GB2419501A - Holographic lithography using geometrical shapes - Google Patents

Abstract

A method and system for generating holographic diffraction patterns is disclosed. The method includes the steps of defining at least one geometrical shape; generating at least one line segment to represent the at least one geometrical shape; calculating a line diffraction pattern on a hologram plane, including calculating the Fresnel diffraction equation for an impulse representing the at least one line segment with a line width control term and a line length control term; and adding vectorially, where there are two or more line segments, the line diffraction patterns to form the holographic diffraction pattern. The method enables generation of two and three-dimensional shapes through the use of the line width and line length control terms and identifying where any lines cross so that the diffraction pattern can be suitably corrected.

Description

1 HOLOGRAPHIC LITHOGRAPHY 3 The present invention relates to holographic 4

lithography and particularly, but not exclusively, to computer generated holographic lithography in 6 three-dimensions. ë ë..

8 Lithography and, in particular, photolithography is i. 9 a well-known technique in semiconductor and printed circuit board (PCB) manufacture for creating 11 electrical components and circuits. Photolithography ë . . 12 involves placing a mask in front of a substrate, ...

. 13 which has been covered by a layer of photoresist, 14 before exposing to light. The areas of photoresist exposed to the light react and change chemical 16 properties compared with the unexposed photoresist.

17 The photoresist is then developed for removing 18 either the exposed portions of photoresist for a 19 positive resist or unexposed portions for a negative resist. The pattern formed in the photoresist allows 21 further process steps to be performed on the 1 substrate, such as, but not limited to, etching, 2 deposition or implantation.

4 The resolution of photolithography is limited by the diffraction of light from the mask features. As the 6 separation between the mask and the substrate 7 increases, so the minimum feature size increases, 8 thus fine-line photolithographic methods are only 9 suitable for flat surfaces. Photolithography on non planar surfaces has been achieved by moulding the 11 mask to the shape of the substrate prior to 12 exposing. This specialized technique is only 13 suitable for large simple shapes.

Holographic masks have been constructed using a A. 16 traditional Total Internal Reflection (TIR) .

17 holographic technique to pattern sub-micron features ë 18 onto large (for example, 15 x 15 inch) flat A. 19 substrates. The holographic mask is much more robust to defects than a standard mask and does not need to 21 be in intimate contact with the substrate in order . . 22 to generate high definition features. Techniques A. . . 23 have also been devised for projecting a pair of TIR 24 holographic masks onto a spherical substrate. The technique involves a complicated optical set-up to 26 generate the holograms.

28 Systems for creating Computer Generated Holograms 29 (CGHs), mainly for use in holographic displays, have also been devised. CGHs are created by defining an 31 object or shape geometrically inside a computer and 32 computing the required patterning of a diffraction 1 mask. A holographic image of that object is created 2 when a suitable light source is emitted towards the 3 diffraction mask.

A CGH system designs the holographic interference 6 pattern which is plotted or printed. A hologram is 7 generated when the pattern is exposed to a 8 monochromatic light source. In common use, and in 9 this context, CGH describes the whole process of creating a hologram from generation of interference 11 pattern within a computer to exposure of the pattern 12 to a light source.

14 Conventionally, CGH patterns for the projection of a light distribution into a 3D volume have been 16 calculated in a number of ways: 17 1.split the volume into a number of slices and <.

18 compute the Fresnel Diffraction Formula (FDF) . 19 for each slice; 2.split the volume into a number of planar 21 segments at various inclinations to the .- . 22 hologram plane and superimpose the results of .

a. 23 the FDF for each planar segment; and 24 3. decompose the object within the volume into line segments and superimpose the results of 26 the FDF for each line segment.

28 The first method requires an optical calculation for 29 every slice through the object volume, each comprising a two- dimensional Fourier Transform and 31 multiplicative factors. Similarly, the second 32 method requires calculation of a two-dimensional 1 Fourier Transform, multiplication by exponential 2 phase factors and a coordinate transform for each 3 plane into which the object has been split.

4 Calculation of diffraction patterns using these methods for large or high-resolution diffraction 6 masks is computationally expensive.

