DEVICE AND METHOD FOR EXPOSURE A LIGHT-SENSITIVE LAYER

Description

[0001] The present invention relates to a method for exposing a photosensitive layer according to claim 1 and a corresponding device according to claim 8.

[0002] Digital micromirror devices (DMDs) have been known in the art for a long time. They are optical elements consisting of a multitude of small mirrors that can be individually moved and aligned. Each mirror can be selectively controlled and oriented electrically. This allows an optical system with a DMD to be used to selectively deflect a wide beam of light with precise spatial resolution. Typically, only two fully deflected positions are considered for each mirror. A mirror either allows the portion of the light beam that falls upon it to pass through the optical system, or it reflects that portion of the light beam so that the light is not transmitted further within the optical system. Each mirror of such a DMD can then be interpreted as a digital light switch.

[0003] Such DMDs are already used in projectors in the state of the art. DMDs are also increasingly finding application in industrial environments, such as in 3D printing, 3D surveying, and maskless lithography.

[0004] The prior art knows only of maskless exposure units that have a single light source illuminating the DMD and thus the surface to be exposed. Such exposure units are shown, for example, in DE 10 2005 054000 A1 and EP 0 467 076 A2.

[0005] It is therefore the object of the present invention to eliminate the disadvantages of the prior art and in particular to achieve an improved exposure result.

[0006] This problem is solved by the subject matter of the dependent claims and the inventive concepts disclosed below. Advantageous embodiments of the invention are specified in the dependent claims.

[0007] The invention demonstrates in particular how an improved, faster, higher-resolution exposure process can be carried out and how exposure can be performed simultaneously at different focus depths.

[0008] The core concept of the invention is to design micromirrors of a micromirror assembly of an optical system such that each mirror intensity profile of the micromirrors is superimposed, or can be superimposed, with its neighboring mirror intensity profiles. Each mirror intensity profile thus consists of a region corresponding to the image point / pixel and a region surrounding the image point / pixel. The mirror intensity profiles are defined at least predominantly, preferably completely, by the optical elements and by the geometric shape of the micromirrors. The micromirrors are preferably rectangular, even more preferably square. Round or triangular micromirrors would also be conceivable. A (DMD) image is formed from the image points / pixels, including their superpositions.

[0009] In particular, an exposure grid is created from several, at least partially superimposed, mirror intensity profiles with a grid intensity profile. The grid intensity profile is thus a sum of the superimposed mirror intensity profiles. The exposure grid preferably exposes a partial area of ​​a layer to be exposed on a substrate, wherein the grid intensity profile of the exposure grid is preferably more homogeneous due to the superimposition than an exposure grid from the prior art without superimposition.

[0010] The mirror intensity profile is preferably selected, adjusted, or controlled such that 50% of the energy deflected by the respective mirror strikes the corresponding pixel. A further 50% of the energy deflected by the respective mirror is distributed among the pixels surrounding the respective pixel. In a more preferred embodiment according to the invention, the mirror intensity profile is selected, adjusted, or controlled such that more than 50%, more preferably more than 60%, and even more preferably more than 70%, of the energy deflected by the respective mirror strikes the corresponding pixel. The remaining energy is distributed among the pixels surrounding the respective pixel.

[0011] In another, less preferred embodiment of the invention, the mirror intensity profile is selected such that less than 50% of the energy deflected by the respective mirror hits the corresponding image point / pixel.

[0012] The invention describes in particular a maskless exposure device or a maskless exposure method. The exposure unit is described in several combinable embodiments. In particular, these are independent but combinable optimization possibilities.

[0013] The most important embodiment consists in the design of the optical system in such a way that the individual image points (pixels generated by the optical system) overlap, in particular with superposition and / or interference. This is achieved in particular by superimposing the intensity profiles of the reflected rays from the adjacent micromirrors.

[0014] In particular, a maskless exposure device with at least two light sources is disclosed, wherein the light beams of the light sources are in particular superimposed, combined, or mixed. A maskless exposure device with integrated measuring technology would also be conceivable according to the invention. Another embodiment relates to a maskless exposure unit that allows simultaneous exposure in different focal planes or spectrally different partial images. According to a further alternative embodiment, an exposure unit with an anisotropic or distorting optical system for achieving higher resolution is disclosed.

[0015] Another implementation, disclosed independently, concerns the method of "de-scanning" for increasing throughput and resolution. Here, the exposure process, which is particularly continuous and / or sequential, is superimposed by a second deflection in one or two directions. This deflection holds the exposure field quasi-statically with a finer positioning accuracy than the pixel size, in order to reduce both motion blur and edge position distortion.

[0016] Furthermore, embodiments are disclosed in which the optical system between the DMD and the photosensitive layer to be exposed (material to be exposed) generates an orthogonally (in particular, parallel to the surface of the layer to be exposed) distorted representation of the original image. The optical distortion of the image can, in particular, increase the positioning accuracy or the overlay. In this specific case, positioning accuracy relative to an existing structure means that the distorted representation of the original image according to the invention increases the resolution of the structures to be written. Advantages of the invention

[0017] No masks are required, allowing the desired structure to be exposed directly onto a photosensitive layer. Using multiple light sources allows for a broader spectrum and increases the light intensity, thus achievable throughput. By tilting the focal plane relative to the surface of the layer being exposed or by individually selecting wavelengths, simultaneous exposure at different depths with full resolution or selective exposure at different levels is possible.

[0018] According to the invention, the device in particular has an optical system with the following features: at least one light source for generating at least one light beam, at least one micromirror device with a plurality of micromirrors, wherein each micromirror serves to illuminate a pixel of an exposure grid with a mirror intensity profile, wherein the optical system is designed such that a superposition of mirror intensity profiles of neighboring micromirrors to a grid intensity profile of the exposure grid as the sum of the mirror intensity profiles of each illuminated pixel of the exposure grid takes place.

[0019] The invention describes a method and a system for irradiating a photosensitive layer using a beam of light, in particular a laser, which is guided through an optical system consisting of at least one, preferably digital, micromirror device (DMD). The irradiation of the layer leads to a physical and / or chemical change in the layer.

[0020] The embodiments according to the invention can in particular be described as full-surface continuous scanning stepwise scanning Maskless exposure units are being designed. "Maskless" refers to the replacement of the imaging of a static template (mask, reticle) with a dynamically optically patterning system.

[0021] A full-area, maskless exposure unit is a system that uses a DMD capable of exposing the entire area to be described without any significant relative displacement between the layer being exposed and the optical system. With current technology, this means that only very small substrates can be exposed, as the size of current DMDs is limited.

