LIQUID CRYSTAL LENS DEVICES BASED ON CARBON NANOTUBES AND THEIR MANUFACTURING METHOD

BACKGROUND TO THE INVENTION

Field of the invention

The present invention relates to optical components, their uses and methods for their manufacture. Particularly, but not exclusively, it relates to micro-optical components.

Related art

S. P. Kotova et al (J. Opt. A: Pure Appl . Opt. 5 (2003) S231- S238 "Technology and electro-optical properties of modal liquid crystal wavefront correctors") disclose a modal liquid crystal (LC) wavefront corrector, comprising a 25 micron thick nematic liquid crystal layer sandwiched between two electrodes. One electrode is a transparent, low resistance (200 Ohm/Square) electrode formed from an ITO layer deposited on a glass substrate and serves as a common electrode. The second electrode is a high resistance (0.05-7 MOhms/Square) layer formed on a glass substrate and further coated to form a dielectric mirror reflective at 630nm. The glass substrate of the second electrode is implanted with 0.5mm diameter molybdenum wire control contacts (correctors) which contact the high resistance electrode. One embodiment also has a

peripheral annular corrector. The alignment of the liquid crystal is initially antiparallel, defined by rubbing a polyamide coating.

The corrector is operated by applying an ac voltage

(amplitude up to 10V, frequency from 20Hz to 256OkHz) between the low-resistance electrode and one or more control contacts. Due to the capacitance and inductance of the LC layer, there is a time lag between the directly applied voltage at the contact and the voltage in the vicinity of the contact. This determines the spatial variation of the electric field amplitude, and therefore the spatial distribution of the refractive index of the liquid crystal layer. The influence of each contact can extend beyond the neighbouring contacts.

A lens is demonstrated by Kotova et al, using a single control contact. Wavefront correction is demonstrated using all 37 correctors. More complex wavefront correctors are realised by varying the relative phases and frequencies of the voltages at each contact. The aperture diameter of the modal wavefront corrector is 30mm or 70mm.

S. Welch et al (S. Welch, P. Doel, A. Greenaway and G. Love, February 2003 Vol.44 1.26-1.29 "Smart optics in astronomy and space") disclose a design for an adaptive LC lens, comprising a 20 micron thick liquid crystal layer sandwiched between two glass plates, each coated with a transparent electrode. One

electrode (the control electrode) has a very high resistance (of the order of MOhms/Sq) . The control electrode has a ring-shaped electrical contact. When an oscillating voltage is applied between the contacts of the two electrodes, the voltage at the centre of the liquid crystal cell will be less than the supply voltage at the edges. This can lead to a voltage profile in the LC layer and thus to a lens-like phase profile for light transmitted by the cell. LC lenses of 5mm diameter and focal length from infinity down to 50cm have been produced.

WO 2006/003441 discloses a device for controlling the polarisation state of transmitted light, the device comprising first and second cell walls enclosing a layer of liquid crystal material having a substantially uniformly aligned helical axis in the absence of an applied field, and electrodes for applying an electric field substantially perpendicularly to the helical axis. The electrodes have a thickness of 10 microns and are arranged around the periphery of the layer to create a transverse field. The liquid crystal used is a chiral nematic liquid crystal, which is chirally birefringent and therefore rotates linear polarisation.

WO 2006/003435 discloses a tuneable laser device comprising first and second cell walls enclosing a layer of liquid crystal material having a helical axis substantially normal

to the inner surfaces of the cell walls in the absence of an applied field.

SUMMARY OF THE INVENTION

The present inventors have realised that the use of nanostructures in liquid crystal devices can provide beneficial effects on the optical properties of the liquid crystal layer.

