomly distributed over the liquid crystal layer, and each picture element of the device will contain a different random pattern of disclination lines. In contrast, in the present invention the formation of an Hup-Hdown disclination line is reproducible, and the liquid crystal layer will contain, in use, one or more pre-determined disclination lines.

The Hup-Hdown disclination line (or at least one disclination line if there is more than one disclination line in the pattern) preferably extends to, or nearly to, one boundary edge of the liquid crystal layer, so that the H-V disclination line 11 is able to propagate as far as possible through the liquid crystal layer by travelling along the Hup-Hdown disclination line.

Where a nucleation region is provided, as in the liquid crystal device of Figures 3 (a) and 3 (b), the or at least one Hup-Hdown disclination line 10 of the pattern is preferably in the vicinity of the nucleation region and particularly preferably extends to the nucleation region 8. The broken line in Figure 3 (b) illustrates the position of Hup-Hdown disclination line 10 before the V-state has nucleated around the nucleation site 8, and it will seen that the disclination line 10 extends to the nucleation region 8. This ensures that as soon as the desired operating state has nucleated around the nucleation site 8, the disclination line 11 enclosing the region of the desired operating state is able to propagate along the Hup-Hdown disclination line 10. In principle, the Hup-Hdown disclination line 10 need not extend to the nucleation site 8, but this would tend to increase the time taken for the desired operating state to grow into the entire liquid crystal layer, since the region 9 of the desired operating state would have to undergo some initial growth before the disclination line 11, enclosing the region 9 in which the desired operating state is stable, would encounter the Hup-Hdown disclination line 10.

In principle, the present invention is not limited to a liquid crystal display device that contains a nucleation site for generating the desired operating state of the liquid crystal layer. If the nucleation site 8 were omitted from the liquid crystal display device of

Figures 3 (a) and 3 (b), some advantage would be obtained over conventional devices since, once the desired operating state had appeared, the subsequent growth of the operating state across the entire liquid crystal layer would be made easier by the presence of the Hup-Hdown disclination line. Omitting the nucleation site 8 from the device of Figures 3 (a) and 3 (b) would tend to increase the time required initially to generate the desired operating state in the liquid crystal layer 1, but this effect could be mitigated by, for example, adding a chiral dopant to the liquid crystal layer so as to facilitate the initial generation of the desired operating state.

Where a nucleation site 8 is provided in the liquid crystal display device, it may be produced by any conventional means.

If the invention is applied to a pixelated display device, it is preferable for the device to be provided with means for generating a pre-determined pattern of disclination lines in each picture element (or pixel), with the pattern in each picture element extending over a substantial part of the picture element. This may be done by providing each picture element of the display device with means for generating a pre-determined pattern of disclination lines in that picture element, with the pattern in each picture element extending over a substantial part of the picture element.

One method of forming the regions in which the Hup-and Hdown-states are stable will now be described. In this method, the regions ofHup-state and Hdown-state are obtained by spatially varying the pre-tilt angle of the liquid crystal molecules in the liquid crystal layer 1 in a controlled fashion.

In this method, the alignment films 4, 4'were each formed from a layer of the polymer alignment material SE7792 (Nissan Chemical Co. ) that was spun, using N-methyl pyrollidone, onto the upper and lower substrates 2,2'at 4,000 rpm for 40 seconds. The substrates 2, 2'were glass substrates, each coated with a transparent electrode layer such as an ITO layer. After the alignment layer had been spun onto the substrate, the

substrates were pre-baked at 90 C for 5 minutes to dry off any excess solvent, and were then baked at 180 C for two hours to cure the polymer. The alignment films were then

uni-directionally rubbed to define a preferred orientation direction. This rubbing process also produced a pre-tilt angle 81 of approximately 7. 4 .

The pre-tilt angle of selected areas of the substrate was then reduced using an irradiation process. The alignment films were partially covered with an opaque photo-mask, and were exposed to deep-UV irradiation from a mercury arc lamp for approximately two hours. The pre-tilt angle of those regions of the alignment film that were irradiated was reduced to a new value 82 of approximately 0. 5 . The pre-tilt angle of the portions of the alignment films that had been covered by the opaque photo-mask was not affected by the irradiation process. In this way, each of the alignment films 4,4 was provided with a region having a pre-tilt 81 = 7. 4 and another region having a pre-tilt 82 = 0. 5 .

The substrates were then disposed such that an area of low-pre-tilt 82 on one substrate was disposed opposite an area of high-tilt 81 on the opposite substrate and vice versa, as shown in Figure 3 (a). In this way, the positions of the Hup and Hdown regions that form when an electric field is applied across the liquid crystal layer can be pre-determined in a controlled manner, so that the location of the disclination line between the Hup region and the Hdown region is also pre-determined.

An alignment film that has a first region having a high pre-tilt 81 and a second region having a lower pre-tilt 82 can be produced by methods other than an irradiation process.

