Method of forming an optical device

Abstract

A method of forming an optical device comprises applying a laser beam to a target area of the surface so as to selectively heat material of the surface thereby to provide transfer of material due to a surface tension gradient, wherein the surface is such that, when liquid, parts of the surface at higher temperatures have a higher surface tension than adjacent parts of the surface at lower temperatures.

Claims

1. A method of forming a diffractive optical device, the method comprising: applying a laser beam to a plurality of target areas located at different positions on a surface of a substrate so as to selectively melt material of the surface; and controlling at least one of atmospheric conditions at the surface and composition of the surface such that the application of the laser beam causes a melt pool at each of the plurality of target areas to exhibit a surface tension gradient, with higher temperature areas of the melt pool having a higher surface tension than lower temperature areas of the melt pool, that results in a transfer of molten material to build up profiles on the surface, the profiles forming a diffractive structure of the diffractive optical device.

2. The method according to claim 1, comprising controlling the atmosphere at the surface such that the higher temperature areas of the melt pool have a higher surface tension than the lower temperature areas of the melt pool.

3. The method according to claim 2, wherein the controlling of the atmosphere comprises providing an atmosphere at the surface rich in a gas, relative to the concentration of the gas in air, that causes at least one of oxidation, phosphorus evaporation, carbide formation and chromium migration.

4. The method according to claim 2, wherein the controlling of the atmosphere comprises providing a CO.sub.2-rich atmosphere, relative to the concentration of CO.sub.2 in air, at the surface during the application of the laser beam to the surface.

5. The method according to claim 1, comprising providing material at the surface having a composition such that the higher temperature areas of the melt pool have a higher surface tension than the lower temperature areas of the melt pool.

6. The method according to claim 5, comprising treating the surface of the substrate with a surface active agent that reacts with material of the substrate to provide the composition.

7. The method according to claim 6, wherein the surface active agent causes at least one of oxidation, phosphorus evaporation, carbide formation and chromium migration.

8. The method according to claim 5, wherein the composition comprises one or more elements from groups 13, 14, 15 and 16 of the periodic table.

9. The method according to claim 8, wherein the composition comprises one or more elements from periods 2 and 3 of the periodic table.

10. The method according to claim 8, wherein the material comprises a high period metal and the composition comprises one or more elements from periods 4 and 5.

11. The method according to claim 5, wherein the composition comprises one or more elements selected from calcium, sulphur, manganese, silicon, titanium, zirconium, aluminium, magnesium, nitrogen, oxygen and phosphorus.

12. The method according to claim 5, wherein the material has a selected oxygen or sulphur content.

13. The method according to claim 1, comprising controlling at least one parameter of the laser beam to obtain the transfer of material.

14. The method according to claim 13, wherein the at least one parameter comprises at least one of intensity, wavelength, pulse length, and pulse repetition time.

15. The method according to claim 1, wherein, for at least one of the plurality of target areas, the laser beam has a greater intensity at a first part of the at least one target area than at a second part of the at least one target area.

16. The method according to claim 15, wherein the laser beam has an intensity above a threshold intensity thereby to reduce the transfer of material to the first part of the at least one target area in comparison to the transfer of material to the second part of the at least one target area.

17. The method according to claim 16, wherein the laser beam has an intensity above the threshold intensity at the first part of the at least one target area and an intensity below the threshold intensity at the second part of the at least one target area.

18. The method according to claim 16, wherein the threshold intensity is an intensity at which radiation of the laser beam ablates the surface.

19. The method according to claim 16, wherein the threshold intensity is a minimum intensity for which radiation of the laser beam acts to oppose the transfer of material due to the surface tension gradient.

20. The method according to claim 16, comprising applying both the laser beam and a further laser beam to the at least one target area, wherein the laser beam has an intensity above the threshold intensity and the further laser beam has a maximum intensity below the threshold intensity.