Nanostructures are of course well known. Some of the most studied forms of nanostructure are carbon nanotubes . These can be grown as dense mats of grass, but have also been shown to be grown in sparse arrays. Single walled carbon nanotubes can have non-uniform conductivity, but multi-walled nanotubes can be grown with well-defined high conductivity. See, for example, Milne et al (W.I. Milne, K. B. K. Teo, M. Chhowalla, G.A.J. Amaratunga, S. B. Lee, D. G. Hasko, H. Ahmed, 0. Groening, P. Legagneux, L. Gangloff, J. P. Schnell, G. Pirio, D. Pribat, M. Castignolles, A. Loiseau, V. Semet, V. T. Binh, "Electrical and Field Emission Investigation of Individual Carbon Nanotubes from Plasma Enhanced Chemical Vapour Deposition", Vol. 12, Issues 3-7, 422-428, Diamond and Related Materials, 2003), the content of which is incorporated herein by reference in its entirety.

In a first aspect, the present invention provides an optical device having: a cover layer ; a substrate; a liquid crystal layer sandwiched between the cover layer and the substrate; at least one substrate electrode formed at the substrate; at least one conductive projecting element electrically connected to and projecting from the substrate electrode and whose electric potential is controllable by the substrate electrode to provide a spatial variation in the optical properties of the liquid crystal layer in a region local to the projecting element, wherein the projecting element has a tip distal from the substrate electrode, said tip having a radius of curvature of 1000 run or less.

The present inventors have found that the effect of such a conductive projecting element is to have a surprisingly profound effect on the electric field profile in the liquid crystal layer. For materials that have dielectric anisotropy as well as optical anisotropy, such an effect on the electric field profile has a corresponding effect on the optical properties of the liquid crystal layer.

In a second aspect, the present invention provides adaptive optical system including an optical device according to the first aspect.

In a third aspect, the present invention provides a method of manufacturing an optical device according to the first aspect, wherein the substrate electrode is formed on the substrate and the projecting element is grown in situ on the substrate electrode.

Preferred and/or optional features will now be set out. These are applicable singly or in any combination with any of the aspects of the invention, unless the context demands otherwise .

Preferably, the radius of curvature of the tip of the projecting element is 500 nm or less, but may be smaller, e.g. 400 nm or less, 300 nm or less, 200 nm or less, 100 nm or less, 90 nm or less, 80 nm or less, 70 nm or less, 60 nm or less, 50 nm or less, 40 nm or less or 30 nm or less.

Preferably, the projecting element is an elongate structure having a width in a direction transverse to the elongate direction of 100 nm or less. The elongate direction is preferably substantially straight along the structure. The width in the direction transverse to the elongate direction is more preferably 90 nm or less, 80 nm or less, 70 nm or

less, 60 nm or less, 50 run or less, 40 nm or less, 30 nm or less or 20 nm or less.

Preferably, the distal tip of the projecting element is located at a distance from the substrate (and/or substrate electrode) significantly greater than the width of the projecting element. Thus, the projecting element preferably has a high aspect ratio (length : width) . This aspect ratio may be at least 100, but is preferably at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900 or at least 1000.

The height of the projecting elements may be at least 0.5 μm or at least 1 μm and is preferably at least 2 μm.

The projecting element may be an elongate nanostructure. Suitable nanostructures include nanotubes, nanorods, nanocylinders and nanowires . Particularly preferred are multiwalled nanotubes.

It is preferred that the projecting element is conductive. For example, carbon nanotubes (CNTs) are preferred, and most especially multiwalled carbon nanotubes. Multiwalled carbon nanotubes have a high probability of low electrical resistance, of the order of metallic electrical resistance. It is considered that single walled carbon nanotubes have around a 33% chance of being suitably conducting, whereas

multiwalled carbon nanotubes are formed of several single walled carbon nanotubes, and so the cumulative likelihood of a multiwalled carbon nanotube being suitably conductive is high.

In the optical device, a group of two or more projecting elements may be provided connected to the same substrate electrode. In this way, the voltage applied to each projecting element in the group may be substantially the same. There may, for example, be four projecting elements provided in the group .

The projecting elements in the group may be aligned substantially parallel to each other. This allows the projecting elements to maintain a separation from each other along their elongate directions.

The spacing between adjacent projecting elements in the group may be 100 μm or less, 50 μm or less, 40 μm or less, 30 μm or less, or 20 μm or less. In preferred embodiments, the spacing may be 15 μm or less, 10 μm or less, 8 μm or less, 6 μm or less, 5 μm or less, 4 μm or less, 3 μm or less, or 2 μm or less.

