The invention relates to a method of forming an optical image in a resist layer, the method comprising the steps of:
The invention also relates to a diffraction element comprising an array of diffraction cells for use with this method, an apparatus for carrying out this method and a method of manufacturing a device using this method.
An array of light valves, or optical shutters, is understood to mean an array of controllable elements, which can be switched between two states. In one of the states, radiation incident on such an element is blocked and, in the other states, the incident radiation is transmitted or reflected to follow a path that is prescribed in the apparatus of which the array forms part.
Such an array may be a transmissive or reflective liquid crystal display (LCD) or a digital mirror device (DMD). A resist layer is a layer of material, which is sensitive to radiation used in optical lithography.
This method and apparatus may be used, inter alia, in the manufacture of devices such as liquid crystalline display (LCD) panels, customized ICs (integrated circuits) and PCBs (printed circuit board). Currently, proximity printing is used in the manufacture of such devices. Proximity printing is a fast and cheap method of forming an image in a radiation-sensitive layer on a substrate of the device, which image comprises features corresponding to device features to be configured in a layer of the substrate. Use is made of a large photo mask that is arranged at a short distance, called the proximity gap, from the substrate and the substrate is illuminated via the photo mask by, for example, ultraviolet (UV) radiation. An important advantage of the method is the large image field, so that large device patterns can be imaged in one image step. The pattern of a conventional photo mask for proximity printing is a true, one-to-one copy, of the image required on the substrate, i.e. each picture element (pixel) of this image is identical to the corresponding pixel in the mask pattern.
Proximity printing has a limited resolution, i.e. the ability to reproduce the points, lines etc., in general the features, in the mask pattern as separate entities in the sensitive layer on the substrate. This is due to the diffractive effects, which occur when the dimensions of the features decrease with respect to the wavelength of the radiation used for imaging. For example, for a wavelength in the near UV range and a proximity gap width of 100 μm, the resolution is 10 μm, which means that pattern features at a mutual distance of 10 sum can be imaged as separate elements.
To increase the resolution in optical lithography, a real projection apparatus, i.e. an apparatus having a real projection system like a lens projection system or a mirror projection system, are used. Examples of such apparatus are wafer steppers or wafer step-and scanners. In a wafer stepper, a whole mask pattern, for example an IC pattern is imaged at one go by a projection lens system on a first IC area of the substrate. Then the mask and substrate are moved (stepped) relative to each other until a second IC area is positioned below the projection lens. The mask pattern is then imaged on the second IC area. These steps are repeated until all IC areas of the substrate are provided with an image of the mask pattern. This is a time-consuming process, due to the sub-steps of moving, aligning and illumination. In a step-and-scanner, only a small portion of the mask pattern is illuminated at once. During illumination, the mask and the substrate are synchronously moved with respect to the illumination beam until the whole mask pattern has been illuminated and a complete image of this pattern has been formed on an IC area of the substrate. Then the mask and substrate are moved relative to each other until the next IC area is positioned under the projection lens and the mask pattern is again scan-illuminated, so that a complete image of the mask pattern is formed on the next IC area. These steps are repeated until all IC areas of the substrate are provided with a complete image of the mask pattern. The step-and-scanning process is even more time-consuming than the stepping process.
If a 1:1 stepper, i.e. a stepper with a magnification of one, is used to print a LCD pattern, a resolution of 3 μm can be obtained, however, at the expense of much time for imaging. Moreover, if the pattern is large and has to be divided into sub-patterns, which are imaged separately, stitching problems may occur, which means that neighboring sub-fields do not fit exactly together.
The manufacture of a photo mask is a time-consuming and cumbersome process, which renders such a mask expensive. If much re-design of a photo mask is necessary or in case customer-specific devices, i.e. a relative small number of the same device, have to be manufactured, the lithographic manufacturing method using a photo mask is an expensive method.