For example, once the alignment film has been rubbed to produce a uniform pre-tilt 81 over the entire area of the alignment film, the pre-tilt of one or more selected regions can be varied by selective exposure of regions of the alignment film to solvents or etchents (such as gasses or plasmas). Alternatively, selected portions of the alignment film may be masked (using, for example, layers of other organic materials such as photo-resistant materials), and the un-masked areas can then be subjected to a further rubbing process so as to change the pre-tilt in the un-masked areas of the alignment film. In principle, any method that can reproducibly create an alignment film having an area of one pre-tilt and another area of a different pre-tilt can be used to produce an alignment film suitable for use in the device of Figure 3 (a) and 3 (b).

In the device shown in Figures 3 (a) and 3 (b), the means for generating the pattern of Hup-Hdown disclination lines comprise the combination of the alignment films defining regions of different pre-tilt and the drive means for applying an electric field across the liquid crystal layer.

Figure 5 is a schematic plan view of a liquid crystal display device according to a further embodiment of the invention. In this embodiment the pre-determined pattern of disclination lines comprises a plurality of disclination lines extending from a nucleation region 8 to the boundary of the liquid crystal region with the disclination lines being substantially linear and having substantially equal angular spacing. Eight Hup-Hdown disclination lines 10 are shown in Figure 5, but the invention is not limited to this and fewer or more than eight Hup-Hdown disclination lines could be provided. Indeed, this embodiment is not limited to a disclination line pattern having approximate rotational symmetry.

Where a single Hup-Hdown disclination line 10 is provided, as in the embodiment of Figures 3 (a) and 3 (b), the speed at which the desired operating state grows in the

direction along the Hup-Hdown disclination line will be increased. However, the speed at which the desired operating state grows in the direction perpendicular to the Hup-Hdown disclination region will remain relatively low. Providing a plurality of Hup-Hdown disclination lines, which extend in different directions, means that the region of the desired operating state can grow quickly along each Hup-Hdown disclination line 10. This reduces the overall time taken for the desired operating state to grow into the entire area of the liquid crystal layer.

If the embodiment of Figure 5 is applied to a pixelated display device, it is preferable for the device to be provided with means for generating one or more disclination lines in pre-determined positions in each picture element. This may be done by providing each picture element of the display device with means for generating one or more disclination lines in predetermined positions in that picture element.

A further method of obtaining a pre-determined pattern of disclination lines is shown in Figure 6. This method generally corresponds to the method of Figures 3 (a) and 3 (b), in that the pre-tilt on the alignment films 4,4'is patterned so as to define the region in which the Hup-state is stable and another region in which the Hdown-state is stable. In Figure 6, however, only one of the alignment films is provided with areas of different pre-tilt, and the other alignment film is provided with a uniform pre-tilt 81. In the display device shown in Figure 6, the alignment film having a uniform pre-tilt angle 81 is the upper alignment film 4, but it would alternatively be possible for the film having a uniform pre-tilt angle to be the lower aligning film 4'.

The aligning film that has areas of different pre-tilt is provided with an area having a pre-tilt 82 > 81 (this region will produce an Hup stable state) and another region has a pre-tilt angle 83 < 81 (this region produces a Hdown alignment, as shown in Figure 6).

The method of Figure 6 has the advantage, compared to Figure 3 (a), that only one aligning film has to be provided with a patterned pre-tilt.

Aligning films 4, 4'suitable for use in the method of Figure 6 can be produced, for example, by fabricating polymer alignment films as described above with regard to Figures 3 (a) and 3 (b), and rubbing both alignment films to produce a pre-tilt 82. The pre-tilt of selected areas of one of the aligning films may then be reduced to 83 by any conventional process for reducing the pre-tilt of an aligning film including, for example, a selective exposure method of the type described with reference to Figures 3 (a) and 3 (b), a selective exposure to solvent or etchant, or a selective rubbing process.

The pre-tilt angle of the upper aligning film is reduced to give a uniform pre-tilt of 8 1, where 82 > 81 > 03, by any suitable process for reducing the pre-tilt angle of an aligning film.

Regions of different pre-tilt angle may also be obtained by increasing the pre-tilt angle of one or more selected areas of an alignment film.

If the embodiment is applied to a pixelated display device, it is preferable for the alignment films in each picture element of the display device to be patterned as shown in Figure 6.

An alternative method of controlling the position of an Hup-Hdown disclination line in a liquid crystal device will now be described with reference to Figure 7 (b).

Figure 7 (a) illustrates the electric field lines generated by applying a voltage between the upper electrode 5 and the lower electrode 5'of a conventional liquid crystal display device. In a conventional liquid crystal display device the area of the device, or even of a picture element, is very much greater than the thickness of the liquid crystal layer.

The electric field that is produced is therefore the electric field generated by two conductive plates whose separation is very much less than the area of the plates. As shown in Figure 7 (a) this sets up lines 12 of electric field that, away from the edges of the electrodes, are approximately uniform and normal to the services of the substrate.

Because the electric field is uniform, the liquid crystal state produced by the electric field is also expected to be uniform throughout the liquid crystal layer.