Preferably, the thickness of the liquid crystal layer is in the range 5 μm to 100 μm. The upper end of this range is thicker than the liquid crystal layer thickness of

conventional liquid crystal devices. The use of a relatively- thick liquid crystal layer allows the device more effectively to create useful optical properties for the liquid crystal layer .

An insulating layer may be provided between the substrate electrode and the liquid crystal layer. This allows the liquid crystal layer to be insulated from the substrate electrode. The thickness of the insulating layer may be at least 10 nm, but is preferably at least 50 nm or at least

100 nm. In some embodiments, the insulating layer may have a thickness of the same order (or substantially the same) as the height of the projecting elements. For example, the tips of the projecting elements may be substantially level with the upper surface of the insulating layer. The advantages of using such a thick insulting layer are that the liquid crystal layer is further insulated from the substrate electrode and also that the projecting elements are mechanically supported by the insulting layer, which improves the robustness of the device.

The insulating layer may be formed after the formation of the projecting elements. This may be, for example, by sputtering or a similar deposition technique. Similarly, the substrate electrode may be formed after formation of the projecting elements (but before the formation of the insulating layer,

if any) . Alternatively, the substrate electrode may be formed before the formation of the projecting elements.

The device preferably has at least one cover layer electrode provided at the cover layer. The cover layer electrode may be located in register with said at least one projecting element . The cover layer electrode may have a patterned shape. This allows further control over the electric field distribution in the liquid crystal layer, in use, between the cover layer electrode and the projecting element(s).

The device may have a second substrate electrode, independent of the first substrate electrode. There may be at least one corresponding conductive projecting element electrically connected to and projecting from the second substrate electrode. In use, control of the electric potential of the respective projecting elements may then allow control of the electric field shape in the liquid crystal layer between the projecting elements. The advantage of providing the electric field distribution in this way is that a more uniform electric field distribution is provided, compared with substantially planar electrodes formed at the substrate surface .

The liquid crystal layer may exhibit a flexoelectric effect controllable by the electric field between the respective projecting elements.

The liquid crystal layer may include vertically aligned nematic liquid crystal.

The liquid crystal layer may include blue phase liquid crystal .

The liquid crystal layer may include a liquid crystal lasing material .

The first and second substrate electrodes may have a plurality of the conductive projecting elements. These may be arranged in a substantially linear formation along each said substrate electrode. For example, they may be formed close to an edge of each said substrate electrode.

Preferably, the liquid crystal has anisotropic optical properties so that its refractive index varies with orientation of the liquid crystal molecules. Furthermore, preferably the liquid crystal molecules have dielectric anisotropy. This, combined with an ability to flow, allows the liquid crystal molecules to orient themselves preferentially with respect to an applied electric field.

The liquid crystal may be a nematic liquid crystal.

The device may have an array of said conductive projecting elements, or an array of said groups of conductive projecting elements . Each array may be considered to provide an optical element. There may also be a corresponding array of substrate electrodes. In this way, the electric potential of each projecting element or each group of projecting elements may be controlled by operation of the substrate electrodes.

The spacing between adjacent conductive projecting elements, or groups of conductive projecting elements, in the array (and similarly the spacing between adjacent substrate electrodes) may be at least 2 μm. This spacing may be at least 3 μm, at least 4 μm, at least 5 μm, at least 6 μm, at least 7 μm, at least 8 μm, at least 9 μm or at least 10 μm.

Preferably, the array has long-range translational symmetry and/or has rotational symmetry.

Preferably, electric potential of each conductive projecting element, or each group of conductive projecting elements, in the array is individually controllable.

Preferably, in use, the electric field shape provided by each conductive projecting element, or group of conductive projecting elements, acts on the surrounding liquid crystals to provide a lenslet to refract electromagnetic radiation.

Preferably, the substrate and/or cover are substantially transparent to the electromagnetic radiation of interest.

In forming the device, preferably a catalyst material is provided on the substrate electrode at the locations at which projecting elements are desired. The catalyst material may be patterned by a lithographic technique.

BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments of the invention will now be set out, with reference to the accompanying drawings, in which: Figs. IA and IB show schematic cross sectional views of a device according to an embodiment of the invention using a single carbon nanotube.

Fig. 2 shows the results of modelling the device of Fig. 1. The modelling results show the upper electrode at +10V and the substrate electrode and carbon nanotube at OV. The greyscale indicates intermediate potentials. Fig. 3 shows a schematic view of the device of Fig. 1 with zero applied voltage.

Figs . 4A and 4B show SEM images of an array of carbon nanotubes for use in an embodiment of the invention. Figs. 5A, 5B, 6A, 6B show optical microscopy images of devices according to the present invention at different applied voltages.

Fig. 7 shows an optical microscopy image of a single carbon nanotube device.

Fig. 8 shows a schematic cross sectional view of a device according to an embodiment of the invention using two carbon nanotubes .

Fig. 9 shows the results of modelling the device of Fig. 8.

This image is taken under similar conditions to that of Fig.

2.

Fig. 10 shows the results of modelling a device having four carbon nanotubes.

Fig. 11 shows a modified version of the device of Fig. 8.

Fig. 12 shows a further modified version of the device of Fig.

8.

Fig. 13 shows the results of modelling a device having five carbon nanotubes at separate locations in an array.

Fig. 14 shows a schematic top view of a device according to an embodiment of the invention.

Fig. 15 shows a schematic cross sectional view of a device according to an embodiment of the invention. Fig. 16A shows a schematic cross sectional view of a in plane electrode device, outside the scope of the present invention.

Figs. 16B and 16C show schematic views of a modification of

Fig. 16A, according to an embodiment of the invention.

Fig. 17 shows a schematic view of a modification of the device of Fig. 16B.

Fig. 18 shows a schematic top view of a modification of the device of Fig. 16C.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS, AND FURTHER PREFERRED AND/OR OPTIONAL FEATURES

We will here describe first the way in which the preferred embodiments of the invention are manufactured, before then discussing the specific preferred uses for these embodiments. The preferred embodiments use carbon nanotubes as the conductive projecting elements, but the invention is not necessarily limited to this. In particular, it is possible to use other conductive projecting nanostructures, such as nanorods, nanowires and nanocylinders . Furthermore, nanostructures of different composition are contemplated, such as carbon nanostructures doped with one or more dopants, or Si, GaN, CdSe, ZnS etc.

A preferred device according to an embodiment of the invention is shown schematically (and not to scale) in Fig. IA. This device has a lower substrate 10 on which is formed an electrode 14 and a sparse array of carbon nanotubes 12 in electrical contact with the electrode 14. In Fig. IA, only one carbon nanotube 12 is shown for the sake of simplicity. Indeed it is possible for the "array" of carbon nanotubes to include only one nanotube 12, as shown in Fig. IA. A cover layer 16 is formed in opposing relation to the lower substrate 10. Cover layer 16 and substrate 10 are held apart by spacers (not shown), typically at the edges of the device.

Cover layer 16 has an electrode 18 formed on its lower surface. A layer of liquid crystal 20 is located sandwiched between the substrate and a cover layer, the carbon nanotubes projecting into the layer of liquid crystal.

Carbon nanotubes (and indeed other nanostructures) have relatively large dimensions when compared to the dimensions of typical liquid crystal molecules. As a result, there can be a strong interaction between the nanotubes and the liquid crystal material. As the skilled person will understand, the molecules of the liquid crystal will arrange themselves, particularly around the end of the nanotube, in an energetically favourable spatial arrangement. This may not be the same spatial arrangement as would be the case if there was no nanotube present. In effect, therefore, the presence of the nanotube provides a defect centre in the liquid crystal layer. This is illustrated in Fig. 3, where most of the reference numbers of the corresponding features to Fig. IA are not repeated.

Since typical liquid crystal materials are optically anisotropic, the interaction between the nanotubes and the liquid crystal molecules can manifest itself as an optical interaction. Thus, at the location of the nanotube itself, there is an optical defect. However, there is also a longer- range effect on the arrangement of liquid crystal molecules, and this can be harnessed to different effect. Furthermore,

the application of a voltage to the nanotube can cause surprising effects on the liquid crystal layer, and these are described in more detail below.