21. The method according to claim 20, wherein the method comprises: applying one of the laser beam and the further laser beam to the at least one target area; allowing the surface at the at least one target area to at least partially solidify; and applying the other of the laser beam and the further laser beam to the at least one target area.

22. The method according to claim 21, wherein application of the further laser beam to the at least one target area fills in with material at least a portion of the profile of surface material formed by application of the laser beam, or vice versa.

23. The method according to claim 22, wherein application of the laser beam forms a two peaked profile of surface material and application of the further laser beam subsequently at least partially fills a well between the two peaks, or application of the further laser beam forms a single peaked profile of surface material and application of the laser beam subsequently at least partially broadens the single peak.

24. The method according to claim 20, wherein the laser beam and the further laser beam have parameters that are controlled such as to produce a profile of material at the at least one target area that has a flat top.

25. The method according to claim 1, comprising applying a further laser beam to a plurality of further target areas, each of the further target areas being at a respective, different position on the surface, thereby to create a further melt pool at each further target area and build up the profiles of the diffractive optical device.

26. The method according to claim 25, wherein the method comprises applying the laser beam to the plurality of target areas in a sequence, wherein at least some of the target areas that are spatially adjacent to each other on the surface are temporally non-adjacent in the sequence.

27. The method according to claim 1, comprising applying the laser beam to at least one target area of the plurality of target areas a plurality of times thereby to build up material at the target area.

28. The method according to claim 27, further comprising: controlling the atmosphere at the surface such that the higher temperature areas of the melt pool have a higher surface tension than the lower temperature areas of the melt pool; and varying a composition of gas above the surface, so that the composition of the gas is different for at least some of the times that the laser beam is applied to the at least one target area than for at least some other of the times that the laser beam is applied to the at least one target area.

29. The method according to claim 1, wherein the method comprises blocking an outer part of the laser beam from reaching the surface.

30. The method according to claim 1, wherein the method comprises passing the laser beam through a beam shaper before applying it to the surface.

31. The method according to claim 30, wherein the beam shaper is configured to redistribute power from higher intensity parts of the laser beam to lower intensity parts of the laser beam.

32. The method according to claim 1, wherein the laser beam has an intensity such as to provide transfer of material due to the surface tension gradient without removal of material from the substrate.

33. The method according to claim 1, wherein the diffractive optical device comprises at least one of a metrological device, a phase scale, a hologram and a diffraction grating.

34. The method according to claim 10, wherein the high period metal is a metal selected from the group of silver, tungsten, platinum and gold.

35. The method according to claim 6, wherein the surface active agent comprises one or more elements selected from calcium, sulphur, manganese, silicon, titanium, zirconium, aluminium, magnesium, nitrogen, oxygen and phosphorus.

36. The method according to claim 25, comprising applying the further laser beam to the plurality of further target areas in a sequence, wherein at least some of the further target areas that are spatially adjacent to each other on the surface are temporally non-adjacent in the sequence.

37. The method according to claim 25, comprising applying the further laser beam to at least one further target area of the plurality of further target areas a plurality of times thereby to build up material at the at least one further target area.

38. The method according to claim 37, comprising: controlling the atmosphere at the surface such that higher temperature areas of the further melt pool have a higher surface tension than lower temperature areas of the further melt pool; and varying a composition of gas above the surface, so that the composition of the gas is different for at least some of the times that the further laser beam is applied to the at least one further target area than for at least some other of the times that the further laser beam is applied to the at least one further target area.