8 Calculations based on the third method are more 9 efficient, because the pattern in the hologram plane can be calculated analytically. This is demonstrated 11 in 'Computer-generated holograms of three 12 dimensional objects composed of line segments" 13 Ch.Frere, D. Lesenberg, O. Bryngdahl, J Optical 14 Society of America 3 (1986) 726-730, where the technique is used in relation to generating a e. 16 holographic display. Unfortunately, this method does sees e 17 not provide adequate means for precisely controlling 18 line width and length and therefore cannot be used 19 for precise applications.

21 According to a first aspect of the present invention .

. 22 there is provided a method of generating a ë I 23 holographic diffraction pattern, the method 24 comprising the steps of: (i) defining at least one geometrical shape; 26 (ii) generating at least one line segment to 27 represent the at least one geometrical shape; 28 (iii) calculating a line diffraction pattern on a 29 hologram plane, including calculating the Fresnel diffraction equation for an impulse 31 representing the at least one line segment 1 with a line width control term and a line 2 length control term; and 3 (iv) adding vectorially, where there are two or 4 more line segments, the line diffraction patterns to form the holographic diffraction 6 pattern.

8 Preferably, the at least one geometrical shape is a 9 three-dimensional geometrical shape.

11 Preferably, the line diffraction pattern is defined 12 by a function H(x,y) and the projected image as a 13 result of a suitable radiation source being 14 diffracted by the line diffraction pattern is defined by a function U(u,v).

17 Preferably, the line width control term is SINC(izY), 18 where w is the width of the line segment and is

19 the wavelength of a suitable radiation source...DTD: 21 Preferably, the line length control term is .

I. 23 I P( j Liz) , where h is the length of the line 24 segment.

26 Preferably, step (iii) comprises calculating a 27 partial holographic pattern representing an area of 28 the holographic plane less than the total area of 29 said holographic plane.

31 Preferably, the method further comprises the step of 32 identifying any point in the at least one 1 geometrical shape wherein two or more line segments 2 will cross, calculating a point diffraction pattern 3 of an identified point and subtracting the point 4 diffraction pattern from the holographic diffraction pattern.

7 According to a second aspect of the present 8 invention, there is provided a holographic 9 lithography system comprising: geometrical shape definition means for defining 11 a geometrical shape; 12 holographic diffraction generation means, 13 wherein the geometrical shape is defined by at least 14 one line segment, a line diffraction pattern is ..

calculated on a hologram plane including a line 16 width control term and a line length control term 17 and, where there is more than one line diffraction 18 pattern, the line diffraction patterns are added 19 vectorially together to form a holographic diffraction pattern; 21 holographic mask generation means, wherein the Ace.

22 complex values of the holographic diffraction 23 pattern are used to generate a holographic mask; 24 photoresist deposition means, wherein photoresist is deposited onto a substrate; and 26 monochromatic light exposure means, wherein the 27 holographic mask is aligned to the substrate and 28 exposed to the monochromatic light.

In this context, monochromatic light is taken to 31 mean narrow band illumination, as typically seen 1 from laser sources, but may include other relatively 2 narrow band light sources.

4 Preferably, the photoresist deposition means deposits electrodepositable photoresist (EDPR) 6 enabling uniform photoresist coverage of the 7 substrate.

9 Subsequent steps after exposure of photoresist use standard processes such as metal build-up, metal 11 etch and exposed or unexposed photoresist removal 12 from the substrate.

14 According a third aspect of the present invention there is provided a computer program product 16 directly loadable into the internal memory of a 17 digital computer comprising software code portions .. e.

18 for performing the method according to the first 19 aspect of the present invention. e 20 .

e 21 Embodiments of the present invention will now be eee.

22 described, by way of example only, with reference to 23 the accompanying drawings, in which;

Fig. 1 illustrates a prior art standard

26 photolithography process on a non-planar 27 substrate; 29 Fig. 2 illustrates an embodiment of a holographic photolithography process according to the present 31 invention on a non-planar substrate; 1 Fig. 3 illustrates an embodiment of a holographic 2 projection of a line segment along the length.