In Fig. IA, the electrode 14 is shown as extending up to and contacting the side of the nanotube 12. However, it is preferred that the nanotube 12 is formed either on top of the electrode 14, or that the electrode surrounds the base of the nanotube, in order to ensure good electrical contact. An electric field can be applied to the device by, for example, applying a voltage of 10V at electrode 14 and keeping electrode 18 grounded. A simplified resultant electric field profile is illustrated by single schematic electric field line 22 in Fig. IA.

A more complete electric field profile is illustrated in Fig. 2. This is a simulated electric field profile (modelled using finite element analysis) surrounding a carbon nanotube at 10V.

As shown in Fig. 3, the nematic liquid crystal molecules 30 (shown schematically) may align planar with the substrate and the cover layer in a known manner when there is zero potential applied across the liquid crystal layer.

However, when a potential is applied across the liquid crystal layer, via the substrate electrode 14 and the carbon

nanotube 12, the liquid crystal molecules 30 align to the electric field due to their dielectric anisotropy and freedom to flow. This is illustrated in Fig. IB. In turn, due to the optical anisotropy of the liquid crystal, the electric field profile dictates the refractive index profile of the liquid crystal layer. Since the electric field profile is substantially Gaussian in shape, so is the refractive index profile. This is therefore in effect a microlens capable of focussing an applied light wave. Furthermore the microlens can be tuned by varying the applied field and thereby varying the optical properties of the microlens.

It is important to note that the illustrations of Figs. 1-3 are two dimensional, but of course the electric field profile and hence the refractive index profile in the describe device are three dimensional in shape.

The cover layer 16 is typically formed of glass. The upper electrode 18 is typically formed of a transparent conductive material, such as indium tin oxide (ITO) . In some embodiments, the substrate is formed of silicon, but in others the substrate is formed of glass, quartz, sapphire or other material that is optically light-transmissive. Silicon is transparent to infra-red radiation, which is "light" for the purposes of the present invention. Where the device is to be used in transmission mode for normal visible light, it is preferred to use an optically transparent substrate.

However, the device can also be used in reflection mode, in which case the substrate need not be transparent. In some circumstances a reflection layer may be used at the substrate, as is well known in LCD devices. The reflection layer may be formed of aluminium, gold or silver, for example. The reflection layer may also be used as the (or one of the) substrate electrodes .

Each nanotube array was grown directly on a silicon wafer by plasma enhanced chemical vapour deposition after employing e- beam lithography to pattern a 5 nm nickel catalyst layer into an array of dots (islands) , each dot being about 100 nm in diameter. Thereafter, single nanotubes were grown on each dot. The substrate was heated by DC current under vacuum of 0.1 mbar to 650 ² C at a ramping rate of 100 ² C per minute. A mild heating process was preferred to protect the catalyst dots from cracking. Ammonia gas was then introduced to etch the surface of the nickel catalyst islands. Acetylene was chosen to be the carbon source, and was imported into the deposition chamber after the temperature reached 690 ² C, followed by a DC voltage of 640 V between the gas shower head and the heating stage to create plasma of 40 W in power. The growth process lasted for 10 to 15 minutes at 725 ² C, to give multiwall carbon nanotubes of around 2 μm in height.

A typical array of individual nanotubes is shown in the electron microscope images of Fig. 4A (low magnification) and

Fig. 4B (higher magnification) . In this case the nanotubes were patterned in small groups of 4 with 1 μm spacing between the nanotubes and 10 μm spacing between the groups.

The sparse array similar to that shown in Figs. 4A and 4B was then fabricated into a liquid crystal device. In order to make the device both reflective and also to provide a common electrical connection to all the nanotubes, 400 nm of aluminium was cold sputtered onto an array of groups of 4 carbon nanotubes on a silicon substrate, the carbon nanotubes in each group being spaced 0.5 μm apart. A 200 nm layer of SiO ₂ was evaporated over the top of the Al layer (this is discussed in more detail below) . The array was then assembled with a top electrode containing ITO on borosilicate glass into a liquid crystal cell with a 5.5 μm cell gap set by spacer balls in UV set glue. No alignment layer was applied to the nanotube array, but the top glass electrode was coated with AM4276 liquid crystal alignment layer and rubbed in the horizontal direction to give planar alignment. The cell was then capillary filled with BLO48 (positive dielectric anisotropy) nematic liquid crystal.