Description

DETAILED DESCRIPTION OF EMBODIMENTS

(1) Embodiments of the invention are now described, by way of non-limiting example, and are illustrated in the following figures, in which:

(2) FIG. 1 is an illustration of an apparatus for forming an optical device according to an embodiment;

(3) FIGS. 2 and 3 are schematic illustrations of the surface of a sample showing the profile of the surface after application of laser beam to a target region

(4) FIG. 4 is a schematic illustration of the surface of a sample showing the profile of the surface after application of a laser beam to a plurality of target regions to form a desired device profile;

(5) FIG. 5 is a schematic illustration of the surface of a sample showing the profile of the surface after application of a laser beam having an intensity in a central part of the beam that is greater than a threshold below which a central peak is formed;

(6) FIG. 6 is a schematic illustration of the surface of a sample showing the profile of the surface after application of a laser beam to a plurality of target regions to form a desired device profile in a further embodiment;

(7) FIG. 7 is a schematic illustration of the surface of a sample showing the surface of the sample after application of a laser beam to a target region; and

(8) FIG. 8 is a schematic illustration of the surface of a sample showing the profile of the surface after application of a laser beam to a plurality of target regions to form a desired device profile having flat topped regions.

(9) An apparatus for forming an optical device on a surface in one embodiment is illustrated schematically in FIG. 1. The apparatus comprises a control unit 2, for example a suitably programmed PC or a dedicated controller, which is connected to, and configured to control operation of a laser 4. The control unit shown in FIG. 1 is a DQB-PS control unit comprising a Renishaw UUC and SPA2+custom interface card. The laser 4 of the apparatus of FIG. 1 is a tripled Nd:YAG laser of 355 nm wavelength, in particular a JDSU Lightwave Q-switched Q301-HD laser, that is controlled to operate with pulse energies of the order of 100 J to 1 mJ, and pulse durations in the microsecond regime.

(10) The laser 4 is aligned with focusing optics, comprising cylindrical lens 6, that in operation focus the laser beam produced by the laser 4 on a sample 8 mounted on a platform 10. Further optical components can be included between the laser 4 and the sample 8 to provide control and focussing of the laser beam as desired, according to known techniques.

(11) According to one mode of operation of the embodiment of FIG. 1, the laser 4 can be controlled to produce a laser pattern on the surface of the sample 8 in the shape of a significantly elongated Gaussian spot, which is formed by passing a high quality circular Gaussian beam from the laser 4 through the cylindrical lens 6 to produce a spot at focus which is of the order of 1.5 mm in breadth by 4 m long. The beam may comprise a pulse or a train of pulses. Stops (not shown) are provided to block an outer part of the beam in this embodiment.

(12) The platform 10 on which the sample is mounted is a motorised platform that is connected to the control unit 2. The motorised platform in this case is a Danaher Dover Fibrebeam precision motion linear motor stage with Elmo controller. The control unit 2 is operable to move the motorised platform 10 relative to any desired position of the laser 4. By controlling movement of the platform 10 and by controlling operation of the laser 4, the control unit 2 is able to apply a laser beam having selected characteristics at a series of selected positions, also referred to as target areas, on the surface of the sample 8, so as to form an optical device having desired properties.

(13) The control unit 2 is also connected to a controllable gas supply 12 that can output gas to flow across the surface of the sample 8. In the embodiment of FIG. 1, the gas supply 12 comprises a cylinder of CO.sub.2, with a controllable flow valve that, in operation, is controlled by the control unit 2 to provide a desired flow rate of CO.sub.2 gas across the surface of the sample 8. The CO.sub.2 may be applied at a target area or across the sample as a whole. In alternative embodiments the sample is provided within a closed housing, sensors are provided to monitor the composition of gas above the sample 8 and the control unit 2 controls operation of the gas supply 12 in dependence on output from the sensors to control precisely the level of CO.sub.2 in the atmosphere above the sample.