Figure 7 (b) is a schematic illustration of a liquid crystal display device according to a further embodiment of the present invention. The device contains an upper substrate 2 on which is disposed an upper electrode 5 and an alignment film 4, and a lower substrate 2'on which is disposed a lower electrode 5'and a lower alignment film 4'.

The upper and lower electrodes 5,5'are connected, in use, to suitable driving circuits (not shown).

In the embodiment of Figure 7 (b), the electrode 5 on the upper substrate 2 does not extend over the entire area of the liquid crystal layer, but a hole 13 is provided in the electrode. The presence of the hole 13 tends to distort the lines of electric field in the vicinity of the hole, producing a non-uniform electric field as shown in Figure 7 (b).

Because of the non-uniform direction of the electric field, the interaction between the electric field and liquid crystal molecules on one side of the hole is not the same as the interaction between the electric field and the liquid crystal molecules on the other side

of the hole. The asymmetry between the interaction between the electric field and the molecules on the two sides of the hole makes it possible for an Hup-state to be the stable

state on one side of the hole and for an Hdown-state to be the stable state on the other side of the hole. Thus, an Hup-Hdown disclination line 10 is set up, even though the pre-tilt angles of the two alignment films are uniform over the entire area of the liquid crystal layer.

If this embodiment of the invention is applied to a pixelated display, it is possible for one electrode in each pixel to be provided with a hole 13, so that an Hup-Hdown disclination line is set up in each pixel.

In a demonstration of the embodiment shown in Figure 7 (b), a pi-cell liquid crystal display device was fabricated in which one pixel had a square hole etched into the electrode on one substrate. In this device an ITO-coated glass substrate was used as the substrate, and a layer of photo-resist SP1813 (Shipley Limited) was spun onto the ITO coating and was then suitably cured. The layer of photo-resist was then patterned using standard photo-lithographic techniques so as to produce a square, un-coated region of the ITO film. The substrate was then etched using a solution ofHCl, so as to etch away the ITO film in the square exposed region. The remaining area of the ITO film was protected by the photo-resist and so was not affected by this etching process.

After the etching process, the remaining photo-resist was removed, and an alignment film was deposited over the electrode and was subjected to a conventional rubbing process. A pi-cell was then assembled using this substrate as one substrate, and using a conventional substrate having a continuous ITO film as the other substrate. It was found that applying an electric field between the electrodes on the two substrates did produce an Hup-Hdown disclination line in the vicinity of the hole in the ITO electrode on one substrate.

It can therefore be seen that disclination lines between two liquid crystal states, such as for example Hup-Hdown disclination lines, can be generated by suitably controlling and shaping the electric field distribution in a liquid crystal display device. This technique

can be used to embody the invention instead of or in addition to the technique of patterning the pre-tilt of the alignment layers.

The electric field distribution can be shaped in other ways than providing one or more holes in one of the electrodes. For example, the electric field can be shaped by depositing one or more regions of a dielectric material 14 on one of the electrodes, as shown in Figure 7 (c).

In principle, the electric field could be shaped by introducing holes into both the upper electrode and the lower electrode, or by disposing regions of a dielectric material on both the upper electrode and the lower electrode.

In the embodiments described above with reference to Figures 7 (b) and 7 (c), the upper and lower alignment films 4,4'can each have a uniform pre-tilt angle over their entire area, and the pre-tilt of the upper alignment film 4 can be the same as the pre-tilt angle of the lower alignment film 4'. In this case, the pattern of disclination lines is generated solely by the electric field lines. In principle, however, it would be possible for one or both of the alignment films 4,4'to have regions of different pre-tilt, for example as in Figure 3 (a) or Figure 6. In this case, there would be two separate mechanisms for generating the desired pre-determined pattern of disclination lines.

Figure 8 (a) is a plan view of a liquid crystal display device according to another embodiment of the invention. In this embodiment regions in which the Hup state is stable and regions in which the Town state is stable are obtained by patterning one or both of the alignment films to provide regions of different pre-tilt.

In the embodiment of Figure 8 (a) the regions in which the Hup state is stable and the regions in which the Hdown state is stable are stripe-shaped, and extend substantially from one side edge of the liquid crystal region to another side edge of the liquid crystal region. The width of each stripe is substantially constant over its length. In Figure 8 (a) the regions in which the Hup state is stable are shown as having approximately the same width as the regions in which the Hdown state is stable. In a modification of this

embodiment, the regions with the higher average pre-tilt have a greater width than the regions with the lower average pre-tilt. This is because a disclination line bounding a region of a V-state moves faster in regions of high pre-tilt than in regions of low pre-tilt.

By making the regions with the higher average pre-tilt cover a larger percentage of the area of the liquid crystal layer (or a selected portion thereof) the time taken for the desired operating state to grow into the entire liquid crystal layer, or selected portion thereof, may be further reduced.