[0022] A scanning, maskless exposure unit is understood to be a device or method in which the exposure field is smaller than the area to be exposed, such that a relative displacement occurs between the optical system and the layer to be exposed in order to expose the entire exposure field. This relative displacement occurs, in particular, continuously along a path. The path is preferably: Meandering, linear with shift to the next line and return to the beginning upon reaching the end, spiral, circular, linear.

[0023] In particular, the aforementioned paths can also be combined. For example, it is conceivable to first use a spiral path for exposure, followed by a linear path that leads, in particular, through the center of the substrate, or to have several initially independent writing threads complement each other. A meandering path is preferred.

[0024] A stepping maskless exposure unit is a system in which the exposure field is smaller than the area to be exposed, the relative displacement between the optical system and the layer to be exposed occurs incrementally, and no exposure takes place between the individual steps. Only when the optical system and the layer to be exposed are in a well-defined position does the optical system expose the layer. This embodiment therefore involves the exposure of several sub-areas of the layer to be exposed. The DMD principle

[0025] A DMD allows the targeted deflection, particularly of multiple parts, of a wide primary beam of light, preferably parallel and / or non-scattering. This enables the creation of a spatially structured, secondary optical beam of light without the need for a mask. Optics, especially projection optics, are usually positioned upstream and / or downstream of the DMD. These optics can manipulate (in particular scale) the primary beam of light incident on the DMD and / or the secondary beams reflected by the DMD, and in microlithography, they can reduce the size of the DMD image. This allows the DMD image to be reduced accordingly.

[0026] The resulting structural resolution of a DMD is in particular between 0.1 µm and 50 µm, preferably between 0.3 µm and 25 µm, and even more preferably between 1 µm and 10 µm.

[0027] Especially for designs where the individual exposure field is smaller than the substrate, it is important that the structures to be created continue seamlessly after a relative movement between the optical system and the layer to be exposed. In stepping, this occurs in two independent directions, while in continuous scanning, it usually only occurs in one. The optical system

[0028] The embodiments according to the invention consist of at least one optical system, which may include several optical elements of different types. At least one of the optical elements is a DMD, although the use of multiple DMDs represents a further embodiment of the invention. In particular, exactly one DMD, preferably at least two DMDs, and more preferably at least four DMDs are located in the optical system. The optical system itself can be used in one or more forms in parallel on a substrate within a system. Parallel exposure of several substrates within the system is also conceivable according to the invention.

[0029] The optical system may also contain the following optical elements: Illumination optics ∘ Coherent light sources, in particular laser light sources, laser diodes, solid-state lasers, excimer lasers ∘ Incoherent light sources, in particular ▪ Gas discharge lamps, in particular mercury vapor lamps ▪ LEDs ∘ Partially coherent light sources ∘ Coherence-modifying components Deflection optics ∘ DMDs ∘ Mirrors, in particular ▪ Cold mirrors ▪ Warm mirrors ∘ Refractive elements, in particular ▪ Prisms ▪ Beam splitters Projection optics ∘ Lenses, in particular ▪ Fresnel lenses ▪ Refractive lenses Convex lenses Concave lenses Biconcave lenses Biconvex lenses Convex-concave lenses Concave-convex lenses Cylindrical lenses Compound lenses ∘ Mirrors, in particular ▪ Cylindrical mirrors ∘ General beam-modifying optical components

[0030] The light sources can be used continuously or pulsed, and in particular can be additionally modulated internally or externally.

[0031] The maximum possible relative velocity between the optical system and the layer to be exposed is limited in particular by the maximum control frequency of the DMD (i.e., the frequency at which the individual mirror elements of a DMD can actually be switched). The relative velocity in a scanning system is further determined by the displacement and / or deflection optics and is in particular between 5 mm / s and 500 mm / s, preferably between 10 mm / s and 250 mm / s, more preferably between 15 mm / s and 200 mm / s, and most preferably between 25 mm / s and 100 mm / s.

[0032] Another important aspect of an embodiment of the invention is to use a higher feed rate than is defined by the screen size and sampling frequency. By selecting a correspondingly higher feed rate, lines are omitted, which are then supplemented by later exposure lines of the DMD.

[0033] To prevent blurring of the exposure in the scan direction, the exposure is interrupted and / or a local "de-scan" mechanism is used to briefly maintain the exposure position. The first possibility will be considered first.

[0034] To compensate for dose loss, particularly caused by higher feed rates, the light output can be increased. This requires greater, especially instantaneous, power from the light source, which often conflicts with increased costs or physical and technical limitations. De-scanning and dynamically distributing the power of a single light source across multiple exposure units can offer more economical solutions to this problem.

[0035] During de-scanning, the relative movement of the image pixels with respect to the substrate is minimized and interrupted by a short, rapid deflection (line jump). Due to the small jump distances, this can be achieved using mechano-, electro-, magneto-, or acousto-optic deflection or displacement units.

[0036] In dynamic light distribution, the light source's power is either dynamically withheld (e.g., cavity-dumped lasers, temperature-limited semiconductor light sources) and focused over time, or distributed between the loads over time by a distribution element, such as a rotating polygon mirror, electro-optical switch, or similar device. The goal is always to keep the long-term power constant while increasing the short-term power via the sampling ratio. Integrated measuring optics

[0037] Preferably, the device includes measuring optics, particularly integrated into the optical system. A beam splitter is especially advantageous for extracting the light reflected from the layer to be exposed from the same optical path used to illuminate the DMD. The measuring optics perform several important tasks, not all of which need to be fulfilled simultaneously: Alignment to align or reshape the exposure field to existing structures on the substrate, calibration and verification of the writing head, in-situ control of the writing process, real-time correction for dynamic changes in the relative image-substrate position.

[0038] The alignment is performed either in relation to alignment marks already applied to the substrate or to deliberately applied or existing structures that serve as alignment marks or registration marks for the structures to be newly exposed. According to the invention, this makes it possible to expose large areas by continuously realigning and / or calibrating the optical system against the already exposed structures.

[0039] Another important aspect of an embodiment of the invention is that, by comparing the currently measured position of the reference marks with the target positions, any distortion of the substrate, particularly caused by upstream processes and / or thermal influences, is calculated, and the image to be written is adapted to this distortion. These distortions can also be of a higher order.

[0040] Especially in avoiding stitching artifacts (errors at transitions between adjacent pixels and / or adjacent exposure grids), alignment and / or real-time correction plays an important role.

[0041] By capturing the structures (which can also be surface-specific noise) of the exposure field during exposure and comparing them with the capture of neighboring exposure structures, a substrate offset can be detected via correlation or similar methods. This offset is applied to the DMD image as an error signal, thus enabling compensation down to the subpixel level.

[0042] The acquisition / measurement is preferably carried out along the same optical path that is also used for exposure, so that a mechanical connection, in particular fixing, can be realized during the acquisition / measurement.