The paper: “Lithographic patterning and confocal imaging with zone plates” of D. Gil et al in: J. Vac. Sci. Technology B 18(6), November/December 2000, pages 2881-2885, describes a lithographic method wherein, instead of a photo mask, a combination of a DMD array and an array of zone plates is used. If the array of zone plates, also called Fresnel lenses, is illuminated, it produces an array of radiation spots, in the experiment described in the paper: an array of 3×3 X-ray spots, on a substrate. The spot size is approximately equal to the minimum feature size, i.e. the outer zone width, of the zone plate. The radiation to each zone plate is separately turned on and off by the micro-mechanic means of the DMD device and by raster scanning the substrate through a zone plate unit cell, arbitrary patterns can be written. In this way, the advantages of maskless lithography are combined with high throughput due to parallel writing with an array of spots. The zone plates of the array are conventional phase zone plates, i.e. they comprise alternating first rings and second rings all first rings and second rings being at a constant first level and a constant second level, respectively. Radiation passing through the first rings undergoes a phase shift of 180° relative to radiation passing through the second rings. In the paper it is remarked that order-sorting apertures are needed to reduce background radiation caused by non-focused diffraction orders.
It is an object of the present invention to provide an accurate and radiation-efficient lithographic imaging method. This method is characterized in that use is made of diffraction lenses in the form of diffraction cells having at least two transmission levels and at least three phase levels.
The diffraction cells are usually, but not necessarily identical. The amplitude level and the phase level of a diffraction cell are measures of the degree, to which a diffraction cell changes the amplitude and phase, respectively, of a beam portion incident on this cell. The phase level of an area of a diffraction cell is determined, for example, by the height or depth of this area with respect to the surface of the total array.
By using more than two phase levels per cell, the diffraction efficiency, i.e. the percentage of the incident radiation that is diffracted in the required diffraction order, for example a first order, increases. This means that the available radiation is optimally used for imaging in the resist layer and that the amount of background radiation, for example zero order, or non-diffracted, radiation is so small that this radiation needs no to be blocked by means of, for example, order filters.
The diffraction efficiency of the cells and the sharpness of the spots formed by these cells increase with the number of phase steps in the cells. Reasonable results can be obtained with a limited number of phase steps, for example four steps differing 90° from each other. Usually, two amplitude levels for each diffraction cell are sufficient. The main part of the cell is “white” and only the border of a cell is black to distinguish the cell from its neighboring cells. “White” means transmitting or reflecting incident radiation to the resist layer and “black” means preventing incident radiation from reaching this layer. The black portions of all diffraction cells may be constituted by one layer of, for example a metal such as chromium, which layer has relative wide openings to accommodate the white portions with the phase structure. Chromium is already widely used in optical lithography. The phase structures may be etched in the diffraction element of, for example, quartz by an ion beam technique.
A first embodiment of the method is characterized in that use is made of an array of diffraction cells each showing a series of rising phase steps and a series of declining phase steps.
A second embodiment of the method is characterized in that use is made of an array of diffraction cells each comprising a number of successive phase structures, each phase structure comprising a number of phase steps rising from a base level to a top level followed by a decline from the top level to the base level.
A third embodiment of the method is characterized in that use is made of an array of diffraction cell each comprising a number of successive phase structures, each phase structure showing a continuous increase from a base level to a top level and an abrupt decline from the top level to the base level.
These embodiments may be further characterized in that use is made of an array comprising collections of diffraction cells, which collections differ from each other in that the focal plane of the diffraction cells of each collection is different from the focal planes of the other collections.
This method allows printing on different planes of the substrate.
The method may be further characterized in that, between successive sub-illuminations, the radiation-sensitive layer and the arrays are displaced relative to each other through a distance which is at most equal to the size of the spots formed in the resist layer.
In this way, image, i.e. pattern, features can be written with a constant intensity across the whole feature. The spots may have a circular, square, diamond or rectangular shape, dependent on the design of the diffraction cells. The size of the spot is the size of the largest dimension within this spot.
If features of the image to be written are very close to each other, these features may broaden and merge with each other, which phenomenon is known as proximity effect. An embodiment of the method, which prevents proximity effects from occurring, is characterized in that the intensity of a spot at the border of an image feature is adapted to the distance between this feature border and a neighboring feature.
The method preferably is characterized in that the illumination step comprises illuminating the array with a beam of monochromatic radiation.
Monochromatic radiation has only one wavelength and is very suitable to be used with a diffraction element, the diffraction property of which is wavelength dependent. A laser may be used for generating the monochromatic radiation.
The method may be further characterized in that the array of light valves is positioned to directly face the array of diffraction cells.