It will be noted that the disclination lines are curved at the edges of the liquid crystal region shown in Figure 8 (a). This curvature is partially caused by the fringing fields that exist at the edge of the liquid crystal region, and partially because the disclination lines prefer to form a closed loop and also prefer not to be perpendicular to the rubbing direction.

The principle of the embodiment of Figure 8 (a) is most easily explained by comparison with the embodiment of Figure 6. In the embodiment of Figure 6, as described above, the disclination lines are obtained by patterning the pre-tilt angle on only one of the substrates, and the other substrate has a uniform pre-tilt angle. In the embodiment of Figure 8 (a), the regions of high and low pre-tilt on the patterned alignment film are arranged so that the disclination lines 10 extend substantially parallel to the rubbing direction of the patterned alignment film. It has been found that it is preferable to align the disclination lines along the rubbing direction of the patterned alignment film, as this provides smooth and continuous disclination lines.

Figure 8 (b) illustrates a further embodiment of the invention which generally corresponds to the embodiment of Figure 8 (a) except that the substrate having different pre-tilt angles is patterned so that the disclination lines 10 extend generally perpendicular to the rubbing direction of the patterned substrates. By arranging the disclination lines in this way, it is still possible to obtain a pattern of disclination lines that is substantially pre-determined, and that will provide improved nucleation in the manner described above. It will, however, be noted from Figure 8 (b) that the disclination lines do not generally extend from one side edge of the liquid crystal region

to another side edge of the liquid crystal region. Instead, some of the disclination lines are broken, and do not extend across the liquid crystal region. It will also be seen that the disclination lines are generally not as smooth as those in the embodiment of Figure 8 (a).

It is preferable that the disclination lines are continuous, as in the example of Figure 8 (a). It is not, however, necessary that the disclination lines are exactly parallel to the rubbing direction in order for them to remain unbroken. It has been found that acceptable results are obtained if the disclination lines are at an angle of up to around 30 to the rubbing direction.

Figure 9 illustrates a further embodiment of the invention, which generally corresponds to that of Figure 8 (a) except that the disclination lines are at an angle of approximately 30 to the rubbing direction of the patterned substrate. It will be seen that the disclination lines are still smooth and continuous, and the embodiment of Figure 9 should therefore be as satisfactory as the embodiment of Figure 8 (b). In contrast, where disclination lines do not extend across the entire width of the liquid crystal region, as in the case of Figure 8 (b), the time taken to put the liquid crystal region into its desired operating state will be greater. This is because, when a disclination line surrounding a region of the desired operating state propagates along one of the Hup-Hdown disclination lines, its propagation will be interrupted because the disclination lines do not extend across the liquid crystal region.

Figure 9 shows a liquid crystal layer having two nucleation sites 8. The nucleation sites are shaped as strips, and extend from one side edge of the liquid crystal region to another side edge of the liquid crystal region. The nucleation regions 8 extend generally parallel to the rubbing direction of the patterned alignment film.

By arranging for the disclination lines to be at a non-zero angle to the rubbing direction of the patterned alignment film, it is possible for a number of destination lines to contact the nucleation regions 8. If the disclination lines were parallel to the rubbing direction, as in the embodiment of Figure 8 (a), only one (at most) disclination line would contact

each of the nucleation sites. By placing the disclination lines at a small angle to the rubbing direction it is possible to increase the number of disclination lines that contact the nucleation regions, while ensuring that the majority of the disclination lines still extend across the liquid crystal region.

Although the embodiments of Figures 8 (a), 8 (b) and 9 have been described with reference to the embodiment of Figure 6, this is only for convenience of description. It would be possible to apply these embodiments to the embodiment of Figure 3 (a) in which the disclination lines are generated by patterning the pre-tilt angle on both the upper and lower alignment films.

As noted above, where the invention is applied to a liquid crystal display device in which the liquid crystal layer comprises a plurality of picture elements, it is preferable that the device is provided with means for generating a non-random pattern of disclination lines in each picture element of the display. However, it is not necessary for every pixel to have the same non-random pattern of disclination lines-two or more different non-random patterns of disclination lines may be used in one device, so that one picture element contains disclination lines arranged in one pre-determined nonrandom pattern and another picture element contains disclination lines arranged in another pre-determined non-random pattern.

Another method of producing a substrate having a first region having high pre-tilt 81 and a second region having a lower pre-tilt 82 will now be described with reference to Figures 10 to 12. In this method the first and second regions of different pre-tilt are generated by means of a patterned step feature on the substrate.

Figure 10 (a) is a cross-sectional view of a substrate 12 to which this method can be applied, and Figure 10 (b) is a plan view of the substrate. As can be seen in Figure 10 (a) the substrate comprises a central portion 12a having a thickness t1 and an end portion 12b having a thickness t2 < ti. As a result, the upper face of the substrate 12 is not flat but contains a step 11 that is defined by a face 11 a and an edge 11 b adjacent the greater thickness portion 12a of the substrate. The height of the step 11 is given by x = t2-tl.