4 Fig. 4 illustrates an embodiment of a holographic projection of a line segment across the width.

7 Fig. 5 illustrates an embodiment of layers of a 8 gray-scale/binary phase mask; Fig. 6 illustrates an embodiment of a holographic 11 photolithography system according to the present 12 invention; 14 Fig. 7a illustrates an embodiment of a first part of a lithography process according to the present 16 invention; ë

I

18 Fig. 7b illustrates an embodiment of a second part 19 of a lithography process according to the present invention) . a 21 I. 22 Fig. 8 illustrates an embodiment of a maskless 23 photolithography process based on a single 24 reflective Spatial Light Modulator (SLM); and 26 Fig. 9 illustrates an embodiment of a maskless 27 photolithography process based on dual transmissive 28 SLMs.

Referring to Fig. 1, a prior art photolithography

31 process has a mask 10, substrate 12, photoresist 14 32 and electromagnetic radiation 16, which in this case 1 is a suitable light source. The substrate 12 is non 2 planar which means that the mask 10 cannot sit 3 directly on the surface of the photoresist 14 on all 4 of the required area. Where there is a gap between the mask 10 and the photoresist 14, the light 16 6 diffracts from the mask 10 before reaching the 7 photoresist 14. The minimum feature size that the 8 mask 10 can produce is affected as the diffraction 9 of the light 16 impinges on a larger area of photoresist than the area of the mask 10.

12 Referring now to Fig. 2, a holographic mask 20 is 13 now used above the non-planar substrate 12. The 14 holographic mask 20 diffracts the light 16 appropriately according to the distance between the 16 holographic mask 20 and the photoresist 14. Exposed ë 17 areas of photoresist are targeted according to how 18 the light 16 is expected to diffract through the 19 holographic mask 20. . .

21 The present invention enables holographic masks to . 22 be generated without creating a physical object to 23 record. The required shapes or patterns are defined 24 in terms of a three-dimensional co-ordinate space and a holographic pattern is generated at a defined 26 distance from the shapes in the co-ordinate space.

27 The holographic pattern is usually termed a Computer 28 Generated Hologram (CGH) as the patterns are 29 normally created within computers. A holographic mask can then be created from the holographic 31 pattern to enable the photolithography of a 32 substrate.

2 Once a shape has been defined in the co-ordinate 3 space, the shape is separated into one or more line 4 segments. For each line segment a line diffraction pattern is calculated.

7 A single line in space is firstly defined as: 8 U(u,v)=(U) (1) that is, an impulse at u=0 extending to + infinity 11 in v-direction, where u and v represent an object 12 co-ordinate plane within which the line, or shape, 13 is defined. ë

.... 15 A hologram co-ordinate plane is defined by x and y 16 with the hologram plane and object plane sharing a 17 common optical axis z. .e..

18 19 For a hologram-image separation z, the line diffraction pattern of the impulse on the hologram .

21 plane H(x,y) is given by:

I

2232 H(x,y)=exl,j (2) where is the wavelength of the illuminating 26 monochromatic light source.

28 The equivalent expression to (2) for a line at an 29 angle to the optical axis is given by: 331O x) 1 Where ax is the distance between the mask and the 2 substrate at a given position along the x-axis.

4 Truncating the analytical distribution H(x,y) to a finite sized mask of sides a and b results in a 6 reconstructed image U(u,v) in the form of equation 7 (4): 8 U(u, v) = f (u). - SINCERE) 11 The image resulting from the truncated distribution 12 H(x,y) takes the form of a SINC function in the v 13 direction, modulated by a Fresnel Integral, flu), in 14 the u-direction. The intensity of the line segment in the image plane therefore varies as SINC2 in the ë . 16 v-direction and can be approximated by a RECT .

17 function for large values of b in the u-direction. . 18

19 SINC, in the context of this invention, is defined as shown in Equation (5).

21 SING(x) = sin sin( ) 22 ( ) 24 RECT, in the context of this invention, is defined as shown in Equation (6).

27Ofor|x|> 28RECT(x) = 2 for|x|=- (6) 29tifor|x|< 31 Consideration of aliasing effects in the mask shows 32 that the minimum achievable line width is equal to 1 twice the sample spacing in the y-direction of the 2 diffraction pattern, independent of the hologram 3 substrate separation, z. Equation (4) suggests that for a fixed substrate 6 mask separation the only mechanism available to 7 adjust the width of a line-segment is the extent of 8 the diffraction pattern in the y-direction.