The assembled device was then viewed in reflection mode under horizontally polarised white light on an optical microscope. The locations of the arrays of nanotubes were clearly visible as black dots. These black dots were not themselves the carbon nanotubes (since carbon nanotubes are many orders of

magnitude too small to be resolved by an optical microscope) but were each the site for a defect in the nematic liquid crystal, caused by the presence of the carbon nanotube tips. These defects could be easily visualised at x20 magnification,

The effect of applying a voltage across the liquid crystal device on the optical properties of the device is particularly difficult to show on paper, but the effects are illustrated by Figs. 5A, 5B, 6A, 6B.

Fig. 5A shows the device with 0 V applied voltage at around 40 times magnification. The analyser is aligned parallel to the rubbing direction of the liquid crystal (horizontal) . The black dots seen in the image are in focus and represent defects in the liquid crystal layer, corresponding to the location of each group of four nanotubes upstanding from the substrate. As the applied field was increased, the nanotube electrodes could be seen to switch clearly at 3.1 Vp-p, which corresponds to the expected Freedrickzs transition electric field. Fig. 5B shows the same device at 4.6 Vp-p applied field and the nanotubes are now all fully switched. The distortion in the liquid crystal director can be seen in this image as the darker regions at the edge of each nanotube group. The distortion seen in the liquid crystal in Fig. 5B is due to the fact that the liquid crystal material was planar aligned in the horizontal direction. Hence as the liquid crystal reorients to the field of the nanotube

electrode it forms a twisted structure centred on each nanotube group .

In Fig. 6A, the same array is shown but with the microscope adjusted out of focus. The image in Fig. 6B is the same area with an applied field of 4.5 Vp-p. The lensing function of the liquid crystal layer can be clearly seen as the defect black dots are now in focus.

Figure 7 shows the profile of a negative dielectric anisotropy (δε) liquid crystal (LC) material (20 μm thickness ZLI4478-000 liquid crystal material homeotropically aligned) about a single carbon nanotube disposed on a silicon substrate. The circles around the CNT show how the LC director is reacting to the field of the CNT and forming a simple micro-optical component. This has the effective performance of a single lenslet on the substrate. When an electric field is applied, the molecules move to align parallel (for positive δε LC materials) or perpendicular (for negative δε LC materials) to it and the lenslet structure is removed from the substrate.

Figs. 1-3 and 7 illustrate the formation and effect of single nanotubes . The present inventor has found that groups of nanostructures can have a useful effect, as demonstrated by Figs. 4-6. This will now be discussed in more detail.

Fig. 8 shows a similar view to Fig. IB except that Fig. 8 shows the effect of multiple carbon nanotubes formed on the substrate. As can be seen in comparison with Fig. IB, the electric field profile is broader in Fig. 8 , as a result of reinforcement of the field from adjacent carbon nanotubes in the same group .

The effect on the electric field profile is modelled in Fig. 9, showing the effect of a 10 V applied field to the device. A comparison of Fig. 9 with Fig. 2 shows that the effective breadth of the electric field profile is increased by using multiple carbon nanotubes in the group.

In Figs. 2 and 9, if the carbon nanotubes were not present, there would be present an ideal parallel electrical field corresponding to a pair of parallel conducting plates. In Figs. 2 and 9, with the carbon nanotubes present, this electric field is modified by the nanotubes. The effective distance of the field from the nanotube is of the same order as the height of the nanotube.