The substrate further comprises a third portion 12 c having a thickness less than the thickness t2 so that the upper surface of the substrate comprises a second step 11'again defined by a face 1 la'and an edge lIb'adjacent the greater thickness portion 12a of the substrate. In this embodiment the first and third portions 12a, 12c of the substrate have the same thickness, but the first and third portions 12a, 12c of the substrate could have different thicknesses. The portion 12c may be omitted if desired, so that the greater thickness portion extends to the right hand end of the substrate.

The portions 12a, 12b and 12c of the substrate can be produced in any suitable way.

For example the substrate could be produced from a quartz, glass or plastic substrate having a thickness equal to the desired thickness t2 of the greater-thickness portion 12a, by reducing the thickness of portions of the substrate to form the low thickness portions 12b, 12c. Alternatively, the substrate could be produced from a substrate having a thickness equal to the desired thickness t ; of the low-thickness portions 12b, 12c by depositing material such as a photoresist on a portion of the substrate. In practice this would be done by depositing a layer of photoresist over the entire upper surface of the substrate, masking a portion of the photoresist layer, and reducing the thickness of (and preferably completely removing) the unmasked photoresist using any suitable etching or irradiation process.

The edge lib defining the first step is not linear, but contains spatial variations. In the embodiment of Figures 10 (a) and 10 (b) the edge lib contains lateral spatial variations, so that the edge 11 b varies about its average position (indicated in broken lines in

Figure 10 (b)). In the embodiment of Figures 10 (a) and 10 (b) the lateral spatial variations in the edge 1 Ib are obtained by providing the face 11 la with a"ziz-zag" profile, when seen in plan view, so that the edge 1 Ib also has the"ziz-zag"profile when seen in plan view.

In principle, the edge lib'defining the second step 11'may also be non-linear.

To manufacture the substrate, an alignment layer is disposed over the upper face of the greater-thickness portion 12a of the substrate 12. (In practice the alignment layer is likely to be deposited over the entire area of the substrate, but in principle the alignment layer need not be deposited over the end portions 12b, 12c of the substrate. ) When the alignment film is rubbed, the substrate is positioned so that the spatially varying edge lib forms the leading edge during the rubbing process. That is to say, a rubbing cloth or other rubbing device is moved over the substrate in the general direction from the

spatially varying edge 11 lb to the other edge 11 lb'. The spatially varying edge 11 lb thus acts as the leading edge, since the rubbing cloth is initially incident on this edge. The substrate is preferably oriented so that the spatially varying edge 11 b is generally perpendicular to the rubbing direction, as indicated in Figure 10 (c).

When the alignment layer is rubbed, the strength of the rubbing on the alignment layer on the upper surface of the greater thickness portion 12a is modified by the spatially varying edge I lb. As the rubbing cloth is brought into contact with the alignment film on the greater thickness portion 12a the cloth will make contact with some parts of the alignment film earlier than other parts of the alignment film, owing to the spatial variations in the edge 1 Ib. The areas of the alignment film that the cloth first makes contact with will be rubbed with a different strength than areas of the alignment film that the cloth subsequently makes contact with. For example, the areas of the alignment film that the cloth first makes contact with may be rubbed more strongly than areas of the alignment film that the cloth subsequently makes contact with. Alternatively, if the edge interferes with the initial contact between the cloth and the alignment film, the areas of the alignment film that the cloth initially makes contact with would be rubbed less strongly than areas of the alignment film that the cloth subsequently makes contact with. Exactly where the areas of harder and softer rubbing will occur will depend on the nature of the materials used, the profile of the step, the nature of the rubbing cloth, and the speed and strength of the rubbing process.

Figure 10 (c) illustrates the results for a case where the areas of the alignment film that the cloth first makes contact with are rubbed more strongly than areas of the alignment film that the cloth subsequently makes contact with. This is shown in Figure 10 (c),

which schematically shows the strength of rubbing for the greater-thickness portion 12a of the substrate. The alignment film has stripe-shaped regions 13 of high rubbing strength alternating with stripe-shaped regions 14 of lower rubbing strength. For a conventional alignment film for which the un-rubbed alignment film has an initial non- zero pre-tilt and rubbing reduces the pre-tilt, the regions 13 of high rubbing strength will have a lower pre-tilt than the regions 14 of lower rubbing strength. Thus, this method provides a substrate having regions of different pre-tilt.

A substrate 12 produced by the method described with reference to Figures 10 (a) to 10 (c) can be used in a liquid crystal display device of the present invention. For example, it may be used as the lower substrate in a device of the general type shown in Figure 6. Figure 1 O (d) is a schematic plan view of such a device, and it will seen that the regions 13 of higher rubbing strength in Figure 10 (c) give rise to Hdown regions in the liquid crystal display device of Figure 10 (d), whereas the regions 14 of lower rubbing strength in Figure 10 (c) give rise to Hup regions in the liquid crystal display device of Figure 10 (d). Thus, a plurality ofHup-Hdown dislocation lines 10 are formed in the liquid crystal layer, with the Hup-Hdown dislocation lines 10 being substantially regularly spaced from one another and extending over a significant part of the liquid crystal layer.