The two arrays are positioned close to each other, without imaging means being arranged between them, so that the method can be performed by compact means. If the array of light valves is an array of LCD cells, which modulate the polarization of incident radiation, a polarization analyzer is arranged between the LCD and the array of diffraction cells.
Alternatively, the method may be characterized in that the array of light valves is imaged on the array of diffraction cells.
Imaging one array on the other by a projection lens provides advantages with respect to stability, thermal effects, and crosstalk
The invention also relates to a diffraction element for use with the method described above and comprising an array of diffraction cells. This diffraction element is characterized in that the diffraction cells have at lest two amplitude levels and at least three phase levels.
A relatively simple embodiment of the diffraction element is characterized in that each diffraction cell has a series of rising phase steps and a series of declining phase steps.
This embodiment may be further characterized in that the diffraction cells have four phase levels, which differ from each other by 90°.
Satisfactory results can be obtained with such a diffraction element.
Even better results are possible with an embodiment of the diffraction element which is characterized in that each diffraction cell comprises a number of successive phase structures, each phase structure comprising a number of phase steps rising from a base level to a top level followed by a decline from the top level to the base level.
An alternative embodiment of the diffraction element is characterized in that each diffraction cell comprises a number of successive phase structures, each phase structure showing a continuous increase from a base level to a top level and an abrupt decline from the top level to the base level.
This embodiment is less easy to manufacture than the preceding one, but provides the best results.
The diffraction element may be further characterized in that it comprises collections of diffraction cells, which collections differ from each other in that the focal plane of the diffraction cells of each collection is different from the focal planes of the other collections.
This diffraction element, which can be used if the spots to be formed in the resist layer are not very small, allows simultaneous imaging of pattern features at different heights in the resist layer so that time can be saved.
The invention also relates to an apparatus for carrying out the method described above. This apparatus comprises:
With this apparatus, arbitrary patterns can be written by scanning the resist layer with a number of sharp spots simultaneously, wherein efficient use is made of the available radiation.
A first embodiment of the apparatus is characterized in that each diffraction cell has a series of rising phase steps and a series of declining phase steps.
This embodiment may be further characterized in that the diffraction cells have four phase levels, which differ from each other by 90°.
Satisfactory results are obtained with such a diffraction element.
A second embodiment of the apparatus is characterized in that each diffraction cell comprises a number of successive phase structures, each phase structure comprising a number of phase steps rising from a base level to a top level followed by a decline from the top level to the base level.
Such a phase profile in each cell allows obtaining a maximum diffraction in one, required, order and maximum sharpness of the spots.
A third embodiment of the apparatus is characterized in that each diffraction cell comprises a number of successive phase structures, each phase structure showing a continuous increase from a base level to a top level and an abrupt decline from the top level to the base level.
All of the above embodiments may be further characterized in that the diffraction element comprises collections of diffraction cells, which collections differ from each other in that the focal plane of the diffraction cells of each collection is different from the focal planes of the other collections.
The apparatus is preferably characterized in that the radiation source is a source of monochromatic radiation.
Under circumstances, also other sources may be used such as a conventional mercury-arc lamp, which emits several wavelength bands.
The apparatus may be further characterized in that the diffraction element is arranged behind the array of light valves without intervening imaging means.
The gap, for example an air gap, may be very small so that this embodiment has a sandwich shape. If the array of light valves is a LCD, a polarization analyzer is arranged between the array of light valves and the array of diffraction cells.
An embodiment of the apparatus, which is alternative to the sandwich embodiment, is characterized in that a projection lens is arranged between the array of light valves and the diffraction element.
The projection lens images each light valve on its associated diffraction cell in the diffraction element so that crosstalk, optical aberrations and temperature effects are eliminated. Moreover, the substrate of the diffraction element may be relatively thick and thus the apparatus will be more stable.
The invention also relates to a method of manufacturing a device in at least one process layer of a substrate, the method comprising the steps of:
Devices, which can be manufactured by means of this method and apparatus, are liquid crystalline display devices, customer-specific ICs, electronic modules, printed circuit boards etc. Examples of such devices are micro-optical-electrical-mechanical (MOEM) modules and integrated optical telecommunication devices comprising a diode laser and/or detector, a light guide and possibly a lens between the light guide and the diode laser, or the detector.