(14) It is a feature of the embodiment of FIG. 1 that by providing a CO.sub.2-enriched atmosphere adjacent to the sample surface, the movement of material from the surface of the sample 8 caused by application of the laser beam can be altered. For example, it has been found that, when treating a metallic surface with a pulse of laser energy below a particular threshold under CO.sub.2 assist gas the material of the surface can be caused to collect at the centre of the laser spot. That is illustrated schematically in FIG. 2, which shows a portion of the sample 8 after application of a laser beam produced by the laser 4. The sample 8 comprises a metal layer 20, in this case stainless steel, for example either martensitic spring steel, such as Sandvik Chromflex 7C27Mo2 Martensitic Stainless Steel (C 0.38%, Si 0.40%, Mn 0.55%, Cr 13.5%, Mo 1.00%), or cold-rolled austenitic-ferritic precision type 316 stainless steel strip. Upon application of the laser beam the metal layer in the region of the laser beam melts and is drawn towards the centre of the target region irradiated by the laser beam. When the laser beam is switched off or moved the metal solidifies and forms a profile having a peak as shown in FIG. 2.

(15) There is no substantive removal of the metal, and the material of the peak is drawn from the lower temperature melt pool either side of the centre of the laser spot. Note that the aspect ratio of this mark is exaggerated for clarity. The manufacture of an 8 m pitch scale might require a central peak of approximately 4 m length, and the height of the feature suitable for a phase scale which uses 800 nm illumination wavelength would be of the order of 200 nm. The profile of FIG. 2 is shown in FIG. 3 with a more realistic aspect ratio. It should be noted that FIG. 3 is also a schematic illustration and is not to scale, but provides a more realistic illustration of the ratio of pitch to depth than FIG. 2 for the described embodiment.

(16) The presence of the CO.sub.2 enriched atmosphere adjacent to the metal layer 20 causes the formation of a peak, rather than a well, at the centre of the laser spot. In the presence of a normal atmospheric composition, without CO.sub.2 enrichment, the metal of the sample of FIGS. 2 and 3 flows away from the centre of the laser spot, and that application of the laser beam at power below an ablation threshold causes formation of a well rather than a peak, again without substantial removal of material.

(17) The control of gas composition adjacent to the sample layer provides a further control parameter that can significantly alter the flow behaviour of sample layer material under action of applied laser radiation and can provided for improved control of formation of optical devices, as will be described in more detail below.

(18) In one embodiment, a succession of the profiles shown in FIGS. 2 and 3, also referred to as marks, are formed adjacent to each other by repeated application of the laser beam at different, selected target areas. It is not necessary to produce adjacent marks in sequence, and there can be thermal advantages to producing a non-adjacent mark after producing a mark, and returning to produce the adjacent mark later. For example, marks can be produced with gaps between which are then filled in later. This method prevents local heat build up from distorting the form of the marks or accuracy of the scale. For instance, it can be ensured that the material of the originally produced mark has fully solidified before producing the adjacent mark and thus unwanted additional flow of the original mark can be avoided.

(19) By suitable control of the laser beam and selection of the target area, marks of the correct size and a spacing which produces a desired degree of overlap can be produced in order to provide an optical device having a desired profile that in turn provides desired optical characteristics. For instance, the pulse energy, and duration can be controlled, along with the position of the surface being marked with respect to the optimum focus of the beam to produce a pattern of the correct dimensions in height, breadth and length.

(20) In some embodiments, multiple lower energy pulses are used to generate small changes of material profile which, together, produce a required mark. This has the advantage of enabling a finer degree of control of the surface form, uses statistical averaging to mitigate the effects of variations between laser pulses, and has a polishing effect on the surface being marked further improving reflectivity and thus scale performance or other optical performance. In some embodiments up to 60 pulses are used to build up a single mark.

(21) FIG. 4 shows a device that has been formed in the surface of the material exposed to the laser beam energy pulses, and has a surface layer 24 that has a close approximation to a sine wave. Again, the aspect ratio of the mark is exaggerated for clarity.

(22) The marks of FIGS. 2 to 4 are shown in side-view and it will be appreciated that they also have a breadth (in the case of these diagrams, the dimension extending into the plane of the page). The combination of marks shown in FIG. 4 has a breadth of the order of 1 to 5 mm, and in the case of a phase scale might usually extend in length (the dimension across the width of the page) from the order of 10 mm up to many meters.