[0043] For detection / measurement, the light signal (light beam from the at least one light source) is coupled from the surface via an optical element, preferably a semi-transparent mirror or prism, and recorded by a suitable detector. The detector (or an evaluation system connected to it) is then able to monitor the surface of the layer to be structured while it is exposed / written. The detector is preferably a camera, more preferably a CCD or CMOS camera. The camera image can capture one or more parts, a larger area, or one or more smaller sub-areas of the irradiation field. The exposure can have its own light source and occur in the same (preferably) or a different wavelength range as the irradiation.

[0044] In a particularly enhanced embodiment, a measuring optic is also provided on the underside of the embodiment according to the invention, with the aid of which alignment marks on the underside of the substrate holder or the substrate can be detected. The principle of underside measurement of a substrate holder is similar to the embodiment disclosed in publication PCT / EP2016 / 070289. The measurement of alignment marks on the underside makes it possible to create structures on both sides that are aligned with each other. Intensity distribution functions

[0045] In a first, particularly preferred, embodiment of the invention, an imaging / secondary optic (for influencing, in particular scaling, the secondary exposure rays reflected by the DMD) is designed such that a sharp image of the DMD mirror in the focal plane is not achieved, but rather an intentionally slightly blurred image that also exposes neighboring pixels. This allows the optic to be constructed much more cost-effectively while still improving the quality of the generated structures. The intensity distribution functions (intensity profiles) across the diameter of the beam can correspond to any mathematical function. Closed-form expressions are used as an approximation to the more complex real distribution. In particular, the following are conceivable: Gaussian distribution, Lorentz distribution, Cauchy distribution, convolutions of different distribution functions.

[0046] The material of the photosensitive layer to be exposed exhibits a defined, preferably nonlinear, response behavior that primarily corresponds to the accumulated amount of deposited light and, in particular, secondarily to the power and thus the exposure behavior. Therefore, an exposure edge forms between high and low doses under a local exposure gradient, which can fluctuate within defined or definable limits. The presented method utilizes this behavior and allows the position of this exposure edge to be controlled with a higher accuracy than the pixel size by overlapping multiple exposure fields (especially with respect to the individual micromirrors).

[0047] For precise positioning, the irradiance is calculated in advance and divided into multiple exposure elements (especially pixels) and / or different exposure steps. These exposure steps are separated by time and / or vary in length, resulting in a target exposure distribution at the raster level (raster intensity profile) and / or at the substrate level (substrate intensity profile). This preferably takes into account the coherent and / or non-coherent overlap of the individual distribution functions (mirror intensity profiles and / or raster intensity profiles) and / or the dynamics of image formation.

[0048] The result is a structuring of the exposure limit, defined primarily as the sum of the exposure processes, but secondarily, in dynamic cases, also taking into account simulated or empirically determined material responses in both their two-dimensional and three-dimensional forms. The concepts of overlap and interference are explained in more detail below.

[0049] Multiple exposures of an individual sub-area can be performed simultaneously or with a time delay. As long as the changes in the exposed material are independent of the exposure power and depend solely on the fluence (accumulated dose per unit area), and no mechanical changes (such as those caused by heating or displacement due to vibrations or shifts) occur, overlapping exposure fields, when simply added together, yield the same results as a field exposed by only one light source.

[0050] In cases where nonlinearities occur, these can be compensated for through appropriate calculations or tests and subsequent adjustments of the partial exposures. In the case of stochastic fluctuations, the overlaps lead to unwanted blurring or fluctuations, which appear as image defects (e.g., as blurring).

[0051] With reproducible or predictable overlaps, edge roughness, jumps, and other undesirable effects can be minimized, and clean stitching of the exposure fields can be guaranteed. Even short-term fluctuations in the exposure position can be compensated for by the method according to the invention.

[0052] Interference effects occur in optics during simultaneous exposure via multiple beam paths when the partial radiation sources exhibit exactly or nearly identical temporal and spatial behavior due to their coherence. Here, the superposition of the fields of electromagnetic partial waves is not simply averaged, but also generates constructive and destructive superpositions in the photosensitive surfaces and volumes—which are relatively slow in the THz range. These effects occur most frequently with superposition of emissions from the same source, where the transit times within the coherence time are similar—meaning the high-frequency oscillation behavior remains interdependent. Photochemical interference effects can also occur with multiple exposures, where the successive exposure steps interact with each other. These are taken into account in the control system. Avoiding overexposure at high speed

[0053] In a second, improved embodiment, the data paths are optimized. A fundamental problem with the maskless exposure process is that very complex and resource-intensive calculations (e.g., rasterization) are performed, resulting in the storage and transmission of enormous amounts of data. The data stored in the computer is synchronized with the relative movement between the optical system and the layer to be exposed, or between the DMD and the substrate. The structural data to be imaged in / on the layer to be exposed is stored in a computer. This structural data defines how the mirrors of the DMD are switched as a function of position to modify the exposure beam so that the desired pattern falls onto the layer to be exposed. In the case of a full-surface or step-wise embodiment, the optical system, and thus the DMD, is always located at a defined, and in particular, fixed, position.Therefore, in this case it is not necessary to establish a dynamic correlation between the position and the structural data.

[0054] However, if a (preferred) scanning embodiment is used, continuous exposure as a function of position takes place, in which the DMD is continuously supplied with new data during the relative movement in order to control the micromirrors as a function of position. The data for the mirror switching must therefore be sent to the DMD quickly enough. Since the calculation of the mirror matrix is ​​computationally intensive, according to an advantageous embodiment, various calculations are divided. In particular, a distinction is made between time-critical and time-non-critical calculations.

[0055] According to the invention, time-critical calculations are understood to be all calculations that must be performed so quickly that the data reaches the DMD before it has changed its relative position with respect to the layer to be exposed.

[0056] The time-critical calculations are particularly important. Image rendering, in particular ∘ DMD control, sensor measurements and feedback calculations, in particular ∘ sampling and adjustments to the current write position, decompression ∘ In particular of data from the DMD and / or to the DMD

[0057] These calculations require relatively high processing speeds, so hardware computing is preferred over software computing. Possible hardware components for this calculation include: FPGA (field programmable grid array) ASIC (Application Specific Integrated Circuit) GPU (Graphics Processing Unit) Signal processor (DSP - digital signal processor)

[0058] Time-independent calculations are those that change only slightly during the exposure process or are not critical to the process, in particular Conversion of vector data to pixel data (rasterization), adjustments to the structure size, large-area distortion compensation, insertion of wafer- or die-specific data (e.g., serial numbers), slicing the data into strips for a single read / write head, rasterization of the vector data

[0059] These calculations are typically very complex, involve very large datasets in a single computation (especially global adjustments), and / or cannot be performed very efficiently by hardware. It makes sense to parallelize these calculations. The process is carried out in at least one thread on the same computer or in at least one thread on at least one other computer. The threads can utilize the resources of the CPUs and / or the GPUs of the graphics cards.