These and other aspects of the invention are apparent from and will be elucidated, by way of non-limitative example, with reference to the embodiments described hereinafter.
In the drawings:
a shows the amplitude structure of a portion of an embodiment of the diffraction element according to the invention;
b shows the phase structure of this embodiment;
c shows the spots formed in the resist layer by means of this embodiment;
IG. 5 shows a second embodiment of such a depth structure;
a-6c show different moments of the printing process in a cross-section of the resist layer;
a-7c show different moments of the printing process in a top view of the resist layer;
a-8c show an array of spots formed with different widths of the gap between the diffraction element and the resist layer, and
The apparatus of
Direct writing of a pattern in the resist layer, for example by an electron beam writer or a laser beam writer, could provide the required flexibility, but is not a real alternative because this process takes too much time.
As the radiation source, the substrate holder and the mask holder are less relevant for understanding the new method, these elements are not shown in
According to the invention, the diffraction cells have two amplitude levels and four phase levels.
c shows the array 60 of spots 62 obtained by an embodiment, having four phase levels, of the diffraction structure of
The phase structure of a diffraction cell may be any phase structure, which introduces the required phase differences in the associated beam portions. From a manufacturing point of view, the phase structure is preferably a depth, or level, structure.
a shows, also in a vertical cross-section, another, preferred embodiment of a diffraction cell with four phase levels. This cell has a relatively broad central portion 80 and two side portions 81,82 at the left side of the central portion and two side portions 83,84 at the right side of the central portion. All portions 80-84 have four different geometrical levels 85-88. If this diffraction cell has to produce a round spot, the area of level 88 of the central portion 80 is round and the areas of the levels 85-87 of the central portions and those of the side portions 81-84 are annular. The thickness d1 of the cell may be of the order of 0.5 μm. In practice, a cell of the kind shown in
The larger the number of phase steps in a diffraction cell is, the finer and brighter the spot that is produced by the cell. A limit to the number of phase steps may be imposed by the manufacturability of a diffraction element with such cells. The phase structure of the diffraction cell of
Under circumstances, the number of phase steps in a diffraction cell may be less than four, for example three. In most cases, two amplitude levels in one diffraction cell will be sufficient, but under circumstances a diffraction cell may be provided with three or more amplitude levels.
The diffraction element with the multilevel phase structure can be manufactured by means of known lithographic techniques. For example, by means of an electron beam pattern generator, the cell pattern can be written in a resist, which is sensitive to electrons, and the different levels can be realized by selective ion etching. Also the so-called Canyon technique can be used. According to this technique an electron beam-sensitive glass, i.e. a glass which changes its transmission in dependence upon the electron beam intensity. The cell pattern is written in this glass as a grey pattern, i.e. amplitude pattern. Then, a three-dimensional cell pattern is formed in a resist layer, coated on a quartz substrate, using the grey pattern as a mask. By reactive ion etching, the resist pattern is transferred to the quartz substrate. After the mask substrate surface 70 has been provided with the multilevel phase structure, this surface is selectively coated with chromium to give the mask the required amplitude structure.
Instead of chromium, other non-transmission materials can be used for the selective coating of the mask. Instead of 100% transmission and 0% transmission for the amplitude levels, the mask may have different amplitude levels.
As demonstrated above, good results are obtained with a diffraction mask structure having two amplitude levels and four phase levels and diffraction cell dimensions of 1×1 μm2. However, the mask structure for the above and other applications may have three or more than four phase levels and/or more than two amplitude levels and/or different dimensions for the diffraction cell areas. In general, it holds that levels the quality of the printed image will enhance with a decreasing cell area dimensions and an increasing number of amplitude and phase.
As shown in
The illumination process of flashing and stepping is illustrated in
a-7c show top views of the resist layer during subsequent sub-illumination steps. In these Figures, the dark spot areas have already been illuminated in preceding sub-illumination steps and the light spot areas are being illuminated in the present illumination step. The portion of the resist layer being illuminated comprises two rows of five light valve areas. In the situation depicted in
a-6c and 7a-7c show how ten light valve areas in the resist layer are simultaneously illuminated in successive steps of displacing the resist layer and opening and closing the ten corresponding light valves. The light valve areas of all light valves of the array are simultaneously illuminated in the same way. As shown in the upper right portion of
Instead of the stepping mode, illustrated in
In a practical embodiment of the proximity printing apparatus shown in
The exposure dose is the amount of illumination radiation energy in a spot area of the resist. The intensity of the illumination beam and the opening time of the light valve determine this dose.