(23) The long axis of the Gaussian laser beam spot can have the unwanted side effect of creating shallower marks near its ends. However, if multiple passes are performed with a beam having a Gaussian profile in the long axis, overlap between adjacent steps being used to increase the depth of the marks away from the centre of the beam.

(24) In some embodiments, including the embodiment of FIG. 1, the effect of shallower marks being produced by end parts of the beam is avoided or mitigated by using the stops to block the end parts of the beam from reaching the surface being marked, which has the advantage of producing marks which are useable all the way across their breadth. However, the use of stops reduces the breadth of the mark produced by each laser beam pass, and again in some embodiments multiple passes over the surface being marked are used with the position of the laser spot relative to the surface stepped along the breadth dimension between passes. That can produce a wider scale or other optical device, which can easier to install and align.

(25) If sufficient power is available the elongated axis of the cross-sectional beam profile can be extended further and then stops used at each end to block the end parts of the beam from reaching the surface being marked. A combination of selectable stops and overlapping passes along the breadth of the scale can be used to make a pattern of practically unlimited breadth, usable all the way to its edge.

(26) In an alternative embodiment, the laser beam is passed through a beam shaper, which transforms the Gaussian form of the beam in the elongated axis in to a flat-top profile. This can take the form of a refractive beam shaper, for instance a -shaper as supplied by Molecular Technology GmbH. This device re-distributes power from the centre of the beam to its edges, but only along the extended axis of the spot. This has the advantage of producing the desired mark form along the whole breadth of the mark without requiring the significant extra laser power needed when stops are used to block the ends of the long axis of the beam. It will be understood that other known means of producing a more uniformly distributed beam along the breadth of the mark may also be used in further alternative embodiments.

(27) The pattern of marks illustrated in FIG. 4 is produced under a CO.sub.2 rich atmosphere which causes peaks rather than wells of material to be formed at the centre of the laser beam. The inventors attribute this change in flow behaviour to a reversal of the surface tension gradient with temperature, caused by the presence of the CO.sub.2 enriched atmosphere, which causes material to flow from lower temperature regions to higher temperature regions (as opposed to the flow from higher temperature regions to lower temperature regions that occurs upon application of the laser beam at normal CO.sub.2 levels).

(28) A similar pattern of marks to that shown in FIG. 4 could also be produced by application of the laser beam at normal CO.sub.2 levels, with the pattern being built up by formation of a series of wells at the centre of the laser beam spot rather than by formation of a series of peaks. It has been found however that marks or patterns of marks that are not otherwise obtainable or that are otherwise obtainable only with increased difficulty or time may be formed under reversed surface tension conditions (for example under a CO.sub.2 rich atmosphere).

(29) For instance, it has been found that when a certain laser beam intensity is exceeded under a CO.sub.2-rich atmosphere, the mark produced by the laser beam can take on the form illustrated schematically in FIG. 5 The inventors believe that the mark is formed by the surface tension gradient with temperature (associated with the CO.sub.2-rich atmosphere) causing material to flow from lower temperature regions to higher temperature regions and possibly a flow reversal encountered towards the end of the melt. In the case of FIG. 5 the intensity is still sufficiently low that little significant, if any, ablation of material occurs and the surface is not significantly roughened. The threshold intensity may be a threshold average power, for example maintained for a threshold period of time. In the case of FIG. 5 the threshold intensity was obtained for a series of 10 pulses, 500 J per pulse. In practice a laser beam of intensity higher than the threshold may be used to form marks, for example a series of 10 pulses, 600 J per pulse in the case of FIG. 5 It will be understood that the threshold intensity may be determined by experimentation for different materials and conditions.

(30) In the case of a further sample of the Chromeflex steel, the transition from surface tension effects alone to a combination of pressure and surface tension effects was found to occur for pulses of energy around 400 J focused on an area 4 mm4 m.