[0060] Communication between the computer and the DMDs takes place via any interface, in particular Wired communication ∘ USB, ∘ Ethernet, ∘ DisplayPort.

[0061] A first improved embodiment is based on the provision of a memory in the hardware for time-critical calculations, which is large enough to store at least two sections of the layer to be exposed (in particular, subsections of the exposure paths, preferably strips, more preferably parallel strips). While one section to be exposed is being written from the buffer / memory, the second section to be exposed is transferred from the computer to the unit for calculating the time-critical tasks. The exposure process of a section to be exposed only begins when it is completely present in the memory. Typically, the second strip is already completely loaded into the memory before the first strip has completed its exposure process. This principle is repeated for the subsequent strips.This effectively prevents situations where data is no longer available within an exposure strip for time-critical calculations, thus preventing incorrect exposures. Real-time correction of write data

[0062] In a third improved embodiment, mechanical errors are at least partially compensated for by optical elements. If the system according to the invention is to expose a continuous structure into the layer to be exposed in a scanning process, it is advantageous to adjust or provide the positioning accuracy of the substrate holder to a higher degree than that of the overlay. Any positional errors originate from imperfectly manufactured mechanical components, play and tolerances, imperfect motor control, etc. It is relatively complex to reduce the positional error by mechanical means. According to the embodiment of the invention, it is now possible to allow larger positional errors than those of the overlay. The positioning error is measured and the DMD image is shifted accordingly, particularly in real time.According to the invention, the positional error is compensated such that it is smaller than the overlay specification at every point during the exposure process. The ratio between the overlay and the uncorrected positional error is particularly less than 1, preferably less than 10, and even more preferably less than 100.

[0063] According to the invention, errors can be compensated normally to the direction of the relative motion and in the direction of the relative motion.

[0064] For errors perpendicular to the direction of relative movement, a slightly narrower exposure strip is used, even though the DMD is designed for exposure of a wider strip. The additional exposure areas to the left and right thus serve as a buffer. If the substrate holder deviates from its position, the exposure strip is shifted left or right in real time so that, despite a positional error, the data is exposed to the correct location on the layer being exposed.

[0065] Errors in the direction of relative movement are compensated analogously. If the substrate holder moves too slowly, the scrolling speed (line rate) is adjusted, and the intensity is compensated by the exposure control. If the substrate holder moves too quickly, the position of the exposed image can also be adjusted by the scrolling speed (line rate), but dose correction by adjusting the exposure time is only possible if the exposure is not at maximum intensity.

[0066] The invention enables the use of cost-effective mechanics while still achieving high positioning accuracy. Stitching

[0067] In a fourth improved embodiment, the layer to be exposed is exposed in such a way that the calculated exposure sections (in particular, stripes) are superimposed. Preferably, areas at the edge of the stripes are double-exposed. To avoid over- or underexposure, the intensity is varied or variable as a function of the position of the optical system, in particular the DMD. Preferably, the intensity profiles then add up to a constant or at least more homogeneous intensity profile. The superimposition of the stripes can be considered seams, which should be exposed in such a way that, ideally, no difference is discernible compared to a full-surface exposure. A more detailed description is given in the figure description of the Figures 6a-cThe stripe superimposition according to the invention can be generated not only by exploiting the drop in intensity at the edge of an exposure strip, but also by specifically employing dynamic illumination control, which causes the intensity to decrease continuously in the edge regions of the exposure strip. According to the invention, this makes it particularly advantageous to control the seams by means of targeted intensity control. Descanning

[0068] In a fifth improved embodiment, the exposure process is carried out such that, in a first step, the optical system or the DMD and the substrate are moved in the same direction, particularly synchronously, and more preferably simultaneously, and the light source is optionally switched on in this process step. In a subsequent, second process step, the optical system (or the DMD) is moved in the opposite direction to the direction of movement of the substrate holder, preferably at a significantly higher speed. In particular, no exposure takes place in this process step. The simultaneous movement followed by the partial return movement of the optical system is then carried out along the entire line, particularly for each subsequent DMD line. The synchronous, in particular collinear, relative movement of the optical system and the substrate produces a quasi-static exposure.This reduces, preferably completely prevents, the blurring of the exposure caused by relative movement during scanning, and a better image can be achieved.

[0069] In a further improved embodiment, the exposure process is carried out such that, in a first step, a relative movement occurs between the optical system and the layer to be exposed in a first direction. In a second process step, which occurs particularly simultaneously, the optical system (or the DMD) is moved in a second direction that differs from the first direction, and is particularly perpendicular to the first direction. Preferably, this movement takes place in smaller steps than the image width of a single pixel, resulting in a significantly finer positioning grid of the image points. The simultaneous movement followed by the lateral movement of the optical system also allows for an increase in positioning accuracy or overlay.The synchronous, and in particular non-collinear, relative movement of the optical system and substrate reduces, preferably completely prevents, the blurring of the exposure caused by the relative movement during scanning, and a better image can be achieved.

[0070] In the embodiment according to the invention, all elements used for image generation, in particular the optical elements, can be adjusted by means of piezoelectric elements. Electro-, acoustic-, and mechano-optical elements can be used in particular. The image is shifted in the x, y, and / or z direction by means of the control signal. Multiple elements can also be used, which can perform translational and / or rotational movements of the image. The image can be tracked at a speed corresponding to the scan speed. Simultaneous exposure of different structures at different focal planes.

[0071] In a sixth improved and preferred embodiment, the focal plane is tilted relative to the layer to be exposed. During exposure, the position of the focal plane and the depth of field are of fundamental importance for the quality of the imaged structure. The focal plane can be tilted in several ways.

[0072] In a first further development of the sixth embodiment, the focal plane is changed by mechanically tilting the entire DMD. This embodiment is the least preferred.

[0073] In a second, more preferred embodiment of the sixth design, the focal plane is influenced by a targeted modification of optical elements located upstream and / or downstream of the DMD. Tilting the focal plane is accomplished by these optical elements. Preferably, the focal plane is set statically once, and the different ranges are used by the DMD to expose different depths in the layer to be exposed.

[0074] In a third further development of the sixth embodiment according to the invention, several light sources of different wavelengths and / or light sources with a broad wavelength spectrum are used to generate several, in particular continuous, focal planes in the layer to be exposed (or through the layer to be exposed).

[0075] In this embodiment, the layer to be exposed is preferably formed from a material, in particular a polymer, which is sensitive to the corresponding radiation over the entire wavelength range. Dynamic exposure control for grayscale lithography

[0076] In a seventh, improved embodiment, the intensity at each position is selectively controlled by one of the following methods. This allows over- and / or underexposure to be compensated for. Furthermore, lower-quality structures can be improved or exposure gradients can be created. The selective control of the intensity at each position allows, in particular, the following: 3D structures, grayscale lithography and / or optical proximity correction be generated.