The end portion 12b, and the end portion 12c, are preferably smaller than the greaterthickness portion 12a of the substrate. The end portions 12b, 12c of the substrate are preferably disposed in a non-display area of the device, such as an inter-pixel gap in the case of a pixelated display device.

It has been found that the regions of high/low rubbing strengths are well defined near

the non-linear edge lib, but that the difference in rubbing strength decreases away from the non-linear edge 1 Ib. The rate at which the difference in rubbing strength between the high and low rubbing areas decreases with increasing distance away from the nonlinear edge lib will depend on the materials used, and on the exact form of the nonlinear edge. To counter this decrease in difference between the rubbing strength of the regions of high rubbing strength and the regions of low rubbing strength, it is preferable

that, in a pixelated device, each pixel is provided with a separate greater-thickness portion. This ensures that there is a non-linear edge associated with each pixel, to ensure that the alignment film for each pixel contains well-defined regions of high rubbing strength and well-defined regions of low rubbing strength.

In a demonstration of the embodiment of Figure 10 (d), a pi-cell liquid crystal display device was fabricated in which a patterned step feature was deposited on one substrate of the device. In this device an ITO-coated glass substrate was used as the substrate, and a layer of photoresist SU-8 manufactured by MicroChem Corporation was spun onto the ITO coating to a thickness of 3 microns, and was then suitably cured. The layer of SU-8 was then patterned using standard photolithographic techniques so as to produce a greater-thickness portion 12a and a low-thickness portion 12b, with the step 11 between the greater-thickness portion 12a and the low-thickness portion 12b feature having a non-linear edge I lb. The SU-8 resist was then fully cured to enable further device processing steps to take place.

An alignment layer of SE7792 (Nissan) was then spun on top of the substrate such that it covered both the greater thickness region of the substrate (formed by remaining SU-8) and the low-thickness regions of the substrate (the regions where the ITO-coated glass had been exposed during the photolithography steps). This alignment layer was then cured and rubbed unidirectionally to induce a preferred direction of liquid crystal alignment. The substrate was positioned during the rubbing step so that the non-linear edge I lb was the leading edge during the rubbing process.

A counter substrate was then fabricated from an ITO-coated glass substrate. The counter substrate was not patterned with SU-8 resist and so had a flat upper surface. An alignment layer of SE7792 was spun, cured and rubbed onto the upper surface of the counter substrate, in the same way as described above for the stepped substrate. A picell was then assembled using these substrates and the liquid crystal E7 (Merck) was filled into the device.

On application of an applied field above the HupIHdo\\l1 threshold voltage, tilt walls were observed to form on top of the SU-8 feature in a striped pattern which corresponded to the uneven edge of the step region.

A substrate produced by the method described above with reference to Figures 10 (a) to 10 (c) may also be used as the upper or lower substrate in a device of the general type shown in Figure 3 (a).

In the method of Figures 10 (a) to 10 (c) it was assumed that the stepped portion 12a had a uniform height. It has been found, however, that production of the spatially varying step 11 can also cause regions of varying step height. This is illustrated schematically in Figures 1 1 (a) to 1 l (c).

Figure 11 (a) is a schematic plan view of a substrate 12 having a greater-thickness portion 12a formed by, for example, depositing a layer of photoresist 15 over a glass plate 16 and subsequently removing the photoresist from end portions 12b, 12c of the substrate so as to expose the glass plate. The step 11 has a spatially varying edge I lb which, in this embodiment, is obtained by providing the step 11 with a stepped face 1 la that contains a series of protrusions lie.

Figures 11 (b) and 1 l (c) show cross-sections through the substrate of Figure 1 l (a) along the lines A-A and B-B respectively. As shown in the Figures, processing the photoresist to produce the spatially varying step face 1 la has resulted in a region 17, near the spatially varying edge lib of the step, in which the thickness of the photoresist varies.

If an alignment layer is disposed over the substrate 12 of Figures 11 (a) to Il (c) and rubbed, the existence of variations in the thickness of the photoresist can induce variations in the change of pretilt of the alignment layer induced by a unidirectional rubbing process in which the spatially varying edge 11 b is the leading edge, since the rubbing will be harder on a region of the alignment film disposed over a thicker portion of photoresist than on a region of the alignment film disposed over a thinner portion of

the photoresist. As shown schematically in Figure 11 (d), the alignment film has stripe- shaped regions 13 of high rubbing strength alternating with stripe-shaped regions 14 of lower rubbing strength. It will be seen that this effect is opposite to the effect in Figs 1 O (a) to (c), since lowest thickness occurs in the protruding portions of the edge, which will be contacted first in the rubbing process (it should be noted that, if the rubbing strength were equal everywhere, the variations in the thickness of the alignment layer may be sufficient to induce variations in the pre-tilt angle.) Thickness variations in the photoresist, along the leading edge of the stepped portion, can also be induced by adding or removing material from the greater-thickness portion 12a of the photoresist. This can be done instead of, or as well as, producing thickness variations through patterning the leading edge of the stepped portion.