(31) It has been found that the mark made by a single application of the laser beam above the necessary threshold to produce a mark having a shape like that shown in FIG. 5 has a greater length (the direction along the plane of the paper) by between 20% and 50% than the corresponding mark produced under the same conditions by a single application of the laser beam having an intensity below this threshold (for example as illustrated in FIG. 2). However, the mark produced by the laser beam having an intensity above the threshold has two peaks rather than a single peak, and thus the use of a laser beam having an intensity above the threshold in a CO.sub.2-rich atmosphere can produce patterns having more closely spaced peaks despite the greater size of the individual mark produced by a single application of the laser beam. Usually the intensity is maintained below a further threshold intensity, for example beyond which ablation effects dominate surface tension effects.

(32) In some embodiments the intensity of the laser beam is controlled to be close to the lower threshold so that that well between the two peaks is of negligible depth, thus effectively forming a single broader peak.

(33) In a further embodiment, a series of marks at different target areas on the surface are produced by application of a laser beam having an intensity above the threshold in a CO.sub.2-rich atmosphere, with the resulting pattern being illustrated schematically in FIG. 6. Each mark is merged with an adjacent mark to produce a continuous pattern which resembles a sinusoid. In this case each mark made by an application of the laser beam produces two peaks. As each mark is only 20% to 50% longer, but includes two peaks rather than a single peak, the pitch of the scale is significantly shorter (between 60% and 75%) than the pitch of scale that would be produced using a laser beam below the threshold. Thus, in FIG. 6 here are 42 peaks, compared with 24 peaks in the same length (as illustrated in FIG. 4 that are produced using a laser beam having an intensity below the threshold at which double peaks are formed).

(34) In the embodiment of FIG. 6 when the spacing of the marks is set such that an even pitch is achieved, and the volume of material is substantially constant (that is, substantially no material is ablated during laser processing) an uneven feature depth can result. In the case where the pattern is used as a phase scale the contrast between the fringes generated by the phase scale can be reduced by uneven feature depth, which can be a source of noise in the electrical signals output by a readhead when reading the phase scale. Nevertheless, a scale of smaller pitch can be produced rapidly using the method. Further processing can also be used, if desired, to even out feature depths, for example using additional ablation or filling steps.

(35) In embodiments that apply a laser beam having an intensity above the threshold it can be particularly useful to use stops or a beam shaper to ensure that the central part of the beam has an intensity above the threshold across the entire breadth of the beam. Otherwise, the edges of the beam may form a single peaked profile rather than a desired double peaked profile.

(36) In a further embodiment, a laser beam having an intensity above the threshold and a laser beam having an intensity below the threshold are applied successively by the laser 4 to the same target area under reversed surface tension conditions (for example under a CO.sub.2 rich atmosphere). It has been found that by applying such successive laser beams above and below the threshold, the different shaped marks formed by the laser beams combine to form particularly useful optical features. It will be understood that the threshold intensity for a particular surface material under selected conditions can be readily determined by the person skilled in the art.

(37) For instance, the double peaked mark produced by a laser beam having an intensity above the threshold can effectively be combined with the single peaked sinusoid-like mark produced by a laser beam having an intensity below the threshold, in order to improve over a pure sinusoid-like profile by flattening out the peaks and troughs of the sinusoid. That makes the form of the mark more closely resemble a square wave, which is the theoretical ideal for high contrast fringes, as illustrated schematically in FIG. 7 for a depth of 200 nm deep and a length of 4 m. Again, the aspect ratio (height to width ratio) is exaggerated for clarity.

(38) The combined mark illustrated in FIG. 7 is similar to that produced by a single laser beam with power below the threshold that produces a double peak, but has a flatter top to the peak. It might be expected that the surface profile formed by the first laser beam would distort the profile formed by the second laser beam such that there would be a distortion of the illustrated additive combination of marks. However, any such distortion, if present, does not seem to cause any significant difficulties.