[0077] A fundamentally important aspect of using DMDs is ensuring that the intensity of the photons striking the layer to be exposed is spatially correctly distributed. Since no light source is an ideal point source, no light source possesses a homogeneous intensity distribution. Consequently, the light beam striking the DMD is also inhomogeneous. Preferably, this light beam is homogenized by optical elements in the optical system before it strikes the DMD.

[0078] According to the invention, preferably not only is the homogeneity of the light source modified, but the DMD (in particular its micromirror) is adjusted to project a reflected beam with the correct intensity distribution onto the layer to be exposed. Several advantageous technical measures exist for correctly controlling such a complex, spatial intensity distribution.

[0079] The power of the light source is in particular between 0.01 watts and 1000 watts, preferably between 0.1 watts and 500 watts, more preferably between 1 watt and 250 watts, most preferably between 5 watts and 100 watts, and most preferably between 9 and 13 watts.

[0080] The intensity is specified in W / m². By focusing the radiation onto a unit area using optical elements, the intensity can be calculated relatively easily. The intensity of the light source can preferably be controlled very precisely. The intensity is changed either by altering the power of the light source, by the duration of the exposure, and / or by optical elements in the optical system. According to the invention, this makes it possible to direct light beams of different intensities onto the DMD within a well-defined period. By means of appropriate mirror control, individual points of the layer to be exposed can be irradiated with defined intensities. The surrounding points of the layer to be exposed can be illuminated with a different intensity.

[0081] The invention particularly enhances the possibility of targeted, point-like (especially pixel-specific) exposure of the layer to be exposed. By means of relative movement between the optical system and the layer to be exposed, each position can be approached multiple times, thus enabling each pixel to be exposed multiple times. Through measurement using a measuring system and targeted analysis of the surface quality, the multiple exposures can be controlled in such a way that optimally homogeneous exposure down to the pixel level can be achieved.

[0082] Thus, on the one hand, intensity inhomogeneities from the primary beam can be compensated if homogeneous pixel exposure is desired, and on the other hand, different exposure doses can be introduced into the layer to be exposed in a location-specific manner.

[0083] The aforementioned embodiments of the invention further enable the exposure sections (in particular strips) to be overlapped. Through this overlap, individual pixels are exposed multiple times. Based on the number of exposures per pixel, the desired intensity level with which the pixel is exposed can be calculated for each individual exposure.

[0084] Insofar as structures with an intensity gradient are to be imaged, the invention particularly specifies how often a pixel is to be exposed with what intensity so that each pixel, after n repetitions, is exposed to the specified absolute intensity at the end of the exposure process. Anisotropic and / or distorted imaging optics to improve overlay error and / or motion blur.

[0085] In a further, eighth embodiment of the invention, horizontal and / or vertical exposure grid lines of the exposure grid are imaged differently by the optical imaging (i.e., not square), so that a different exposure grid resolution is set in the vertical and horizontal directions. The calculation / control of the exposure is compensated for the deviation.

[0086] In another embodiment, the axes of the image or the exposure grid lines are not arranged orthogonally, but obliquely. The use of such a projection, particularly an affine distortion (especially a shear), enables simple calculation of the illumination positions and linear guidance with precise placement of the illumination points below the grid resolution (subpixel accuracy) for highly accurate shaping of the exposure edges.

[0087] In further embodiments according to the invention, the horizontal and / or vertical exposure grid lines of the exposure grid are not equidistant along the vertical and / or horizontal direction.

[0088] The different exposure grids can be generated from a homogeneously isotropic image of the DMD by the optical elements upstream and / or downstream of the DMD, and / or are a direct result of an anisotropically and / or inhomogeneously constructed DMD.

[0089] In the following, some possibilities are shown for generating a corresponding projection according to the invention, in particular shearing of the exposure grid.

[0090] In a first possible embodiment according to the invention, at least one cylindrical lens with a cylindrical axis is used as the imaging optics to effect a change in the exposure grid. In particular, exactly two cylindrical lenses are used. The cylindrical axes of the cylindrical lenses are preferably parallel to the surface to be exposed. To achieve a shear effect according to the invention, an angle of less than 90°, preferably less than 70°, more preferably less than 50°, and most preferably less than 20° is set between the two cylindrical axes. However, the most preferred angle results from the shear angle to be generated.

[0091] In another embodiment of the invention, the optical system consists of only a single so-called compound lens. A compound lens is understood to be a lens whose surfaces have been ground in such a way that the optical properties are identical to those that would be obtained by combining two lenses.

[0092] The embodiments described above can be achieved with both regular and irregular irradiation of the substrate. In the special case of a regular, but not exactly integer, exposure repetition rate matching the traverse speed, the exposure structures are still activated precisely at the current writing position. This results in non-integer offsets to subpixel placements in the traverse direction, leading to improved placement accuracy and edge roughness.

[0093] A combination of the geometric and / or temporal shifts described above produces sub-pixel resolution in all directions and reduces the susceptibility to errors caused by the failure of individual exposure components. To characterize the correct exposure dose distribution, knowledge of distortions caused by imaging errors and / or artificially induced distortions is required. Linearly distorted or rotated images offer the advantage of simpler calculations and easier light source control. Reduction of motion blur

[0094] In a further, ninth embodiment of the invention, the relative speed and / or motion between the optical system and the layer to be exposed is changed such that several pixels within a single pixel size overlap. This overlap can be interpreted as a long exposure of a moving object. This results in a blurring of the generated pixel in the direction of relative motion. By defining when one of the pixels is written or not written, the pixel size in the direction of relative motion can be precisely controlled. For illustration, see the Figures 8a-b or the corresponding character description.

[0095] In other words, two sequentially illuminated exposure grids are shifted / moved with a relative displacement between the micromirror device and the light-sensitive layer by less than one pixel width, preferably less than half a pixel width, more preferably less than a quarter of the pixel width.

[0096] If the reduction or enlargement of the optics in the direction of the scan movement is significantly less than in the direction orthogonal to it, the effect of motion blur resulting from the finite exposure time can be partially compensated. This makes it possible to achieve an isotropic image. Compression in the direction of movement does not reduce the motion blur itself, but only the overall size of the exposure spot in the direction of travel.

[0097] All the embodiments and processes according to the invention mentioned above can be combined with one another as desired, but are described individually. Where process features are described, these shall also be deemed disclosed as device features, and vice versa.