Even if the central portion 12a of the photoresist has a uniform thickness, it is possible that the alignment layer may not be deposited with a uniform thickness. It is possible that thickness variations will occur in the alignment layer near the spatially varying edge of the photoresist when the alignment layer is deposited (for example by spin coating).

In principle, it would be possible for the leading edge lib of the step 11 to be linear, and to obtain regions of high and low rubbing strength by making the thickness of the greater-thickness portion 12b of the substrate non-uniform. This approach would, however, require that the regions of different thickness of the substrate extended over a majority of the display area of the liquid crystal layer, or over a majority of the area of a pixel, and this could degrade the quality of the display.

Both of the techniques have the advantage that providing a step portion feature with a spatially varying boundary can cause tilt walls to form in the active pixel region and can also form a nucleation site. As described in co-pending UK Patent Applications Nos.

0002733.4 and 0024636.3, regions of increased thickness of a liquid crystal layer can act as nucleation sites for aiding the nucleation of a bend state.

In addition to spatially varying the leading edge of the step portion, the slope of the step can also be varied in order to affect the strength of the rubbing. This is demonstrated in Figures 12 (a) to 12 (d).

Figure 12 (a) is a schematic plan view of a substrate 12 having a greater thickness portion 12a formed by, for example, depositing a layer of photoresist 15 over a glass plate 16 and subsequently removing the photoresist from end portions 12b, 12c of the substrate.

The step 11 has a spatially varying edge I lb which, in this embodiment, is obtained by providing the step 11 with a stepped face 1 la that contains a series of protrusions lie.

Figures 12 (b) and 12 (c) show cross-sections through the substrate of Figure 12 (a) along the lines al-al and a2-a2 respectively. As shown in Figure 12 (b), the front faces of the protrusions 11c are sloped at an angle 8 to the surface of the plate 16. When an alignment layer is disposed over the substrate and is rubbed in a unidirectional rubbing process in which the spatially varying edge 11 lb is the leading edge, the slope of the protrusions causes the strength of the rubbing to vary.

In Figure 12 (d) it is assumed that the regions of greater rubbing strength correspond to the recessed portions of the edge. However, it is possible that the regions corresponding to the sloped portions of the edge will get a harder rub, since it is possible that, at the non-sloped, recessed portions of the edge the rubbing cloth may"jump"over the edge and consequently will not make good contact with the alignment layer.

The methods described with reference to Figures 10 (a) to 12 (c) above are not limited to use in a liquid crystal display device of the present invention. They may be used in any liquid crystal display device which requires a substrate having regions of different pretilt.

Claims

CLAIMS:

1. A liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; and means for generating a pre-determined pattern of one or more disclination lines in a selected portion of the liquid crystal layer, the or each disclination line separating a first region in which a first asymmetric H-state is stable from a second region in which a second asymmetric H-state different from the first asymmetric H-state is stable, the pre-determined pattern extending over a substantial part of the selected portion of the liquid crystal layer.

2. A liquid crystal display device as claimed in claim 1 and further comprising a nucleation region for nucleating a desired liquid crystal state, wherein a part of at least one disclination line of the pattern is in the vicinity of the nucleation region.

3. A liquid crystal display device as claimed in claim 2, wherein the at least one disclination line of the pattern extends into the nucleation region.

4. A liquid crystal display device as claimed in claim 2 or 3 wherein the nucleation region is for nucleating a V-state.

5. A liquid crystal display device as claimed in any preceding claim wherein the first asymmetric H-state is an Hup-state and the second asymmetric H-state is a Hdown- state.

6. A liquid crystal display device as claimed in any preceding claim wherein the means for generating the pattern of one or more disclination lines are adapted to generate a plurality of disclination lines in the selected portion of the liquid crystal region, each disclination line extending over a substantial part of the selected portion of the liquid crystal layer.

7. A liquid crystal display device as claimed in any preceding claim wherein at least one disclination line of the pattern extends to a boundary side of the selected portion of the liquid crystal layer.

8. A liquid crystal display device as claimed in any preceding claim wherein the means for generating the pattern of one or more disclination lines comprise means for inducing a first pre-tilt angle in liquid crystal molecules adjacent a first region of the first substrate and for inducing a second pre-tilt angle different from the first pre-tilt angle in liquid crystal molecules adjacent a second region of the first substrate.

9. A liquid crystal display device as claimed in claim 8 wherein the means for generating the pattern of one or more disclination lines comprise drive means for applying an electric field across the selected portion of the liquid crystal layer.

10. A liquid crystal display device as claimed in any of claims 1 to 7 wherein the means for generating the pattern of one or more disclination lines comprise drive means for applying an electric field across the selected portion of the liquid crystal layer, the drive means being adapted to produce an electric field distribution so shaped as to define the pattern of one or more disclination lines in the selected portion of the liquid crystal layer.

11. A liquid crystal display device as claimed in claim 10 wherein the drive means comprises a first electrode disposed on the first substrate and a second electrode disposed on the second substrate, one of the first and second electrodes being shaped so as to produce the electric field distribution.