One of the main limitations of the microlenses formed using the structures of Figs. 1-3 and 7 is that they have a tight aperture due to the localisation of the electric field around the nanotube electrodes. However, as shown in Fig. 9, the aperture can be broadened using two adjacent nanotubes in a group, due to overlap of the field profiles. Further

simulations show that if small groups of 3 or 4 nanotubes are grown within a micron of each other, then the additional field profiles overlap making a still larger optical aperture for each lenslet. Fig. 10 shows a combination of 4 nanotubes to make a single optical structure.

Modelling also shows that, for a single nanotube, the field profile is circularly symmetric about the axis of the nanotube. The curved electrical field profile in 3D can be used to control the behaviour of the liquid crystal molecules as the field intensity is strong enough to reorient them with an applied voltage of around 5V. Varying the applied voltage changes the intensity of the field, the area over which the electric field extends, and hence the alignment of liquid crystals within the area concerned. This makes the fabrication of a focus-tunable lens possible.

Fig. 11 shows a modified structure to Fig. 8. Similar features to Fig. 8 will therefore not be described again.

The device of Fig. 11 includes an aluminium substrate electrode, formed on the upper surface of the silicon substrate. As described above, this substrate electrode may be formed by cold sputtering, after formation of the carbon nanotubes. The aluminium substrate electrode also functions as an optically reflective layer. A silicon dioxide layer 42 is formed over the aluminium substrate electrode, typically

by evaporation. The effect of the silicon dioxide layer is to isolate (to some extent) the liquid crystal layer from the substrate electrode.

It is possible to provide a thicker insulating layer over the substrate electrode. An example of this is shown in Fig. 12. Here instead of (or in addition to) the SiO ₂ layer 42, a thick insulting layer 44 formed of a polymer is provided that has a height of the order of the height of the carbon nanotubes . The effect of this is to separate bulk switching of the liquid crystal (due to the substrate electrode) from localised switching due to the carbon nanotube field, the effect being even greater than for the structure shown in Fig. 11. Substantially, the only electric field now "seen" by the liquid crystal is from the carbon nanotubes. Furthermore, if the thick insulating layer is the same height as the carbon nanotubes, then the liquid crystal sits on a substantially flat surface, so that there is little or no physical interaction between the carbon nanotube and the liquid crystal. This allows the device to be physically more robust. Suitable polymer insulating layers are BCB (benzocyclobutene- based polymers) and/or Su8 (a photo-imagable epoxy, available from MicroChem Corp., Newton, MA, USA) .

Fig. 13 illustrates the electric field profile for an array of groups of single carbon nanotubes upstanding from a common substrate electrode, again modelled via finite element

analysis. The carbon nanotubes are spaced 10 μm apart in this model. Given that the carbon nanotubes have a height of 2-5 μm, this spacing amounts substantially to at least twice the nanotube height. This spacing reduces the likelihood of electrostatic interaction between the carbon nanotubes. This does not mean that there cannot be more than one carbon nanotube at each location in the array. As indicated above, it is preferred for each location in the array to have up to 4 carbon nanotubes, or more.

Fig. 14 shows in schematic form a plan view of the device from above. This is to illustrate that the upper electrode 58 need not necessarily be plain, but can itself be patterned to assist in the formation of a suitable electric field profile in the liquid crystal layer. In Fig. 14, the upper electrode is formed with annular sections 59, located so as to be centred on the axes 60 of the corresponding locations of the carbon nanotube array.

Fig. 15 shows in schematic form the electric field profile formed by applying a voltage to the carbon nanotube array. The profile due to adjacent locations in the array can overlap. Furthermore, different voltages may be applied to different locations in the array (although this is not shown in Fig. 15) . In this way, the electric field profile can be tailored and varied to suit a particular application. For example, a 2 dimensional array of carbon nanotubes, each

location in the array being individually addressable to apply a desired voltage at that location, can be used to provide a tunable optical powered device in an adaptive optical system such as a wavefront sensor (i.e. a Shack-Hartmann sensor) or digitial video camera.

Other applications include enhanced pixel structures for displays and the combination of lenslet elements with a fixed aperture array to give a very efficient high contrast display.