[0098] Further advantages, features, and details of the invention will become apparent from the following description of preferred embodiments and from the drawings. These show: Figure 1 a first embodiment of the device according to the invention, Figure 2 a second embodiment of the device according to the invention, Figure 3 a third embodiment of the device according to the invention, Figure 4a a schematic, not to scale, representation of a DMD (micromirror device) with an enlarged section with micromirrors in a first position, Figure 4b a schematic, not to scale, representation of a DMD (micromirror device) with an enlarged section with micromirrors in a second position, Figure 5a a schematic, not to scale, view of a top view of a first exposed section of a layer to be exposed, Figure 5b a schematic, not to scale, view of a second exposed section of the layer to be exposed, slightly offset from the first, Figure 6a a schematic,Figure 6a: schematic, non-scale view of a substrate with a first exposed section; Figure 6b: schematic, non-scale view of a substrate with a second exposed section; Figure 6c: schematic, non-scale view of a substrate with the superimposed first and second exposed sections; Figure 7a: schematic, non-scale view of a further embodiment of the device according to the invention; Figure 7b: schematic, non-scale view of a further embodiment of the device according to the invention; Figure 7c: schematic, non-scale view of a further embodiment of the device according to the invention; Figure 8a: schematic, non-scale view of a further exposed section of a layer to be exposed according to a first embodiment of the method according to the invention.Figure 8a is a schematic, not-to-scale, top view of a further exposed section of a layer to be exposed according to a second embodiment of the method according to the invention; Figure 9a is a schematic, not-to-scale, enlarged view of a first embodiment of an exposure grid according to the invention; Figure 9b is a schematic, not-to-scale, enlarged view of a first embodiment of an exposure grid according to the invention; Figure 10 is a schematic view of an intensity distribution of two adjacent pixels with three different intensity levels; Figure 11 is a schematic view of an exposure grid with several exposed pixels; and Figure 12 is a schematic view of an exposure grid distorted by optical elements.

[0099] In the figures, identical components or components with the same function are marked with the same reference symbols.

[0100] Figure 1 Figure 1 shows a first embodiment comprising an optical system 8 with at least one light source 7 and at least one DMD 1 (micromirror device), and a substrate holder 11. The substrate holder 11 can be moved relative to a coordinate system K3.

[0101] A substrate 10 is fixed to the substrate holder 11 using fixing agents 13. A light-sensitive layer 9 made of an exposeable material is located on the substrate and is exposed by means of the device.

[0102] The origin of a sample-fixed (i.e., fixed to the substrate 10 or the layer 9 to be exposed) coordinate system K2 is preferably placed in the center of the surface 90 of the layer 9.

[0103] A light beam 6 (primary light beam) emitted by the light source 7, which can pass through several optical elements (not shown) on its way to the DMD 1, is converted by the DMD 1 into a structured light beam 6' (secondary light beam). This can pass through several optical elements (not shown) on its way to layer 9.

[0104] By means of a semi-transparent mirror 14", a detector 19, in particular a camera, more preferably a CCD or CMOS camera, can directly detect and / or measure the surface 90 of the layer 9 to be exposed. The measurement results are preferably used for direct control of the process and / or calibration of the device. For the sake of clarity, the depiction of such measuring means is omitted in the subsequent figure descriptions and figures. However, the measuring means according to the invention can be used in any of the embodiments mentioned.

[0105] Figure 2 Figure 1 shows a second embodiment, wherein the optical system 8 is equipped with two light sources 7, 7'. Light rays 6 are emitted by the two light sources 7, 7'. One of the light rays 6 is redirected by a mirror 14 onto a beam splitter 14' and combined by this with the light ray 6 of the second light source 7'.

[0106] The combined light beam 6 is directed onto the DMD 1 and converted by it into a structured light beam 6', which in turn can pass through several optical elements (not shown) on its way to layer 9.

[0107] One particularly independent aspect of the invention is that the two light sources 7 can differ in radiation intensity, wavelength, coherence length and optionally other properties or parameters, so that a laser beam 6 with a multitude of different optical parameters can be generated.

[0108] According to the invention, more than 2, more preferably more than 5, even more preferably more than 10, and most preferably more than 20 light sources 7, 7' can be used. Each light source can preferably also be an LED array or a laser diode array.

[0109] Figure 3 Figure 1 shows a third embodiment consisting of an optical system 8 with at least one light source 7 and two DMDs 1.

[0110] A light beam 6 is emitted from the light source 7 and split by a beam splitter 14'. A first split light beam 6.1 is modified by a first DMD 1 to a first modified beam 6.1'. The layer 9 is exposed by the first modified beam 6.1'. The second split light beam 6.2 is redirected by a mirror 14 to a second DMD 1 and forwarded as a second modified beam 6.2' to the layer 9. Preferably, a different position of the layer 9 is exposed by the second modified beam 6.2' than by the first modified light beam 6.1'. All of the aforementioned light beams can pass through several optical elements (not shown).

[0111] One particularly independent aspect of the invention is that at least two DMDs 1 are used, by means of which exposure can be carried out simultaneously at two different positions of the layer 9, preferably using a single, and in particular combined, light beam for the illumination of the DMDs. This leads in particular to a widening of the exposure area, in particular exposure strip, and thus to an increase in throughput.

[0112] Figure 4aFigure 1 shows DMD 1 with a mirror surface 2. The enlarged view of a portion of the mirror surface 2 shows several (16) mirrors 3 of a plurality of mirrors 3. The mirrors are arranged in an untilted orientation, referred to as the initial position. A coordinate system K1 is assigned to DMD 1. The Z-axis of K1 (i.e., K1z) is perpendicular to the mirror surface 2, and the x- and y-coordinates are parallel to the mirror surface edges 2kx and 2ky of the mirror surface 2 and define a mirror plane.

[0113] Figure 4b The same DMD 1 is shown, with one of the mirrors 3 arranged in a position tilted or rotated about the x-axis. The part of the light beam 6 that strikes the tilted mirror 3 is therefore reflected in a direction that is not identical to the direction of reflection of the light beam 6 reflected by the non-tilted mirrors 3.

[0114] Figure 5aFigure 1 shows a schematic view of the mirror surface 2 with a central, in particular strip-shaped, writing area 4 and two buffer areas 5 adjacent to the edge of the writing area 4, preferably parallel.

[0115] Instead of the mirrors 3, the pixels 23 of an exposure grid 24 reflected by the mirror surface 2 are shown, which correspond to structures 12 to be formed at position K2y=12 on the layer 9 to be exposed (possibly modified by optical elements between the DMD 1 and the layer 9 to be exposed).

[0116] According to an advantageous embodiment of the invention, only the mirrors 3 arranged in the writing area 4 are used for exposure, so that a writing buffer is formed by the buffer areas 5, which is explained below. The center line D runs through the defined center point 0 of the K2x axis of the sample coordinate system K2.