The mercury arc discharge lamp emits radiation, 40% of which has a wavelength of 365 nm, 20% has a wavelength of 405 nm and 40% has a wavelength of 436 nm. The effective contribution to the image formation of this lamp radiation is 60% by the 365 nm component, 15% by the 405 nm component and 25% by the 436 nm component due to the absorption in the resist layer. A general problem with diffraction elements is that their performance is wavelength-dependent. For the present method and apparatus, this means that beam components of the mercury arc lamp with different wavelengths would be focused in different planes. However, in the case where somewhat broader spots are allowed, some freedom of design of the diffraction cells remains. This freedom can be used to correct the wavelength dependence and to design the diffraction cells in such a way that the beam components with different wavelengths will be focused in the same plane. This allows use of the mercury arc discharge lamp, which has proved its benefits for conventional proximity printing, also in the new method and apparatus.
Nevertheless, it is preferred to use a monochromatic source, for example a YAG laser emitting radiation at a wavelength of 350 nm, because no wavelength corrections are needed.
The invention can also be implemented with other radiation sources, preferably lasers, especially lasers used currently or to be used in the near future in wafer steppers and wafer step-and-scanners, emitting radiation at wavelengths of 248, 193 and 157 nm, respectively. Lasers provide the advantage that they emit a beam, which is collimated to the required degree. Essential for the present imaging method is that the illumination beam is substantially a collimated beam. The best results are obtained with a fully collimated beam, i.e. a beam having an aperture angle of 0°. However, satisfactory results can also be obtained with a beam having an aperture angle smaller than 1°.
The required movement with respect to each other of the resist layer, on the one hand, and the array of light valves and the diffraction element, on the other hand, is most practically performed by movement of the substrate stage. Substrate stages currently used in wafer steppers are very suitable for this purpose, because they are more than accurate enough. It will be clear that movement of the substrate stage, for either the stepping mode or the scanning mode, should be synchronized with the switching of the light valve. To that end, the stage movement may be controlled by the computer, 30 in
An image pattern which is larger than the illumination field of one array of light valves and one array of diffraction cells can be produced by dividing, in the software, such a pattern into sub-patterns and successively transferring the sub-patterns to neighboring resist areas having the size of the image field. By using an accurate substrate stage, the sub-image patterns can be put together precisely so that one non-interrupted large image is obtained.
A large image pattern can also be produced by using a composed light valve array and a composed diffraction cell array. The composed light valve array comprises, for example, five LCDs, each having 1000×1000 light valves. The LCDs are arranged in a series to cover, for example, the width of the image pattern to be produced. The composed diffraction element is constructed in a corresponding way to fit to the composed light valve array. The image pattern is produced by first scanning and illuminating a resist area having a length covered by a single array of light valves and a width covered by the series of light valve arrays. Subsequently, the substrate with the resist layer and the series of arrays are displaced relative to each other in the longitudinal direction through a distance covered by a single array. A second resist area, which now faces the composed arrays is scanned and illuminated, etc until the whole image pattern has been produced.
An essential parameter for the imaging process is the gap width 44 (
For an apparatus with a larger design gap width, for example 250 μm, the requirements for the real gap width can be mitigated. With an increasing design gap width, the NA of the sub-beams (101-105 in
The minimum size of the spots is also related to the gap width. If the gap width is reduced, this size can be decreased, for example below 1 μm. A smaller gap width requires a better control of this width.
A feature of the present diffraction element is that it allows the creation of multiple focal planes within a single image field. The outlay of this diffraction elements allows its design, or computation on a cell-by-cell basis, in which the distance between the diffraction element and the resist layer, or the focal distance, may be taken as one of the input parameters. This allows design of a diffraction element wherein one or more area(s), comprising a number of diffraction cells, is (are) intended for a different focal distance than the remaining portion of the diffraction element. This multiple focal diffraction element can be used for the manufacture of a device composed of sub-devices positioned at different levels. Such a device may be a pure electronic device or a device that comprises two or more different kinds of features from the range of electrical, mechanical or optical systems. An example of such a system is a micro-optical-electrical-mechanical (MOEM) module or a device comprising a diode laser or a detector and a light guide and possibly lens means to couple light from the laser into the light guide or from the light guide to the detector. The lens means may be planar diffraction means. For the manufacture of a multilevel device, a substrate is used that has a resist layer deposited on different levels. By using a multiple focal diffraction element, all sub-images can be printed simultaneously on the relevant levels, so that a lot of time can be saved.