The ability to address individual locations in the array thus allows the creation of a new class of modally addressed devices to be fabricated in which adjacent locations in the array distort the liquid crystal director and interact with their nearest neighbours to create an arbitrary modal wavefront as shown in Fig. 15. This process is further enhanced by patterning the upper electrode to suit the desired application, as illustrated in Fig. 14. This creates more flexibility in the type and structure of the electric field and therefore the optical element. This means that, in effect, any complex wavefront can be generated. This has application in adaptive optical interconnects, astronomy, telecoms and holographic projectors.

One particular application of a 2 dimensional array is as a diffuser for a display, e.g. a heads-up display. Known diffusers aim to improve the perception of a pixelated image

to a viewer. Some known diffusers rely on static optical elements (e.g. ground glass) but these can provide stationary- visible artefacts, which are undesirable. These may be removed by moving (e.g. rotating) the optical element. However, using a moving optical element is not preferred, particularly in a head-up display that is incorporated in a helmet. Thus, a 2 dimensional variable tunable optical element according to an embodiment of the invention can provide a variable degree of focus and defocus with respect to time and area. Assuming a location spacing in the array of 10 μm as suggested above, a 10 mm by 10 mm device can have a 1000 by 1000 array. This provides a suitable basis for a diffuser. Smaller location spacings are possible, e.g. 7 μm.

Using a similar construction, an embodiment of the invention may be used as part of an autofocus system in an optical apparatus. Using suitable voltage control, the focal length of the lenslets in the array can be varied. This is of use in determining the optimum focus for a diffuse sample viewed under a light microscope.

The experimental details above indicate the formation of carbon nanotubes on a silicon substrate. Carbon nanotubes can also be grown on top of metals such as gold, and thus they can be grown connected to an electrode (optionally one which has been patterned before CNT growth) . One application of this technique is where in-plane electric fields are

required with liquid crystal materials such as flexoelectrics, vertically aligned nematics and blue phase materials. CNTs may be grown in an array, each location in the array being individually addressable to create a lateral field between the carbon nanotubes .

A typical form of in-plane electrode structure (not within the scope of the invention) is shown in Fig. 16A where there are provided two individually addressable substrate electrodes 62, 64. Due to the low-profile nature of the electrodes, the electric field profile has a distorted shape, and this affects the properties of the device. A modified version of this device is shown in Fig. 16B, in which carbon nanotube 66 is formed at the edge of substrate electrode 62 and carbon nanotube 68 is formed at the edge of substrate electrode 64. Due to the height of the carbon nanotubes compared with the height of the substrate electrodes, the electric field profile is far more uniform and linear.

Using the structure of Fig. 16B, many different field profiles can be created by varying the carbon nanotube voltages (Vl, V2 ) as well as the voltage on an upper electrode (V3 ) (upper electrode not shown in Fig. 16B) to make more complex field structures. As before this upper electrode may also be patterned to give further design options to be optimised in the device design. A top view of the device of Fig. 16B is shown in Fig. 16C.

In order to simplify the fabrication of the device of Figs. 16B and 16C, it is possible to grow dense grass CNTs at the edges of the substrate electrodes to make them physically bigger and also more robust. This is shown in Fig. 17.

The number of electrodes is not necessarily limited to two. Fig. 18 shows a 4 electrode CNT device where the electric field is set between the 4 electrodes and can be used to manipulate LC molecules in 3D. Each CNT (or groups of CNTs) is addressed by a separate voltage Vl to V4. This is useful for flexoelectric and blue phase materials where helical orientation of the LC can be manipulated by electric field. An LC lasing material may also be added to create a LC laser which can be electrically directed or tuned. The top electrode field is not shown in Fig. 18 but it can be included to enhance the switching performance between the 4 CNT electrodes and it can also be pattered such as in Fig. 14

Suitable blue phase materials are disclosed in Coles and

Pivnenko (Coles H.J. and Pivnenko M.N. , "Liquid crystal 'blue phases' with a wide temperature range", Nature, 2005 August 18; 436(7053) : 997-1000), the content of which is hereby incorporated by reference in its entirety.

Preferred embodiments of the invention have been described by way of example. On reading this disclosure, modifications of

these embodiments, further embodiments and modifications thereof will be apparent to the skilled person and as such are within the scope of the present invention.