[0117] The Figure 5b The image shows the same DMD 1, but after a relative movement of 0.5 au in the K2y direction. DMD 1 is therefore located at position K2y = 12.5 au. Simultaneously, it can be seen that a relative movement of approximately 2 au has also occurred in the K2x direction. This relative movement is unintentional and results, for example, from a mounting error. It is clearly visible how structure 12 has been shifted to the left relative to the DMD coordinate system K1 in order to expose it correctly relative to the K2y coordinate system. The write buffer is thus used.

[0118] The targeted programming of the DMD therefore allows for the correction of mechanical errors. The substrate holder 9 thus does not move the substrate 10, and therefore layer 9, in a perfectly straight line towards K2y, but is slightly shifted towards K2x during the movement towards K2y.

[0119] According to the invention (independent aspect of the invention), preferably the mechanics of the substrate holder 9 are not used for error correction, but rather the writing area 4 and the buffer areas 5 are electronically programmed / controlled so that the structures 12 to be exposed are shifted accordingly (here in the negative K1x direction). Thus, the electronics and / or the mechanics of the DMD 1 compensate for writing errors, in this case the mechanical error of the substrate holder 11.

[0120] The Figure 6a The figure shows a top view of a layer 9 exposed along a first strip 15. The strip 15 corresponds to the area of ​​layer 9 exposed by the writing area 4 of the DMD 1 after a movement of the DMD 1 relative to the exposed layer 9 in the K2y direction.

[0121] In an area between intensity variation ranges 161, 16r of strip 15, the illuminated pixels 23 are illuminated with the most homogeneous intensity possible.

[0122] In contrast, the illuminated pixels 23 in the intensity variation ranges 161, 16r are controlled such that the intensity of the reflected light beam 6' from the writing area 4 towards the edge of the DMD 1 decreases, in particular continuously, preferably proportionally to the distance from the writing area 4. A corresponding raster intensity profile, from which the intensity profile as a function of position can be read, is shown (Intensity / Position). The intensity with which the layer 9 is exposed thus exhibits a maximum, in particular in the area of ​​the strip 15, and decreases laterally and preferably continuously and / or linearly to zero.

[0123] The Figure 6b shows an analogous representation to Figure 7aRegarding a second strip 15', which in particular follows directly after the exposure of the first strip 15. This strip is shifted so far to the right compared to the first strip 15 that the right intensity variation range 16r is removed from the Figure 6a with the left intensity variation range 161' from the Figure 7b overlapping, preferably congruent. The pixels 23 are switched in the right intensity variation range 16r in the same way as in the left intensity variation range 161', whereby the intensity of the corresponding pixels 23 are each summed to a pixel intensity that corresponds to the intensity in strip 15.

[0124] The Figure 6cThe image shows a top view where the intensity variation range 16r of the first strip 15 has been aligned with the intensity variation range 161' of the second strip 15', resulting in a constant intensity profile. The exposure is therefore homogeneous, as the intensities of the first strip 15 and the intensities of the second strip 15' are summed.

[0125] The Figure 7aFigure 1 shows an enlarged section of an embodiment according to the invention with a DMD 1, optical paths of the primary light beam 6 and the secondary light beam 6', and the layer 9 to be exposed. The primary light beam 6 and the secondary light beam are symbolized here only by the optical paths 6, 6' and are preferably large enough to illuminate the entire DMD 1. The secondary light beam 6' is perpendicular to the layer 9 to be exposed. The focal plane 17 is, in particular, parallel to, and preferably on, the surface 90. A depth-of-field area 18 indicates the depth within which a sharp image of the pixels 23 can be produced.

[0126] The Figure 7b shows an enlarged section of a preferred embodiment compared to the embodiment according to Figure 7aIn the modified embodiment according to the invention, the secondary light beam 6' is reflected at an angle α onto the layer 9 to be exposed. The focal plane 17 thus intersects the layer 9 to be exposed at an angle α. It lies outside the layer 9 to be exposed on the left side and within the layer 9 to be exposed on the right side. The depth of field 18 therefore penetrates deeper into the layer 9 on the right side and can thus be used to produce sharp structures in three-dimensional structures in depressions without moving the focal plane 17 in the K2z direction by moving the DMD 1.

[0127] The need to shift DMD 1 in the K2z direction to achieve sharper imaging of deeper structures can be circumvented, in particular, by tilting DMD 1. Dynamic shifts of the exposure areas on DMD 1 thus allow for targeted, sharp exposure at specific depths. A significant advantage over multiple exposures is the ability to create highly precise structures in all spatial directions without mechanical errors.

[0128] The Figure 7c shows an enlarged section of a more preferred embodiment compared to the embodiments according to Figure 7a and / or Figure 7b in the modified embodiment according to the invention. The secondary light beam 6' is deflected by optical elements (not shown) located between the DMD 1 and the layer 9 to be exposed, such that the focal plane 17 is inclined to the substrate surface 90 by an angle α.

[0129] The Figure 8aFigure 1 shows a top view of a section of layer 9 to be exposed. A pixel 23 (smallest unit of an exposure grid 24) was illuminated by the DMD 1 along a length 1 by means of a relative movement between the DMD 1 (not shown) and the layer 9 to be exposed, and by illuminating pixel 23 throughout the entire travel path along length 1. The exposed areas correspond to the pixels 23, with the exposure intensity controlled by the illumination intensity. In the case of overlaps, the intensities in the overlapping areas sum.

[0130] The exposure took place from K2y position 0 to K2y position 3. A blurring effect in the direction of relative movement is created by taking multiple exposures within the range of approximately -2 to 5 during the relative movement. The intensity profiles of pixel 23 overlap and produce a strong increase in intensity along the path 1.

[0131] The Figure 8b shows alternative embodiment to Figure 8a , the difference being to Figure 8aThe shorter length 1' is achieved by having the corresponding mirror 3 of the DMD 1 (both not shown) begin exposure only at K2y position 1 and end exposure at K2y position 2. By selectively controlling the mirrors 3, particularly by switching them on later or off earlier, it is possible to increase the resolution in the K2y direction. In this specific case, the exposure of pixel 23 is therefore delayed by 33.33% and ended 33.33% earlier. This is because it is not possible to expose pixel 23 square-shaped, as each beam has an intensity profile that deviates from a step shape. The exposure grids 24, 24', 24", 24‴ described below are shown as guidelines that can be visualized as lying over the layer 9 to be exposed. Preferably, the grids correspond to the dimensions of pixel 23.

[0132] The Figure 9aFigure 1 shows a first, less preferred exposure grid 24 according to the invention, which has equidistant spacing of the exposure grid lines 27 in the two mutually orthogonal directions K2x, K2y. The exposure grid 24 is thus isotropic and homogeneous in both directions K2x and K2y.