9 Maintaining a constant width of the main lobe of the SINC function for different z values results in 11 variations in the intensity of the line. It is 12 therefore difficult to maintain a constant line 13 cross-section over a large z range or to effectively 14 alter the width of a line segment. .

. 16 Equation (2)is improved in two ways from the point ë 17 of view of lithography; a line width control term is :.'.: 18 introduced to allow effective control of the width 19 of the line-segment and an integral term representing line length control is added to improve 21 the drop-off at the ends of the lines and to smooth . 22 the intensity along it.

24 The distribution for a line of width w and length h in the line diffraction pattern then becomes: 27 H(x,y)=exp(jfi> tINC(WY) | exp(j(X-u) )du (7) The line length control integral in equation (7) is 31 a function of only one variable and therefore can be 32 numerically evaluated relatively quickly, 1 alternatively it can be expressed as a Fresnel 2 Integral whose values are tabulated in most 3 mathematical software packages.

4 An equivalent expression for a line at an angle to the optical axis is given by: 0(x y) = exjj Rev tINC( wy) j ex it (1 + tan2 ())(x u)2) Fig. 3 shows a line diffraction pattern 30 and a 11 image plane 32 separated by a distance z. When a 12 light source 34 is diffracted by the diffraction 13 pattern 30, a line segment 36 is formed by the light 14 source 34 on the image place 32. An intensity plot 38 shows the intensity of light across the image 16 plane 32 corresponding to the length of the line 17 segment 36. A threshold 40 shows the light intensity .

18 required to activate a photoresist and therefore 19 shows the definition of the line segment. .

21 Fig. 4 has substantially the same features as Fig. 3 ë 22 and therefore has like reference numerals for like 23 items. In Fig. 4 the diffraction pattern 30 has been 24 rotated through 90 thereby rotating the line segment 36 through 90 . An intensity plot 42 across 26 the image plane 32 is now formed with a single thin 27 peak corresponding to the width of the line segment 28 36.

By superimposing a number of the line-segment 31 diffraction patterns described in equations (7) and 32 (8) a holographic diffraction pattern is created for 1 the geometric shape of interest. This enables 2 circuit tracks to be created over an arbitrary 3 piecewise planar surface. In this context, a 4 "piecewise planar surface" is a surface constructed from planar surfaces at various angles and 6 orientations.

8 Superimposing, in this case, involves vectorially 9 adding the complex solution of the line diffraction patterns to generate the holographic diffraction 11 pattern. 'Complex" in this sense relates to complex 12 numbers, as the solution of the line diffraction 13 patterns will have both amplitude and phase 14 components.

16 When the line-segment diffraction patterns are .

17 superimposed, points at which two or more line . 18 segments intersect generate an image in which the 19 intensity at the intersection is much greater than the rest of the line-segments. This can result in . 21 over exposure of the photoresist in this area. To 22 eliminate this effect, the diffraction pattern from 23 a single point located at the intersection of the 24 two lines is calculated and the result subtracted from the original mask. The diffraction of light 26 from the point pattern will destructively interfere 27 with the line segment patterns and reduce the 28 intensity at the intersection appropriately.

Where the line-segment diffractions patterns are 31 better described as rectangles than lines, that is 32 they have significant width, a diffraction pattern 1 from the overlapping section must be subtracted. For 2 example, when two intersecting line-segments of 3 equal width intersect a square diffraction pattern 4 representing the intersection must be generated and subtracted from the overall image.

7 To further reduce the complexity of calculations 8 required, and therefore reduce processing time on a 9 computer, each line-segment diffraction pattern is localized on the hologram plane. This is achieved by 11 calculating a partial line-segment diffraction 12 pattern for an area on the hologram plane less than 13 the total area of the hologram plane. The partial 14 line segments are then added vectorially whilst maintaining their relative positions in the hologram a16 plane to create the holographic diffraction pattern. Afar

.. 18 Localisation relies on the superposition of 19 individual line-segment CGH calculated over an area smaller than the area of the mask. The size of the 21 calculation area determines the quality of the .e 22 resulting line image on the substrate and is limited 23 by the resolution of the CGH mask. The calculation 24 area can be chosen as the largest area allowable for the given CGH mask resolution.