(39) In a further mode of operation, the apparatus of FIG. 1 is operated to form a scale with flattened peaks and troughs by making at least two passes of the laser beam over the surface to be marked to form marks at each of a plurality of target areas, directing pulses below the threshold that produces the double peak and pulses above this threshold at the surface to each target area. A schematic representation of the resulting pattern of material that is generated is shown in FIG. 8. Again the aspect ratio is exaggerated for clarity.

(40) The combined multi-pass method described in relation to FIG. 8 provides additional process parameters which can be optimised to achieve a desired depth and width of the optical features, and the flattening of peaks and troughs. For instance, the pulse power and degree of focus can adjust a feature depth and width; the degree to which the pulse is above threshold at which a double peak is formed can be used to control the depth of the central trough in that feature, which controls the degree of flattening of the peak; the width of the marks and their separation can be adjusted to control the flattening of the trough. The overall depth of the marks can be increased by using multiple pulses. Also the direction of flow of molten material can be reversed by means of performing marking with or without CO.sub.2 cover gas. Using these parameters the form of the marks can be accurately controlled.

(41) It will be understood that the laser beam parameters needed to produce particular effects or structures can vary for different materials and conditions, and that suitable laser beam parameters can be determined by experimentation.

(42) For instance in further embodiments it has been found that to produce single peaked marks on a spar of cold rolled austenitic-ferritic precision type 316 stainless steel a series of 10 pulses, each 30 s long at powers between 300 J and 400 J per pulse can be used, with a cylindrical lens focal length=40 mm, and a nominally Gaussian input beam of diameter around 3.6 mm. With 600 J, 60 s long pulses the centre is depressed by ablation pressure, giving the double peaked structure. For the limit of 500 J per pulse with 10 pulses mixed results are obtained, for example protrusions can be found next to dips on the same sample, or a wobbly surface can be produced. For energy per pulse 600 uJ dips/grooves are obtained without visible protrusions.

(43) The dimensions of lines produced in various experiments using laser parameters from the preceding paragraph include:

(44) In air on the stainless steel sparbetween 2 m and 8 m wide, 1 mm to 2 mm long. Depth:up to 1.5 m deep in high energy regime (surface tension and recoil pressure effects), more usually 200 nm to 300 nm deep, but up to 500 nm in low energy regime (surface tension effects only).

(45) In CO.sub.2 on the stainless steel spar:line width: approximately 4 m wide, length: approximately 1 to 2 mm long, peak to trough depth: up to 120 nm.

(46) In air and on Nickel: protrusions having a line width of approximately 4 m, length 2 mm, peak to trough depth: up to 350 nm.

(47) It will be understood that the nature and dimensions of lines or other patterns that can be produced are not limited to those mentioned in the preceding four paragraphs. A wide range of desired patterns and dimensions can be obtained with suitable choice of laser parameters.

(48) In further embodiments, a CO.sub.2-rich atmosphere is provided at to the sample surface for different laser beam pulses and the level of additional CO.sub.2 adjacent to the surface is controlled to be different for different pulses, thus providing an additional control parameter for controlling the shaping of the surface material.

(49) In other embodiments, the composition of the atmosphere adjacent to the surface is controlled by controlling the amount of other gases, as well as or instead of CO.sub.2, in order to vary surface tension effects, for example in order to provide a reversal of the surface temperature gradient with temperature. In some such embodiments, H.sub.2S gas is provided at the surface. The pressure of the atmosphere at the surface, and the relative pressure of components of the atmosphere is also controlled in some embodiments.