[0133] The Figure 9b Figure 24 shows a second, more preferred exposure grid 24' according to the invention, which has a separate, and in particular equidistant, spacing between the exposure grid lines 27 for each direction. The exposure grid 24' is thus anisotropic, but homogeneous in each of the directions K2x and K2y.

[0134] It is also conceivable that the exposure is due to Exposure grid line intersection points 25 and / or exposure grid sub-areas 26 and not within the individual grid areas.

[0135] The different exposure grids 24, 24', 24", 24''. can be generated / modified, in particular, by optical elements (not shown) located upstream and / or downstream of the DMD 1 (not shown). The DMD 1 (not shown) would preferably be isotropic and homogeneous, wherein the optical elements (not shown), in particular only the downstream ones, are designed to produce an anisotropic and / or non-homogeneous imaging of the DMD.

[0136] The Figure 10Figure 1 shows a schematic cross-sectional view of two mirrors 3, as well as the mirror intensity profiles 22, 22', 22' (especially Gaussian profiles) resulting from three different sets of parameters and / or structural modifications of the mirrors 3, and the pixels 23, 23', 23" generated by these mirror intensity profiles 22, 22', 22". It can be seen that the intensity distributions overlap more and more with increasing parameters characterizing the distribution functions, especially the full width at half maximum (FWHM), FWHM', FWHM" (FWHM), so that two adjacent pixels 23, 23', 23" become more blurred. The result is a very homogeneous exposure grid at the most overlapping pixels 23".

[0137] The Figure 11Figure 1 shows a schematic, not-to-scale top view of several pixels 23 on a 5x5 exposure grid 24. A pattern 28 is visible, which is illuminated by the targeted switching of the corresponding mirrors 3. The mirror intensity profiles of the mirrors 3 are so sharp that the intensity maxima are clearly recognizable and the intensity fall-off is so steep that the intensity distribution per mirror 3 (not shown) is very strongly limited to the associated exposure grid sub-area 26. In particularly preferred embodiments of the invention, the mirror intensity profiles are partially superimposed beyond the exposure grid sub-area 26, as is the case with the implementation of the pixels 23" according to [reference missing]. Figure 10 that would be the case.

[0138] The Figure 12Figure 1 shows a schematic, not-to-scale top view of an exposure grid 24", distorted in particular by optical elements of the optical system 8. The optical elements reflect the partial rays reflected by the mirrors 3 of the DMD 1 orthogonally onto the layer 9 to be exposed; however, distortion preferably occurs exclusively within the K2x-K2y plane. This method according to the invention allows an exposure grid 24" to be optically created, resulting in an increase in the overlay according to the invention. In this embodiment, the DMD 1 is preferably not tilted, but rather the original image of the DMD 1 is affinely distorted to effect the tilting of the exposure grid 24". Reference symbol list

[0139] 1DMD 2Mirror surface 2kx, 2kyMirror surface edges 3,3'Mirror 4Writing area 5Buffer area 6Light beam 6'Modified / structured light beam 6.1'First modified beam 6.2'Second modified beam 7, 7'Light sources 8Optical system 9Layer 10Substrate 11Substrate holder 12, 12', 12"Structures 13Fixing agent 14Mirror 14'Beam splitter 14"Semitransparent mirror 15, 15',15"Stripes 161,16r,161',16r',16r"Intensity variation range 17Focus plane 18Depth of field 19Detector 20Point grid 22, 22', 22"Mirror intensity profiles 23, 23'Pixel 24, 24', 24", 24‴ Exposure grid 25 Exposure grid line intersection 26 Exposure grid sub-area 27 Exposure grid line 28 Pattern 1, 1' Length b Width D Travel direction v Vertical dot grid spacing h Horizontal dot grid spacing r Exposure point radius p Mirror center distance

Claims

1. A method for exposing a light-sensitive layer (9) with a optical system (8), wherein at least one light beam (6, 6') is produced by at least one light source (7) and pixels (23) of a exposure raster (24, 24', 24", 24‴) are illuminated by at least one micromirror device (1) with a plurality of micromirrors (3) each with a mirror intensity profile (22, 22', 22") and a continuous and / or sequential scanning, maskless imagesetter, characterised by the fact that a superposition of mirror intensity profiles (22, 22', 22") of adjacent micromirrors (3) to form a raster intensity profile of the exposure raster (24, 24' 24" 24‴) takes place as the sum of the mirror intensity profiles (22, 22', 22") of each illuminated pixel (23) of the exposure raster (24, 24', 24" 24‴).

2. The method according to claim 1, wherein the optical system (8) is shifted into a second direction deviating from a first direction of the continuously and / or sequentially scanning, maskless imagesetter, in particular arranged normal to the first direction.

3. The method according to claim 1 or 2, wherein the mirror intensity profiles, in particular due to the geometric shape of the micromirrors (3), generate pixels (23) that are larger than individual grids of the exposure raster (24, 24', 24", 24"').

4. The method according to any one of the preceding claims, wherein at least two exposure rasters (24, 24', 24", 24"') are illuminated, in particular sequentially, and the raster intensity profiles of the exposure rasters (24, 24', 24", 24"') are superimposed as a sum to form an exposure intensity profile of the light-sensitive layer (9).

5. The method according to any one of the preceding claims, wherein the superposition takes place by blurred image of the micromirrors (3) in the exposure raster (24, 24', 24", 24"').

6. The method according to any one of claims 4 or 5, wherein two exposure rasters (24, 24', 24", 24‴) illuminated sequentially one after another are displaced with a relative displacement between the micromirror device (1) and the light-sensitive layer (9) of less than one pixel width, preferably less than half a pixel width, even more preferably less than a quarter of a pixel width.

7. The method according to any one of the preceding claims, wherein exposure raster grid lines of the exposure raster (24, 24', 24", 24‴) that are horizontal and, in particular, running parallel to one, and / or that are vertical and, in particular, running parallel to one another are arranged running distorted and / or are, in particular affine.

8. A device for exposing a light-sensitive layer (9) with an optical system (8) comprising: - at least one light source (7) for generating of at least one light beam (6, 6'), - a continuously and / or sequentially scanning, maskless imagesetter, - at least one micromirror device (1) with a plurality of micromirrors (3), wherein each micromirror (3) serves to illuminate one pixel (23) of an exposure raster (24, 24', 24", 24"') with a mirror intensity profile (22, 22', 22"), characterised in that the optical system (8) is designed in such a way that a superposition of mirror intensity profiles (22, 22', 22") of adjacent micromirrors (3) to form a raster intensity profile of the exposure raster (24, 24', 24", 24"') takes place as the sum of the mirror intensity profiles (22, 22', 22") of each illuminated pixel (23) of the exposure raster (24, 24', 24", 24‴).