27 Localisation of the CGH calculation area to the 28 region directly above the corresponding substrate 29 area means smaller mask-substrate separations can be achieved for a given mask resolution. It is 31 important to ensure that higher diffraction orders 32 from each line segment are attenuated sufficiently 1 such that when combined their intensity is less than 2 the fixing threshold of the photo-resist.

4 The holographic diffraction pattern can then be converted to a format suitable for fabrication. The 6 real part of equation (7) is quantized into n 7 greyscale levels. As shown in Fig. 5 the amplitude 8 of the resulting matrix is then encoded into a 9 greyscale layer as a rectangle of area proportional to the greyscale value contained within each sample 11 cell. This pattern can then be transferred onto a 12 mask blank using standard processing techniques 13 ( i. e. selective etching of chrome coating on mask 14 blank).

c 16 It should be appreciated that this is only one of a eee.

I 17 number of possible ways of making a mask. It could 18 be envisaged that a continuous tone mask could be 19 used in place of a mask with a number of greyscale levels. 21

I 22 In alternative (maskless) embodiments, a fixed . 23 holographic mask may be replaced by one or more 24 Spatial Light Modulator (SLM). Fig. 8 shows an embodiment where the fixed mask is replaced by a 26 single reflective SLM. Fig. 9 shows an embodiment 27 where the fixed mask is replaced by two transmissive 28 SLM's. The SLMs control light amplitude and or 29 phase in a programmable and time-varying way, in an analogous fashion to fixed glass masks. A single 31 SLM may be used in conjunction with a light source 32 114 beam shaper 116 and computer 118 in order to 1 realise binary amplitude modulation type CGH's. A 2 dual SLM system (shown in Fig. 9) can be used to 3 realise full complex CGH mask designs. No 4 fundamental alterations are required to the CGH calculations previously described, although it may 6 be desirable to optimise a CGH pattern for best 7 performance under SLM modulation. SLM' s have 8 already been employed to project holograms enabling 9 ultra high speed data storage, but they have not been used for holographic photolithography.

11 Replacement of the mask with an SLM allows 12 numerically controlled holograms to be rapidly 13 produced and projected. It allows sequences of such 14 holograms to be projected sequentially and enables step-and-repeat, step-and-scan and multiple exposure .

.: 16 methods. This in turn enables exposure over large .

, . 17 surfaces, with finer detail and the creation of more 18 complex images. ..

This system also allows a mechanical scanning system :. 21 (not shown) to scan the SLM's over large area .

. 22 substrates, thus removing the limitation of mask .

23 size. In this regard, such a mechanical scanning 24 system actually exploits the inherent localization of the previously described CGH designs, such that 26 the active area of the SLM is matched to the local 27 area of the CGH. It should be noted that other 28 embodiments of maskless systems are possible besides 29 those depicted in Figs. 8 and 9.

When generating two phase levels, the sign of the 31 real part of equation (7J can be realised using a 32 technique employed to generate phase-shift masks. r

1 This could mean either the selective etching of the 2 mask substrate or the patterning of a transparent 3 layer deposited on the mask. The depth of the 4 pattern corresponds to a 180 phase-shift to the transmitted light.

7 It is possible that additional phase levels can be 8 generated and used within the mask to further 9 improve the CGH. For example, the embodiment of two levels of phase could be extended to four or more 11 levels or even to a continuous range of phase 12 shifts.

14 The amplitude and phase-shift layers are co-aligned.

It is possible to deposit the chrome either before ë ,: . 16 or after the patterning of the phase-shift layer. .

. 17 18 Referring now to Fig. 6, a monochromatic light 19 source 60, which in this case is a laser, emits monochromatic light towards a collimator 62. The 21 collimator 62 generates a parallel beam 64 which is . 22 directed towards a holographic diffraction pattern .