(50) It has also been found that a reversal of the usual surface temperature gradient with temperature can also be provided by providing a suitable surface material, for example stainless steel that has a sufficiently high oxygen content. A further example of such a material that is used in some embodiments is nickel deposited as a layer on a thicker layer of copper. In embodiments that use such materials, surface tension gradient-induced flow from lower temperature regions to higher temperature regions can be obtained even for normal atmospheric compositions, for example in air and without a CO.sub.2-rich atmosphere being provided adjacent to the surface of the material. Such embodiments can provide for formation of optical devices using a combination of laser pulses having intensities above and below particular thresholds, similar to those illustrated in FIGS. 7 and 8.

(51) The extent to which surface tension is related to temperature for stainless steel has been shown to, at least partially, depend on the presence of elements, such as sulphur (see Equation to estimate the surface tensions of stainless steel. Y. Su, Z. Li, K. C. Mills. s.l.: Journal of Materials Science, 2005, Vols. 40 (2005) 2201-2205 and Determination of the surface tension of liquid stainless steel. Z. Li, K. Mukai, M. Zeze, K. C. Mills. s.l.: Journal of Materials Science, 2005, Vols. 40 (2005) 2191-2195). Material flow in liquid iron films appears to be affected by oxygen contamination (see Thermacapillary flow excited by focused nanosecond laser pulses in contaminated thin iron films. V. Yu. Balandin, D. Otte, O. Bostanjoglo. 3 pages 2037-2044, s.l.: Journal of Applied Physics, 2005, Vol. 78). Nickel-phosphorous surfaces demonstrate comparable behaviour with phosphorous concentration (see Melting and surface deformation in pulsed laser surface micromodification of Ni-P disks. S. C. Chen, D. G. Cahill, C. P. Grigoropoulous. s.l.: Journal of heat transfer, 2000, Vols. 122 pages 107-112) an effect which may be casued by phosphorous evaporation (see Marangoni mechanism in pulsed laser texturing of magnetic disk substrates. T. D. Bennett, D. J. Krajnovich, C. P. Grigoropoulos, P. Baumgart, A. C. Tam. s.l.: Journal of Heat Transfer, 1997, Vols. 199 pages 589-596). Observations from the welding industry reinforce the importance of surface elements for surface tension and bulk fluid motion in a melt pool (see Welding 304L stainless steel tubing having variable penetration characteristics. I. Grant, M. J. Tinkler, G. Mizuno, C. Gluck and Marangoni effects in welding. K. C. Mills, B. J. Keene, R. F. Brooks, A. Shirali. 1739 pp. 911-925, s.l.: Philosophical Transactions: Mathematical, Physical and Engineering Science, 1998, Vol. 356) with sulphur, oxygen and nitrogen being the surface active elements. As few as ten parts per million of the active element can have a large effect. For stainless steel, calcium, sulphur, manganese, silicon, titanium, zirconium, aluminium, magnesium and oxygen appear to affect the surface tension. For nickel, phosphorus appears to have a significant effect, although oxygen can also play a role.

(52) Any suitable materials can be used for the sample in alternative embodiments, as well as or instead of stainless steel. For phase scale applications it is generally desirable that the surface layer of the sample is as reflective as possible and does not tarnish easily. The surface layer of the sample may be metallic, either a pure metal or an alloy. The sample can comprise a surface layer of a first material formed on a substrate of a second, different material, for example glass or polymer, or can be formed of a single material. Suitable materials for the sample can include steel, chrome, nickel, gold, silver or indium. In some embodiments the marks are formed to provide a pattern having a desired profile in a non-reflective material for example a thermoplastic material such as a suitable polymer, which is subsequently coated with a reflective coating, such as a chrome coating, in a subsequent process.

(53) The method can be used for the formation of any suitable optical device, including diffraction gratings, reflective holograms or metrological devices such as amplitude or phase scales.

(54) Further features are described in Appendix 1 and 2.

(55) It will be understood that the present invention has been described above purely by way of example, and modifications of detail can be made within the scope of the invention.

(56) Each feature disclosed in the description, and (where appropriate) the claims and drawings may be provided independently or in any appropriate combination.