A multiple diffraction element can only be used for the production of a device showing the multiple level structure corresponding to the multiple focus structure of the diffraction element. The imaging apparatus may, however, be designed, in such a way that diffraction elements can easily be placed in and removed from the apparatus. This allows production of different multiple level devices by means of different appropriate multiple focus diffraction elements.
Multiple level devices can also be produced by means of a general, single focus, diffraction element. In the software, the total image pattern is divided into a number of sub-images each belonging to a different level of the device to be produced. In a first sub-imaging process, the first sub-image is produced, with the resist layer being positioned at a first level. The first sub-imaging process is performed according to the, scanning or stepping, method and by the means described hereinbefore. Then the resist layer is positioned at a second level, and in a second sub-imaging process the sub-image belonging to the second level is produced. The shifts of the resist layer in the Z-direction and the sub-imaging processes are repeated until all sub-images of the multiple level device are transferred to the resist layer.
The method of the invention can be carried out with a robust apparatus that is, moreover, quite simple as compared with a stepper or step-and-scan lithographic projection apparatus.
In the apparatus, schematically shown in
To reduce the distance between the light valves and the diffraction cells and to prevent annoying crosstalk, the diffraction element may be arranged on the lower surface of the polarizer and/or the polarizer may be arranged on the light valve structure.
The left part of
The LCD 20 may have a pixel size of 20 μm and the projection lens may image the LCD pixel structure on the diffraction element with a magnification of 5×. For such imaging, no large numerical aperture (NA) for the projection lens is required. To achieve that the illumination beam incident on the diffraction element is a parallel beam, a collimator lens 262 is arranged in front of the diffraction element. For example, a diaphragm opening of 1 mm will be imaged by the projection lens and a diffraction cell in a spot with a dimension of 1 μm. As the operation of the LCD is based on changing the polarization state of incident radiation, a polarizer, which gives the radiation the required initial polarization state, and a polarization analyzer, which converts the polarization state into an intensity, are needed. This polarizer and analyzer are denoted by reference numerals 250 and 258, respectively The polarizer and analyzer are adapted to the wavelength of the illumination beam. They are not shown in
As the image of the LCD pixel structure is focused on the diffraction element, practically no crosstalk will occur in an apparatus with a projection lens. Moreover, the diffraction element may comprise a thick substrate so that it is more stable. When in use, a LCD shutter array absorbs radiation and produces heat, which may cause thermal effects in the apparatus. Such effects are considerably reduced in an apparatus with a projection lens, because the LCD is arranged at a relatively large distance from the diffraction element. Moreover, the design allows separate cooling of the LCD. A LCD light valve array may comprise spacers in the form of small, for example 4 μm, spheres of a polymer material. Such spheres may cause optical disturbances. In an apparatus with a projection lens, the effects of the spacers are reduced because the projection lens with a relatively small NA functions as a spatial filter for the high-frequency disturbances.
When using a projection lens, it will be easy to replace a transmission light valve array by a reflective array, such as a reflective LCD or a digital mirror device.
The apparatus of
In practice, the method of the invention will be applied as one step in a process for manufacturing a device having device features in at least one process layer of a substrate. After the image has been printed in the resist layer on top of the process layer, material is removed from, or added to, areas of the process layer which areas are delineated by the printed image. These process steps of imaging and material removing or adding are repeated for all process layers until the whole device is finished. In those cases where sub-devices are to be formed at different levels and use can be made of multiple level substrates, a multiple focal diffraction element can be used for image printing.
The invention can be used for printing patterns of, and thus for manufacturing display devices like LCD, Plasma Display Panels and PolyLed Displays, printed circuit boards (PCB) and micro-multiple function systems (MOEMS).
Number | Date | Country | Kind |
---|---|---|---|
012049334 | Dec 2001 | EP | regional |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/IB02/05342 | 12/11/2002 | WO |