23 66. The monochromatic light is diffracted by the 24 holographic diffraction pattern onto a non-planar substrate 68. The substrate 68 has a layer of 26 photoresist (not shown) from which exposed or 27 unexposed areas may be removed to enable other 28 processes, such as etching, to be performed.

A non-planar and possibly complicated substrate 31 geometry still requires an even layer of 32 photoresist. When patterning a nominally 'flat' 1 substrate, such as a semiconductor wafer or a 2 printed circuit board, the photoresist is generally 3 applied by either spinning a liquid precursor or 4 laminating a dry-film. These methods are not suitable for grossly non- planar surfaces since they 6 do not enable a uniform-thickness layer of 7 photoresist to be deposited. Spray nozzles have been 8 developed that allow modest topographies to be 9 covered uniformly, but for grossly non-planar surfaces the preferred method is to use an electro 11 depositable photoresist (EDPR).

13 Therefore, photolithography can be applied to 14 grossly non-planar surfaces through the generation of a holographic mask and use of EPDR. Complicated ë .: 16 circuits can now be patterned on non-planar

I

. 17 substrates without re-shaping masks but through the 18 use of the present invention.

Referring now to Fig. 7a and Fig. 7b, a computer I:. 21 generated hologram photolithography system 70 for ë . 22 etching a substrate 92, firstly, has a number of .

23 geometric shapes inputted 72. The geometric shapes 24 72 are then defined in three-dimensional sampled space 74. A CGH 76, or holographic diffraction 26 pattern, is generated from the sampled space 74 as 27 described previously. The CGH 76 is converted into a 28 holographic mask by: 29 binarising the CGH 78; generating greyscale and phase distributions 31 80; 32 creating machine-format patterns 82; 1 adding alignment marks 84; 2 plotting the master greyscale and phase masks 3 86; 4 create binary amplitude phase mask 88; create greyscale and phase mask from masters 6 90; 7 aligning a photoresist coated substrate to a 8 CGH mask, the CGH mask being the greyscale and phase 9 mask (step 92) and binary amplitude phase mask (step 88) combined; 11 exposing the substrate with a suitable light 12 source 94; and 13 developing exposed photoresist 96.

If a negative acting photoresist is being used the .,.. 16 process continues by: 17 etching exposed seed layer 98; 18 stripping photoresist 100; and ë 19 an optional step of increasing thickness of circuitry by electro or electroless plating 102. :. 21 ë

a. 22 If a positive acting photoresist is being used the ease 23 process continues by: 24 increasing thickness of exposed circuitry 104; stripping photoresist 106; and 26 etching exposed seed layer 108.

28 For example, the following applications are enabled 29 through the present invention: Conical Spiral Antennae - using a conical log 31 spiral geometry for an antenna results in a 32 wideband receiver with a highly directional 1 beam pattern and very little backscatter. This 2 type of antenna is useful in GPS and radar 3 applications and possibly the emerging Ultra 4 Wideband (UWB) technology. Conventionally such an antenna is difficult to construct accurately 6 and involves bending pre-etched flexible arms 7 onto a former. The present invention has 8 enabled photolithography which can produce high 9 resolution tracks directly imaged onto the antenna substrate.

11 Novel Print Head Architectures - the present 12 invention can be used in the etching of a novel 13 print head assembly involving tracks running 14 over stepped piezoelectric surfaces, a process that typically requires a direct laser writing 16 procedure. .

17 System assembly - sensors, integrated chips and 18 discrete components may be connected together 19 on a system board, having two or three dimensions, using the present invention. That . 21 is, the present invention enables deposition of 22 interconnects via photolithography/etching 23 rather than high temperature soldering, which 24 is advantageous for delicate sensors and integrated circuits. Even for a flat system 26 board, there is still a need for laying down 27 fine pitch interconntects between, essentially, 28 non-planar components.

Although the present invention has been described 31 with particular reference to generating holograms 32 for use with a three-dimensional surface, generation 1 of holograms for two-dimensional surfaces is as 2 equally applicable.