12. A liquid crystal display device as claimed in claim 11 wherein one or more apertures are provided in the one of the first and second electrodes.

13. A liquid crystal display device as claimed in claim 10 wherein the drive means comprises: a first electrode disposed on the first substrate; a second electrode disposed

on the second substrate; and a dielectric material disposed on one of the first and second electrodes so as to produce the electric field distribution.

14. A liquid crystal display device as claimed in any preceding claim wherein the pattern of disclination lines comprises a plurality of disclination lines which are substantially parallel to one another.

15. A liquid crystal display device as claimed in claim 14 wherein the disclination lines are substantially parallel to the rubbing direction of one of the substrates.

16. A liquid crystal display device as claimed in claim 14 wherein the disclination lines are at an angle of up to 30 to the rubbing direction of one of the substrates.

17. A liquid crystal display device as claimed in any preceding claim wherein the selected portion of the liquid crystal layer is a picture element.

18. A liquid crystal display device as claimed in claim 17 and comprising: an array of picture elements: and means for generating, in each of the picture elements, a predetermined pattern of one or more disclination lines, the or each disclination line separating a first region in which a first asymmetric H-state is stable from a second region in which a second asymmetric H-state different from the first asymmetric H-state is stable.

19. A liquid crystal display device as claimed in claim 18 wherein each picture element of the array of picture elements comprises means for generating a predetermined pattern of one or more disclination lines in the respective picture element.

20. A liquid crystal display device substantially as described herein with reference to Figures 3 (a) and 3 (b), or to Figure 5, or to Figure 6, or to Figure 7 (b), or to Figure 7 (c), or to Figure 8 (a) or to Figure 9 of the accompanying drawings.

21. A method of operating a liquid crystal display device as defined in any preceding claim, the method comprising generating a pre-determined pattern of one or more disclination lines in the selected region of the liquid crystal layer, the pattern including at least a first disclination line separating a first region in which the first asymmetric H-state is stable and a second region in which the second asymmetric H- state is stable; generating a desired liquid crystal state in a region of the liquid crystal layer; and propagating a second disclination line bounding the desired liquid crystal state along the first disclination line.

22. A liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; a nucleation region defined in a selected portion of the liquid crystal layer; and means for generating a pre-determined pattern of one or more disclination lines in the selected portion of the liquid crystal layer between first and second liquid crystal states, a part of at least one disclination line of the pattern being in the vicinity of the nucleation region.

23. A liquid crystal display device comprising: a first substrate and a second substrate; a liquid crystal layer disposed between the first substrate and the second substrate; drive means for applying an electric field across a selected portion of the liquid crystal layer; wherein the drive means is adapted to produce an electric field distribution so shaped as to define a pre-determined pattern of one or more disclination lines in the selected portion of the liquid crystal layer.

24. A liquid crystal display device as claimed in claim 23 wherein the drive means comprises a first electrode disposed on the first substrate and a second electrode

disposed on the second substrate, one of the first and second electrodes being shaped so as to produce the electric field distribution.

25. A liquid crystal display device as claimed in claim 24 wherein one or more apertures are provided in the one of the first and second electrodes.

26. A liquid crystal display device as claimed in claim 23 wherein the drive means comprises: a first electrode disposed on the first substrate; a second electrode disposed on the second substrate; and a dielectric material disposed on one of the first and second electrodes so as to produce the electric field distribution.

27. A liquid crystal display device as claimed in any of claims 23 to 26 wherein the selected portion of the liquid crystal layer is a picture element.

28. A liquid crystal display device as claimed in any of claims 1 to 20 and 22 to 27 wherein the liquid crystal display device is a surface mode liquid crystal display device.

29. A liquid crystal display device as claimed in claim 28 wherein the liquid crystal display device is a pi-cell.

30. A method of manufacturing a substrate for a liquid crystal display device, the method comprising providing a substrate having a first portion having a first thickness and a second portion having a second thickness greater than the first thickness whereby an upper surface of the substrate includes a step, the edge of the step adjacent the second portion of the substrate being non-linear; disposing an alignment layer over the upper surface of the substrate; and rubbing the alignment layer along a rubbing direction; wherein the substrate is arranged during the rubbing step such that the non-linear edge of the step is the leading edge during the rubbing step.

31. A method as claimed in claim 30 wherein the substrate is positioned during the rubbing step such that the non-linear edge of the step is generally perpendicular to the rubbing direction.

32. A method as claimed in claim 30 or 31 wherein the non-linear edge of the step contains lateral variations.

33. A method as claimed in claim 30,31 or 32 wherein the height, above the first portion of the substrate, of the non-linear edge of the step varies over along the edge.

34. A method as claimed in any of claims 30 to 33 wherein the step of providing the substrate comprises: disposing a layer of photoresist over a substrate; masking a selected portion of the layer of photoresist; and reducing the thickness of the unmasked portion of the layer of photoresist.