4 As the substrate size moves to liquid crystal flat panel display dimensions and beyond, the need for 6 100% defect free quality is paramount. The yield 7 achieved for these sorts of dimensions is extremely 8 low, for many prior art lithography systems due to 9 the size of the substrate. Holographic lithography in general improves yield as dust particles and 11 other imperfections are less destructive to the 12 creation of circuits due to multiple light paths.

14 The present invention obviates or mitigates the traditional restrictions in size associated with aces

16 prior art holographic lithography systems. In ë 17 particular, with prior art holographic lithography e 18 systems, large

two-dimensional substrates suffer 19 from spherical aberrations towards the edges of the substrates. The present invention includes A. ë 21 corrections for spherical aberrations inherently in .e I. 22 the creation of a diffraction pattern.

24 The technique is not restricted to photolithography and may be applied to other types of lithography in 26 substantially the same manner.

28 For the purpose of clarity, the diagrams and 29 equations associated with the present invention represent zero order holograms. It should be 31 appreciated that both the diagrams and equations can 1 be modified to represent nth order holograms without 2 departing from the scope of the invention.

4 Improvements and modifications may be incorporated without departing from the scope of the present 6 invention. caas eae. Abbe

I e. e Abbe

Claims (9)

1 CLAIMS 3 1. A method of generating a holographic diffraction 4 pattern,

the method comprising the steps of: (i) defining at least one geometrical shape; 6 (ii) generating at least one line segment to 7 represent the at least one geometrical 8 shape; 9 (iii) calculating a line diffraction pattern on a hologram plane, including calculating the 11 Fresnel diffraction equation for an impulse 12 representing the at least one line segment 13 with a line width control term and a line 14 length control term; and (iv) adding vectorially, where there are two or 16 more line segments, the line diffraction 17 patterns to form the holographic 18 diffraction pattern.

2. A method as claimed in claim 1, wherein the at 21 least one geometrical shape is a three-dimensional 22 geometrical shape.

24

3. A method as claimed in claim 1 or claim 2, wherein the line diffraction pattern is defined by a 26 function H(x,y) and the projected image as a result 27 of a suitable radiation source being diffracted by 28 the line diffraction pattern is defined by a 29 function U(u,v).

31 4. A method as claimed in any of claims 1 to 3, 32 wherein the line width control term is tPlNC.'( - ), 1 where w is the width of the line segment and is 2 the wavelength of a suitable radiation source.

4

5. A method as claimed in any of claims 1 to 4, wherein the line length control term is 7 h] P i AZ) ' where h is the length of the line 8 segment.

6. A method as claimed in any of claims 1 to 5, 11 wherein step (iii) comprises calculating a partial 12 holographic pattern representing an area of the 13 holographic plane less than the total area of said 14 holographic plane.

16

7. A method as claimed in any of claims 1 to 6, 17 wherein the method further comprises the step of 18 identifying any point in the at least one 19 geometrical shape wherein two or more line segments will cross, calculating a point diffraction pattern 21 of an identified point and subtracting the point 22 diffraction pattern from the holographic diffraction 23 pattern.

8. A holographic lithography system comprising: 26 geometrical shape definition means for defining 27 a geometrical shape; 28 holographic diffraction generation means, 29 wherein the geometrical shape is defined by at least one line segment, a line diffraction pattern is 31 calculated on a hologram plane including a line 32 width control term and a line length control term 1 and, where there is more than one line diffraction 2 pattern, the line diffraction patterns are added 3 vectorially together to form a holographic 4 diffraction pattern; holographic mask generation means, wherein the 6 complex values of the holographic diffraction 7 pattern are used to generate a holographic mask; 8 photoresist deposition means, wherein 9 photoresist is deposited onto a substrate; and monochromatic light exposure means, wherein the 11 holographic mask is aligned to the substrate and 12 exposed to the monochromatic light.

14 8. A system as claimed in claim 7, wherein the photoresist deposition means deposits electro 16 depositable photoresist (EDPR) enabling uniform 17 photoresist coverage of the substrate.

19

9. A computer program product directly loadable into the internal memory of a digital computer comprising 21 software code portions for performing the method 22 according any of claims 1 to 6.